From sipping-bounces@ietf.org Thu Feb 01 06:05:20 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HCZkA-0003J5-Cq; Thu, 01 Feb 2007 06:04:42 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HCZk9-0003IW-Ej
	for sipping@ietf.org; Thu, 01 Feb 2007 06:04:41 -0500
Received: from wip-ec-wd.wipro.com ([203.91.193.32])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HCZk6-0006vj-HK
	for sipping@ietf.org; Thu, 01 Feb 2007 06:04:41 -0500
Received: from wip-ec-wd.wipro.com (localhost.wipro.com [127.0.0.1])
	by localhost (Postfix) with ESMTP id B5060205E2
	for <sipping@ietf.org>; Thu,  1 Feb 2007 16:27:05 +0530 (IST)
Received: from blr-ec-bh01.wipro.com (blr-ec-bh01.wipro.com [10.201.50.91])
	by wip-ec-wd.wipro.com (Postfix) with ESMTP id A0A73205D9
	for <sipping@ietf.org>; Thu,  1 Feb 2007 16:27:05 +0530 (IST)
Received: from blr-m2-msg.wipro.com ([10.116.50.99]) by blr-ec-bh01.wipro.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 1 Feb 2007 16:34:36 +0530
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Thu, 1 Feb 2007 16:39:16 +0530
Message-ID: <532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdFSTeGOeXbq/LdTDaGeEGh4VH5iQAp3zYQ
References: <45C0AEF2.8010101@ericsson.com>
From: <ravishankar.shiroor@wipro.com>
To: <Gonzalo.Camarillo@ericsson.com>,
	<sipping@ietf.org>
X-OriginalArrivalTime: 01 Feb 2007 11:04:36.0774 (UTC)
	FILETIME=[C2E71460:01C745F0]
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 00e94c813bef7832af255170dca19e36
Cc: mary.barnes@nortel.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hello,

IMO, Service Identification is definitely needed and should be taken up
as a WG as proposed below.

Further, I think Service Identification scope should be as per the
requirement in Jonathan's draft, as I still do not understand the need
for having a Communication Service Identifier and another Application
Identifier as given in the 3gpp-ics-requirements document.

Regards,
Ravi.

--
=20
Ravishankar. G. Shiroor
Wipro Technologies, Bangalore.
=20
ravishankar.shiroor@wipro.com
--
=20

> -----Original Message-----
> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]=20
> Sent: Wednesday, January 31, 2007 8:30 PM
> To: sipping
> Cc: Mary Barnes
> Subject: [Sipping] Service Identification in SIP
>=20
> Folks,
>=20
> during the last year, there have been a number of threads on=20
> the mailing list discussing service identification in SIP. In=20
> particular, the following thread discussed 3GPP requirements=20
> on this area:
>=20
> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
>=20
> At a later point, those 3GPP requirements were documented in=20
> the following (expired by now) Internet Draft, which was=20
> presented in a face-to-face IETF meeting.
>=20
> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
> cs-requirements-00.txt
>=20
> Lately, the following draft has been submitted to the archives:
>=20
> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
> rvice-identification-00.txt
>=20
> This draft proposes a solution that intends to meet known=20
> requirements in this area.
>=20
> Given the amount of discussions on issues related to service=20
> identification we have seen during the last year, we believe=20
> this is a fairly important topic for the SIPPING WG to=20
> tackle. At some point, we would like to have a milestone with=20
> its associated WG item dealing with this issue.
>=20
> Therefore, we would like to encourage people to read this=20
> draft and send comments to the list. Additionally, if you=20
> believe that this work is (or is not) important, we would=20
> like to hear that as well.
>=20
> Cheers,
>=20
> Gonzalo
> SIPPING co-chair
>=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 01 15:51:12 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HCit6-0003r8-LQ; Thu, 01 Feb 2007 15:50:32 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HCisk-0002bi-1y; Thu, 01 Feb 2007 15:50:10 -0500
Received: from ns1.neustar.com ([2001:503:c779:1a::9c9a:108a])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HCisd-0006qv-JE; Thu, 01 Feb 2007 15:50:10 -0500
Received: from stiedprstage1.ietf.org (stiedprstage1.va.neustar.com
	[10.31.47.10]) by ns1.neustar.com (Postfix) with ESMTP id 220B6272B4;
	Thu,  1 Feb 2007 20:50:03 +0000 (GMT)
Received: from ietf by stiedprstage1.ietf.org with local (Exim 4.43)
	id 1HCisc-00086X-WA; Thu, 01 Feb 2007 15:50:03 -0500
Content-Type: Multipart/Mixed; Boundary="NextPart"
Mime-Version: 1.0
To: i-d-announce@ietf.org
From: Internet-Drafts@ietf.org
Message-Id: <E1HCisc-00086X-WA@stiedprstage1.ietf.org>
Date: Thu, 01 Feb 2007 15:50:02 -0500
X-Spam-Score: -2.5 (--)
X-Scan-Signature: b7b9551d71acde901886cc48bfc088a6
Cc: sipping@ietf.org
Subject: [Sipping] I-D ACTION:draft-ietf-sipping-policy-package-03.txt 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

--NextPart

A New Internet-Draft is available from the on-line Internet-Drafts 
directories.
This draft is a work item of the Session Initiation Proposal Investigation Working Group of the IETF.

	Title		: A Session Initiation Protocol (SIP) Event Package for Session-Specific Session Policies.
	Author(s)	: V. Hilt, G. Camarillo
	Filename	: draft-ietf-sipping-policy-package-03.txt
	Pages		: 18
	Date		: 2007-2-1
	
This specification defines a Session Initiation Protocol (SIP) event
   package for session-specific policies.  This event package enables
   user agents to subscribe to session policies for a SIP session and to
   receive notifications if these policies change.

A URL for this Internet-Draft is:
http://www.ietf.org/internet-drafts/draft-ietf-sipping-policy-package-03.txt

To remove yourself from the I-D Announcement list, send a message to 
i-d-announce-request@ietf.org with the word unsubscribe in the body of 
the message. 
You can also visit https://www1.ietf.org/mailman/listinfo/I-D-announce 
to change your subscription settings.

Internet-Drafts are also available by anonymous FTP. Login with the 
username "anonymous" and a password of your e-mail address. After 
logging in, type "cd internet-drafts" and then 
"get draft-ietf-sipping-policy-package-03.txt".

A list of Internet-Drafts directories can be found in
http://www.ietf.org/shadow.html 
or ftp://ftp.ietf.org/ietf/1shadow-sites.txt

Internet-Drafts can also be obtained by e-mail.

Send a message to:
	mailserv@ietf.org.
In the body type:
	"FILE /internet-drafts/draft-ietf-sipping-policy-package-03.txt".
	
NOTE:	The mail server at ietf.org can return the document in
	MIME-encoded form by using the "mpack" utility.  To use this
	feature, insert the command "ENCODING mime" before the "FILE"
	command.  To decode the response(s), you will need "munpack" or
	a MIME-compliant mail reader.  Different MIME-compliant mail readers
	exhibit different behavior, especially when dealing with
	"multipart" MIME messages (i.e. documents which have been split
	up into multiple messages), so check your local documentation on
	how to manipulate these messages.

Below is the data which will enable a MIME compliant mail reader
implementation to automatically retrieve the ASCII version of the
Internet-Draft.

--NextPart
Content-Type: Multipart/Alternative; Boundary="OtherAccess"

--OtherAccess
Content-Type: Message/External-body; access-type="mail-server";
	server="mailserv@ietf.org"

Content-Type: text/plain
Content-ID: <2007-2-1142350.I-D@ietf.org>

ENCODING mime
FILE /internet-drafts/draft-ietf-sipping-policy-package-03.txt

--OtherAccess
Content-Type: Message/External-body;
	name="draft-ietf-sipping-policy-package-03.txt";
	site="ftp.ietf.org"; access-type="anon-ftp";
	directory="internet-drafts"

Content-Type: text/plain
Content-ID: <2007-2-1142350.I-D@ietf.org>


--OtherAccess--

--NextPart
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--NextPart--





From sipping-bounces@ietf.org Tue Feb 06 06:54:54 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HEOsw-0006Q0-0U; Tue, 06 Feb 2007 06:53:18 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HEOsu-0006Mp-G7
	for sipping@ietf.org; Tue, 06 Feb 2007 06:53:16 -0500
Received: from mailgate.siemenscomms.co.uk ([195.171.110.225]
	helo=bemg01.siemenscomms.co.uk)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HEOsr-00076C-Or
	for sipping@ietf.org; Tue, 06 Feb 2007 06:53:16 -0500
Received: from ntht207e.uksgcs.siemenscomms.co.uk ([137.223.247.82])
	by siemenscomms.co.uk (PMDF V6.0-24 #40642)
	with ESMTP id <0JD10053LICCR5@siemenscomms.co.uk> for sipping@ietf.org;
	Tue, 06 Feb 2007 11:53:00 +0000 (GMT)
Received: by ntht207e.uksgcs.siemenscomms.co.uk with Internet Mail Service
	(5.5.2657.72)	id <D2K3T41M>; Tue, 06 Feb 2007 11:53:00 +0000
Content-return: allowed
Date: Tue, 06 Feb 2007 11:52:58 +0000
From: "Elwell, John" <john.elwell@siemens.com>
Subject: RE: [Sipping] draft-ietf-sipping-config-framework-10
To: Sumanth Channabasappa <sumanth@cablelabs.com>
Message-id: <50B1CBA96870A34799A506B2313F26670AF57275@ntht201e.siemenscomms.co.uk>
MIME-version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
Content-type: text/plain
Content-transfer-encoding: 7BIT
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 32029c790f79bd4a84a26bd2915c54b9
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Sumanth,

Sorry I missed your deadline by one day. Also I have not reviewed the entire
document, but only the initial sections.

In general I find it hard to read and understand, and I can't help feeling
that this is an over-complex solution. This is due in part to the fact that
we have three different profile types, and it is not clear to me that this
is needed in a lot of deployments. The use cases you give are certainly not
all that compelling, and I can't help feeling that a much simpler solution
would serve a large proportion of cases and would be a lot more attractive
to implement.

Secondly it seems to me that we haven't got the terminology consistent yet.
In particular, I see various uses of "service" and "subscription" (and
variations), which need to be straightened out.

Specific comments:

Section 1, last paragraph. We start talking about modifying profiles here,
without having used the term profile in the preceding paragraph on
provisioning.

Section 2, definition of client. This definition seems to include things
like gateways and media servers. Is this the intention? The concept of user
may differ for gateways, for example.

Section 3, definition of device.  I see no point in using two terms (device
and client) for the same thing. I prefer "device", because SIP has its own
use for "client".

Section 3.1 "Profile Notification Component (PNC), responsible for enrolling
Clients in Profile event subscriptions..." Are enrollment and subscription
two separate procedures or the same? If the same, use a single term, e.g.,
"...responsible for accepting profile event subscriptions from clients..."
However, we should only use the term "subscription" in the context of a SIP
subscription, otherwise it gets confusing.

Section 3.1, figure. I don't think it is helpful to introduce different
profile types in the figure annotations without having described them first.

Section 3.1, figure, "Provides 'Client'(e.g. allowed Users) & 'User'(such as
Services) profile" and "Provides 'Local Network' profile data (e.g. STUN
Server Address)" The definition of service provider is fairly general, but
these annotations in the figure seems to imply it is a provider of profiles.
I guess this is made clear in the text much further down, but it is
confusing to see it in the figure without it already having been explained.

Section 3.1 "The observations above show a need for a clear distinction
between different Profile Types".  I think what it shows is the need for a
device to obtain configuration information from different sources. We can
choose to call a set of configuration information from a particular source a
profile. The device needs therefore to obtain multiple profiles.

Section 3.2, descriptions of Profile Discovery etc. These sound like
definitions, but we already have definitions of most of these terms in the
terminology section. So if the formal definition is in the terminology
section, here we should word them in a way that does not suggest they are
formal definitions.

Section 3.2 "Profile Enrollment:  The process by which a Client makes itself
known to a PDS.  While enrolling, the Client provides identity information
and requested Profile Type(s) for profile retrieval."  Again it is not clear
to me why we need to separate enrollment and subscription.

Section 3.2 "SUBSCRIBE session". What is a "SUBSCRIBE session"? Can't we
just use "subscription" or "SUBSCRIBE-initiated dialog"? In SIP, the word
"session" has a different meaning.

Section 3.3. In contrast to 3.2 we have some definitions that are not
present in the Terminology section.

Section 3.3 "User Profile:  Contains configuration data related to the
specific User, as required to reflect that User's preferences and the
particular services subscribed to."  I am not sure about the concept of
subscription here (in SIP, subscription has a particular meaning). Can't we
just replace "and the particular services subscribed to" by "and
entitlements".

Section 4.1. I don't think this example is compelling. Why would a data
service provider provide a SIP-based PDS when it is not a SIP service
provider?

Section 4.1, item B, "The Client proceeds to request the 'local-network'
Profile Type. The PDS in the local network responds, allowing the Client to
retrieve the local-network profile". We should describe here how the client
address the request to the PDS.

Section 4.1, item C, "based on the subscription status". What subscription
status is this? Why would a device (as opposed to a user) have to
'subscribe' prior to being able to be used? How does the device identify
itself to the PDS so that its 'subscription' can be verified?

Section 4.1, item D, "The Client retrieves the User profile (this can
contain information such as service profiles to be retrieved". What is a
service profile? I thought we just had network, device and user profiles.

Section 4.1, item D, "The Client then starts providing services." What is
meant by "providing services". Normally a server provides services, not the
client. Does it just mean that the device commences normal operation (e.g.,
registering with a SIP server)?

Section 4.2 "SIP Service Provider A provides the local network connectivity,
'local-network' and 'device' profiles for the Client.  The Service Provider
also provides 'user' profiles for existing subscribers".  Why is this
necessarily a SIP service provider? I would imagine that the device would
perform SIP registration with Service Provider B (on a per user basis) and
would not register with Service Provider A. So to call it a SIP Service
Provider is misleading. OK, in this case if there are some users who happen
to be supported by this provider, then I guess it is a SIP service provider,
but if all users are supported by other service providers then it is not.

Section 4.2, figure "User A attempts services". I think it is attempting
subscription to the event package - that is all.

Section 4.2, item C, "The Client initializes the User interface for
services". What services?

Section 5.1 "The first step to obtaining a profile is PDS Discovery.  This
is accomplished by creating a profile subscription using the Event Package
described in Section 6, and preparing for Profile Enrollment." This is
circular - you have to discover the PDS before you can submit a subscription
request to it.

Section 5.1, second paragraph "Client identifier". I think we need to say
something here about what this is and how a PDS would use it. Taking a MAC
address as an example, is this likely to be meaningful to the Device Profile
PDS?

Section 5.1.1 "the user part of the "From" field MUST be the Identifier that
the Client will use to request the 'device' Profile Type". Why is this
needed if we are only requesting network profile? A network PDS is unlikely
to have any knowledge of particular devices.

Section 5.1.1 "a user AOR, if known to the Client MUST be provided in the
"network-user" event header parameter". Again, why is this needed if we are
only requesting the network profile?

Section 5.1.1 "the uniqueness of the From field provides an important
capability" As far as the network is concerned, isn't the IP address a
sufficient distinction between devices?

John

> -----Original Message-----
> From: Sumanth Channabasappa [mailto:sumanth@cablelabs.com] 
> Sent: 30 January 2007 20:15
> To: sipping
> Subject: RE: [Sipping] draft-ietf-sipping-config-framework-10
> 
> Folks,
> 
> As indicated earlier, -v10 was a draft update and we have received
> limited comments thus far (thanks to those who provided comments). If
> there are additional comments that participants want to 
> share, it would
> be beneficial to receive them by Feb 5, 2007 to address them 
> in a timely
> manner. If you have not reviewed -v10 as yet, this email is a 
> request to
> please do so (as a note, please check the open items for things not
> completely addressed in this rev).
> 
> If you wish to participate in offline discussions, please 
> send and email
> to the co-chairs, the author or myself.
> 
> Thanks!
> - S
> 
> 
> 
> -----Original Message-----
> From: Sumanth Channabasappa 
> Sent: Wednesday, January 24, 2007 1:55 PM
> To: sipping
> Subject: [Sipping] draft-ietf-sipping-config-framework-10
> 
> Folks,
> 
> Now that v10 is available online
> (http://www.ietf.org/internet-drafts/draft-ietf-sipping-config
> -framework
> -10.txt), here is a brief summary of the changes in this version
> (reproduced in part from section 12.1).
> 
>    Changes incorporated by the author:
>    o  Incorporated numerous edits and corrections from 
> CableLabs review
>    o  Used better ascii art picture of overview from Josh Littlefield
>    o  Fixed the normative text for network-user so that it is now
>       consistant: MAY provide for device profile, MUST provide for
>       local-network profile
> 
>    Changes incorporated by the editor:
>    o  Restructured document based on a) Keith's recommendations in
>       the email dated 11/09/2006 and responses (Peter, 
> Sumanth, Josh) b)
>       subsequent discussions by the ad-hoc group consisting of the
>       editor, the author, expert contributors (Peter Blatherwick, Josh
>       Littlefield, Alvin Jiang, Jason Fischl, Martin Dolly, Cullen
>       Jennings) and the co-chairs 
>    o  Modified the contents of various sections as part of the
> restructuring 
>       efforts (notably: Overview, Introduction, Framework 
> Requirements &
> 
>       Security)
>    o  Revised Use cases and added detailed examples
> 
> This (-v10) is preliminary update and we plan to address the 
> open items
> (listed in section 11; for example, the security section could use
> expert reviews) in the upcoming weeks. This is being 
> presented to obtain
> feedback from the WG. Additionally, if folks wish to participate in
> offline discussions or sign up as expert reviewers please let myself,
> the author or the co-chairs know. 
> 
> Comments and feedback will be appreciated (thanks to Cullen 
> and Henning
> for early feedback on the mailing list).
> 
> Thanks!
> 
> - S
> 
> 
> 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use
> sip-implementors@cs.columbia.edu for questions on current sip Use
> sip@ietf.org for new developments of core SIP
> 
> 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 06 06:56:00 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HEOvP-0007yS-K7; Tue, 06 Feb 2007 06:55:51 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HEOvO-0007yM-Dc
	for sipping@ietf.org; Tue, 06 Feb 2007 06:55:50 -0500
Received: from host10.216.41.24.conversent.net ([216.41.24.10]
	helo=acmepacket.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HEOvN-0007Zp-5v
	for sipping@ietf.org; Tue, 06 Feb 2007 06:55:50 -0500
Date: Tue,  6 Feb 2007 06:55:46 -0500
Message-Id: <10702060655.AA64943@acmepacket.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
From: "Patrick Timmons" <ptimmons@acmepacket.com>
To: sipping@ietf.org
X-Mailer: <SMTP32 v9.10>
Precedence: bulk
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 7bac9cb154eb5790ae3b2913587a40de
Subject: [Sipping] automated response
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Reply-To: ptimmons@acmepacket.com
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

IMPORTANT NOTE: if you are receiving this auto responder after accessing Acme Packet's web portal or sending an email to a mailing list, rest assured that your message has been received and will be acted upon by the appropriate personnel.

I am taking Friday, February 2 through Tuesday, February 6 off and will be mostly unavailable during this time.  I will not have access to a computer, but will be reachable via my cell phone in the event of emergencies.

For support issues, please contact support@acmepacket.com.  For non-support issues, please contact Erin Medeiros, Vice President of Professional Services, at emedeiros@acmepacket.com.

Thank you,
Patrick Timmons
Vice President, Systems Engineering
Acme Packet, Inc.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 07 03:09:00 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HEhpd-0005Gr-MN; Wed, 07 Feb 2007 03:07:09 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HEhpc-0005Gj-4H
	for sipping@ietf.org; Wed, 07 Feb 2007 03:07:08 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HEhpP-00012K-EC
	for sipping@ietf.org; Wed, 07 Feb 2007 03:07:08 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	7C3B0201F1; Wed,  7 Feb 2007 09:06:44 +0100 (CET)
X-AuditID: c1b4fb3e-b1ed7bb0000007e1-c9-45c98894cb77 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	59758200DE; Wed,  7 Feb 2007 09:06:44 +0100 (CET)
Received: from esealmw128.eemea.ericsson.se ([153.88.254.176]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 7 Feb 2007 09:06:44 +0100
Received: from mail.lmf.ericsson.se ([131.160.11.50]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 7 Feb 2007 09:06:43 +0100
Received: from nomadiclab.lmf.ericsson.se (nomadiclab.lmf.ericsson.se
	[131.160.33.3])
	by mail.lmf.ericsson.se (Postfix) with ESMTP id 6A91F2403;
	Wed,  7 Feb 2007 10:06:43 +0200 (EET)
Received: from nomadiclab.lmf.ericsson.se (localhost [127.0.0.1])
	by nomadiclab.lmf.ericsson.se (Postfix) with ESMTP id F2CD04DB55;
	Wed,  7 Feb 2007 10:06:42 +0200 (EET)
Received: from localhost.localdomain (localhost [IPv6:::1])
	by nomadiclab.lmf.ericsson.se (Postfix) with ESMTP id 827A34DA95;
	Wed,  7 Feb 2007 10:06:42 +0200 (EET)
Subject: RE: [Sipping] Service Identification in SIP
From: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>
To: ravishankar.shiroor@wipro.com
In-Reply-To: <532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
References: <45C0AEF2.8010101@ericsson.com>
	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
Content-Type: text/plain
Date: Wed, 07 Feb 2007 10:07:22 +0200
Message-Id: <1170835642.29763.54.camel@n95.nomadiclab.com>
Mime-Version: 1.0
X-Mailer: Evolution 2.2.3 (2.2.3-4.fc4) 
Content-Transfer-Encoding: 7bit
X-Virus-Scanned: ClamAV using ClamSMTP
X-OriginalArrivalTime: 07 Feb 2007 08:06:43.0746 (UTC)
	FILETIME=[E7C2C020:01C74A8E]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 5ebbf074524e58e662bc8209a6235027
Cc: sipping@ietf.org, Gonzalo.Camarillo@ericsson.com, mary.barnes@nortel.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

hi Ravi,

in IMS a Service is a "type of communication defined by a service
definition that specifies the rules and procedures and allowed media, if
any, for a specific type of communication and that utilises the IMS
enabler. An IMS communication service defines what is possible within a
single SIP session or standalone transaction."
So a Service implies some combination of capabilities and ways to use
those capabilities.
If you want the Service specifies the context for a communication.

Instead the Application define the specific communication (or if you
want application) that you want run in the specific context. So you
should have several different applications designed to run on a specific
environment or that have different performance or behavior in different
context (running on top of different service).

Lets go to Jonathan's example of the Chess.
In his example the Chess is a service, for IMS instead is an application
that you can run on top of different service.
For example you can run Chess application on top of PoC service or on
top of the "standard" service. So let say you run Chess on top of PoC
service, then you can use only voice in the way that PoC specifies.

/sal




On Thu, 2007-02-01 at 16:39 +0530, ravishankar.shiroor@wipro.com wrote:
> Hello,
> 
> IMO, Service Identification is definitely needed and should be taken up
> as a WG as proposed below.
> 
> Further, I think Service Identification scope should be as per the
> requirement in Jonathan's draft, as I still do not understand the need
> for having a Communication Service Identifier and another Application
> Identifier as given in the 3gpp-ics-requirements document.
> 
> Regards,
> Ravi.
> 
> --
>  
> Ravishankar. G. Shiroor
> Wipro Technologies, Bangalore.
>  
> ravishankar.shiroor@wipro.com
> --
>  
> 
> > -----Original Message-----
> > From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com] 
> > Sent: Wednesday, January 31, 2007 8:30 PM
> > To: sipping
> > Cc: Mary Barnes
> > Subject: [Sipping] Service Identification in SIP
> > 
> > Folks,
> > 
> > during the last year, there have been a number of threads on 
> > the mailing list discussing service identification in SIP. In 
> > particular, the following thread discussed 3GPP requirements 
> > on this area:
> > 
> > http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
> > 
> > At a later point, those 3GPP requirements were documented in 
> > the following (expired by now) Internet Draft, which was 
> > presented in a face-to-face IETF meeting.
> > 
> > http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
> > cs-requirements-00.txt
> > 
> > Lately, the following draft has been submitted to the archives:
> > 
> > http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
> > rvice-identification-00.txt
> > 
> > This draft proposes a solution that intends to meet known 
> > requirements in this area.
> > 
> > Given the amount of discussions on issues related to service 
> > identification we have seen during the last year, we believe 
> > this is a fairly important topic for the SIPPING WG to 
> > tackle. At some point, we would like to have a milestone with 
> > its associated WG item dealing with this issue.
> > 
> > Therefore, we would like to encourage people to read this 
> > draft and send comments to the list. Additionally, if you 
> > believe that this work is (or is not) important, we would 
> > like to hear that as well.
> > 
> > Cheers,
> > 
> > Gonzalo
> > SIPPING co-chair
> > 
> > 
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP 
> > Use sip-implementors@cs.columbia.edu for questions on current 
> > sip Use sip@ietf.org for new developments of core SIP
> > 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 07 12:15:56 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HEqMf-0006KF-Dd; Wed, 07 Feb 2007 12:13:49 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HEqMe-0006Ju-GJ
	for sipping@ietf.org; Wed, 07 Feb 2007 12:13:48 -0500
Received: from sj-iport-3-in.cisco.com ([171.71.176.72]
	helo=sj-iport-3.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HEqMY-0004JK-Qh
	for sipping@ietf.org; Wed, 07 Feb 2007 12:13:48 -0500
Received: from sj-dkim-4.cisco.com ([171.71.179.196])
	by sj-iport-3.cisco.com with ESMTP; 07 Feb 2007 09:13:42 -0800
X-IronPort-AV: i="4.13,296,1167638400"; 
	d="scan'208"; a="462888726:sNHT56924732"
Received: from sj-core-5.cisco.com (sj-core-5.cisco.com [171.71.177.238])
	by sj-dkim-4.cisco.com (8.12.11/8.12.11) with ESMTP id l17HDcTe024161; 
	Wed, 7 Feb 2007 09:13:38 -0800
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by sj-core-5.cisco.com (8.12.10/8.12.6) with ESMTP id l17HDeV2001299;
	Wed, 7 Feb 2007 09:13:41 -0800 (PST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 7 Feb 2007 12:13:23 -0500
Received: from [161.44.183.228] ([161.44.183.228]) by
	xfe-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 7 Feb 2007 12:13:24 -0500
Message-ID: <45CA08B2.4050603@cisco.com>
Date: Wed, 07 Feb 2007 12:13:22 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>
Subject: Re: [Sipping] Service Identification in SIP
References: <45C0AEF2.8010101@ericsson.com>	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
	<1170835642.29763.54.camel@n95.nomadiclab.com>
In-Reply-To: <1170835642.29763.54.camel@n95.nomadiclab.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 07 Feb 2007 17:13:24.0373 (UTC)
	FILETIME=[46753C50:01C74ADB]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=5302; t=1170868418;
	x=1171732418; c=relaxed/simple; s=sjdkim4002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20; bh=xq4ZKej4hrlx7R51JvU94dwx32nygMQ2iwww3kXJLJ8=;
	b=A35KOi/VmE51q+98XBsu27q/idd9tmMO9wvPjVQ2szMbKjvD4tQxquH/ppU9Nd93uLxX85h3
	EP7iIzXNUfXfs39LLxthUy3ga9PFgnqWL+6DVdNSltanhhdeKgve4GTi;
Authentication-Results: sj-dkim-4; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim4002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 68ba2b07ef271dba6ee42a93832cfa4c
Cc: mary.barnes@nortel.com, sipping@ietf.org, ravishankar.shiroor@wipro.com,
	Gonzalo.Camarillo@ericsson.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I still don't get the intent of identifying the application. It seems to 
me this is evil - akin to sending an email message and identifying that 
the recipient must use Outlook to receive it, not Thunderbird, Eudora, 
gmail, or whatever.

I also wish somebody could give better examples of a communication 
service. Chess is too artificial to be believable, and IMO PoC is only a 
service because of the way the implementation has been done. 
Conceptually PoC could be nothing but a special floor control protocol 
accompanying a voice session. In that case none of the Service mechanism 
would be needed.

	Paul

Salvatore Loreto (JO/LMF) wrote:
> hi Ravi,
> 
> in IMS a Service is a "type of communication defined by a service
> definition that specifies the rules and procedures and allowed media, if
> any, for a specific type of communication and that utilises the IMS
> enabler. An IMS communication service defines what is possible within a
> single SIP session or standalone transaction."
> So a Service implies some combination of capabilities and ways to use
> those capabilities.
> If you want the Service specifies the context for a communication.
> 
> Instead the Application define the specific communication (or if you
> want application) that you want run in the specific context. So you
> should have several different applications designed to run on a specific
> environment or that have different performance or behavior in different
> context (running on top of different service).
> 
> Lets go to Jonathan's example of the Chess.
> In his example the Chess is a service, for IMS instead is an application
> that you can run on top of different service.
> For example you can run Chess application on top of PoC service or on
> top of the "standard" service. So let say you run Chess on top of PoC
> service, then you can use only voice in the way that PoC specifies.
> 
> /sal
> 
> 
> 
> 
> On Thu, 2007-02-01 at 16:39 +0530, ravishankar.shiroor@wipro.com wrote:
>> Hello,
>>
>> IMO, Service Identification is definitely needed and should be taken up
>> as a WG as proposed below.
>>
>> Further, I think Service Identification scope should be as per the
>> requirement in Jonathan's draft, as I still do not understand the need
>> for having a Communication Service Identifier and another Application
>> Identifier as given in the 3gpp-ics-requirements document.
>>
>> Regards,
>> Ravi.
>>
>> --
>>  
>> Ravishankar. G. Shiroor
>> Wipro Technologies, Bangalore.
>>  
>> ravishankar.shiroor@wipro.com
>> --
>>  
>>
>>> -----Original Message-----
>>> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com] 
>>> Sent: Wednesday, January 31, 2007 8:30 PM
>>> To: sipping
>>> Cc: Mary Barnes
>>> Subject: [Sipping] Service Identification in SIP
>>>
>>> Folks,
>>>
>>> during the last year, there have been a number of threads on 
>>> the mailing list discussing service identification in SIP. In 
>>> particular, the following thread discussed 3GPP requirements 
>>> on this area:
>>>
>>> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
>>>
>>> At a later point, those 3GPP requirements were documented in 
>>> the following (expired by now) Internet Draft, which was 
>>> presented in a face-to-face IETF meeting.
>>>
>>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
>>> cs-requirements-00.txt
>>>
>>> Lately, the following draft has been submitted to the archives:
>>>
>>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
>>> rvice-identification-00.txt
>>>
>>> This draft proposes a solution that intends to meet known 
>>> requirements in this area.
>>>
>>> Given the amount of discussions on issues related to service 
>>> identification we have seen during the last year, we believe 
>>> this is a fairly important topic for the SIPPING WG to 
>>> tackle. At some point, we would like to have a milestone with 
>>> its associated WG item dealing with this issue.
>>>
>>> Therefore, we would like to encourage people to read this 
>>> draft and send comments to the list. Additionally, if you 
>>> believe that this work is (or is not) important, we would 
>>> like to hear that as well.
>>>
>>> Cheers,
>>>
>>> Gonzalo
>>> SIPPING co-chair
>>>
>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP 
>>> Use sip-implementors@cs.columbia.edu for questions on current 
>>> sip Use sip@ietf.org for new developments of core SIP
>>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP
>> Use sip-implementors@cs.columbia.edu for questions on current sip
>> Use sip@ietf.org for new developments of core SIP
> 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 07 12:36:07 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HEqhi-0002fs-Tj; Wed, 07 Feb 2007 12:35:34 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HEqhh-0002fn-NB
	for sipping@ietf.org; Wed, 07 Feb 2007 12:35:33 -0500
Received: from ondar.cablelabs.com ([192.160.73.61])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HEqhg-0001zy-8W
	for sipping@ietf.org; Wed, 07 Feb 2007 12:35:33 -0500
Received: from kyzyl.cablelabs.com (kyzyl.cablelabs.com [10.253.0.7])
	by ondar.cablelabs.com (8.13.8/8.13.8) with ESMTP id l17HZTRv020613;
	Wed, 7 Feb 2007 10:35:29 -0700 (MST)
Received: from srvxchg.cablelabs.com (10.5.0.20)
	by kyzyl.cablelabs.com (F-Secure/fsigk_smtp/488/kyzyl.cablelabs.com);
	Wed, 7 Feb 2007 10:35:28 -0700 (MST)
X-Virus-Status: clean(F-Secure/fsigk_smtp/488/kyzyl.cablelabs.com)
X-MimeOLE: Produced By Microsoft Exchange V6.0.6603.0
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] draft-ietf-sipping-config-framework-10
Date: Wed, 7 Feb 2007 10:35:28 -0700
Message-ID: <CD6CE349CFD30D40BF5E13B3E0D84804021B638A@srvxchg.cablelabs.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] draft-ietf-sipping-config-framework-10
Thread-Index: AcdJ5V3wIDRS3ukESPKkSywuiIXmkwA+LmFg
References: A<50B1CBA96870A34799A506B2313F26670AF57275@ntht201e.siemenscomms.co.uk>
From: "Sumanth Channabasappa" <sumanth@cablelabs.com>
To: "Elwell, John" <john.elwell@siemens.com>
X-Approved: ondar
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 29dc808194f5fb921c09d0040806d6eb
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

John,

The posted version (-v10) was a first draft from the design team and was
meant to obtain feedback from the WG, so your email helps - thanks (the
deadline was mostly a request to help us address comments in a timely
manner)! I have attempted to address a few of comments in this email and
will be requesting other members of the design team to address the rest
(and to keep me honest)!

Specific responses can be found inline (marked with [S])...


In general I find it hard to read and understand, and I can't help
feeling that this is an over-complex solution. This is due in part to
the fact that we have three different profile types, and it is not clear
to me that this is needed in a lot of deployments. The use cases you
give are certainly not all that compelling, and I can't help feeling
that a much simpler solution would serve a large proportion of cases and
would be a lot more attractive to implement.
[S] The Use Cases present a couple of complex deployments with the hope
that other Use Cases can easily be envisioned, especially simple ones.
They are compelling based on the feedback obtained from the design team
(IMHO) - and hopefully helpful to understanding the general framework -
but we are open to suggestions and any simplification that can also take
care of the presented Use Cases (and are actively pursuing that). See in
particular open items listed in sec 11, items 2 and 4. I would strongly
recommend that you provide more details to the design team in this
regard.

Now, the profile types are indicative of profile data categorization
(and the profile data sources). The intention is not to force the
retrieval of all the profile types, but to provide a framework that
allows for the necessary profiles to be retrieved (e.g., a client may
not be provided with a local-network profile). The I-D also allows folks
to define new profile types.



Section 2, definition of client. This definition seems to include things
like gateways and media servers. Is this the intention? The concept of
user may differ for gateways, for example.
[S] Yes, this was the intention. We may want to explain this clearly.


Section 3, definition of device.  I see no point in using two terms
(device and client) for the same thing. I prefer "device", because SIP
has its own use for "client".
[S] The usage of 'client' (specific usage in the I-D) was to allow for a
wider variety of entities (e.g., software-based entities) to be
represented. I will check with the design team on this.
=20

Section 4.1. I don't think this example is compelling. Why would a data
service provider provide a SIP-based PDS when it is not a SIP service
provider?
[S] There were a few Use Cases presented for this. For example, if an
enterprise network allows users to use SIP based devices (such as
dual-mode mobile devices) within their network, but do not provide SIP
services themselves, they may still want to provide local network
configuration information (e.g., STUN/TURN addresses and such) to enable
such services within their perimeter. The alternative is WiFi hotspots
or even home networks that have special configuration for SIP based
communication to be realized. This is not always the case, however, and
a deployment can use it as need be (this may have to be stated upfront,
will check on that).


Section 4.1, item B, "The Client proceeds to request the 'local-network'
Profile Type. The PDS in the local network responds, allowing the Client
to retrieve the local-network profile". We should describe here how the
client address the request to the PDS.
[S] This is detailed in the later sections when we describe each step in
further detail. We did not want to duplicate the information.


Section 4.1, item C, "based on the subscription status". What
subscription status is this? Why would a device (as opposed to a user)
have to 'subscribe' prior to being able to be used? How does the device
identify itself to the PDS so that its 'subscription' can be verified?
[S] Perhaps we need to explain that more clearly or eliminate nuances.
To answer your specific question, there are architectures where the
'device' and the 'user' are two separate entities and have separate
identities (and possibly credentials). Examples I can think of are GSM
devices which have different identifiers for the handset and the User
(identity tied to the SIM) or PacketCable clients where the devices are
identified independent of the user. In simple cases, this is true of
certain enterprise networks where IP phones have their own identity (and
credentials) and can cater to any user who can 'login'. Thus, the
'device' identifies itself to get IP access and provide basic services,
but anyone who wants them to act on their behalf (i.e., behave as if a
IP Phone was their own) can do so.=20


Section 4.1, item D, "The Client retrieves the User profile (this can
contain information such as service profiles to be retrieved". What is a
service profile? I thought we just had network, device and user
profiles.
[S] This may need to be clarified. There is a separate I-D for the
'application' profile type (defined outside of the framework) and should
probably not be mentioned here. We will look at rewording this.


Section 4.1, item D, "The Client then starts providing services." What
is meant by "providing services". Normally a server provides services,
not the client. Does it just mean that the device commences normal
operation (e.g., registering with a SIP server)?
[S] This may need some clean up. The objective was to say that client
can start normal operation.=20


Section 4.2 "SIP Service Provider A provides the local network
connectivity, 'local-network' and 'device' profiles for the Client.  The
Service Provider also provides 'user' profiles for existing
subscribers".  Why is this necessarily a SIP service provider? I would
imagine that the device would perform SIP registration with Service
Provider B (on a per user basis) and would not register with Service
Provider A. So to call it a SIP Service Provider is misleading. OK, in
this case if there are some users who happen to be supported by this
provider, then I guess it is a SIP service provider, but if all users
are supported by other service providers then it is not.
[S] This is just the assumption in the Use Case and is not reflective or
all scenarios.=20


Section 5.1 "The first step to obtaining a profile is PDS Discovery.
This is accomplished by creating a profile subscription using the Event
Package described in Section 6, and preparing for Profile Enrollment."
This is circular - you have to discover the PDS before you can submit a
subscription request to it.
[S] The design team is attempting to clean this up. Discovery of the PDS
is not an act in itself. It involves creating a subscription that is
aimed at a PDS (i.e., profile source).


Section 5.1, second paragraph "Client identifier". I think we need to
say something here about what this is and how a PDS would use it. Taking
a MAC address as an example, is this likely to be meaningful to the
Device Profile PDS?
[S] This is explained in detail in 5.1.2.1, no?


Again, your detailed comments are much appreciated. I would also request
you to participate in the design team discussions to add value to this
effort.

Thanks!
- S



_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 07 12:36:45 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HEqir-0003M5-La; Wed, 07 Feb 2007 12:36:45 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HEqip-0003Ly-SX
	for sipping@ietf.org; Wed, 07 Feb 2007 12:36:43 -0500
Received: from exprod6og52.obsmtp.com ([64.18.1.185])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1HEqil-0001zZ-7N
	for sipping@ietf.org; Wed, 07 Feb 2007 12:36:43 -0500
Received: from source ([192.150.20.142]) by exprod6ob52.postini.com
	([64.18.5.12]) with SMTP; Wed, 07 Feb 2007 09:35:24 PST
Received: from inner-relay-3.eur.adobe.com (inner-relay-3b [10.128.4.236])
	by outbound-smtp-2.corp.adobe.com (8.12.10/8.12.10) with ESMTP id
	l17HZkIY019006; Wed, 7 Feb 2007 09:35:46 -0800 (PST)
Received: from fe1.corp.adobe.com (fe1.corp.adobe.com [10.8.192.70])
	by inner-relay-3.eur.adobe.com (8.12.10/8.12.9) with ESMTP id
	l17HZ1eU024179; Wed, 7 Feb 2007 09:35:13 -0800 (PST)
Received: from namail5.corp.adobe.com ([10.8.192.88]) by fe1.corp.adobe.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 7 Feb 2007 09:35:12 -0800
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Wed, 7 Feb 2007 09:33:55 -0800
Message-ID: <24CCCC428EFEA2469BF046DB3C7A8D224530FD@namail5.corp.adobe.com>
In-Reply-To: <45CA08B2.4050603@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdK2+Kqf8XxF4RlRBKez3RRQPczsAAAHOqw
References: <45C0AEF2.8010101@ericsson.com>	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com><1170835642.29763.54.camel@n95.nomadiclab.com>
	<45CA08B2.4050603@cisco.com>
From: "Henry Sinnreich" <hsinnrei@adobe.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>,
	"Salvatore Loreto \(JO/LMF\)" <salvatore.loreto@ericsson.com>
X-OriginalArrivalTime: 07 Feb 2007 17:35:12.0118 (UTC)
	FILETIME=[51EF4960:01C74ADE]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a92270ba83d7ead10c5001bb42ec3221
Cc: Gonzalo.Camarillo@ericsson.com, sipping@ietf.org, mary.barnes@nortel.com,
	ravishankar.shiroor@wipro.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

> I still don't get the intent of identifying the application. It seems
to > me this is evil - akin to sending an email message and identifying
that > the recipient must use Outlook to receive it, not Thunderbird,
Eudora,=20
> gmail, or whatever.

This goes back to the old discussion about the IN versus the "dumb
network" and we know how this has played out.

Placing intelligence (AKA services) in the network is just "dumb design"
IMHO for all the good reasons, but makes no sense to discuss on this
list.

SIP was designed as a rendezvous protocol for multimedia communications
on the Internet and not to replicate ISDN, etc., as has been so well
documented in RFC 4485:

   "SIP is not meant to be used as a strict Public Switched Telephone
   Network (PSTN) signaling replacement.  It is not a superset of the
   Integrated Services Digital Network (ISDN) User Part (ISUP).
   Although it can support gatewaying of PSTN signaling and can provide
   many features present in the PSTN, the mere existence of a feature or
   capability in the PSTN is not a justification for its inclusion in
   SIP."

Which makes me wonder how many folks have read RFC 4485 and taken it to
heart or may have chosen to forget it.

Thanks, Henry

-----Original Message-----
From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
Sent: Wednesday, February 07, 2007 11:13 AM
To: Salvatore Loreto (JO/LMF)
Cc: mary.barnes@nortel.com; sipping@ietf.org;
ravishankar.shiroor@wipro.com; Gonzalo.Camarillo@ericsson.com
Subject: Re: [Sipping] Service Identification in SIP

I still don't get the intent of identifying the application. It seems to

me this is evil - akin to sending an email message and identifying that=20
the recipient must use Outlook to receive it, not Thunderbird, Eudora,=20
gmail, or whatever.

I also wish somebody could give better examples of a communication=20
service. Chess is too artificial to be believable, and IMO PoC is only a

service because of the way the implementation has been done.=20
Conceptually PoC could be nothing but a special floor control protocol=20
accompanying a voice session. In that case none of the Service mechanism

would be needed.

	Paul

Salvatore Loreto (JO/LMF) wrote:
> hi Ravi,
>=20
> in IMS a Service is a "type of communication defined by a service
> definition that specifies the rules and procedures and allowed media,
if
> any, for a specific type of communication and that utilises the IMS
> enabler. An IMS communication service defines what is possible within
a
> single SIP session or standalone transaction."
> So a Service implies some combination of capabilities and ways to use
> those capabilities.
> If you want the Service specifies the context for a communication.
>=20
> Instead the Application define the specific communication (or if you
> want application) that you want run in the specific context. So you
> should have several different applications designed to run on a
specific
> environment or that have different performance or behavior in
different
> context (running on top of different service).
>=20
> Lets go to Jonathan's example of the Chess.
> In his example the Chess is a service, for IMS instead is an
application
> that you can run on top of different service.
> For example you can run Chess application on top of PoC service or on
> top of the "standard" service. So let say you run Chess on top of PoC
> service, then you can use only voice in the way that PoC specifies.
>=20
> /sal
>=20
>=20
>=20
>=20
> On Thu, 2007-02-01 at 16:39 +0530, ravishankar.shiroor@wipro.com
wrote:
>> Hello,
>>
>> IMO, Service Identification is definitely needed and should be taken
up
>> as a WG as proposed below.
>>
>> Further, I think Service Identification scope should be as per the
>> requirement in Jonathan's draft, as I still do not understand the
need
>> for having a Communication Service Identifier and another Application
>> Identifier as given in the 3gpp-ics-requirements document.
>>
>> Regards,
>> Ravi.
>>
>> --
>> =20
>> Ravishankar. G. Shiroor
>> Wipro Technologies, Bangalore.
>> =20
>> ravishankar.shiroor@wipro.com
>> --
>> =20
>>
>>> -----Original Message-----
>>> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]=20
>>> Sent: Wednesday, January 31, 2007 8:30 PM
>>> To: sipping
>>> Cc: Mary Barnes
>>> Subject: [Sipping] Service Identification in SIP
>>>
>>> Folks,
>>>
>>> during the last year, there have been a number of threads on=20
>>> the mailing list discussing service identification in SIP. In=20
>>> particular, the following thread discussed 3GPP requirements=20
>>> on this area:
>>>
>>> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
>>>
>>> At a later point, those 3GPP requirements were documented in=20
>>> the following (expired by now) Internet Draft, which was=20
>>> presented in a face-to-face IETF meeting.
>>>
>>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
>>> cs-requirements-00.txt
>>>
>>> Lately, the following draft has been submitted to the archives:
>>>
>>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
>>> rvice-identification-00.txt
>>>
>>> This draft proposes a solution that intends to meet known=20
>>> requirements in this area.
>>>
>>> Given the amount of discussions on issues related to service=20
>>> identification we have seen during the last year, we believe=20
>>> this is a fairly important topic for the SIPPING WG to=20
>>> tackle. At some point, we would like to have a milestone with=20
>>> its associated WG item dealing with this issue.
>>>
>>> Therefore, we would like to encourage people to read this=20
>>> draft and send comments to the list. Additionally, if you=20
>>> believe that this work is (or is not) important, we would=20
>>> like to hear that as well.
>>>
>>> Cheers,
>>>
>>> Gonzalo
>>> SIPPING co-chair
>>>
>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP=20
>>> Use sip-implementors@cs.columbia.edu for questions on current=20
>>> sip Use sip@ietf.org for new developments of core SIP
>>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP
>> Use sip-implementors@cs.columbia.edu for questions on current sip
>> Use sip@ietf.org for new developments of core SIP
>=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 07 13:38:52 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HErfs-00088E-TU; Wed, 07 Feb 2007 13:37:44 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HErfr-00087u-NL
	for sipping@ietf.org; Wed, 07 Feb 2007 13:37:43 -0500
Received: from smtp.nokia.com ([131.228.20.171] helo=mgw-ext12.nokia.com)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HErfo-00046C-29
	for sipping@ietf.org; Wed, 07 Feb 2007 13:37:43 -0500
Received: from esebh107.NOE.Nokia.com (esebh107.ntc.nokia.com [172.21.143.143])
	by mgw-ext12.nokia.com (Switch-3.2.5/Switch-3.2.5) with ESMTP id
	l17IZKhN022644; Wed, 7 Feb 2007 20:35:24 +0200
Received: from esebh103.NOE.Nokia.com ([172.21.143.33]) by
	esebh107.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 7 Feb 2007 20:36:37 +0200
Received: from esebe104.NOE.Nokia.com ([172.21.143.44]) by
	esebh103.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 7 Feb 2007 20:36:57 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Wed, 7 Feb 2007 20:36:34 +0200
Message-ID: <921445D056CC0E4FAFA3F0566793E8270307B6EB@esebe104.NOE.Nokia.com>
In-Reply-To: <1170835642.29763.54.camel@n95.nomadiclab.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdKj57vjpj5y+qcQPmVCV+xwzqTfQAVoPEw
References: <45C0AEF2.8010101@ericsson.com><532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
	<1170835642.29763.54.camel@n95.nomadiclab.com>
From: <Jari.Mutikainen@nokia.com>
To: <salvatore.loreto@ericsson.com>, <ravishankar.shiroor@wipro.com>
X-OriginalArrivalTime: 07 Feb 2007 18:36:57.0798 (UTC)
	FILETIME=[F2B12E60:01C74AE6]
X-eXpurgate-Category: 1/0
X-eXpurgate-ID: 149371::070207203525-53FF0BB0-2C84B134/0-0/0-1
X-Nokia-AV: Clean
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 87a3f533bb300b99e2a18357f3c1563d
Cc: sipping@ietf.org, Gonzalo.Camarillo@ericsson.com, mary.barnes@nortel.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hello Sal,

In your example of Chess application running on top of PoC service,
since the service (here PoC) defined what is possible within a single
SIP session, and I assume Chess was not defined to be possible as part
of PoC service, then how it is possible to run Chess on top of PoC
within single dialog, and if it is not possible, then what does it mean
to "run Chess on top of PoC" if it is only possible with separate SIP
dialogs?=20

BR,
Jari

-----Original Message-----
From: ext Salvatore Loreto (JO/LMF)
[mailto:salvatore.loreto@ericsson.com]=20
Sent: 07 February, 2007 10:07
To: ravishankar.shiroor@wipro.com
Cc: sipping@ietf.org; Gonzalo.Camarillo@ericsson.com;
mary.barnes@nortel.com
Subject: RE: [Sipping] Service Identification in SIP

hi Ravi,

in IMS a Service is a "type of communication defined by a service
definition that specifies the rules and procedures and allowed media, if
any, for a specific type of communication and that utilises the IMS
enabler. An IMS communication service defines what is possible within a
single SIP session or standalone transaction."
So a Service implies some combination of capabilities and ways to use
those capabilities.
If you want the Service specifies the context for a communication.

Instead the Application define the specific communication (or if you
want application) that you want run in the specific context. So you
should have several different applications designed to run on a specific
environment or that have different performance or behavior in different
context (running on top of different service).

Lets go to Jonathan's example of the Chess.
In his example the Chess is a service, for IMS instead is an application
that you can run on top of different service.
For example you can run Chess application on top of PoC service or on
top of the "standard" service. So let say you run Chess on top of PoC
service, then you can use only voice in the way that PoC specifies.

/sal




On Thu, 2007-02-01 at 16:39 +0530, ravishankar.shiroor@wipro.com wrote:
> Hello,
>=20
> IMO, Service Identification is definitely needed and should be taken=20
> up as a WG as proposed below.
>=20
> Further, I think Service Identification scope should be as per the=20
> requirement in Jonathan's draft, as I still do not understand the need

> for having a Communication Service Identifier and another Application=20
> Identifier as given in the 3gpp-ics-requirements document.
>=20
> Regards,
> Ravi.
>=20
> --
> =20
> Ravishankar. G. Shiroor
> Wipro Technologies, Bangalore.
> =20
> ravishankar.shiroor@wipro.com
> --
> =20
>=20
> > -----Original Message-----
> > From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
> > Sent: Wednesday, January 31, 2007 8:30 PM
> > To: sipping
> > Cc: Mary Barnes
> > Subject: [Sipping] Service Identification in SIP
> >=20
> > Folks,
> >=20
> > during the last year, there have been a number of threads on the=20
> > mailing list discussing service identification in SIP. In=20
> > particular, the following thread discussed 3GPP requirements on this

> > area:
> >=20
> > http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
> >=20
> > At a later point, those 3GPP requirements were documented in the=20
> > following (expired by now) Internet Draft, which was presented in a=20
> > face-to-face IETF meeting.
> >=20
> > http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
> > cs-requirements-00.txt
> >=20
> > Lately, the following draft has been submitted to the archives:
> >=20
> > http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
> > rvice-identification-00.txt
> >=20
> > This draft proposes a solution that intends to meet known=20
> > requirements in this area.
> >=20
> > Given the amount of discussions on issues related to service=20
> > identification we have seen during the last year, we believe this is

> > a fairly important topic for the SIPPING WG to tackle. At some=20
> > point, we would like to have a milestone with its associated WG item

> > dealing with this issue.
> >=20
> > Therefore, we would like to encourage people to read this draft and=20
> > send comments to the list. Additionally, if you believe that this=20
> > work is (or is not) important, we would like to hear that as well.
> >=20
> > Cheers,
> >=20
> > Gonzalo
> > SIPPING co-chair
> >=20
> >=20
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP Use=20
> > sip-implementors@cs.columbia.edu for questions on current sip Use=20
> > sip@ietf.org for new developments of core SIP
> >=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 08 03:19:56 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HF4TX-0000AF-ML; Thu, 08 Feb 2007 03:17:51 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HF4TW-0000A1-CA
	for sipping@ietf.org; Thu, 08 Feb 2007 03:17:50 -0500
Received: from mailgate.siemenscomms.co.uk ([195.171.110.225]
	helo=bemg01.siemenscomms.co.uk)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HF4TR-0002nP-PZ
	for sipping@ietf.org; Thu, 08 Feb 2007 03:17:50 -0500
Received: from ntht207e.uksgcs.siemenscomms.co.uk ([137.223.247.82])
	by siemenscomms.co.uk (PMDF V6.0-24 #40642)
	with ESMTP id <0JD400D1DXP3ME@siemenscomms.co.uk> for sipping@ietf.org;
	Thu, 08 Feb 2007 08:17:27 +0000 (GMT)
Received: by ntht207e.uksgcs.siemenscomms.co.uk with Internet Mail Service
	(5.5.2657.72)	id <D2K34X8X>; Thu, 08 Feb 2007 08:17:25 +0000
Content-return: allowed
Date: Thu, 08 Feb 2007 08:17:24 +0000
From: "Elwell, John" <john.elwell@siemens.com>
Subject: RE: [Sipping] draft-ietf-sipping-config-framework-10
To: Sumanth Channabasappa <sumanth@cablelabs.com>
Message-id: <50B1CBA96870A34799A506B2313F26670AFAF435@ntht201e.siemenscomms.co.uk>
MIME-version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
Content-type: text/plain
Content-transfer-encoding: 7BIT
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 140baa79ca42e6b0e2b4504291346186
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Sumanth,

Thanks for your explanations. I am sorry my comments were not detailed
enough, but that was due to lack of time. In the next version, in addition
to providing clarification in line with some of my detailed comments, I hope
there will be some up-front explanation of simpler scenarios (even start
with simpler scenarios and then explain some of the additional capabilities
aimed at more complex scenarios). Perhaps that is already there, but it
didn't come over at first read. I look forward to your next revision, which
I will try to review in more detail and in the light of the explanations you
have given.

John

> -----Original Message-----
> From: Sumanth Channabasappa [mailto:sumanth@cablelabs.com] 
> Sent: 07 February 2007 17:35
> To: Elwell, John
> Cc: sipping@ietf.org
> Subject: RE: [Sipping] draft-ietf-sipping-config-framework-10
> 
> John,
> 
> The posted version (-v10) was a first draft from the design 
> team and was
> meant to obtain feedback from the WG, so your email helps - 
> thanks (the
> deadline was mostly a request to help us address comments in a timely
> manner)! I have attempted to address a few of comments in 
> this email and
> will be requesting other members of the design team to 
> address the rest
> (and to keep me honest)!
> 
> Specific responses can be found inline (marked with [S])...
> 
> 
> In general I find it hard to read and understand, and I can't help
> feeling that this is an over-complex solution. This is due in part to
> the fact that we have three different profile types, and it 
> is not clear
> to me that this is needed in a lot of deployments. The use cases you
> give are certainly not all that compelling, and I can't help feeling
> that a much simpler solution would serve a large proportion 
> of cases and
> would be a lot more attractive to implement.
> [S] The Use Cases present a couple of complex deployments 
> with the hope
> that other Use Cases can easily be envisioned, especially simple ones.
> They are compelling based on the feedback obtained from the 
> design team
> (IMHO) - and hopefully helpful to understanding the general 
> framework -
> but we are open to suggestions and any simplification that 
> can also take
> care of the presented Use Cases (and are actively pursuing 
> that). See in
> particular open items listed in sec 11, items 2 and 4. I 
> would strongly
> recommend that you provide more details to the design team in this
> regard.
> 
> Now, the profile types are indicative of profile data categorization
> (and the profile data sources). The intention is not to force the
> retrieval of all the profile types, but to provide a framework that
> allows for the necessary profiles to be retrieved (e.g., a client may
> not be provided with a local-network profile). The I-D also 
> allows folks
> to define new profile types.
> 
> 
> 
> Section 2, definition of client. This definition seems to 
> include things
> like gateways and media servers. Is this the intention? The concept of
> user may differ for gateways, for example.
> [S] Yes, this was the intention. We may want to explain this clearly.
> 
> 
> Section 3, definition of device.  I see no point in using two terms
> (device and client) for the same thing. I prefer "device", because SIP
> has its own use for "client".
> [S] The usage of 'client' (specific usage in the I-D) was to 
> allow for a
> wider variety of entities (e.g., software-based entities) to be
> represented. I will check with the design team on this.
>  
> 
> Section 4.1. I don't think this example is compelling. Why 
> would a data
> service provider provide a SIP-based PDS when it is not a SIP service
> provider?
> [S] There were a few Use Cases presented for this. For example, if an
> enterprise network allows users to use SIP based devices (such as
> dual-mode mobile devices) within their network, but do not provide SIP
> services themselves, they may still want to provide local network
> configuration information (e.g., STUN/TURN addresses and 
> such) to enable
> such services within their perimeter. The alternative is WiFi hotspots
> or even home networks that have special configuration for SIP based
> communication to be realized. This is not always the case, 
> however, and
> a deployment can use it as need be (this may have to be 
> stated upfront,
> will check on that).
> 
> 
> Section 4.1, item B, "The Client proceeds to request the 
> 'local-network'
> Profile Type. The PDS in the local network responds, allowing 
> the Client
> to retrieve the local-network profile". We should describe 
> here how the
> client address the request to the PDS.
> [S] This is detailed in the later sections when we describe 
> each step in
> further detail. We did not want to duplicate the information.
> 
> 
> Section 4.1, item C, "based on the subscription status". What
> subscription status is this? Why would a device (as opposed to a user)
> have to 'subscribe' prior to being able to be used? How does 
> the device
> identify itself to the PDS so that its 'subscription' can be verified?
> [S] Perhaps we need to explain that more clearly or eliminate nuances.
> To answer your specific question, there are architectures where the
> 'device' and the 'user' are two separate entities and have separate
> identities (and possibly credentials). Examples I can think of are GSM
> devices which have different identifiers for the handset and the User
> (identity tied to the SIM) or PacketCable clients where the 
> devices are
> identified independent of the user. In simple cases, this is true of
> certain enterprise networks where IP phones have their own 
> identity (and
> credentials) and can cater to any user who can 'login'. Thus, the
> 'device' identifies itself to get IP access and provide basic 
> services,
> but anyone who wants them to act on their behalf (i.e., behave as if a
> IP Phone was their own) can do so. 
> 
> 
> Section 4.1, item D, "The Client retrieves the User profile (this can
> contain information such as service profiles to be 
> retrieved". What is a
> service profile? I thought we just had network, device and user
> profiles.
> [S] This may need to be clarified. There is a separate I-D for the
> 'application' profile type (defined outside of the framework) 
> and should
> probably not be mentioned here. We will look at rewording this.
> 
> 
> Section 4.1, item D, "The Client then starts providing services." What
> is meant by "providing services". Normally a server provides services,
> not the client. Does it just mean that the device commences normal
> operation (e.g., registering with a SIP server)?
> [S] This may need some clean up. The objective was to say that client
> can start normal operation. 
> 
> 
> Section 4.2 "SIP Service Provider A provides the local network
> connectivity, 'local-network' and 'device' profiles for the 
> Client.  The
> Service Provider also provides 'user' profiles for existing
> subscribers".  Why is this necessarily a SIP service provider? I would
> imagine that the device would perform SIP registration with Service
> Provider B (on a per user basis) and would not register with Service
> Provider A. So to call it a SIP Service Provider is misleading. OK, in
> this case if there are some users who happen to be supported by this
> provider, then I guess it is a SIP service provider, but if all users
> are supported by other service providers then it is not.
> [S] This is just the assumption in the Use Case and is not 
> reflective or
> all scenarios. 
> 
> 
> Section 5.1 "The first step to obtaining a profile is PDS Discovery.
> This is accomplished by creating a profile subscription using 
> the Event
> Package described in Section 6, and preparing for Profile Enrollment."
> This is circular - you have to discover the PDS before you 
> can submit a
> subscription request to it.
> [S] The design team is attempting to clean this up. Discovery 
> of the PDS
> is not an act in itself. It involves creating a subscription that is
> aimed at a PDS (i.e., profile source).
> 
> 
> Section 5.1, second paragraph "Client identifier". I think we need to
> say something here about what this is and how a PDS would use 
> it. Taking
> a MAC address as an example, is this likely to be meaningful to the
> Device Profile PDS?
> [S] This is explained in detail in 5.1.2.1, no?
> 
> 
> Again, your detailed comments are much appreciated. I would 
> also request
> you to participate in the design team discussions to add value to this
> effort.
> 
> Thanks!
> - S
> 
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 08 11:39:55 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HFCI4-00012D-0m; Thu, 08 Feb 2007 11:38:32 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HFCHG-0000W9-4C
	for sipping@ietf.org; Thu, 08 Feb 2007 11:37:42 -0500
Received: from ihemail4.lucent.com ([135.245.0.39])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HFCFR-00049F-Gc
	for sipping@ietf.org; Thu, 08 Feb 2007 11:35:56 -0500
Received: from ilexp02.ndc.lucent.com (h135-3-39-2.lucent.com [135.3.39.2])
	by ihemail4.lucent.com (8.13.8/IER-o) with ESMTP id l18GZeLn014203;
	Thu, 8 Feb 2007 10:35:40 -0600 (CST)
Received: from DEEXP02.DE.lucent.com ([135.248.187.66]) by
	ilexp02.ndc.lucent.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 8 Feb 2007 10:35:40 -0600
Received: from DEEXC1U01.de.lucent.com ([135.248.187.28]) by
	DEEXP02.DE.lucent.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 8 Feb 2007 17:35:24 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Thu, 8 Feb 2007 17:35:24 +0100
Message-ID: <5D1A7985295922448D5550C94DE29180C65124@DEEXC1U01.de.lucent.com>
In-Reply-To: <24CCCC428EFEA2469BF046DB3C7A8D224530FD@namail5.corp.adobe.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdK2+Kqf8XxF4RlRBKez3RRQPczsAAAHOqwABhI3MA=
References: <45C0AEF2.8010101@ericsson.com>	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com><1170835642.29763.54.camel@n95.nomadiclab.com><45CA08B2.4050603@cisco.com>
	<24CCCC428EFEA2469BF046DB3C7A8D224530FD@namail5.corp.adobe.com>
From: "Drage, Keith \(Keith\)" <drage@alcatel-lucent.com>
To: "Henry Sinnreich" <hsinnrei@adobe.com>,
	"Paul Kyzivat" <pkyzivat@cisco.com>,
	"Salvatore Loreto \(JO/LMF\)" <salvatore.loreto@ericsson.com>
X-OriginalArrivalTime: 08 Feb 2007 16:35:24.0679 (UTC)
	FILETIME=[22114970:01C74B9F]
X-Scanned-By: MIMEDefang 2.57 on 135.245.2.39
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 6a45e05c1e4343200aa6b327df2c43fc
Cc: ravishankar.shiroor@wipro.com, sipping@ietf.org,
	Gonzalo.Camarillo@ericsson.com, mary.barnes@nortel.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I don't claim to understand much of this, but I did understand that one
of the functions of the application identifier was to identify the
appropriate API to call up in the terminal. Therefore the in that sense
it could be used to "rendezvous" the APIs of the two communicating
terminals.

However, not being a terminal builder, maybe one of that select breed
could identify if that is a useful function.

Regards

Keith=20

> -----Original Message-----
> From: Henry Sinnreich [mailto:hsinnrei@adobe.com]=20
> Sent: 07 February 2007 17:34
> To: Paul Kyzivat; Salvatore Loreto (JO/LMF)
> Cc: Gonzalo.Camarillo@ericsson.com; sipping@ietf.org;=20
> mary.barnes@nortel.com; ravishankar.shiroor@wipro.com
> Subject: RE: [Sipping] Service Identification in SIP
>=20
> > I still don't get the intent of identifying the=20
> application. It seems
> to > me this is evil - akin to sending an email message and=20
> identifying that > the recipient must use Outlook to receive=20
> it, not Thunderbird, Eudora,=20
> > gmail, or whatever.
>=20
> This goes back to the old discussion about the IN versus the=20
> "dumb network" and we know how this has played out.
>=20
> Placing intelligence (AKA services) in the network is just=20
> "dumb design"
> IMHO for all the good reasons, but makes no sense to discuss=20
> on this list.
>=20
> SIP was designed as a rendezvous protocol for multimedia=20
> communications on the Internet and not to replicate ISDN,=20
> etc., as has been so well documented in RFC 4485:
>=20
>    "SIP is not meant to be used as a strict Public Switched Telephone
>    Network (PSTN) signaling replacement.  It is not a superset of the
>    Integrated Services Digital Network (ISDN) User Part (ISUP).
>    Although it can support gatewaying of PSTN signaling and=20
> can provide
>    many features present in the PSTN, the mere existence of a=20
> feature or
>    capability in the PSTN is not a justification for its inclusion in
>    SIP."
>=20
> Which makes me wonder how many folks have read RFC 4485 and=20
> taken it to heart or may have chosen to forget it.
>=20
> Thanks, Henry
>=20
> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
> Sent: Wednesday, February 07, 2007 11:13 AM
> To: Salvatore Loreto (JO/LMF)
> Cc: mary.barnes@nortel.com; sipping@ietf.org;=20
> ravishankar.shiroor@wipro.com; Gonzalo.Camarillo@ericsson.com
> Subject: Re: [Sipping] Service Identification in SIP
>=20
> I still don't get the intent of identifying the application.=20
> It seems to
>=20
> me this is evil - akin to sending an email message and=20
> identifying that the recipient must use Outlook to receive=20
> it, not Thunderbird, Eudora, gmail, or whatever.
>=20
> I also wish somebody could give better examples of a=20
> communication service. Chess is too artificial to be=20
> believable, and IMO PoC is only a
>=20
> service because of the way the implementation has been done.=20
> Conceptually PoC could be nothing but a special floor control=20
> protocol accompanying a voice session. In that case none of=20
> the Service mechanism
>=20
> would be needed.
>=20
> 	Paul
>=20
> Salvatore Loreto (JO/LMF) wrote:
> > hi Ravi,
> >=20
> > in IMS a Service is a "type of communication defined by a service=20
> > definition that specifies the rules and procedures and=20
> allowed media,
> if
> > any, for a specific type of communication and that utilises the IMS=20
> > enabler. An IMS communication service defines what is=20
> possible within
> a
> > single SIP session or standalone transaction."
> > So a Service implies some combination of capabilities and=20
> ways to use=20
> > those capabilities.
> > If you want the Service specifies the context for a communication.
> >=20
> > Instead the Application define the specific communication=20
> (or if you=20
> > want application) that you want run in the specific context. So you=20
> > should have several different applications designed to run on a
> specific
> > environment or that have different performance or behavior in
> different
> > context (running on top of different service).
> >=20
> > Lets go to Jonathan's example of the Chess.
> > In his example the Chess is a service, for IMS instead is an
> application
> > that you can run on top of different service.
> > For example you can run Chess application on top of PoC=20
> service or on=20
> > top of the "standard" service. So let say you run Chess on=20
> top of PoC=20
> > service, then you can use only voice in the way that PoC specifies.
> >=20
> > /sal
> >=20
> >=20
> >=20
> >=20
> > On Thu, 2007-02-01 at 16:39 +0530, ravishankar.shiroor@wipro.com
> wrote:
> >> Hello,
> >>
> >> IMO, Service Identification is definitely needed and=20
> should be taken
> up
> >> as a WG as proposed below.
> >>
> >> Further, I think Service Identification scope should be as per the=20
> >> requirement in Jonathan's draft, as I still do not understand the
> need
> >> for having a Communication Service Identifier and another=20
> Application=20
> >> Identifier as given in the 3gpp-ics-requirements document.
> >>
> >> Regards,
> >> Ravi.
> >>
> >> --
> >> =20
> >> Ravishankar. G. Shiroor
> >> Wipro Technologies, Bangalore.
> >> =20
> >> ravishankar.shiroor@wipro.com
> >> --
> >> =20
> >>
> >>> -----Original Message-----
> >>> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
> >>> Sent: Wednesday, January 31, 2007 8:30 PM
> >>> To: sipping
> >>> Cc: Mary Barnes
> >>> Subject: [Sipping] Service Identification in SIP
> >>>
> >>> Folks,
> >>>
> >>> during the last year, there have been a number of threads on the=20
> >>> mailing list discussing service identification in SIP. In=20
> >>> particular, the following thread discussed 3GPP=20
> requirements on this=20
> >>> area:
> >>>
> >>>=20
> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
> >>>
> >>> At a later point, those 3GPP requirements were documented in the=20
> >>> following (expired by now) Internet Draft, which was=20
> presented in a=20
> >>> face-to-face IETF meeting.
> >>>
> >>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
> >>> cs-requirements-00.txt
> >>>
> >>> Lately, the following draft has been submitted to the archives:
> >>>
> >>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
> >>> rvice-identification-00.txt
> >>>
> >>> This draft proposes a solution that intends to meet known=20
> >>> requirements in this area.
> >>>
> >>> Given the amount of discussions on issues related to service=20
> >>> identification we have seen during the last year, we=20
> believe this is=20
> >>> a fairly important topic for the SIPPING WG to tackle. At some=20
> >>> point, we would like to have a milestone with its=20
> associated WG item=20
> >>> dealing with this issue.
> >>>
> >>> Therefore, we would like to encourage people to read this=20
> draft and=20
> >>> send comments to the list. Additionally, if you believe that this=20
> >>> work is (or is not) important, we would like to hear that as well.
> >>>
> >>> Cheers,
> >>>
> >>> Gonzalo
> >>> SIPPING co-chair
> >>>
> >>>
> >>> _______________________________________________
> >>> Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> >>> This list is for NEW development of the application of SIP Use=20
> >>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> >>> sip@ietf.org for new developments of core SIP
> >>>
> >> _______________________________________________
> >> Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> >> This list is for NEW development of the application of SIP Use=20
> >> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> >> sip@ietf.org for new developments of core SIP
> >=20
> >=20
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP Use=20
> > sip-implementors@cs.columbia.edu for questions on current sip Use=20
> > sip@ietf.org for new developments of core SIP
> >=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 08 12:16:10 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HFCrk-0006h3-JQ; Thu, 08 Feb 2007 12:15:24 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HFCrh-0006g9-Ux
	for sipping@ietf.org; Thu, 08 Feb 2007 12:15:21 -0500
Received: from expint.psytechnics.com ([217.33.106.234])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HFCrf-0007gV-KF
	for sipping@ietf.org; Thu, 08 Feb 2007 12:15:21 -0500
Importance: normal
Priority: normal
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1896
Content-Class: urn:content-classes:message
MIME-Version: 1.0
Date: Thu, 8 Feb 2007 17:15:46 -0000
Message-ID: <257B18BFF522214E92EB52F3478B014E09466D@psy01exch.intra.psytechnics.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: draft-ietf-sipping-rtcp-summary-01
thread-index: AcdLpMVi3gc5UnMfRnu3Skrb1cHJWQ==
From: "Ravi Raviraj" <ravi.raviraj@psytechnics.com>
To: <sipping@ietf.org>
X-Spam-Score: 0.4 (/)
X-Scan-Signature: 1449ead51a2ff026dcb23465f5379250
Subject: [Sipping] draft-ietf-sipping-rtcp-summary-01
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1364155334=="
Errors-To: sipping-bounces@ietf.org

This is a multi-part message in MIME format.

--===============1364155334==
Content-Transfer-Encoding: 7bit
Content-Class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C74BA4.C5C04882"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C74BA4.C5C04882
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

Hi all,

With reference to voice quality reporting event package, the document
talks about items like MOSLQ and MOSCQ without any definition.

These quality parameters are described in ITU-T P.800.1 and further
defined by P.562, P.563, and P.564 recommendations.

Shouldn't we give reference to these documents directly like the R
factor (used with network planning) reference G.107.

=20

Number of people raised this doubt with me.

=20

Ravi.

=20





This email and attachment are confidential. If you have received this =
message in error, please notify the sender immediately and delete the =
message from your computer. Psytechnics Ltd has taken steps to ensure =
that this email is virus free but the recipient should take necessary =
steps to make certain.




------_=_NextPart_001_01C74BA4.C5C04882
Content-Type: text/html;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

<html xmlns:v=3D"urn:schemas-microsoft-com:vml" =
xmlns:o=3D"urn:schemas-microsoft-com:office:office" =
xmlns:w=3D"urn:schemas-microsoft-com:office:word" =
xmlns:st1=3D"urn:schemas-microsoft-com:office:smarttags" =
xmlns=3D"http://www.w3.org/TR/REC-html40">

<head>
<meta http-equiv=3DContent-Type content=3D"text/html; =
charset=3Dus-ascii">
<meta name=3DGenerator content=3D"Microsoft Word 11 (filtered medium)">
<o:SmartTagType =
namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"
 name=3D"place"/>
<!--[if !mso]>
<style>
st1\:*{behavior:url(#default#ieooui) }
</style>
<![endif]-->
<style>
<!--
 /* Style Definitions */
 p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0cm;
	margin-bottom:.0001pt;
	font-size:12.0pt;
	font-family:"Times New Roman";}
a:link, span.MsoHyperlink
	{color:blue;
	text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
	{color:purple;
	text-decoration:underline;}
span.EmailStyle17
	{mso-style-type:personal-compose;
	font-family:Arial;
	color:windowtext;}
@page Section1
	{size:612.0pt 792.0pt;
	margin:72.0pt 90.0pt 72.0pt 90.0pt;}
div.Section1
	{page:Section1;}
-->
</style>

</head>

<body lang=3DEN-US link=3Dblue vlink=3Dpurple>

<div class=3DSection1>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>Hi all,<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>With reference to voice quality reporting =
event
package, the document talks about items like MOSLQ and MOSCQ without any
definition.<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>These quality parameters are described in =
ITU-T
P.800.1 and further defined by P.562, P.563, and P.564 =
recommendations.<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>Shouldn&#8217;t we give reference to these =
documents
directly like the R factor (used with network planning) reference =
G.107.<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>Number of people raised this doubt with =
me.<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><st1:place w:st=3D"on"><font size=3D2 =
face=3DArial><span
 lang=3DEN-GB =
style=3D'font-size:10.0pt;font-family:Arial'>Ravi</span></font></st1:plac=
e><font
size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:10.0pt;font-family:Arial'>.<o:p></o:p></span></font></=
p>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'><o:p>&nbsp;</o:p></span></font></p>

</div>

</body>

</html>




This email and attachment are confidential. If you have received this =
message in error, please notify the sender immediately and delete the =
message from your computer. Psytechnics Ltd has taken steps to ensure =
that this email is virus free but the recipient should take necessary =
steps to make certain.




------_=_NextPart_001_01C74BA4.C5C04882--


--===============1364155334==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1364155334==--




From sipping-bounces@ietf.org Thu Feb 08 14:48:09 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HFFEa-0006dd-4D; Thu, 08 Feb 2007 14:47:08 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HFFEY-0006bD-5F; Thu, 08 Feb 2007 14:47:06 -0500
Received: from mx12.bbn.com ([128.33.0.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HFFEW-0001QH-Qh; Thu, 08 Feb 2007 14:47:06 -0500
Received: from dommiel.bbn.com ([192.1.122.15] helo=[127.0.0.1])
	by mx12.bbn.com with esmtp (Exim 4.60)
	(envelope-from <rbarnes@bbn.com>)
	id 1HFFEN-0004XN-3U; Thu, 08 Feb 2007 14:46:58 -0500
Message-ID: <45CB7E2F.1070706@bbn.com>
Date: Thu, 08 Feb 2007 14:46:55 -0500
From: Richard Barnes <rbarnes@bbn.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: sip@ietf.org,  sipping@ietf.org
Content-Type: multipart/mixed; boundary="------------050001030003050007020902"
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 3f5efd9d415d254aec2f521f63a5384c
Cc: 
Subject: [Sipping] New draft on early media
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

This is a multi-part message in MIME format.
--------------050001030003050007020902
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

After the big discussions on SIPPING last summer about how to fix early 
media, and the reaction at the last couple of IETFs to Brian Stucker's 
draft on how to cope with it, I've put together a draft to try to 
consolidate all of the discussions and move toward a solution.

draft-barnes-sip-em-ps-req-sol includes three important parts:

1. Problem statement: Although there's lots of definitions of "early 
media" in existing documents, I think there's broad agreement that the 
stuff that causes problems is media that arrives before an SDP 
offer/answer exchange has completed.  The problem is therefore to enable 
the SDP exchange(s) in the INVITE transaction to complete more quickly, 
and make it more difficult (if not impossible) to send media before it 
has completed.  We are explicitly NOT trying to prevent the sending of 
media before a 200/OK or other closing of an INVITE transaction.

2. Requirements: Here, we try to create a consolidated list of 
requirements for an early media fix.

3. Analysis of Solutions: Lots of folks have put forward proposals to 
fix the early media problem.  This section tries to enumerate most of 
them and evaluate them against the requirements in section 2.

It hasn't shown up yet on the I-D tracker, so I've attached a copy.

--Richard

--------------050001030003050007020902
Content-Type: text/plain;
 name="draft-barnes-sip-em-ps-req-sol-00.txt"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="draft-barnes-sip-em-ps-req-sol-00.txt"




Network Working Group                                          R. Barnes
Internet-Draft                                               M. Lepinski
Intended status: Informational                          BBN Technologies
Expires: July 19, 2007                                  January 15, 2007


  Early Media in SIP: Problem Statement, Requirements, and Analysis of
                               Solutions
                   draft-barnes-sip-em-ps-req-sol-00

Status of this Memo

   By submitting this Internet-Draft, each author represents that any
   applicable patent or other IPR claims of which he or she is aware
   have been or will be disclosed, and any of which he or she becomes
   aware will be disclosed, in accordance with Section 6 of BCP 79.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF), its areas, and its working groups.  Note that
   other groups may also distribute working documents as Internet-
   Drafts.

   Internet-Drafts are draft documents valid for a maximum of six months
   and may be updated, replaced, or obsoleted by other documents at any
   time.  It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress."

   The list of current Internet-Drafts can be accessed at
   http://www.ietf.org/ietf/1id-abstracts.txt.

   The list of Internet-Draft Shadow Directories can be accessed at
   http://www.ietf.org/shadow.html.

   This Internet-Draft will expire on July 19, 2007.

Copyright Notice

   Copyright (C) The IETF Trust (2007).

Abstract

   The Session Initiation Protocol (SIP) enables and permits endpoints
   in an INVITE transaction to send media packets before an SDP offer/
   answer exchange has completed; we refer to such media packets as
   early media.  The use of early media is common in current SIP
   implementations, and causes several problems, which have been
   documented elsewhere.  This document puts forward the goal of
   modifying SIP so that compliant implementations will complete an SDP



Barnes & Lepinski         Expires July 19, 2007                 [Page 1]

Internet-Draft                em-ps-req-sol                 January 2007


   exchange before sending media, and lays out high-level requirements
   for such a solution.  The set of possible solutions is then laid out,
   and each possibility examined in light of these requirements.


Table of Contents

   1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  3
   2.  Requirements . . . . . . . . . . . . . . . . . . . . . . . . .  5
     2.1.  Elmination of Early Media  . . . . . . . . . . . . . . . .  5
     2.2.  Additional Messaging . . . . . . . . . . . . . . . . . . .  6
     2.3.  Interoperability with current SIP standards  . . . . . . .  6
     2.4.  Interoperability with the PSTN . . . . . . . . . . . . . .  6
     2.5.  Denial of Service  . . . . . . . . . . . . . . . . . . . .  6
     2.6.  IPR Constraints  . . . . . . . . . . . . . . . . . . . . .  6
   3.  Early Media Solutions  . . . . . . . . . . . . . . . . . . . .  6
     3.1.  Minimizing SDP completion time . . . . . . . . . . . . . .  7
       3.1.1.  Fast completion of INVITE transaction  . . . . . . . .  7
       3.1.2.  Elimination of offer from INVITE . . . . . . . . . . .  8
       3.1.3.  Provisional Responses  . . . . . . . . . . . . . . . .  8
       3.1.4.  Separate INVITE transaction  . . . . . . . . . . . . .  9
       3.1.5.  Non-INVITE SIP Mechanisms  . . . . . . . . . . . . . .  9
       3.1.6.  Non-SIP mechanisms . . . . . . . . . . . . . . . . . . 10
     3.2.  Handling remaining media . . . . . . . . . . . . . . . . . 11
       3.2.1.  Unreliable SDP answer  . . . . . . . . . . . . . . . . 11
       3.2.2.  Backwards-compatibility  . . . . . . . . . . . . . . . 11
   4.  IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 11
   5.  Security Considerations  . . . . . . . . . . . . . . . . . . . 12
   6.  Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 12
   7.  References . . . . . . . . . . . . . . . . . . . . . . . . . . 12
     7.1.  Normative References . . . . . . . . . . . . . . . . . . . 12
     7.2.  Informative References . . . . . . . . . . . . . . . . . . 12
   Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 13
   Intellectual Property and Copyright Statements . . . . . . . . . . 14

















Barnes & Lepinski         Expires July 19, 2007                 [Page 2]

Internet-Draft                em-ps-req-sol                 January 2007


1.  Introduction

   A primary function of the Session Initiation Protocol (SIP,
   [RFC3261]) is to carry SDP messages that form an SDP offer/answer
   exchange [RFC3264].  The offer/answer exchange serves to negotiate
   the parameters necessary for the two end hosts to send multimedia to
   each other, including transports, port numbers, codecs, and security
   parameters (e.g., MIKEY [RFC3830]).  In the most common SIP INVITE
   transaction, the calling party (UAC) is the SDP offerer and the
   called party (UAS) is the SDP answerer; in this scenario, the SDP
   offer is carried in the SIP INVITE request, and the SDP answer in the
   200/OK response.

   Because there is often a substantial delay between the sending of an
   INVITE and the receipt a 200/OK, this assignment of SDP messages to
   SIP messages creates a substantial delay between an SDP offer and an
   SDP answer.  Consequently, there is a period of time between the two
   messages in which the two parties have asymmetric information: The
   called party has knowledge of both the SDP offered by the calling
   party and of his own intended response, while the calling party knows
   only the offer he sent.  Given this information, the called party has
   all the information he needs to begin sending media to the calling
   party (because this was included in the SDP offer); moreover, he can
   send these media without sending an SDP answer to the calling party.
   We define early media as media sent before the completion of an SDP
   offer/answer exchange.

   Ours is a different definition than is used in RFC 3960 [RFC3960],
   which defines early media as "media ... that is exchanged before a
   particular session is accepted by the called user."  The definition
   in RFC 3960 is consistent with the current usage of SDP within the
   SIP INVITE transaction, but at the same time conflates two functions
   of SIP: Negotiation of session-layer parameters (through the SDP
   offer/answer) and communication of an application-layer request and
   acceptance of a session (by the INVITE and 200/OK messages).  As
   suggested above, it is this assignment of SDP messages to SIP
   messages that creates the possibility of early media.

   Early media is known to create several problems for SIP operation:

   o  Call failure: When endpoints use early media for an extended
      period of time, timers can fire to close the UAC INVITE
      transaction before the early media finish and the UAS sends a
      200/OK.  This causes the session proposed in the initial INVITE to
      be dropped by the UAC, and in some implementations, the flow of
      media is cut off.





Barnes & Lepinski         Expires July 19, 2007                 [Page 3]

Internet-Draft                em-ps-req-sol                 January 2007


   o  Confidentiality: Techniques such as MIKEY [RFC3830]and Security
      Descriptions [RFC4568], which use SDP to establish keys for media
      security protocols, require a full SDP offer/answer exchange to
      negotiate keys.  Such keying techniques are therefore incompatible
      with early media.

   o  Access Control: In the presence of early media, a UAC must
      anticipate that media will arrive without any signaling, and thus
      is obliged to receive (and presumably render) media received from
      any host on the Internet.

   o  Interaction with forking/retargeting: When an INVITE from a UAC is
      forked or retarget to UASs that send early media, because these
      UASs send no signaling, the UAC has no information besides the
      media itself on which to base a decision as to which media stream
      to render.

   o  Denial of Service: Early media is the basis of the "voice hammer"
      attack, described in [draft-ietf-mmusic-ice].

   To date, attempts to deal with early media have failed to solve all
   of the problems caused by early media.  Such work typically has
   started from the assumption that early media is a necessary component
   of SIP (largely due to RFC 3398 [RFC3398] and RFC 3578 [RFC3578]) and
   has generally avoided the issue of whether key use-cases (such as
   PSTN interoperability) can be satisfied without the use of early
   media.  RFC 3960 presents two models for managing early media
   streams, which mitigate early-media related problems in some cases.
   However, the document neither specifies the models in sufficient
   detail nor makes their implementation a mandatory part of SIP.  More
   recently, Stucker has written a draft extending 3960 by detailing a
   more current set of early media coping mechanisms, and proposing a
   solution that uses additional SDP attributes to mitigate the problems
   caused by early media [draft-stucker-sipping-early-media-coping].
   Although Stucker's proposed solution addresses several of the above
   problems, it stops short of requiring an SDP exchange before media
   transmission.  Ejzak has also authored a recent draft regarding early
   media [draft-ejzak-sipping-p-em-auth].  However, his draft proposes
   the use of a p-header to prevent fraudulent uses of early media,
   which are largely orthogonal to the early media problems considered
   in this draft.

   In contrast to previous approaches, this draft proposes to address
   the problems of early media by normatively modifying the SIP protocol
   to preclude the transmission of early media.  In order to remove the
   possibility of early media, this modification must ensure (to the
   greatest extent possible) that the SDP offer/answer exchange has
   completed before any media are sent.  At a high level, this goal can



Barnes & Lepinski         Expires July 19, 2007                 [Page 4]

Internet-Draft                em-ps-req-sol                 January 2007


   be approached from two directions: To complete the SDP exchange as
   quickly and reliably as possible, and to forbid sending of media
   before completion.  A successful solution will combine these two
   approaches, since as progress is made on the former, the latter
   becomes less onerous.

   The goal of this draft is three-fold: First, in the above, we have
   provided a consolidated statement of the early media problem, namely
   that the presence of media before the completion of an SDP offer/
   answer exchange is harmful to the functioning of SIP calls.  Second,
   Section 2 specifies requirements for an update to SIP that eliminates
   early media, in an effort to consolidate the requirements that have
   been presented to date and enable discussion of solutions.  Third,
   Section 3 lays out the space of possible solutions, enumerating the
   trade-offs entailed by each so that consensus can be reached on the
   direction that should be taken by the SIP working group.


2.  Requirements

   There have been several informal proposals to date that would
   mitigate or eliminate early media, but each was disqualified in turn
   by requirements as perceived by the SIP and SIPPING working groups.
   In this section, we seek to create a consolidated set of requirements
   so that a consistent evaluation of solutions can move forward.
   Below, we use the term "proposed solution" to refer to a protocol
   that normatively updates the relevant standards (e.g., RFCs 3261,
   3264, and 3398) in such a way as to eliminate early media, while we
   use the term "current SIP" to refer to SIP as defined in RFC 3261 and
   related documents.

2.1.  Elmination of Early Media

   The proposed solution must define a normative modification to the SIP
   protocol (and other relevant protocols) that minimizes or eliminates
   the possibility of early media in SIP sessions; i.e., it should
   minimize or eliminate the possibility that a compliant implementation
   could send media before an SDP offer/answer exchange has completed.
   Solutions in which receipt of SDP-bearing messages is explicitly
   acknowledged (e.g., an INVITE/200/ACK or INVITE/180/PRACK exchange)
   are preferred, since these acknowledgements are the only way to
   completely eliminate early media (otherwise the answerer cannot know
   with certainty that the offerer has received his answer).  If a
   proposed solution does not include such acknowledgements, it must
   define UAS behaviors that minimize the probability of the UAS sending
   media before the UAC has received the SDP answer.





Barnes & Lepinski         Expires July 19, 2007                 [Page 5]

Internet-Draft                em-ps-req-sol                 January 2007


2.2.  Additional Messaging

   In order to minimize the impact of the proposed solution on call-
   setup times, the proposed solution should minimize the number of
   signaling messages that are required in addition to those used by
   current SIP.  Proposed solutions should accommodate the fact that
   messages sent "on the media path" (i.e. directly between the two SIP
   endpoints) will often be unable to reach their destinations due to
   current media gating techniques.

2.3.  Interoperability with current SIP standards

   The proposed solution must define how entities implementing it will
   interact with entities that implement the current SIP protocol.  The
   proposed solution should minimize the set of conditions under which a
   call would fail in such a hybrid scenario, but not in a scenario in
   which both endpoints use current SIP.

2.4.  Interoperability with the PSTN

   The proposed solution must be compatible with the creation of
   gateways to the PSTN, and may define normative updates as necessary
   to RFC 3398 and RFC 3578.

2.5.  Denial of Service

   The proposed solution must minimize the possibility of new denial of
   service attacks by minimizing the amount of state that is kept by a
   UAS prior to the completion of an offer/answer exchange.  Ideally, no
   state in addition to the SIP INVITE state machine will be required
   (i.e., the initial SDP offer/answer will be conducted statelessly).

2.6.  IPR Constraints

   The proposed solution must, to the extent possible, avoid dependence
   on technologies on which there are known IPR constraints, as this
   tends to hinder adoption.


3.  Early Media Solutions

   The fundamental goal of a system to eliminate early media is to
   ensure that an SDP offer/answer exchange has completed before media
   packets are sent.  Such a system will combine two approaches:
   Completing the SDP exchange as quickly and reliably as possible, and
   providing restrictions on the sending and receiving of media before
   such completion.  A priori, neither approach is at odds with the
   requirements stated above, so we explore the possible mechanisms for



Barnes & Lepinski         Expires July 19, 2007                 [Page 6]

Internet-Draft                em-ps-req-sol                 January 2007


   accomplishing each.  Many of the solutions below have been suggested
   on the SIP and SIPPING mailing lists already, sometimes by multiple
   people independently; we have not made an effort to attribute them.

3.1.  Minimizing SDP completion time

   The simplest way to preclude early media is for the SIP endpoints to
   ensure that an SDP exchange has completed before media packets are
   generated.  In order to convey the SDP messages comprising this
   exchange, it is necessary to modify the way that SDP messages are
   carried within SIP session establishment.  There are a variety of
   mechanisms for achieving this goal, including modifications to
   messages within the INVITE transaction, other non-INVITE SIP
   mechanisms, and protocols outside of SIP.

3.1.1.  Fast completion of INVITE transaction

   Perhaps the most straightforward approach to quickly completing an
   SDP offer/answer exchange would be to simply complete the SDP
   exchange already inherent in the SIP INVITE transaction more quickly:
   To do this, the UAS would send a 200/OK response as soon as an INVITE
   message is received, rather than awaiting an indication from a higher
   layer to do so (formally, this moves both the client and server
   INVITE transactions to the Terminated state as quickly as possible).
   In essence, this is a change to the semantic meaning of the "200/OK"
   response; rather than indicating that an application-layer session
   has begun, it would indicate that an media association had been
   established, in the sense that all necessary parameters are
   established for media to flow.  A separate message (for instance, an
   UPDATE [RFC3311]) could be used to indicate that an application-layer
   user is ready to begin exchanging media.

   Quickly completing the INVITE transaction has the potential to
   eliminate early media entirely: Because the 200 response is
   explicitly acknowledged by the ACK message, both endpoints can be
   sure that the SDP exchange is complete when the ACK is transmitted.
   In ISUP terms, a 200/OK response used in this way would correspond to
   an ACM (rather than to an ANM, as in RFC 3398), and the UPDATE (or
   similar message) would correspond to the ANM.  The failing of this
   approach is its interaction with SIP forking: Since a 200/OK message
   moves both endpoints to the Terminated state, the first UAS to
   respond would automatically cut off all other branches of the call.
   Although this issue could likely be resolved with a modification to
   rules for SIP forking, it could inhibit backward compatibility in the
   interim.






Barnes & Lepinski         Expires July 19, 2007                 [Page 7]

Internet-Draft                em-ps-req-sol                 January 2007


3.1.2.  Elimination of offer from INVITE

   One way to ensure that a UAS cannot send media until an SDP exchange
   has completed would be to forbid a UAC from including an SDP offer in
   its INVITE message, so that the SDP offer would be sent by the UAS in
   either a reliable provisional response or the 200/OK message, and the
   answer in the UAC's PRACK or ACK message.  This behavior is allowed
   by current SIP, but does not eliminate early media because it is not
   mandatory.  At first blush, reusing the existing INVITE transaction
   in this way seems appealing: Because it is a subset of behaviors
   allowed by current SIP, it will interact well with current SIP
   endpoints, and it introduces no additional messaging or state
   machinery.  However, if either endpoint does not support reliable
   provisional responses [RFC3262], this solution would prevent any
   media flows before the sending of a 200/OK (distinct from an SDP
   answer), which would cause many existing SIP applications to fail
   (including a large number of PSTN applications as they pass through
   gateways).  In addition, the ACK message is not a reliable mechanism
   for transporting SDP answers, since the UAC (as an SDP answerer)
   receives no notification that the UAS has received its answer.

3.1.3.  Provisional Responses

   Within the INVITE transaction, the only other messages sent from UAS
   to UAC (besides the INVITE, 200, and ACK messages) that could carry
   SDP information are provisional (101-199) responses (response type
   100 is excepted because it is often not retransmitted by proxies).
   Upon receipt of an INVITE message, a UAS would respond with a 1xx
   message that would complete the SDP offer/answer.  Some have
   suggested use of the 183 (Call Progress) message for this purpose, as
   is done in ICE, though another existing or new response type could in
   principle be used.

   This approach has the advantage that if the UAS does not support this
   use of provisional messages, then the call will proceed in the same
   way as with current SIP.  For PSTN gateways, this approach only
   requires the sending or understanding a single additional message.
   However, it would be difficult to use provisional responses to
   eliminate early media, since this would require them to be reliable.
   Though there exist both SIP (PRACK) and non-SIP (ICE binding request)
   messages that can acknowledge provisional responses, AT&T holds IPR
   related to reliable provisional responses.  This has limited the
   deployment of PRACK, and may have the same effect on relevant
   portions of ICE.







Barnes & Lepinski         Expires July 19, 2007                 [Page 8]

Internet-Draft                em-ps-req-sol                 January 2007


3.1.4.  Separate INVITE transaction

   The Application Server model of RFC 3960 provides an SDP exchange for
   early media by conducting separate INVITE transactions for an "Early"
   session and for the "Final" session.  In a recently proposed
   realization and extension of this model
   [draft-stucker-sipping-early-media-indirection], any entity along the
   signaling path can include an EMIND (Early Media Indirection) header
   that will prompt the UAC to initiate a separate INVITE transaction
   with a third entity, e.g., a media server to provide ring-back tones
   or balance announcements.  This transaction establishes an "Early
   Session," which is terminated when the orignal INVITE transaction is
   completed and the "Final Session" begun.

   Using a separate INVITE transaction for early media has several
   advantages: Since the answer in a 2xx response triggers an ACK in the
   early INVITE transaction, it can be used to completely prevent early
   media.  Prospects for interoperability with current SIP
   implementations are good, since no new mechanisms are employed; in
   particular, the EMIND header will be ignored by implementations that
   do not support it.  And the entire burden of state maintenance is
   placed on the UAC, which must manage both INVITE transactions.
   However, using two separate INVITE transactions incurs significant
   siganling overhead, and it is unclear that this approach, at least as
   currently specified, can eliminate early media from PSTN gateways
   that are unable to separate early and regular media.

3.1.5.  Non-INVITE SIP Mechanisms

   There are other SIP messages that could be used to carry an SDP
   answer that would be formally outside of the INVITE transaction, but
   could be contemporaneous with it.  Because these messages would be
   outside the INVITE transaction, a separate mechanism may be required
   to associate the transmitted SDP answer with the SDP offer in the
   INVITE request.

3.1.5.1.  INFO and UPDATE

   The SIP INFO method [RFC2976] is currently used to convey additional
   signaling information in the middle of a session, after the INVITE
   transaction has terminated.  However, it could also be used to convey
   information relevant to an INVITE transaction, subject to the
   constraint in RFC 2976 that an INFO request must not change the state
   of a SIP call.  While receipt of an SDP answer in an INFO request
   would change state at the TU layer (namely, in the UAC core), it
   would not change the transaction-layer state of the call.  INFO
   messages carrying SDP answers can be associated with the
   corresponding SDP offer by virtue of the fact that the INFO message



Barnes & Lepinski         Expires July 19, 2007                 [Page 9]

Internet-Draft                em-ps-req-sol                 January 2007


   carrying the answer will carry the same Call-ID as the INVITE message
   carrying the offer.  The SIP UPDATE method [RFC3311]could be likewise
   used (in fact, RFC 3311 lists early media as the main motivation for
   UPDATE, and RFC 3960 uses UPDATE messages as the basis for its
   gateway model).

   Current standards require that the UAS in an INFO or UPDATE
   transaction to send a response to an INFO or UPDATE message, so these
   messages would provide a reliable transport for SDP answers.  Just as
   with provisional responses, using INFO or UPDATE messages to convey
   SDP answers is backwards-compatible, and requires minimal
   modification to PSTN gateways.  One potential drawback is that as
   currently specified, UPDATE and INFO messages cannot be sent before a
   dialog is established (i.e., a 101-199 response is sent).  If not
   changed or lifted by the proposed solution, this restriction could
   limit the efficacy of UPDATE (or INFO) for quickly concluding an SDP
   exchange, since the required 1xx message would introduce additional
   delay in the transmission of the SDP answer.

3.1.5.2.  Other Response Codes

   In current SIP, if a UAS wishes to send early media in response to an
   INVITE, it will send the media without any SIP response.  As an
   alternative, the UAS wishing to send early media could send a 401
   response requiring http-auth with null authentication.  The UAC would
   then signal its willingness to receive the media by sending a new
   INVITE (with the nonce for null authentication).  In this model, the
   UAS's SDP answer would be sent in its 401 response, and the UAC's re-
   INVITE would confirm receipt of this answer.

   Although this proposal does provide a reliable transport for SDP
   answers, it has significant drawbacks: In addition to adding
   significant messaging overhead, it requires both endpoints to
   maintain state across INVITE transactions (in addition to the nonce
   required by null authentication).  And the prospects for backward
   compatibility with this solution are bleak: All calls to current SIP
   endpoints will fail unless the endpoints support the indicated
   authentication mechanisms, and forked calls would show unpredictable
   behavior due to the Heterogeneous Error Response Forking Problem
   (HERFP).

3.1.6.  Non-SIP mechanisms

   There are some proposals to use protocols other than SIP to resolve
   the problem of early media in SIP.  For instance, a lower-level
   protocol could be defined that could negotiate and SDP session that
   could then be used by SIP, much as TLS negotiations are conducted
   before HTTPS data are sent.  However, such a protocol would have to



Barnes & Lepinski         Expires July 19, 2007                [Page 10]

Internet-Draft                em-ps-req-sol                 January 2007


   duplicate many SIP features, and any such solution would require SIP
   entities to implement an additional protocol stack, which would have
   negative consequences for backward-compatability and PSTN
   interoperability.

3.2.  Handling remaining media

   Early media can be completely eliminated only if neither party sends
   media until it is assured that the SDP offer/answer exchange is
   completed.  Of course, this is only possible when both endpoints
   support the proposed solution and the SDP answer is carried in a
   reliable message: In this optimal case, both the offerer and the
   answerer must not send media until they have received assurance that
   the SDP exchange has completed.  The proposed solution will specify
   the two remaining cases.

3.2.1.  Unreliable SDP answer

   If both endpoints support the solution, but the solution uses an
   unreliable message to carry the SDP answer, then the offerer can be
   assured that the SDP exchange has completed (because he has received
   the answer), but the answerer cannot.  In this case, the offerer must
   not send media until the SDP exchange has completed.  The proposed
   solution should provide an explicit, deterministic criterion for when
   the answerer may send media: For instance, the answerer may be free
   to send media after it receives media from the offerer, or after a
   timer fires that is set when the answer is sent.

3.2.2.  Backwards-compatibility

   Depending on the nature of the proposed solution, one party may not
   be able to know whether the other supports the solution.  If a party
   cannot know, then it must behave as if the other party does support
   the solution (otherwise, the solution would never be used).  If a
   party can be aware of the other party's support, then the solution
   should allow a supporting party to knowingly send early media to an
   non-supporting party, since to do otherwise -- to mandate that a
   supporting party behave as it would if the other party supported the
   solution -- is incompatible with existing SIP applications that use
   early media.  Nonetheless, an implementation of the solution should
   always avoid the use of early media; it should only send early media
   when it has received early media from a non-supporting endpoint.


4.  IANA Considerations

   This document makes no request of IANA.




Barnes & Lepinski         Expires July 19, 2007                [Page 11]

Internet-Draft                em-ps-req-sol                 January 2007


   Note to RFC Editor: this section may be removed on publication as an
   RFC.


5.  Security Considerations

   The only new security concern that arises from this draft is the
   possibility of introducing new denial of service attacks.  The
   requirement addressing this concern is stated in Section 2.5 above.
   The performance of each candidate solution against this requirement
   is examined in the section where that solution's is proposed.


6.  Acknowledgements


7.  References

7.1.  Normative References

   [RFC3261]  "SIP: Session Initiation Protocol", June 2002.

   [RFC3264]  "An Offer/Answer Model with the Session Description
              Protocol (SDP)", June 2002.

7.2.  Informative References

   [RFC2976]  "The SIP INFO Method", October 2000.

   [RFC3262]  2, "Reliability of Provisional Responses in the Session
              Initiation Protocol (SIP)", June 2002.

   [RFC3311]  "The Session Initiation Protocol (SIP) UPDATE Method",
              September 2002.

   [RFC3398]  "Integrated Services Digital Network (ISDN) User Part
              (ISUP) to Session Initiation Protocol (SIP) Mapping",
              December 2002.

   [RFC3578]  "Mapping of Integrated Services Digital Network (ISDN)
              User Part (ISUP) Overlap Signalling to the Session
              Initiation Protocol (SIP)", August 2003.

   [RFC3830]  "MIKEY: Multimedia Internet KEYing", August 2004 2004.

   [RFC3960]  "Early Media and Ringing Tone Generation in the Session
              Initiation Protocol (SIP)", December 2004.




Barnes & Lepinski         Expires July 19, 2007                [Page 12]

Internet-Draft                em-ps-req-sol                 January 2007


   [RFC4568]  "Session Description Protocol (SDP) Security Descriptions
              for Media Streams", July 2006.

   [draft-ejzak-sipping-p-em-auth]
              "Private Header (P-Header) Extension to the Session
              Initiation Protocol (SIP) for Authorization of Early
              Media", January 2007.

   [draft-ietf-mmusic-ice]
              "Interactive Connectivity Establishment (ICE): A
              Methodology for Network Address Translator (NAT) Traversal
              for Offer/Answer Protocols", October 2006.

   [draft-stucker-sipping-early-media-coping]
              "Coping with Early Media in the Session Initiation
              Protocol (SIP)", October 2006.

   [draft-stucker-sipping-early-media-indirection]
              "Early Media INDirection (EMIND) in the Session Initiation
              Protocol (SIP)", October 2006.


Authors' Addresses

   Richard Barnes
   BBN Technologies
   9861 Broken Land Pkwy
   Columbia, Maryland  21046
   USA

   Phone: +1-410-290-6169
   Email: rbarnes@bbn.com


   Matt Lepinski
   BBN Technologies
   10 Moulton St.
   Cambridge, Massachusetts  02138
   USA

   Phone: +1-617-873-5939
   Email: mlepinsk@bbn.com









Barnes & Lepinski         Expires July 19, 2007                [Page 13]

Internet-Draft                em-ps-req-sol                 January 2007


Full Copyright Statement

   Copyright (C) The IETF Trust (2007).

   This document is subject to the rights, licenses and restrictions
   contained in BCP 78, and except as set forth therein, the authors
   retain all their rights.

   This document and the information contained herein are provided on an
   "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS
   OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND
   THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS
   OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF
   THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED
   WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.


Intellectual Property

   The IETF takes no position regarding the validity or scope of any
   Intellectual Property Rights or other rights that might be claimed to
   pertain to the implementation or use of the technology described in
   this document or the extent to which any license under such rights
   might or might not be available; nor does it represent that it has
   made any independent effort to identify any such rights.  Information
   on the procedures with respect to rights in RFC documents can be
   found in BCP 78 and BCP 79.

   Copies of IPR disclosures made to the IETF Secretariat and any
   assurances of licenses to be made available, or the result of an
   attempt made to obtain a general license or permission for the use of
   such proprietary rights by implementers or users of this
   specification can be obtained from the IETF on-line IPR repository at
   http://www.ietf.org/ipr.

   The IETF invites any interested party to bring to its attention any
   copyrights, patents or patent applications, or other proprietary
   rights that may cover technology that may be required to implement
   this standard.  Please address the information to the IETF at
   ietf-ipr@ietf.org.


Acknowledgment

   Funding for the RFC Editor function is provided by the IETF
   Administrative Support Activity (IASA).





Barnes & Lepinski         Expires July 19, 2007                [Page 14]



--------------050001030003050007020902
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--------------050001030003050007020902--





From sipping-bounces@ietf.org Fri Feb 09 01:58:44 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HFPgN-0006gt-GJ; Fri, 09 Feb 2007 01:56:31 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HFPgM-0006go-9t
	for sipping@ietf.org; Fri, 09 Feb 2007 01:56:30 -0500
Received: from wip-ec-wd.wipro.com ([203.91.193.32])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HFPgB-0005Uc-FE
	for sipping@ietf.org; Fri, 09 Feb 2007 01:56:30 -0500
Received: from wip-ec-wd.wipro.com (localhost.wipro.com [127.0.0.1])
	by localhost (Postfix) with ESMTP id 44949205DD
	for <sipping@ietf.org>; Fri,  9 Feb 2007 12:18:33 +0530 (IST)
Received: from blr-ec-bh02.wipro.com (blr-ec-bh02.wipro.com [10.201.50.92])
	by wip-ec-wd.wipro.com (Postfix) with ESMTP id 2C13C205DA
	for <sipping@ietf.org>; Fri,  9 Feb 2007 12:18:33 +0530 (IST)
Received: from blr-m2-msg.wipro.com ([10.116.50.99]) by blr-ec-bh02.wipro.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 9 Feb 2007 12:26:15 +0530
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Fri, 9 Feb 2007 12:31:07 +0530
Message-ID: <532B18E13CF9E64380EF5FDDE265E071B6192E@blr-m2-msg.wipro.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdK21NyBYG3D/ErQ02kd6cS6HKcCABPEoYQ
References: <45C0AEF2.8010101@ericsson.com>	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
	<1170835642.29763.54.camel@n95.nomadiclab.com>
	<45CA08B2.4050603@cisco.com>
From: <ravishankar.shiroor@wipro.com>
To: <pkyzivat@cisco.com>,
	<salvatore.loreto@ericsson.com>
X-OriginalArrivalTime: 09 Feb 2007 06:56:15.0361 (UTC)
	FILETIME=[6445E710:01C74C17]
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 2bf730a014b318fd3efd65b39b48818c
Cc: sipping@ietf.org, Gonzalo.Camarillo@ericsson.com, mary.barnes@nortel.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi Paul,

IMO, service identification is different from identifying the tool that
should be used for executing that service on the end device. In the
analogy that you have given, e-mail is the service. Eudora/gmail/... Are
the tools that are used to provide the service.

Regards,
Ravi.


--
=20
Ravishankar. G. Shiroor
Wipro Technologies, Bangalore.
=20
ravishankar.shiroor@wipro.com
--
=20

> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
> Sent: Wednesday, February 07, 2007 10:43 PM
> To: Salvatore Loreto (JO/LMF)
> Cc: RaviShankar Shiroor (WT01 - Product Engineering=20
> Solutions); sipping@ietf.org; Gonzalo.Camarillo@ericsson.com;=20
> mary.barnes@nortel.com
> Subject: Re: [Sipping] Service Identification in SIP
>=20
> I still don't get the intent of identifying the application.=20
> It seems to me this is evil - akin to sending an email=20
> message and identifying that the recipient must use Outlook=20
> to receive it, not Thunderbird, Eudora, gmail, or whatever.
>=20
> I also wish somebody could give better examples of a=20
> communication service. Chess is too artificial to be=20
> believable, and IMO PoC is only a service because of the way=20
> the implementation has been done.=20
> Conceptually PoC could be nothing but a special floor control=20
> protocol accompanying a voice session. In that case none of=20
> the Service mechanism would be needed.
>=20
> 	Paul
>=20
> Salvatore Loreto (JO/LMF) wrote:
> > hi Ravi,
> >=20
> > in IMS a Service is a "type of communication defined by a service=20
> > definition that specifies the rules and procedures and=20
> allowed media,=20
> > if any, for a specific type of communication and that=20
> utilises the IMS=20
> > enabler. An IMS communication service defines what is=20
> possible within=20
> > a single SIP session or standalone transaction."
> > So a Service implies some combination of capabilities and=20
> ways to use=20
> > those capabilities.
> > If you want the Service specifies the context for a communication.
> >=20
> > Instead the Application define the specific communication=20
> (or if you=20
> > want application) that you want run in the specific context. So you=20
> > should have several different applications designed to run on a=20
> > specific environment or that have different performance or=20
> behavior in=20
> > different context (running on top of different service).
> >=20
> > Lets go to Jonathan's example of the Chess.
> > In his example the Chess is a service, for IMS instead is an=20
> > application that you can run on top of different service.
> > For example you can run Chess application on top of PoC=20
> service or on=20
> > top of the "standard" service. So let say you run Chess on=20
> top of PoC=20
> > service, then you can use only voice in the way that PoC specifies.
> >=20
> > /sal
> >=20
> >=20
> >=20
> >=20
> > On Thu, 2007-02-01 at 16:39 +0530,=20
> ravishankar.shiroor@wipro.com wrote:
> >> Hello,
> >>
> >> IMO, Service Identification is definitely needed and=20
> should be taken=20
> >> up as a WG as proposed below.
> >>
> >> Further, I think Service Identification scope should be as per the=20
> >> requirement in Jonathan's draft, as I still do not understand the=20
> >> need for having a Communication Service Identifier and another=20
> >> Application Identifier as given in the=20
> 3gpp-ics-requirements document.
> >>
> >> Regards,
> >> Ravi.
> >>
> >> --
> >> =20
> >> Ravishankar. G. Shiroor
> >> Wipro Technologies, Bangalore.
> >> =20
> >> ravishankar.shiroor@wipro.com
> >> --
> >> =20
> >>
> >>> -----Original Message-----
> >>> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
> >>> Sent: Wednesday, January 31, 2007 8:30 PM
> >>> To: sipping
> >>> Cc: Mary Barnes
> >>> Subject: [Sipping] Service Identification in SIP
> >>>
> >>> Folks,
> >>>
> >>> during the last year, there have been a number of threads on the=20
> >>> mailing list discussing service identification in SIP. In=20
> >>> particular, the following thread discussed 3GPP=20
> requirements on this=20
> >>> area:
> >>>
> >>>=20
> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
> >>>
> >>> At a later point, those 3GPP requirements were documented in the=20
> >>> following (expired by now) Internet Draft, which was=20
> presented in a=20
> >>> face-to-face IETF meeting.
> >>>
> >>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
> >>> cs-requirements-00.txt
> >>>
> >>> Lately, the following draft has been submitted to the archives:
> >>>
> >>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
> >>> rvice-identification-00.txt
> >>>
> >>> This draft proposes a solution that intends to meet known=20
> >>> requirements in this area.
> >>>
> >>> Given the amount of discussions on issues related to service=20
> >>> identification we have seen during the last year, we=20
> believe this is=20
> >>> a fairly important topic for the SIPPING WG to tackle. At some=20
> >>> point, we would like to have a milestone with its=20
> associated WG item=20
> >>> dealing with this issue.
> >>>
> >>> Therefore, we would like to encourage people to read this=20
> draft and=20
> >>> send comments to the list. Additionally, if you believe that this=20
> >>> work is (or is not) important, we would like to hear that as well.
> >>>
> >>> Cheers,
> >>>
> >>> Gonzalo
> >>> SIPPING co-chair
> >>>
> >>>
> >>> _______________________________________________
> >>> Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> >>> This list is for NEW development of the application of SIP Use=20
> >>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> >>> sip@ietf.org for new developments of core SIP
> >>>
> >> _______________________________________________
> >> Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> >> This list is for NEW development of the application of SIP Use=20
> >> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> >> sip@ietf.org for new developments of core SIP
> >=20
> >=20
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP Use=20
> > sip-implementors@cs.columbia.edu for questions on current sip Use=20
> > sip@ietf.org for new developments of core SIP
> >=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 09 03:11:21 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HFQpB-0000FP-WE; Fri, 09 Feb 2007 03:09:42 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HFQpA-0000FK-Jp
	for sipping@ietf.org; Fri, 09 Feb 2007 03:09:40 -0500
Received: from wip-ec-wd.wipro.com ([203.91.193.32])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HFQp6-0004NG-Oz
	for sipping@ietf.org; Fri, 09 Feb 2007 03:09:40 -0500
Received: from wip-ec-wd.wipro.com (localhost.wipro.com [127.0.0.1])
	by localhost (Postfix) with ESMTP id 86A53205DA
	for <sipping@ietf.org>; Fri,  9 Feb 2007 13:31:46 +0530 (IST)
Received: from blr-ec-bh01.wipro.com (blr-ec-bh01.wipro.com [10.201.50.91])
	by wip-ec-wd.wipro.com (Postfix) with ESMTP id 6FDAB205D9
	for <sipping@ietf.org>; Fri,  9 Feb 2007 13:31:46 +0530 (IST)
Received: from blr-m2-msg.wipro.com ([10.116.50.99]) by blr-ec-bh01.wipro.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 9 Feb 2007 13:39:28 +0530
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Fri, 9 Feb 2007 13:44:20 +0530
Message-ID: <532B18E13CF9E64380EF5FDDE265E071B61978@blr-m2-msg.wipro.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdKjuwlCQhcBBrAT7W0aZIf2BKFrQBktcDA
References: <45C0AEF2.8010101@ericsson.com>
	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
	<1170835642.29763.54.camel@n95.nomadiclab.com>
From: <ravishankar.shiroor@wipro.com>
To: <salvatore.loreto@ericsson.com>
X-OriginalArrivalTime: 09 Feb 2007 08:09:28.0519 (UTC)
	FILETIME=[9ECC9D70:01C74C21]
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 6ba8aaf827dcb437101951262f69b3de
Cc: sipping@ietf.org, Gonzalo.Camarillo@ericsson.com, mary.barnes@nortel.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi Salvatore,

As an end user, poc is also an application that I use and chess is also
an application that I can use. The way I see it, in your example, it
will be running chess application on top of PoC application.

So the service urn for this would be something like:

Urn:service:voice.poc.chess

Regards,
Ravi.

--
=20
Ravishankar. G. Shiroor
Wipro Technologies, Bangalore.
=20
ravishankar.shiroor@wipro.com
--
=20

> -----Original Message-----
> From: Salvatore Loreto (JO/LMF)=20
> [mailto:salvatore.loreto@ericsson.com]=20
> Sent: Wednesday, February 07, 2007 1:37 PM
> To: RaviShankar Shiroor (WT01 - Product Engineering Solutions)
> Cc: Gonzalo.Camarillo@ericsson.com; sipping@ietf.org;=20
> mary.barnes@nortel.com
> Subject: RE: [Sipping] Service Identification in SIP
>=20
> hi Ravi,
>=20
> in IMS a Service is a "type of communication defined by a=20
> service definition that specifies the rules and procedures=20
> and allowed media, if any, for a specific type of=20
> communication and that utilises the IMS enabler. An IMS=20
> communication service defines what is possible within a=20
> single SIP session or standalone transaction."
> So a Service implies some combination of capabilities and=20
> ways to use those capabilities.
> If you want the Service specifies the context for a communication.
>=20
> Instead the Application define the specific communication (or=20
> if you want application) that you want run in the specific=20
> context. So you should have several different applications=20
> designed to run on a specific environment or that have=20
> different performance or behavior in different context=20
> (running on top of different service).
>=20
> Lets go to Jonathan's example of the Chess.
> In his example the Chess is a service, for IMS instead is an=20
> application that you can run on top of different service.
> For example you can run Chess application on top of PoC=20
> service or on top of the "standard" service. So let say you=20
> run Chess on top of PoC service, then you can use only voice=20
> in the way that PoC specifies.
>=20
> /sal
>=20
>=20
>=20
>=20
> On Thu, 2007-02-01 at 16:39 +0530,=20
> ravishankar.shiroor@wipro.com wrote:
> > Hello,
> >=20
> > IMO, Service Identification is definitely needed and should=20
> be taken=20
> > up as a WG as proposed below.
> >=20
> > Further, I think Service Identification scope should be as per the=20
> > requirement in Jonathan's draft, as I still do not=20
> understand the need=20
> > for having a Communication Service Identifier and another=20
> Application=20
> > Identifier as given in the 3gpp-ics-requirements document.
> >=20
> > Regards,
> > Ravi.
> >=20
> > --
> > =20
> > Ravishankar. G. Shiroor
> > Wipro Technologies, Bangalore.
> > =20
> > ravishankar.shiroor@wipro.com
> > --
> > =20
> >=20
> > > -----Original Message-----
> > > From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
> > > Sent: Wednesday, January 31, 2007 8:30 PM
> > > To: sipping
> > > Cc: Mary Barnes
> > > Subject: [Sipping] Service Identification in SIP
> > >=20
> > > Folks,
> > >=20
> > > during the last year, there have been a number of threads on the=20
> > > mailing list discussing service identification in SIP. In=20
> > > particular, the following thread discussed 3GPP=20
> requirements on this=20
> > > area:
> > >=20
> > >=20
> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
> > >=20
> > > At a later point, those 3GPP requirements were documented in the=20
> > > following (expired by now) Internet Draft, which was=20
> presented in a=20
> > > face-to-face IETF meeting.
> > >=20
> > > http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
> > > cs-requirements-00.txt
> > >=20
> > > Lately, the following draft has been submitted to the archives:
> > >=20
> > > http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
> > > rvice-identification-00.txt
> > >=20
> > > This draft proposes a solution that intends to meet known=20
> > > requirements in this area.
> > >=20
> > > Given the amount of discussions on issues related to service=20
> > > identification we have seen during the last year, we=20
> believe this is=20
> > > a fairly important topic for the SIPPING WG to tackle. At some=20
> > > point, we would like to have a milestone with its=20
> associated WG item=20
> > > dealing with this issue.
> > >=20
> > > Therefore, we would like to encourage people to read this=20
> draft and=20
> > > send comments to the list. Additionally, if you believe that this=20
> > > work is (or is not) important, we would like to hear that as well.
> > >=20
> > > Cheers,
> > >=20
> > > Gonzalo
> > > SIPPING co-chair
> > >=20
> > >=20
> > > _______________________________________________
> > > Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> > > This list is for NEW development of the application of SIP Use=20
> > > sip-implementors@cs.columbia.edu for questions on current sip Use=20
> > > sip@ietf.org for new developments of core SIP
> > >=20
> >=20
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP Use=20
> > sip-implementors@cs.columbia.edu for questions on current sip Use=20
> > sip@ietf.org for new developments of core SIP
>=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 09 08:31:36 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HFVpS-0004pD-9f; Fri, 09 Feb 2007 08:30:18 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HFVpN-0004mx-Mi
	for sipping@ietf.org; Fri, 09 Feb 2007 08:30:13 -0500
Received: from sj-iport-2-in.cisco.com ([171.71.176.71]
	helo=sj-iport-2.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HFVmI-0007dR-IA
	for sipping@ietf.org; Fri, 09 Feb 2007 08:27:07 -0500
Received: from sj-dkim-1.cisco.com ([171.71.179.21])
	by sj-iport-2.cisco.com with ESMTP; 09 Feb 2007 05:27:02 -0800
X-IronPort-AV: i="4.13,306,1167638400"; 
	d="scan'208"; a="360209269:sNHT60059280"
Received: from sj-core-1.cisco.com (sj-core-1.cisco.com [171.71.177.237])
	by sj-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l19DR2uv008657; 
	Fri, 9 Feb 2007 05:27:02 -0800
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by sj-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l19DQiDw025766;
	Fri, 9 Feb 2007 05:26:57 -0800 (PST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 9 Feb 2007 08:26:48 -0500
Received: from [161.44.183.228] ([161.44.183.228]) by
	xfe-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 9 Feb 2007 08:26:48 -0500
Message-ID: <45CC7697.1060204@cisco.com>
Date: Fri, 09 Feb 2007 08:26:47 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: ravishankar.shiroor@wipro.com
Subject: Re: [Sipping] Service Identification in SIP
References: <45C0AEF2.8010101@ericsson.com>	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
	<1170835642.29763.54.camel@n95.nomadiclab.com>
	<45CA08B2.4050603@cisco.com>
	<532B18E13CF9E64380EF5FDDE265E071B6192E@blr-m2-msg.wipro.com>
In-Reply-To: <532B18E13CF9E64380EF5FDDE265E071B6192E@blr-m2-msg.wipro.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 09 Feb 2007 13:26:48.0253 (UTC)
	FILETIME=[F35D76D0:01C74C4D]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=7696; t=1171027622;
	x=1171891622; c=relaxed/simple; s=sjdkim1004;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20 |To:=20ravishankar.shiroor@wipro.com
	|Content-Type:=20text/plain=3B=20charset=3DISO-8859-1=3B=20format=3Dflowe
	d |Content-Transfer-Encoding:=207bit |MIME-Version:=201.0;
	bh=uwP37mYO4+2enV9HJimx3grxD/T7YahYQJV84Ar03nU=;
	b=eC/RS2t6bD1ayMW5zO/ROc921X6UtqaJli5uJtNdfzShNH8hpNS6bBnCUmdj7D81UT3IqiUR
	WphWIW805oHyj5uMPMAPInd77M4+1Gl+89awI+pxDxFdfy+gBayihMxP6U5AeEj9HgWFA0gXgU
	V9YYbPZ4lynfCGDiBp9fd4YDo=;
Authentication-Results: sj-dkim-1; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim1004 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 311e798ce51dbeacf5cdfcc8e9fda21b
Cc: sipping@ietf.org, Gonzalo.Camarillo@ericsson.com, mary.barnes@nortel.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



ravishankar.shiroor@wipro.com wrote:
> Hi Paul,
> 
> IMO, service identification is different from identifying the tool that
> should be used for executing that service on the end device. In the
> analogy that you have given, e-mail is the service. Eudora/gmail/... Are
> the tools that are used to provide the service.

The comment below was about identifying the application in the end 
system, not the protocol.

In the case of email, the "service" is determined by the protocol and 
the address/port of the destination. Nobody has suggested adding headers 
to mail messages to identify the service/application in the end system 
that should process the mail message. But here we see proposals to add a 
header to a sip message to identify the service/application in the UAS.

	Paul

> Regards,
> Ravi.
> 
> 
> --
>  
> Ravishankar. G. Shiroor
> Wipro Technologies, Bangalore.
>  
> ravishankar.shiroor@wipro.com
> --
>  
> 
>> -----Original Message-----
>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
>> Sent: Wednesday, February 07, 2007 10:43 PM
>> To: Salvatore Loreto (JO/LMF)
>> Cc: RaviShankar Shiroor (WT01 - Product Engineering 
>> Solutions); sipping@ietf.org; Gonzalo.Camarillo@ericsson.com; 
>> mary.barnes@nortel.com
>> Subject: Re: [Sipping] Service Identification in SIP
>>
>> I still don't get the intent of identifying the application. 
>> It seems to me this is evil - akin to sending an email 
>> message and identifying that the recipient must use Outlook 
>> to receive it, not Thunderbird, Eudora, gmail, or whatever.
>>
>> I also wish somebody could give better examples of a 
>> communication service. Chess is too artificial to be 
>> believable, and IMO PoC is only a service because of the way 
>> the implementation has been done. 
>> Conceptually PoC could be nothing but a special floor control 
>> protocol accompanying a voice session. In that case none of 
>> the Service mechanism would be needed.
>>
>> 	Paul
>>
>> Salvatore Loreto (JO/LMF) wrote:
>>> hi Ravi,
>>>
>>> in IMS a Service is a "type of communication defined by a service 
>>> definition that specifies the rules and procedures and 
>> allowed media, 
>>> if any, for a specific type of communication and that 
>> utilises the IMS 
>>> enabler. An IMS communication service defines what is 
>> possible within 
>>> a single SIP session or standalone transaction."
>>> So a Service implies some combination of capabilities and 
>> ways to use 
>>> those capabilities.
>>> If you want the Service specifies the context for a communication.
>>>
>>> Instead the Application define the specific communication 
>> (or if you 
>>> want application) that you want run in the specific context. So you 
>>> should have several different applications designed to run on a 
>>> specific environment or that have different performance or 
>> behavior in 
>>> different context (running on top of different service).
>>>
>>> Lets go to Jonathan's example of the Chess.
>>> In his example the Chess is a service, for IMS instead is an 
>>> application that you can run on top of different service.
>>> For example you can run Chess application on top of PoC 
>> service or on 
>>> top of the "standard" service. So let say you run Chess on 
>> top of PoC 
>>> service, then you can use only voice in the way that PoC specifies.
>>>
>>> /sal
>>>
>>>
>>>
>>>
>>> On Thu, 2007-02-01 at 16:39 +0530, 
>> ravishankar.shiroor@wipro.com wrote:
>>>> Hello,
>>>>
>>>> IMO, Service Identification is definitely needed and 
>> should be taken 
>>>> up as a WG as proposed below.
>>>>
>>>> Further, I think Service Identification scope should be as per the 
>>>> requirement in Jonathan's draft, as I still do not understand the 
>>>> need for having a Communication Service Identifier and another 
>>>> Application Identifier as given in the 
>> 3gpp-ics-requirements document.
>>>> Regards,
>>>> Ravi.
>>>>
>>>> --
>>>>  
>>>> Ravishankar. G. Shiroor
>>>> Wipro Technologies, Bangalore.
>>>>  
>>>> ravishankar.shiroor@wipro.com
>>>> --
>>>>  
>>>>
>>>>> -----Original Message-----
>>>>> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
>>>>> Sent: Wednesday, January 31, 2007 8:30 PM
>>>>> To: sipping
>>>>> Cc: Mary Barnes
>>>>> Subject: [Sipping] Service Identification in SIP
>>>>>
>>>>> Folks,
>>>>>
>>>>> during the last year, there have been a number of threads on the 
>>>>> mailing list discussing service identification in SIP. In 
>>>>> particular, the following thread discussed 3GPP 
>> requirements on this 
>>>>> area:
>>>>>
>>>>>
>> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
>>>>> At a later point, those 3GPP requirements were documented in the 
>>>>> following (expired by now) Internet Draft, which was 
>> presented in a 
>>>>> face-to-face IETF meeting.
>>>>>
>>>>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
>>>>> cs-requirements-00.txt
>>>>>
>>>>> Lately, the following draft has been submitted to the archives:
>>>>>
>>>>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
>>>>> rvice-identification-00.txt
>>>>>
>>>>> This draft proposes a solution that intends to meet known 
>>>>> requirements in this area.
>>>>>
>>>>> Given the amount of discussions on issues related to service 
>>>>> identification we have seen during the last year, we 
>> believe this is 
>>>>> a fairly important topic for the SIPPING WG to tackle. At some 
>>>>> point, we would like to have a milestone with its 
>> associated WG item 
>>>>> dealing with this issue.
>>>>>
>>>>> Therefore, we would like to encourage people to read this 
>> draft and 
>>>>> send comments to the list. Additionally, if you believe that this 
>>>>> work is (or is not) important, we would like to hear that as well.
>>>>>
>>>>> Cheers,
>>>>>
>>>>> Gonzalo
>>>>> SIPPING co-chair
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Sipping mailing list  
>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>> This list is for NEW development of the application of SIP Use 
>>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>>> sip@ietf.org for new developments of core SIP
>>>>>
>>>> _______________________________________________
>>>> Sipping mailing list  
>> https://www1.ietf.org/mailman/listinfo/sipping
>>>> This list is for NEW development of the application of SIP Use 
>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>> sip@ietf.org for new developments of core SIP
>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP Use 
>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>> sip@ietf.org for new developments of core SIP
>>>
> 
> 
> The information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments. 
> 
> WARNING: Computer viruses can be transmitted via email. The recipient should check this email and any attachments for the presence of viruses. The company accepts no liability for any damage caused by any virus transmitted by this email.
>  
> www.wipro.com
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 09 09:39:35 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HFWtS-0007rY-B1; Fri, 09 Feb 2007 09:38:30 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HFWtR-0007r1-A6
	for sipping@ietf.org; Fri, 09 Feb 2007 09:38:29 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HFWtK-0002Lw-BU
	for sipping@ietf.org; Fri, 09 Feb 2007 09:38:29 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	834AD2057C; Fri,  9 Feb 2007 15:38:11 +0100 (CET)
X-AuditID: c1b4fb3c-b17cbbb0000007de-a2-45cc8753e6aa 
Received: from esealmw127.eemea.ericsson.se (unknown [153.88.254.122])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	64212204AE; Fri,  9 Feb 2007 15:38:11 +0100 (CET)
Received: from esealmw127.eemea.ericsson.se ([153.88.254.175]) by
	esealmw127.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 9 Feb 2007 15:38:11 +0100
Received: from mail.lmf.ericsson.se ([131.160.11.50]) by
	esealmw127.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 9 Feb 2007 15:38:10 +0100
Received: from nomadiclab.lmf.ericsson.se (nomadiclab.lmf.ericsson.se
	[131.160.33.3])
	by mail.lmf.ericsson.se (Postfix) with ESMTP id CC0DC2358;
	Fri,  9 Feb 2007 16:38:10 +0200 (EET)
Received: from nomadiclab.lmf.ericsson.se (localhost [127.0.0.1])
	by nomadiclab.lmf.ericsson.se (Postfix) with ESMTP id 5BDCB4DBB4;
	Fri,  9 Feb 2007 16:38:10 +0200 (EET)
Received: from localhost.localdomain (localhost [IPv6:::1])
	by nomadiclab.lmf.ericsson.se (Postfix) with ESMTP id 9118D4DA95;
	Fri,  9 Feb 2007 16:38:09 +0200 (EET)
Subject: Re: [Sipping] Service Identification in SIP
From: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>
To: Paul Kyzivat <pkyzivat@cisco.com>
In-Reply-To: <45CC7697.1060204@cisco.com>
References: <45C0AEF2.8010101@ericsson.com>
	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
	<1170835642.29763.54.camel@n95.nomadiclab.com>
	<45CA08B2.4050603@cisco.com>
	<532B18E13CF9E64380EF5FDDE265E071B6192E@blr-m2-msg.wipro.com>
	<45CC7697.1060204@cisco.com>
Content-Type: text/plain
Date: Fri, 09 Feb 2007 16:38:50 +0200
Message-Id: <1171031930.3464.92.camel@localhost.localdomain>
Mime-Version: 1.0
X-Mailer: Evolution 2.2.3 (2.2.3-4.fc4) 
Content-Transfer-Encoding: 7bit
X-Virus-Scanned: ClamAV using ClamSMTP
X-OriginalArrivalTime: 09 Feb 2007 14:38:10.0868 (UTC)
	FILETIME=[EC00A740:01C74C57]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a1f9797ba297220533cb8c3f4bc709a8
Cc: mary.barnes@nortel.com, sipping@ietf.org, ravishankar.shiroor@wipro.com,
	Gonzalo.Camarillo@ericsson.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi,

In my opinion we should also consider the user expectation.
I fully agree with you about the mail service,
but if I have look at the Instant Messaging service where at present
there are different and incompatible programs, then maybe the necessity
of a identification for the specific IM service become a little bit more
clear.


I think that everyone here agree that there is the possibility to use a
media in different ways, and that we need a distinction about what the
capabilities are and how they are used, whatever will be the mechanism
that at the end we will identify to do that.
What we do need is a way to identify a preference.

then what I don't completely understand, surely it is my fault, that a
combination of capabilities and ways to use those capabilities it is not
itself a capability, perhaps something we can call a *macro-capability*
Even if I agree on the opportunity of don't create a real Tower of 
Babel and avoid . So definitely this macro-capability should be well
defined and standardized.


Of course we need also discuss the opportunity to use the collerpref
mechanism to address this capability.
I agree with Paul that "callerprefs doesn't support well, if at all, is
the inclusion of intermediate servers into the call path."
This of course is not a problem from 3GPP point of view, because in a
IMS network it is present the IFC mechanism that resolve this
"callerpref" problem. 
But I do think we need a general mechanism that works also in a pure SIP
environment.



/sal






On Fri, 2007-02-09 at 08:26 -0500, Paul Kyzivat wrote:
> 
> ravishankar.shiroor@wipro.com wrote:
> > Hi Paul,
> > 
> > IMO, service identification is different from identifying the tool that
> > should be used for executing that service on the end device. In the
> > analogy that you have given, e-mail is the service. Eudora/gmail/... Are
> > the tools that are used to provide the service.
> 
> The comment below was about identifying the application in the end 
> system, not the protocol.
> 
> In the case of email, the "service" is determined by the protocol and 
> the address/port of the destination. Nobody has suggested adding headers 
> to mail messages to identify the service/application in the end system 
> that should process the mail message. But here we see proposals to add a 
> header to a sip message to identify the service/application in the UAS.
> 
> 	Paul
> 
> > Regards,
> > Ravi.
> > 
> > 
> > --
> >  
> > Ravishankar. G. Shiroor
> > Wipro Technologies, Bangalore.
> >  
> > ravishankar.shiroor@wipro.com
> > --
> >  
> > 
> >> -----Original Message-----
> >> From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
> >> Sent: Wednesday, February 07, 2007 10:43 PM
> >> To: Salvatore Loreto (JO/LMF)
> >> Cc: RaviShankar Shiroor (WT01 - Product Engineering 
> >> Solutions); sipping@ietf.org; Gonzalo.Camarillo@ericsson.com; 
> >> mary.barnes@nortel.com
> >> Subject: Re: [Sipping] Service Identification in SIP
> >>
> >> I still don't get the intent of identifying the application. 
> >> It seems to me this is evil - akin to sending an email 
> >> message and identifying that the recipient must use Outlook 
> >> to receive it, not Thunderbird, Eudora, gmail, or whatever.
> >>
> >> I also wish somebody could give better examples of a 
> >> communication service. Chess is too artificial to be 
> >> believable, and IMO PoC is only a service because of the way 
> >> the implementation has been done. 
> >> Conceptually PoC could be nothing but a special floor control 
> >> protocol accompanying a voice session. In that case none of 
> >> the Service mechanism would be needed.
> >>
> >> 	Paul
> >>
> >> Salvatore Loreto (JO/LMF) wrote:
> >>> hi Ravi,
> >>>
> >>> in IMS a Service is a "type of communication defined by a service 
> >>> definition that specifies the rules and procedures and 
> >> allowed media, 
> >>> if any, for a specific type of communication and that 
> >> utilises the IMS 
> >>> enabler. An IMS communication service defines what is 
> >> possible within 
> >>> a single SIP session or standalone transaction."
> >>> So a Service implies some combination of capabilities and 
> >> ways to use 
> >>> those capabilities.
> >>> If you want the Service specifies the context for a communication.
> >>>
> >>> Instead the Application define the specific communication 
> >> (or if you 
> >>> want application) that you want run in the specific context. So you 
> >>> should have several different applications designed to run on a 
> >>> specific environment or that have different performance or 
> >> behavior in 
> >>> different context (running on top of different service).
> >>>
> >>> Lets go to Jonathan's example of the Chess.
> >>> In his example the Chess is a service, for IMS instead is an 
> >>> application that you can run on top of different service.
> >>> For example you can run Chess application on top of PoC 
> >> service or on 
> >>> top of the "standard" service. So let say you run Chess on 
> >> top of PoC 
> >>> service, then you can use only voice in the way that PoC specifies.
> >>>
> >>> /sal
> >>>
> >>>
> >>>
> >>>
> >>> On Thu, 2007-02-01 at 16:39 +0530, 
> >> ravishankar.shiroor@wipro.com wrote:
> >>>> Hello,
> >>>>
> >>>> IMO, Service Identification is definitely needed and 
> >> should be taken 
> >>>> up as a WG as proposed below.
> >>>>
> >>>> Further, I think Service Identification scope should be as per the 
> >>>> requirement in Jonathan's draft, as I still do not understand the 
> >>>> need for having a Communication Service Identifier and another 
> >>>> Application Identifier as given in the 
> >> 3gpp-ics-requirements document.
> >>>> Regards,
> >>>> Ravi.
> >>>>
> >>>> --
> >>>>  
> >>>> Ravishankar. G. Shiroor
> >>>> Wipro Technologies, Bangalore.
> >>>>  
> >>>> ravishankar.shiroor@wipro.com
> >>>> --
> >>>>  
> >>>>
> >>>>> -----Original Message-----
> >>>>> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
> >>>>> Sent: Wednesday, January 31, 2007 8:30 PM
> >>>>> To: sipping
> >>>>> Cc: Mary Barnes
> >>>>> Subject: [Sipping] Service Identification in SIP
> >>>>>
> >>>>> Folks,
> >>>>>
> >>>>> during the last year, there have been a number of threads on the 
> >>>>> mailing list discussing service identification in SIP. In 
> >>>>> particular, the following thread discussed 3GPP 
> >> requirements on this 
> >>>>> area:
> >>>>>
> >>>>>
> >> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
> >>>>> At a later point, those 3GPP requirements were documented in the 
> >>>>> following (expired by now) Internet Draft, which was 
> >> presented in a 
> >>>>> face-to-face IETF meeting.
> >>>>>
> >>>>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
> >>>>> cs-requirements-00.txt
> >>>>>
> >>>>> Lately, the following draft has been submitted to the archives:
> >>>>>
> >>>>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
> >>>>> rvice-identification-00.txt
> >>>>>
> >>>>> This draft proposes a solution that intends to meet known 
> >>>>> requirements in this area.
> >>>>>
> >>>>> Given the amount of discussions on issues related to service 
> >>>>> identification we have seen during the last year, we 
> >> believe this is 
> >>>>> a fairly important topic for the SIPPING WG to tackle. At some 
> >>>>> point, we would like to have a milestone with its 
> >> associated WG item 
> >>>>> dealing with this issue.
> >>>>>
> >>>>> Therefore, we would like to encourage people to read this 
> >> draft and 
> >>>>> send comments to the list. Additionally, if you believe that this 
> >>>>> work is (or is not) important, we would like to hear that as well.
> >>>>>
> >>>>> Cheers,
> >>>>>
> >>>>> Gonzalo
> >>>>> SIPPING co-chair
> >>>>>
> >>>>>
> >>>>> _______________________________________________
> >>>>> Sipping mailing list  
> >> https://www1.ietf.org/mailman/listinfo/sipping
> >>>>> This list is for NEW development of the application of SIP Use 
> >>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
> >>>>> sip@ietf.org for new developments of core SIP
> >>>>>
> >>>> _______________________________________________
> >>>> Sipping mailing list  
> >> https://www1.ietf.org/mailman/listinfo/sipping
> >>>> This list is for NEW development of the application of SIP Use 
> >>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
> >>>> sip@ietf.org for new developments of core SIP
> >>>
> >>> _______________________________________________
> >>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> >>> This list is for NEW development of the application of SIP Use 
> >>> sip-implementors@cs.columbia.edu for questions on current sip Use 
> >>> sip@ietf.org for new developments of core SIP
> >>>
> > 
> > 
> > The information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments. 
> > 
> > WARNING: Computer viruses can be transmitted via email. The recipient should check this email and any attachments for the presence of viruses. The company accepts no liability for any damage caused by any virus transmitted by this email.
> >  
> > www.wipro.com
> > 


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 09 10:04:49 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HFXII-0001Bc-TL; Fri, 09 Feb 2007 10:04:10 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HFXIH-0001BU-IS
	for sipping@ietf.org; Fri, 09 Feb 2007 10:04:09 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HFXI7-0006kw-4F
	for sipping@ietf.org; Fri, 09 Feb 2007 10:04:09 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	8182721273; Fri,  9 Feb 2007 16:03:52 +0100 (CET)
X-AuditID: c1b4fb3e-afed3bb0000007e1-52-45cc8d58f31c 
Received: from esealmw127.eemea.ericsson.se (unknown [153.88.254.122])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	611B020564; Fri,  9 Feb 2007 16:03:52 +0100 (CET)
Received: from esealmw129.eemea.ericsson.se ([153.88.254.177]) by
	esealmw127.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 9 Feb 2007 16:03:51 +0100
Received: from mail.lmf.ericsson.se ([131.160.11.50]) by
	esealmw129.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 9 Feb 2007 16:03:51 +0100
Received: from nomadiclab.lmf.ericsson.se (nomadiclab.lmf.ericsson.se
	[131.160.33.3])
	by mail.lmf.ericsson.se (Postfix) with ESMTP id 564B42358;
	Fri,  9 Feb 2007 17:03:51 +0200 (EET)
Received: from nomadiclab.lmf.ericsson.se (localhost [127.0.0.1])
	by nomadiclab.lmf.ericsson.se (Postfix) with ESMTP id 2AB5F4DBB4;
	Fri,  9 Feb 2007 17:03:51 +0200 (EET)
Received: from localhost.localdomain (localhost [IPv6:::1])
	by nomadiclab.lmf.ericsson.se (Postfix) with ESMTP id E25D74DA95;
	Fri,  9 Feb 2007 17:03:49 +0200 (EET)
Subject: RE: [Sipping] Service Identification in SIP
From: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>
To: ravishankar.shiroor@wipro.com
In-Reply-To: <532B18E13CF9E64380EF5FDDE265E071B61978@blr-m2-msg.wipro.com>
References: <45C0AEF2.8010101@ericsson.com>
	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
	<1170835642.29763.54.camel@n95.nomadiclab.com>
	<532B18E13CF9E64380EF5FDDE265E071B61978@blr-m2-msg.wipro.com>
Date: Fri, 09 Feb 2007 17:04:30 +0200
Message-Id: <1171033470.3464.115.camel@localhost.localdomain>
Mime-Version: 1.0
X-Mailer: Evolution 2.2.3 (2.2.3-4.fc4) 
X-Virus-Scanned: ClamAV using ClamSMTP
X-OriginalArrivalTime: 09 Feb 2007 15:03:51.0643 (UTC)
	FILETIME=[826042B0:01C74C5B]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 876202f9cbc0933cffbc58102e40f8f2
Cc: sipping@ietf.org, Gonzalo.Camarillo@ericsson.com, mary.barnes@nortel.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============0339608214=="
Errors-To: sipping-bounces@ietf.org


--===============0339608214==
Content-Type: multipart/alternative; boundary="=-B5j+JNFloK3SXnfaXgsI"


--=-B5j+JNFloK3SXnfaXgsI
Content-Type: text/plain
Content-Transfer-Encoding: 7bit

Hi Ravi,

if you consider the Jonathan draft, he says that 

        "...chess game that
        
        utilizes voice commands to move pieces.  When one player says "Queen
        to D7", the software on their phone recognizes this and moves the
        piece.  At the same time, the user's voice is sent to the other
        player, where it is both rendered to the user, and interpreted
        locally in order to move the piece."


and he suggests that the following urn

   urn:service:games:voice-chess 

could be used to identify the voice chess application.


I guess that with voice he supposes audio, but nowadays we can also have voice
using PoC. Then we need a way to specify which kind of voice we want: a standard
full-duplex or a PoC. 
Maybe the chess game works well with, or it is designed for a standard audio but 
not for a PoC audio.
So we need to address also the voice preference we want.

I like your proposal, having something hierarchical was something I was thinking too
I don't know if it is your idea, but with the service urn you are proposing:

Urn:service:voice.poc.chess

I imaging you are asking to play a Chess game, that work utilizing voice commands, 
but specifically voice over PoC.
This is also compliant to the draft-ietf-ecrit-service-urn (see section 3)


        "Declaration of syntactic structure: The URN consists of a
        hierarchical service identifier, with a sequence of labels
        separated by periods.  The left-most label is the most significant
        one and is called 'top-level service', while names to the right
        are called 'sub-services'.  The set of allowable characters is the
        same as that for domain names [1] and a subset of the labels
        allowed in [5].  Labels are case-insensitive and MUST be specified
        in all lower-case.  For any given service URN, service-identifiers
        can be removed right-to-left and the resulting URN is still valid,
        referring a more generic service.  In other words, if a service
        'x.y.z' exists, the URNs 'x' and 'x.y' are also valid service
        URNs."


Regards
sal
 


On Fri, 2007-02-09 at 13:44 +0530, ravishankar.shiroor@wipro.com wrote:

> Hi Salvatore,
> 
> As an end user, poc is also an application that I use and chess is also
> an application that I can use. The way I see it, in your example, it
> will be running chess application on top of PoC application.
> 
> So the service urn for this would be something like:
> 
> Urn:service:voice.poc.chess
> 
> Regards,
> Ravi.
> 
> --
>  
> Ravishankar. G. Shiroor
> Wipro Technologies, Bangalore.
>  
> ravishankar.shiroor@wipro.com
> --
>  
> 
> > -----Original Message-----
> > From: Salvatore Loreto (JO/LMF) 
> > [mailto:salvatore.loreto@ericsson.com] 
> > Sent: Wednesday, February 07, 2007 1:37 PM
> > To: RaviShankar Shiroor (WT01 - Product Engineering Solutions)
> > Cc: Gonzalo.Camarillo@ericsson.com; sipping@ietf.org; 
> > mary.barnes@nortel.com
> > Subject: RE: [Sipping] Service Identification in SIP
> > 
> > hi Ravi,
> > 
> > in IMS a Service is a "type of communication defined by a 
> > service definition that specifies the rules and procedures 
> > and allowed media, if any, for a specific type of 
> > communication and that utilises the IMS enabler. An IMS 
> > communication service defines what is possible within a 
> > single SIP session or standalone transaction."
> > So a Service implies some combination of capabilities and 
> > ways to use those capabilities.
> > If you want the Service specifies the context for a communication.
> > 
> > Instead the Application define the specific communication (or 
> > if you want application) that you want run in the specific 
> > context. So you should have several different applications 
> > designed to run on a specific environment or that have 
> > different performance or behavior in different context 
> > (running on top of different service).
> > 
> > Lets go to Jonathan's example of the Chess.
> > In his example the Chess is a service, for IMS instead is an 
> > application that you can run on top of different service.
> > For example you can run Chess application on top of PoC 
> > service or on top of the "standard" service. So let say you 
> > run Chess on top of PoC service, then you can use only voice 
> > in the way that PoC specifies.
> > 
> > /sal
> > 
> > 
> > 
> > 
> > On Thu, 2007-02-01 at 16:39 +0530, 
> > ravishankar.shiroor@wipro.com wrote:
> > > Hello,
> > > 
> > > IMO, Service Identification is definitely needed and should 
> > be taken 
> > > up as a WG as proposed below.
> > > 
> > > Further, I think Service Identification scope should be as per the 
> > > requirement in Jonathan's draft, as I still do not 
> > understand the need 
> > > for having a Communication Service Identifier and another 
> > Application 
> > > Identifier as given in the 3gpp-ics-requirements document.
> > > 
> > > Regards,
> > > Ravi.
> > > 
> > > --
> > >  
> > > Ravishankar. G. Shiroor
> > > Wipro Technologies, Bangalore.
> > >  
> > > ravishankar.shiroor@wipro.com
> > > --
> > >  
> > > 
> > > > -----Original Message-----
> > > > From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
> > > > Sent: Wednesday, January 31, 2007 8:30 PM
> > > > To: sipping
> > > > Cc: Mary Barnes
> > > > Subject: [Sipping] Service Identification in SIP
> > > > 
> > > > Folks,
> > > > 
> > > > during the last year, there have been a number of threads on the 
> > > > mailing list discussing service identification in SIP. In 
> > > > particular, the following thread discussed 3GPP 
> > requirements on this 
> > > > area:
> > > > 
> > > > 
> > http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
> > > > 
> > > > At a later point, those 3GPP requirements were documented in the 
> > > > following (expired by now) Internet Draft, which was 
> > presented in a 
> > > > face-to-face IETF meeting.
> > > > 
> > > > http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
> > > > cs-requirements-00.txt
> > > > 
> > > > Lately, the following draft has been submitted to the archives:
> > > > 
> > > > http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
> > > > rvice-identification-00.txt
> > > > 
> > > > This draft proposes a solution that intends to meet known 
> > > > requirements in this area.
> > > > 
> > > > Given the amount of discussions on issues related to service 
> > > > identification we have seen during the last year, we 
> > believe this is 
> > > > a fairly important topic for the SIPPING WG to tackle. At some 
> > > > point, we would like to have a milestone with its 
> > associated WG item 
> > > > dealing with this issue.
> > > > 
> > > > Therefore, we would like to encourage people to read this 
> > draft and 
> > > > send comments to the list. Additionally, if you believe that this 
> > > > work is (or is not) important, we would like to hear that as well.
> > > > 
> > > > Cheers,
> > > > 
> > > > Gonzalo
> > > > SIPPING co-chair
> > > > 
> > > > 
> > > > _______________________________________________
> > > > Sipping mailing list  
> > https://www1.ietf.org/mailman/listinfo/sipping
> > > > This list is for NEW development of the application of SIP Use 
> > > > sip-implementors@cs.columbia.edu for questions on current sip Use 
> > > > sip@ietf.org for new developments of core SIP
> > > > 
> > > 
> > > _______________________________________________
> > > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > > This list is for NEW development of the application of SIP Use 
> > > sip-implementors@cs.columbia.edu for questions on current sip Use 
> > > sip@ietf.org for new developments of core SIP
> > 
> > 
> 
> 
> The information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments. 
> 
> WARNING: Computer viruses can be transmitted via email. The recipient should check this email and any attachments for the presence of viruses. The company accepts no liability for any damage caused by any virus transmitted by this email.
>  
> www.wipro.com

--=-B5j+JNFloK3SXnfaXgsI
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: 7bit

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 TRANSITIONAL//EN">
<HTML>
<HEAD>
  <META HTTP-EQUIV="Content-Type" CONTENT="text/html; CHARSET=UTF-8">
  <META NAME="GENERATOR" CONTENT="GtkHTML/3.6.2">
</HEAD>
<BODY>
Hi Ravi,<BR>
<BR>
if you consider the Jonathan draft, he says that <BR>
<BLOCKQUOTE>
    &quot;<I>...chess game that</I>
<PRE>
<I>utilizes voice commands to move pieces.&nbsp; When one player says &quot;Queen</I>
<I>to D7&quot;, the software on their phone recognizes this and moves the</I>
<I>piece.&nbsp; At the same time, the user's voice is sent to the other</I>
<I>player, where it is both rendered to the user, and interpreted</I>
<I>locally in order to move the piece.</I>&quot;
</PRE>
</BLOCKQUOTE>
<BR>
and he suggests that the following urn<BR>
<BR>
&nbsp;&nbsp; urn:service:games:voice-chess <BR>
<BR>
could be used to identify the voice chess application.<BR>
<BR>
<PRE>
I guess that with voice he supposes audio, but nowadays we can also have voice
using PoC. Then we need a way to specify which kind of voice we want: a standard
full-duplex or a PoC. 
Maybe the chess game works well with, or it is designed for a standard audio but 
not for a PoC audio.
So we need to address also the voice preference we want.

I like your proposal, having something hierarchical was something I was thinking too
I don't know if it is your idea, but with the service urn you are proposing:

<FONT COLOR="#000000">Urn:service:voice.poc.chess</FONT>

I imaging you are asking to play a Chess game, that work utilizing voice commands, 
but specifically voice over PoC.
This is also compliant to the draft-ietf-ecrit-service-urn (see section 3)

</PRE>
<BLOCKQUOTE>
<PRE>
&quot;<I>Declaration of syntactic structure: The URN consists of a</I>
<I>hierarchical service identifier, with a sequence of labels</I>
<I>separated by periods.&nbsp; The left-most label is the most significant</I>
<I>one and is called 'top-level service', while names to the right</I>
<I>are called 'sub-services'.&nbsp; The set of allowable characters is the</I>
<I>same as that for domain names [1] and a subset of the labels</I>
<I>allowed in [5].&nbsp; Labels are case-insensitive and MUST be specified</I>
<I>in all lower-case.&nbsp; For any given service URN, service-identifiers</I>
<I>can be removed right-to-left and the resulting URN is still valid,</I>
<I>referring a more generic service.&nbsp; In other words, if a service</I>
<I>'x.y.z' exists, the URNs 'x' and 'x.y' are also valid service</I>
<I>URNs.</I>&quot;
</PRE>
</BLOCKQUOTE>
<PRE>

Regards
sal
 
</PRE>
<BR>
On Fri, 2007-02-09 at 13:44 +0530, ravishankar.shiroor@wipro.com wrote:
<BLOCKQUOTE TYPE=CITE>
<PRE>
<FONT COLOR="#000000">Hi Salvatore,</FONT>

<FONT COLOR="#000000">As an end user, poc is also an application that I use and chess is also</FONT>
<FONT COLOR="#000000">an application that I can use. The way I see it, in your example, it</FONT>
<FONT COLOR="#000000">will be running chess application on top of PoC application.</FONT>

<FONT COLOR="#000000">So the service urn for this would be something like:</FONT>

<FONT COLOR="#000000">Urn:service:voice.poc.chess</FONT>

<FONT COLOR="#000000">Regards,</FONT>
<FONT COLOR="#000000">Ravi.</FONT>

<FONT COLOR="#000000">--</FONT>
<FONT COLOR="#000000"> </FONT>
<FONT COLOR="#000000">Ravishankar. G. Shiroor</FONT>
<FONT COLOR="#000000">Wipro Technologies, Bangalore.</FONT>
<FONT COLOR="#000000"> </FONT>
<FONT COLOR="#000000"><A HREF="mailto:ravishankar.shiroor@wipro.com">ravishankar.shiroor@wipro.com</A></FONT>
<FONT COLOR="#000000">--</FONT>
<FONT COLOR="#000000"> </FONT>

<FONT COLOR="#000000">&gt; -----Original Message-----</FONT>
<FONT COLOR="#000000">&gt; From: Salvatore Loreto (JO/LMF) </FONT>
<FONT COLOR="#000000">&gt; [mailto:<A HREF="mailto:salvatore.loreto@ericsson.com">salvatore.loreto@ericsson.com</A>] </FONT>
<FONT COLOR="#000000">&gt; Sent: Wednesday, February 07, 2007 1:37 PM</FONT>
<FONT COLOR="#000000">&gt; To: RaviShankar Shiroor (WT01 - Product Engineering Solutions)</FONT>
<FONT COLOR="#000000">&gt; Cc: <A HREF="mailto:Gonzalo.Camarillo@ericsson.com">Gonzalo.Camarillo@ericsson.com</A>; <A HREF="mailto:sipping@ietf.org">sipping@ietf.org</A>; </FONT>
<FONT COLOR="#000000">&gt; <A HREF="mailto:mary.barnes@nortel.com">mary.barnes@nortel.com</A></FONT>
<FONT COLOR="#000000">&gt; Subject: RE: [Sipping] Service Identification in SIP</FONT>
<FONT COLOR="#000000">&gt; </FONT>
<FONT COLOR="#000000">&gt; hi Ravi,</FONT>
<FONT COLOR="#000000">&gt; </FONT>
<FONT COLOR="#000000">&gt; in IMS a Service is a &quot;type of communication defined by a </FONT>
<FONT COLOR="#000000">&gt; service definition that specifies the rules and procedures </FONT>
<FONT COLOR="#000000">&gt; and allowed media, if any, for a specific type of </FONT>
<FONT COLOR="#000000">&gt; communication and that utilises the IMS enabler. An IMS </FONT>
<FONT COLOR="#000000">&gt; communication service defines what is possible within a </FONT>
<FONT COLOR="#000000">&gt; single SIP session or standalone transaction.&quot;</FONT>
<FONT COLOR="#000000">&gt; So a Service implies some combination of capabilities and </FONT>
<FONT COLOR="#000000">&gt; ways to use those capabilities.</FONT>
<FONT COLOR="#000000">&gt; If you want the Service specifies the context for a communication.</FONT>
<FONT COLOR="#000000">&gt; </FONT>
<FONT COLOR="#000000">&gt; Instead the Application define the specific communication (or </FONT>
<FONT COLOR="#000000">&gt; if you want application) that you want run in the specific </FONT>
<FONT COLOR="#000000">&gt; context. So you should have several different applications </FONT>
<FONT COLOR="#000000">&gt; designed to run on a specific environment or that have </FONT>
<FONT COLOR="#000000">&gt; different performance or behavior in different context </FONT>
<FONT COLOR="#000000">&gt; (running on top of different service).</FONT>
<FONT COLOR="#000000">&gt; </FONT>
<FONT COLOR="#000000">&gt; Lets go to Jonathan's example of the Chess.</FONT>
<FONT COLOR="#000000">&gt; In his example the Chess is a service, for IMS instead is an </FONT>
<FONT COLOR="#000000">&gt; application that you can run on top of different service.</FONT>
<FONT COLOR="#000000">&gt; For example you can run Chess application on top of PoC </FONT>
<FONT COLOR="#000000">&gt; service or on top of the &quot;standard&quot; service. So let say you </FONT>
<FONT COLOR="#000000">&gt; run Chess on top of PoC service, then you can use only voice </FONT>
<FONT COLOR="#000000">&gt; in the way that PoC specifies.</FONT>
<FONT COLOR="#000000">&gt; </FONT>
<FONT COLOR="#000000">&gt; /sal</FONT>
<FONT COLOR="#000000">&gt; </FONT>
<FONT COLOR="#000000">&gt; </FONT>
<FONT COLOR="#000000">&gt; </FONT>
<FONT COLOR="#000000">&gt; </FONT>
<FONT COLOR="#000000">&gt; On Thu, 2007-02-01 at 16:39 +0530, </FONT>
<FONT COLOR="#000000">&gt; <A HREF="mailto:ravishankar.shiroor@wipro.com">ravishankar.shiroor@wipro.com</A> wrote:</FONT>
<FONT COLOR="#000000">&gt; &gt; Hello,</FONT>
<FONT COLOR="#000000">&gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; IMO, Service Identification is definitely needed and should </FONT>
<FONT COLOR="#000000">&gt; be taken </FONT>
<FONT COLOR="#000000">&gt; &gt; up as a WG as proposed below.</FONT>
<FONT COLOR="#000000">&gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; Further, I think Service Identification scope should be as per the </FONT>
<FONT COLOR="#000000">&gt; &gt; requirement in Jonathan's draft, as I still do not </FONT>
<FONT COLOR="#000000">&gt; understand the need </FONT>
<FONT COLOR="#000000">&gt; &gt; for having a Communication Service Identifier and another </FONT>
<FONT COLOR="#000000">&gt; Application </FONT>
<FONT COLOR="#000000">&gt; &gt; Identifier as given in the 3gpp-ics-requirements document.</FONT>
<FONT COLOR="#000000">&gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; Regards,</FONT>
<FONT COLOR="#000000">&gt; &gt; Ravi.</FONT>
<FONT COLOR="#000000">&gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; --</FONT>
<FONT COLOR="#000000">&gt; &gt;  </FONT>
<FONT COLOR="#000000">&gt; &gt; Ravishankar. G. Shiroor</FONT>
<FONT COLOR="#000000">&gt; &gt; Wipro Technologies, Bangalore.</FONT>
<FONT COLOR="#000000">&gt; &gt;  </FONT>
<FONT COLOR="#000000">&gt; &gt; <A HREF="mailto:ravishankar.shiroor@wipro.com">ravishankar.shiroor@wipro.com</A></FONT>
<FONT COLOR="#000000">&gt; &gt; --</FONT>
<FONT COLOR="#000000">&gt; &gt;  </FONT>
<FONT COLOR="#000000">&gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; -----Original Message-----</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; From: Gonzalo Camarillo [mailto:<A HREF="mailto:Gonzalo.Camarillo@ericsson.com">Gonzalo.Camarillo@ericsson.com</A>]</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; Sent: Wednesday, January 31, 2007 8:30 PM</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; To: sipping</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; Cc: Mary Barnes</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; Subject: [Sipping] Service Identification in SIP</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; Folks,</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; during the last year, there have been a number of threads on the </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; mailing list discussing service identification in SIP. In </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; particular, the following thread discussed 3GPP </FONT>
<FONT COLOR="#000000">&gt; requirements on this </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; area:</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; <A HREF="http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html">http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html</A></FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; At a later point, those 3GPP requirements were documented in the </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; following (expired by now) Internet Draft, which was </FONT>
<FONT COLOR="#000000">&gt; presented in a </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; face-to-face IETF meeting.</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; <A HREF="http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i">http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i</A></FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; cs-requirements-00.txt</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; Lately, the following draft has been submitted to the archives:</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; <A HREF="http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se">http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se</A></FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; rvice-identification-00.txt</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; This draft proposes a solution that intends to meet known </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; requirements in this area.</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; Given the amount of discussions on issues related to service </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; identification we have seen during the last year, we </FONT>
<FONT COLOR="#000000">&gt; believe this is </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; a fairly important topic for the SIPPING WG to tackle. At some </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; point, we would like to have a milestone with its </FONT>
<FONT COLOR="#000000">&gt; associated WG item </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; dealing with this issue.</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; Therefore, we would like to encourage people to read this </FONT>
<FONT COLOR="#000000">&gt; draft and </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; send comments to the list. Additionally, if you believe that this </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; work is (or is not) important, we would like to hear that as well.</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; Cheers,</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; Gonzalo</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; SIPPING co-chair</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; _______________________________________________</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; Sipping mailing list  </FONT>
<FONT COLOR="#000000">&gt; <A HREF="https://www1.ietf.org/mailman/listinfo/sipping">https://www1.ietf.org/mailman/listinfo/sipping</A></FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; This list is for NEW development of the application of SIP Use </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; <A HREF="mailto:sip-implementors@cs.columbia.edu">sip-implementors@cs.columbia.edu</A> for questions on current sip Use </FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; <A HREF="mailto:sip@ietf.org">sip@ietf.org</A> for new developments of core SIP</FONT>
<FONT COLOR="#000000">&gt; &gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; </FONT>
<FONT COLOR="#000000">&gt; &gt; _______________________________________________</FONT>
<FONT COLOR="#000000">&gt; &gt; Sipping mailing list  <A HREF="https://www1.ietf.org/mailman/listinfo/sipping">https://www1.ietf.org/mailman/listinfo/sipping</A></FONT>
<FONT COLOR="#000000">&gt; &gt; This list is for NEW development of the application of SIP Use </FONT>
<FONT COLOR="#000000">&gt; &gt; <A HREF="mailto:sip-implementors@cs.columbia.edu">sip-implementors@cs.columbia.edu</A> for questions on current sip Use </FONT>
<FONT COLOR="#000000">&gt; &gt; <A HREF="mailto:sip@ietf.org">sip@ietf.org</A> for new developments of core SIP</FONT>
<FONT COLOR="#000000">&gt; </FONT>
<FONT COLOR="#000000">&gt; </FONT>


<FONT COLOR="#000000">The information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments. </FONT>

<FONT COLOR="#000000">WARNING: Computer viruses can be transmitted via email. The recipient should check this email and any attachments for the presence of viruses. The company accepts no liability for any damage caused by any virus transmitted by this email.</FONT>
<FONT COLOR="#000000"> </FONT>
<FONT COLOR="#000000"><A HREF="http://www.wipro.com">www.wipro.com</A></FONT>
</PRE>
</BLOCKQUOTE>
</BODY>
</HTML>

--=-B5j+JNFloK3SXnfaXgsI--



--===============0339608214==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============0339608214==--





From sipping-bounces@ietf.org Fri Feb 09 12:14:36 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HFZIm-0002Z3-JN; Fri, 09 Feb 2007 12:12:48 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HFZIj-0002Yy-Kc
	for sipping@ietf.org; Fri, 09 Feb 2007 12:12:45 -0500
Received: from rtp-iport-1.cisco.com ([64.102.122.148])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HFZIb-0005JW-UD
	for sipping@ietf.org; Fri, 09 Feb 2007 12:12:45 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by rtp-iport-1.cisco.com with ESMTP; 09 Feb 2007 09:12:38 -0800
X-IronPort-AV: i="4.13,307,1167638400"; 
	d="scan'208"; a="52616786:sNHT69614528"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l19HCb0g011590; 
	Fri, 9 Feb 2007 12:12:37 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l19HCbVT013704; 
	Fri, 9 Feb 2007 12:12:37 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 9 Feb 2007 12:12:37 -0500
Received: from [161.44.183.228] ([161.44.183.228]) by
	xfe-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 9 Feb 2007 12:12:37 -0500
Message-ID: <45CCAB84.4040404@cisco.com>
Date: Fri, 09 Feb 2007 12:12:36 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>
Subject: Re: [Sipping] Service Identification in SIP
References: <45C0AEF2.8010101@ericsson.com>	
	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>	
	<1170835642.29763.54.camel@n95.nomadiclab.com>
	<45CA08B2.4050603@cisco.com>	
	<532B18E13CF9E64380EF5FDDE265E071B6192E@blr-m2-msg.wipro.com>	
	<45CC7697.1060204@cisco.com>
	<1171031930.3464.92.camel@localhost.localdomain>
In-Reply-To: <1171031930.3464.92.camel@localhost.localdomain>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 09 Feb 2007 17:12:37.0537 (UTC)
	FILETIME=[7F5E2110:01C74C6D]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=13887; t=1171041157;
	x=1171905157; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20
	|To:=20=22Salvatore=20Loreto=20(JO/LMF)=22=20<salvatore.loreto@ericsson.c
	om>; bh=ATLdtVILwYOyhlGJbineKxIp6MTjC1ZGVjLv5BIjA1Q=;
	b=e9VQAeeiCom4BbJnWKrGrn+cnBMZrORECyrGNYCpgoYwPXgUQl1JjdPoEqxNkeGXXzmiNr39
	nAofSekcjr4DHwdGSzWpfelRJb16eDYiixw8P1VxSBMKc7xeAiR5xB5S;
Authentication-Results: rtp-dkim-2; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: bd8a74b81c71f965ca7918b90d1c49c0
Cc: mary.barnes@nortel.com, sipping@ietf.org, ravishankar.shiroor@wipro.com,
	Gonzalo.Camarillo@ericsson.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Salvatore Loreto (JO/LMF) wrote:
> Hi,
> 
> In my opinion we should also consider the user expectation.
> I fully agree with you about the mail service,
> but if I have look at the Instant Messaging service where at present
> there are different and incompatible programs, then maybe the necessity
> of a identification for the specific IM service become a little bit more
> clear.

I don't understand you here. Once upon a time there were different and 
incompatible mail programs. But the introduction of SMTP forced them to 
become compatible. Similarly, it seems to me that SIMPLE should result 
in the various IM clients being a user preference rather than a 
prerequisite to communication.

> I think that everyone here agree that there is the possibility to use a
> media in different ways, and that we need a distinction about what the
> capabilities are and how they are used, whatever will be the mechanism
> that at the end we will identify to do that.

When stated in such a general way I guess almost everyone will agree. 
That is why I would like more examples, because I have the feeling there 
are a number of cases where some think there is a distinction that must 
be made, and others who think that making such a distinction is bad and 
will result in balkanization. Personally I am yet to be convinced that 
there are practical cases where the distinction cannot be made without a 
special mechanism.

> What we do need is a way to identify a preference.

I'm especially concerned with about a preference on one end implying 
that the same preference should be applied on the other end.

> then what I don't completely understand, surely it is my fault, that a
> combination of capabilities and ways to use those capabilities it is not
> itself a capability, perhaps something we can call a *macro-capability*
> Even if I agree on the opportunity of don't create a real Tower of 
> Babel and avoid . So definitely this macro-capability should be well
> defined and standardized.

This is one place where I think we really disagree. And this goes back 
several years to a similar discussion about RPIDS and the representation 
of "services" there.

The problem is that you may find some combination of capabilities to be 
interesting, and want to embody that combination in an application. 
Then, when you call me, I guess you want to simply name that 
combination, and expect me to understand it and use my application that 
supports the same combination.

But I may not know about that particular combination and have an 
application that supports it. But I may have an application that 
supports a different set of capabilities that is a superset of what you 
have requested. Or I may not have anything that supports the complete 
combination that you want, but be able to support a useful subset of 
those that should interoperate with you, if only in a degraded manner.

The compatibility between these things can be more easily achieved if we 
negotiate by enumerating the basic capabilities, rather than by using 
names for collections.

I think it would be fine to name macro-capabilities as a shorthand for 
communicating between a particular user and his UA, or perhaps even 
between a UA and its own service provider. But I think the two ends of 
the call should be able to decide for themselves that macro-capabilities 
they support, with interoperation being decided by the compatibility of 
the supported basic capabilities.

For instance, your service provider may provide basic telephony (audio) 
and multimedia (audio/video) services. Mine may have only an 
advanced-telephony (audio/video/im) service. You should be able to call 
me using either of your services even though I know nothing about them.

> Of course we need also discuss the opportunity to use the collerpref
> mechanism to address this capability.
> I agree with Paul that "callerprefs doesn't support well, if at all, is
> the inclusion of intermediate servers into the call path."
> This of course is not a problem from 3GPP point of view, because in a
> IMS network it is present the IFC mechanism that resolve this
> "callerpref" problem. 

I have been concerned about the proposal to use callerprefs in IMS to 
provide service identification, and my concern has a lot to do with the 
idea that the IFC mechanism, applied to callerprefs, can solve the 
problem of including appropriate intermediate servers into the call 
path. This is another case where I would very much like to see an 
example of what is intended.

I know there can be a simple filter criterion that can test for the 
presence of the name of a particular feature tag in an Accept-Contact 
header, and then cause the request to be routed to a server that handles 
the "service" denoted by that feature tag.

But that assumes a simplistic use of callerprefs. If the callerprefs are 
structured to specify more complex criteria, such as establishing a 
priority rating among feaures, then a particular feature tag may be 
present and yet not imply that the corresponding server is needed. (E.g. 
I might prefer full duplex audio/video, full duplex audio less so, and 
PoC as a last resort. If an IMS system unconditionally routed that call 
to a PoC server it wouldn't be adhering to the intent of the caller.)

> But I do think we need a general mechanism that works also in a pure SIP
> environment.

We agree here. Again I think it would be very useful to have a bunch of 
use cases to consider. Then we could see how they might interoperate 
with different assumptions about the environment of the caller and callee.

	Paul

> /sal
> 
> 
> 
> 
> 
> 
> On Fri, 2007-02-09 at 08:26 -0500, Paul Kyzivat wrote:
>> ravishankar.shiroor@wipro.com wrote:
>>> Hi Paul,
>>>
>>> IMO, service identification is different from identifying the tool that
>>> should be used for executing that service on the end device. In the
>>> analogy that you have given, e-mail is the service. Eudora/gmail/... Are
>>> the tools that are used to provide the service.
>> The comment below was about identifying the application in the end 
>> system, not the protocol.
>>
>> In the case of email, the "service" is determined by the protocol and 
>> the address/port of the destination. Nobody has suggested adding headers 
>> to mail messages to identify the service/application in the end system 
>> that should process the mail message. But here we see proposals to add a 
>> header to a sip message to identify the service/application in the UAS.
>>
>> 	Paul
>>
>>> Regards,
>>> Ravi.
>>>
>>>
>>> --
>>>  
>>> Ravishankar. G. Shiroor
>>> Wipro Technologies, Bangalore.
>>>  
>>> ravishankar.shiroor@wipro.com
>>> --
>>>  
>>>
>>>> -----Original Message-----
>>>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
>>>> Sent: Wednesday, February 07, 2007 10:43 PM
>>>> To: Salvatore Loreto (JO/LMF)
>>>> Cc: RaviShankar Shiroor (WT01 - Product Engineering 
>>>> Solutions); sipping@ietf.org; Gonzalo.Camarillo@ericsson.com; 
>>>> mary.barnes@nortel.com
>>>> Subject: Re: [Sipping] Service Identification in SIP
>>>>
>>>> I still don't get the intent of identifying the application. 
>>>> It seems to me this is evil - akin to sending an email 
>>>> message and identifying that the recipient must use Outlook 
>>>> to receive it, not Thunderbird, Eudora, gmail, or whatever.
>>>>
>>>> I also wish somebody could give better examples of a 
>>>> communication service. Chess is too artificial to be 
>>>> believable, and IMO PoC is only a service because of the way 
>>>> the implementation has been done. 
>>>> Conceptually PoC could be nothing but a special floor control 
>>>> protocol accompanying a voice session. In that case none of 
>>>> the Service mechanism would be needed.
>>>>
>>>> 	Paul
>>>>
>>>> Salvatore Loreto (JO/LMF) wrote:
>>>>> hi Ravi,
>>>>>
>>>>> in IMS a Service is a "type of communication defined by a service 
>>>>> definition that specifies the rules and procedures and 
>>>> allowed media, 
>>>>> if any, for a specific type of communication and that 
>>>> utilises the IMS 
>>>>> enabler. An IMS communication service defines what is 
>>>> possible within 
>>>>> a single SIP session or standalone transaction."
>>>>> So a Service implies some combination of capabilities and 
>>>> ways to use 
>>>>> those capabilities.
>>>>> If you want the Service specifies the context for a communication.
>>>>>
>>>>> Instead the Application define the specific communication 
>>>> (or if you 
>>>>> want application) that you want run in the specific context. So you 
>>>>> should have several different applications designed to run on a 
>>>>> specific environment or that have different performance or 
>>>> behavior in 
>>>>> different context (running on top of different service).
>>>>>
>>>>> Lets go to Jonathan's example of the Chess.
>>>>> In his example the Chess is a service, for IMS instead is an 
>>>>> application that you can run on top of different service.
>>>>> For example you can run Chess application on top of PoC 
>>>> service or on 
>>>>> top of the "standard" service. So let say you run Chess on 
>>>> top of PoC 
>>>>> service, then you can use only voice in the way that PoC specifies.
>>>>>
>>>>> /sal
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> On Thu, 2007-02-01 at 16:39 +0530, 
>>>> ravishankar.shiroor@wipro.com wrote:
>>>>>> Hello,
>>>>>>
>>>>>> IMO, Service Identification is definitely needed and 
>>>> should be taken 
>>>>>> up as a WG as proposed below.
>>>>>>
>>>>>> Further, I think Service Identification scope should be as per the 
>>>>>> requirement in Jonathan's draft, as I still do not understand the 
>>>>>> need for having a Communication Service Identifier and another 
>>>>>> Application Identifier as given in the 
>>>> 3gpp-ics-requirements document.
>>>>>> Regards,
>>>>>> Ravi.
>>>>>>
>>>>>> --
>>>>>>  
>>>>>> Ravishankar. G. Shiroor
>>>>>> Wipro Technologies, Bangalore.
>>>>>>  
>>>>>> ravishankar.shiroor@wipro.com
>>>>>> --
>>>>>>  
>>>>>>
>>>>>>> -----Original Message-----
>>>>>>> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
>>>>>>> Sent: Wednesday, January 31, 2007 8:30 PM
>>>>>>> To: sipping
>>>>>>> Cc: Mary Barnes
>>>>>>> Subject: [Sipping] Service Identification in SIP
>>>>>>>
>>>>>>> Folks,
>>>>>>>
>>>>>>> during the last year, there have been a number of threads on the 
>>>>>>> mailing list discussing service identification in SIP. In 
>>>>>>> particular, the following thread discussed 3GPP 
>>>> requirements on this 
>>>>>>> area:
>>>>>>>
>>>>>>>
>>>> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
>>>>>>> At a later point, those 3GPP requirements were documented in the 
>>>>>>> following (expired by now) Internet Draft, which was 
>>>> presented in a 
>>>>>>> face-to-face IETF meeting.
>>>>>>>
>>>>>>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
>>>>>>> cs-requirements-00.txt
>>>>>>>
>>>>>>> Lately, the following draft has been submitted to the archives:
>>>>>>>
>>>>>>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
>>>>>>> rvice-identification-00.txt
>>>>>>>
>>>>>>> This draft proposes a solution that intends to meet known 
>>>>>>> requirements in this area.
>>>>>>>
>>>>>>> Given the amount of discussions on issues related to service 
>>>>>>> identification we have seen during the last year, we 
>>>> believe this is 
>>>>>>> a fairly important topic for the SIPPING WG to tackle. At some 
>>>>>>> point, we would like to have a milestone with its 
>>>> associated WG item 
>>>>>>> dealing with this issue.
>>>>>>>
>>>>>>> Therefore, we would like to encourage people to read this 
>>>> draft and 
>>>>>>> send comments to the list. Additionally, if you believe that this 
>>>>>>> work is (or is not) important, we would like to hear that as well.
>>>>>>>
>>>>>>> Cheers,
>>>>>>>
>>>>>>> Gonzalo
>>>>>>> SIPPING co-chair
>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> Sipping mailing list  
>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>>>> This list is for NEW development of the application of SIP Use 
>>>>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>>>>> sip@ietf.org for new developments of core SIP
>>>>>>>
>>>>>> _______________________________________________
>>>>>> Sipping mailing list  
>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>>> This list is for NEW development of the application of SIP Use 
>>>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>>>> sip@ietf.org for new developments of core SIP
>>>>> _______________________________________________
>>>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>>>> This list is for NEW development of the application of SIP Use 
>>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>>> sip@ietf.org for new developments of core SIP
>>>>>
>>>
>>> The information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments. 
>>>
>>> WARNING: Computer viruses can be transmitted via email. The recipient should check this email and any attachments for the presence of viruses. The company accepts no liability for any damage caused by any virus transmitted by this email.
>>>  
>>> www.wipro.com
>>>
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sat Feb 10 15:33:37 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HFysI-0005lM-2p; Sat, 10 Feb 2007 15:31:10 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HFysG-0005lC-Ob
	for sipping@ietf.org; Sat, 10 Feb 2007 15:31:08 -0500
Received: from jalapeno.cc.columbia.edu ([128.59.29.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HFysD-0007xS-8k
	for sipping@ietf.org; Sat, 10 Feb 2007 15:31:08 -0500
Received: from [192.168.0.41] (pool-141-153-190-218.mad.east.verizon.net
	[141.153.190.218]) (user=hgs10 mech=PLAIN bits=0)
	by jalapeno.cc.columbia.edu (8.13.7/8.13.6) with ESMTP id
	l1AKUtg3011007
	(version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NOT);
	Sat, 10 Feb 2007 15:30:56 -0500 (EST)
In-Reply-To: <1171033470.3464.115.camel@localhost.localdomain>
References: <45C0AEF2.8010101@ericsson.com>
	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
	<1170835642.29763.54.camel@n95.nomadiclab.com>
	<532B18E13CF9E64380EF5FDDE265E071B61978@blr-m2-msg.wipro.com>
	<1171033470.3464.115.camel@localhost.localdomain>
Mime-Version: 1.0 (Apple Message framework v752.2)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <FE059A45-A940-4CAE-9B9C-19EFE99F1D66@cs.columbia.edu>
Content-Transfer-Encoding: 7bit
From: Henning Schulzrinne <hgs@cs.columbia.edu>
Subject: Re: [Sipping] Service Identification in SIP
Date: Sat, 10 Feb 2007 15:30:57 -0500
To: Salvatore Loreto (JO/LMF) <salvatore.loreto@ericsson.com>
X-Mailer: Apple Mail (2.752.2)
X-No-Spam-Score: Local
X-Scanned-By: MIMEDefang 2.48 on 128.59.29.5
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 827a2a57ca7ab0837847220f447e8d56
Cc: sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I have to admit that I find a discussion that uses a single example a  
bit peculiar, particularly since the example is rather contrived.  
(Why would you use voice, plus speech-to-text presumably, to talk to  
a chess application?)

In a networking sense, the issue seems to be that we're using voice  
as a transport protocol to carry different distinct applications. (In  
the old days, a modem would have been such an application.)

I'll repeat what I've said before: SIP doesn't exactly suffer from a  
lack of complexity and mechanisms. At this point, before adding more  
machinery, we might want to ask as to whether there is truly a broad- 
based need for justifying additional machinery.

With regards to service URNs, I happen to believe that they are very  
reasonable as "modern" replacements for tel URIs, essentially  
extending the concept of 800 numbers. But that's a different problem  
from what I understand this discussion to be about. (In that usage,  
service URNs would be mapped to SIP AORs based on geography, time-of- 
day and other factors.)

Henning


On Feb 9, 2007, at 10:04 AM, Salvatore Loreto (JO/LMF) wrote:

> Hi Ravi,
>
> if you consider the Jonathan draft, he says that
> "...chess game that utilizes voice commands to move pieces.  When  
> one player says "Queen to D7", the software on their phone  
> recognizes this and moves the piece.  At the same time, the user's  
> voice is sent to the other player, where it is both rendered to the  
> user, and interpreted locally in order to move the piece."
> and he suggests that the following urn
>
>    urn:service:games:voice-chess
>
> could be used to identify the voice chess application.
>
> I guess that with voice he supposes audio, but nowadays we can also  
> have voice using PoC. Then we need a way to specify which kind of  
> voice we want: a standard full-duplex or a PoC. Maybe the chess  
> game works well with, or it is designed for a standard audio but  
> not for a PoC audio. So we need to address also the voice  
> preference we want. I like your proposal, having something  
> hierarchical was something I was thinking too I don't know if it is  
> your idea, but with the service urn you are proposing:  
> Urn:service:voice.poc.chess I imaging you are asking to play a  
> Chess game, that work utilizing voice commands, but specifically  
> voice over PoC. This is also compliant to the draft-ietf-ecrit- 
> service-urn (see section 3) "Declaration of syntactic structure:  
> The URN consists of a hierarchical service identifier, with a  
> sequence of labels separated by periods.  The left-most label is  
> the most significant one and is called 'top-level service', while  
> names to the right are called 'sub-services'.  The set of allowable  
> characters is the same as that for domain names [1] and a subset of  
> the labels allowed in [5].  Labels are case-insensitive and MUST be  
> specified in all lower-case.  For any given service URN, service- 
> identifiers can be removed right-to-left and the resulting URN is  
> still valid, referring a more generic service.  In other words, if  
> a service 'x.y.z' exists, the URNs 'x' and 'x.y' are also valid  
> service URNs." Regards sal
> On Fri, 2007-02-09 at 13:44 +0530, ravishankar.shiroor@wipro.com  
> wrote:
>> Hi Salvatore, As an end user, poc is also an application that I  
>> use and chess is also an application that I can use. The way I see  
>> it, in your example, it will be running chess application on top  
>> of PoC application. So the service urn for this would be something  
>> like: Urn:service:voice.poc.chess Regards, Ravi. -- Ravishankar.  
>> G. Shiroor Wipro Technologies, Bangalore.  
>> ravishankar.shiroor@wipro.com -- > -----Original Message----- >  
>> From: Salvatore Loreto (JO/LMF) >  
>> [mailto:salvatore.loreto@ericsson.com] > Sent: Wednesday, February  
>> 07, 2007 1:37 PM > To: RaviShankar Shiroor (WT01 - Product  
>> Engineering Solutions) > Cc: Gonzalo.Camarillo@ericsson.com;  
>> sipping@ietf.org; > mary.barnes@nortel.com > Subject: RE:  
>> [Sipping] Service Identification in SIP > > hi Ravi, > > in IMS a  
>> Service is a "type of communication defined by a > service  
>> definition that specifies the rules and procedures > and allowed  
>> media, if any, for a specific type of > communication and that  
>> utilises the IMS enabler. An IMS > communication service defines  
>> what is possible within a > single SIP session or standalone  
>> transaction." > So a Service implies some combination of  
>> capabilities and > ways to use those capabilities. > If you want  
>> the Service specifies the context for a communication. > > Instead  
>> the Application define the specific communication (or > if you  
>> want application) that you want run in the specific > context. So  
>> you should have several different applications > designed to run  
>> on a specific environment or that have > different performance or  
>> behavior in different context > (running on top of different  
>> service). > > Lets go to Jonathan's example of the Chess. > In his  
>> example the Chess is a service, for IMS instead is an >  
>> application that you can run on top of different service. > For  
>> example you can run Chess application on top of PoC > service or  
>> on top of the "standard" service. So let say you > run Chess on  
>> top of PoC service, then you can use only voice > in the way that  
>> PoC specifies. > > /sal > > > > > On Thu, 2007-02-01 at 16:39  
>> +0530, > ravishankar.shiroor@wipro.com wrote: > > Hello, > > > >  
>> IMO, Service Identification is definitely needed and should > be  
>> taken > > up as a WG as proposed below. > > > > Further, I think  
>> Service Identification scope should be as per the > > requirement  
>> in Jonathan's draft, as I still do not > understand the need > >  
>> for having a Communication Service Identifier and another >  
>> Application > > Identifier as given in the 3gpp-ics-requirements  
>> document. > > > > Regards, > > Ravi. > > > > -- > > > >  
>> Ravishankar. G. Shiroor > > Wipro Technologies, Bangalore. > > > >  
>> ravishankar.shiroor@wipro.com > > -- > > > > > > > -----Original  
>> Message----- > > > From: Gonzalo Camarillo  
>> [mailto:Gonzalo.Camarillo@ericsson.com] > > > Sent: Wednesday,  
>> January 31, 2007 8:30 PM > > > To: sipping > > > Cc: Mary Barnes >  
>> > > Subject: [Sipping] Service Identification in SIP > > > > > >  
>> Folks, > > > > > > during the last year, there have been a number  
>> of threads on the > > > mailing list discussing service  
>> identification in SIP. In > > > particular, the following thread  
>> discussed 3GPP > requirements on this > > > area: > > > > > > >  
>> http://www1.ietf.org/mail-archive/web/sipping/current/ 
>> msg10562.html > > > > > > At a later point, those 3GPP  
>> requirements were documented in the > > > following (expired by  
>> now) Internet Draft, which was > presented in a > > > face-to-face  
>> IETF meeting. > > > > > > http://www.watersprings.org/pub/id/draft- 
>> loreto-sipping-3gpp-i > > > cs-requirements-00.txt > > > > > >  
>> Lately, the following draft has been submitted to the archives: >  
>> > > > > > http://www.ietf.org/internet-drafts/draft-rosenberg- 
>> sipping-se > > > rvice-identification-00.txt > > > > > > This  
>> draft proposes a solution that intends to meet known > > >  
>> requirements in this area. > > > > > > Given the amount of  
>> discussions on issues related to service > > > identification we  
>> have seen during the last year, we > believe this is > > > a  
>> fairly important topic for the SIPPING WG to tackle. At some > > >  
>> point, we would like to have a milestone with its > associated WG  
>> item > > > dealing with this issue. > > > > > > Therefore, we  
>> would like to encourage people to read this > draft and > > > send  
>> comments to the list. Additionally, if you believe that this > > >  
>> work is (or is not) important, we would like to hear that as well.  
>> > > > > > > Cheers, > > > > > > Gonzalo > > > SIPPING co-chair > >  
>> > > > > > > > _______________________________________________ > >  
>> > Sipping mailing list > https://www1.ietf.org/mailman/listinfo/ 
>> sipping > > > This list is for NEW development of the application  
>> of SIP Use  > > > sip-implementors@cs.columbia.edu for questions  
>> on current sip Use > > > sip@ietf.org for new developments of core  
>> SIP > > > > > > > _______________________________________________  
>> > > Sipping mailing list https://www1.ietf.org/mailman/listinfo/ 
>> sipping > > This list is for NEW development of the application of  
>> SIP Use > > sip-implementors@cs.columbia.edu for questions on  
>> current sip Use > > sip@ietf.org for new developments of core SIP  
>> > > The information contained in this electronic message and any  
>> attachments to this message are intended for the exclusive use of  
>> the addressee(s) and may contain proprietary, confidential or  
>> privileged information. If you are not the intended recipient, you  
>> should not disseminate, distribute or copy this e-mail. Please  
>> notify the sender immediately and destroy all copies of this  
>> message and any attachments. WARNING: Computer viruses can be  
>> transmitted via email. The recipient should check this email and  
>> any attachments for the presence of viruses. The company accepts  
>> no liability for any damage caused by any virus transmitted by  
>> this email. www.wipro.com
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sat Feb 10 23:51:49 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HG6eo-0005Rz-8M; Sat, 10 Feb 2007 23:49:46 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HG6em-0005Rt-Q6
	for sipping@ietf.org; Sat, 10 Feb 2007 23:49:44 -0500
Received: from alnrmhc16.comcast.net ([206.18.177.56])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HG6el-0004hy-HF
	for sipping@ietf.org; Sat, 10 Feb 2007 23:49:44 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (alnrmhc16) with ESMTP
	id <20070211044942b1600i819oe>; Sun, 11 Feb 2007 04:49:43 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1B4nfAM020333
	for <sipping@ietf.org>; Sat, 10 Feb 2007 23:49:41 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1B4nfk2020329;
	Sat, 10 Feb 2007 23:49:41 -0500
Date: Sat, 10 Feb 2007 23:49:41 -0500
Message-Id: <200702110449.l1B4nfk2020329@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <1171031930.3464.92.camel@localhost.localdomain>
	(salvatore.loreto@ericsson.com)
Subject: Re: [Sipping] Service Identification in SIP
References: <45C0AEF2.8010101@ericsson.com>
	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
	<1170835642.29763.54.camel@n95.nomadiclab.com>
	<45CA08B2.4050603@cisco.com>
	<532B18E13CF9E64380EF5FDDE265E071B6192E@blr-m2-msg.wipro.com>
	<45CC7697.1060204@cisco.com>
	<1171031930.3464.92.camel@localhost.localdomain>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: e8a67952aa972b528dd04570d58ad8fe
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I've attempted to re-read the threads that Gonzalo referred to, and
the major limitation that I see is the lack of any concrete example of
why these identifiers are needed.  Probably the IMS people have
discussed this so many times that they've forgotten the detailed
rationale behind the requirement.  But I'm a net-head, (and most of
the people here are net-heads) and the SIP systems that I've dealt
with show no evidence of needing "service identification", they're
pure proxies, and requests are passed transparently end-to-end.

Could somebody please lay out an example showing the utility of
service identification?

   From: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>

   I think that everyone here agree that there is the possibility to use a
   media in different ways, and that we need a distinction about what the
   capabilities are and how they are used, whatever will be the mechanism
   that at the end we will identify to do that.

Respectfully, I disagree.  We do need ways of identifying the
capabilities of SIP agents and transports, but I see no necessity of
creating specified bundles of capabailities, and then constraining
dialogs to require an initial agreement as to which bundle of
capabilities will be used.

One of the central features of Internet architecture is that agents
negotiate compatibility on a feature-by-feature basis; there are no
"profiles" which must be known on a network-wide basis.  (And I have
observed that this feature is of significant commercial value.)

In regard to identifying an "application" to process the dialog (and I
use "application" in a standard computing way), I see no reason that a
user cannot select URI-parameters to append to his AOR, and arrange
for his UA to route invites with those URI-parameters to the
appropriate application.  (And given the rich mechanisms for
redirection in SIP, the application need not even reside at a UA which
is registered for the AOR.)  These URI-parameters can be formally or
informally standardized, or established ad-hoc by each user.

In regard to PoC (as someone said, "a specialized floor-control
protocol") and similar features, SIP currently has mechanisms for
indicating support of features and requiring that a recipient support
a feature before accepting a call (the Supported and Requires headers,
and simply using the feature).  The only limitation I can see that
might need to be addressed is that there is no way to create an option
tag for a new feature that one has invented that is not yet thoroughly
standardized.

But I now discover RFC 3261 section 8.1.1.9:

   The option tags listed MUST only refer to extensions defined in
   standards-track RFCs.  This is to prevent servers from insisting that
   clients implement non-standard, vendor-defined features in order to
   receive service.  Extensions defined by experimental and
   informational RFCs are explicitly excluded from usage with the
   Supported header field in a request, since they too are often used to
   document vendor-defined extensions.

So I see that as a fundamental architectural decision, SIP does not
permit sending INVITEs that require specialized recipients, excepting
if the specialization is standards-track.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sun Feb 11 09:59:00 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGG8p-0005YS-AH; Sun, 11 Feb 2007 09:57:23 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGG8n-0005YF-WF
	for sipping@ietf.org; Sun, 11 Feb 2007 09:57:22 -0500
Received: from exprod6og54.obsmtp.com ([64.18.1.189])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1HGG8k-0003Ob-BB
	for sipping@ietf.org; Sun, 11 Feb 2007 09:57:21 -0500
Received: from source ([192.150.11.134]) by exprod6ob54.postini.com
	([64.18.5.12]) with SMTP; Sun, 11 Feb 2007 06:57:13 PST
Received: from inner-relay-1.corp.adobe.com ([153.32.1.51])
	by outbound-smtp-1.corp.adobe.com (8.12.10/8.12.10) with ESMTP id
	l1BEvAvV010635; Sun, 11 Feb 2007 06:57:10 -0800 (PST)
Received: from fe1.corp.adobe.com (fe1.corp.adobe.com [10.8.192.70])
	by inner-relay-1.corp.adobe.com (8.12.10/8.12.10) with ESMTP id
	l1BEvDPF026894; Sun, 11 Feb 2007 06:57:13 -0800 (PST)
Received: from namail5.corp.adobe.com ([10.8.192.88]) by fe1.corp.adobe.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Sun, 11 Feb 2007 06:57:12 -0800
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Sun, 11 Feb 2007 06:56:44 -0800
Message-ID: <24CCCC428EFEA2469BF046DB3C7A8D224538F2@namail5.corp.adobe.com>
In-Reply-To: <200702110449.l1B4nfk2020329@dragon.ariadne.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdNmKVUAU7ffDneTA+ELvuVBIOo/wAVCJog
References: <45C0AEF2.8010101@ericsson.com><532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com><1170835642.29763.54.camel@n95.nomadiclab.com><45CA08B2.4050603@cisco.com><532B18E13CF9E64380EF5FDDE265E071B6192E@blr-m2-msg.wipro.com><45CC7697.1060204@cisco.com><1171031930.3464.92.camel@localhost.localdomain>
	<200702110449.l1B4nfk2020329@dragon.ariadne.com>
From: "Henry Sinnreich" <hsinnrei@adobe.com>
To: <Dale.Worley@comcast.net>, <sipping@ietf.org>
X-OriginalArrivalTime: 11 Feb 2007 14:57:12.0865 (UTC)
	FILETIME=[E9831110:01C74DEC]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 1a1bf7677bfe77d8af1ebe0e91045c5b
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Dale Worley wrote:

>Could somebody please lay out an example showing the utility of service
>identification?

Sure: The service identification is most likely a new charging
contraption invented in a marketing department. As such, any protocol
significance cannot be found nor explained.

One has to wonder if users who have the choice of superior quality
wideband audio are willing to see new billing items for what they can
have for free and with richer features, as long as they pay for the only
service that matters: Internet access.

Proposals such as the "Service Identification in SIP" could be better
handled in a marketing WG, if there were such a thing in the IETF :-)

It would be helpful if such schemas devised in other organizations and
submitted here for the IETF blessing would not burden the Internet
standards work. Maybe a guidance paper from the IESG is in order?=20
My two cents.=20

Thanks, Henry

-----Original Message-----
From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net]=20
Sent: Saturday, February 10, 2007 10:50 PM
To: sipping@ietf.org
Subject: Re: [Sipping] Service Identification in SIP

I've attempted to re-read the threads that Gonzalo referred to, and
the major limitation that I see is the lack of any concrete example of
why these identifiers are needed.  Probably the IMS people have
discussed this so many times that they've forgotten the detailed
rationale behind the requirement.  But I'm a net-head, (and most of
the people here are net-heads) and the SIP systems that I've dealt
with show no evidence of needing "service identification", they're
pure proxies, and requests are passed transparently end-to-end.

Could somebody please lay out an example showing the utility of
service identification?

   From: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>

   I think that everyone here agree that there is the possibility to use
a
   media in different ways, and that we need a distinction about what
the
   capabilities are and how they are used, whatever will be the
mechanism
   that at the end we will identify to do that.

Respectfully, I disagree.  We do need ways of identifying the
capabilities of SIP agents and transports, but I see no necessity of
creating specified bundles of capabailities, and then constraining
dialogs to require an initial agreement as to which bundle of
capabilities will be used.

One of the central features of Internet architecture is that agents
negotiate compatibility on a feature-by-feature basis; there are no
"profiles" which must be known on a network-wide basis.  (And I have
observed that this feature is of significant commercial value.)

In regard to identifying an "application" to process the dialog (and I
use "application" in a standard computing way), I see no reason that a
user cannot select URI-parameters to append to his AOR, and arrange
for his UA to route invites with those URI-parameters to the
appropriate application.  (And given the rich mechanisms for
redirection in SIP, the application need not even reside at a UA which
is registered for the AOR.)  These URI-parameters can be formally or
informally standardized, or established ad-hoc by each user.

In regard to PoC (as someone said, "a specialized floor-control
protocol") and similar features, SIP currently has mechanisms for
indicating support of features and requiring that a recipient support
a feature before accepting a call (the Supported and Requires headers,
and simply using the feature).  The only limitation I can see that
might need to be addressed is that there is no way to create an option
tag for a new feature that one has invented that is not yet thoroughly
standardized.

But I now discover RFC 3261 section 8.1.1.9:

   The option tags listed MUST only refer to extensions defined in
   standards-track RFCs.  This is to prevent servers from insisting that
   clients implement non-standard, vendor-defined features in order to
   receive service.  Extensions defined by experimental and
   informational RFCs are explicitly excluded from usage with the
   Supported header field in a request, since they too are often used to
   document vendor-defined extensions.

So I see that as a fundamental architectural decision, SIP does not
permit sending INVITEs that require specialized recipients, excepting
if the specialization is standards-track.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 12 08:43:27 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGbSA-0005I4-8s; Mon, 12 Feb 2007 08:42:46 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGbS8-0005HZ-71; Mon, 12 Feb 2007 08:42:44 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HGbRx-0006tC-I5; Mon, 12 Feb 2007 08:42:42 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	E60212041F; Mon, 12 Feb 2007 14:42:28 +0100 (CET)
X-AuditID: c1b4fb3c-b17cbbb0000007de-f9-45d06ec4aae6 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	CDBC4203CB; Mon, 12 Feb 2007 14:42:28 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 12 Feb 2007 14:42:28 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Date: Mon, 12 Feb 2007 14:42:28 +0100
Message-ID: <7374777208BDC7449D5620EF9423256701FF1078@esealmw113.eemea.ericsson.se>
In-Reply-To: <45CB7E2F.1070706@bbn.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sip] New draft on early media
Thread-Index: AcdLujpNMPZNpklsRX687LKDDU19oQC8L/tQ
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Richard Barnes" <rbarnes@bbn.com>,
	<sip@ietf.org>,
	<sipping@ietf.org>
X-OriginalArrivalTime: 12 Feb 2007 13:42:28.0408 (UTC)
	FILETIME=[A2FAFB80:01C74EAB]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: f4c2cf0bccc868e4cc88dace71fb3f44
Cc: 
Subject: [Sipping] RE: [Sip] New draft on early media
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,

I haven't had time to read the document yet, but I am not sure I agree
with the statement saying that the biggest problem is media arriving
before the SDP answer. In some environments such media will not even
reach the terminal, due to gating etc.

Many implementations DO send an SDP answer (or, at least an unreliable
"preview" of the SDP answer) before (or at the same time as) early media
is sent.

And, in the cases where implementations do NOT send an SDP answer before
they start sending early media - what behavior would you be able to
standardize?

Regards,

Christer=20


> -----Original Message-----
> From: Richard Barnes [mailto:rbarnes@bbn.com]=20
> Sent: 8. helmikuuta 2007 21:47
> To: sip@ietf.org; sipping@ietf.org
> Subject: [Sip] New draft on early media
>=20
> After the big discussions on SIPPING last summer about how to=20
> fix early media, and the reaction at the last couple of IETFs=20
> to Brian Stucker's draft on how to cope with it, I've put=20
> together a draft to try to consolidate all of the discussions=20
> and move toward a solution.
>=20
> draft-barnes-sip-em-ps-req-sol includes three important parts:
>=20
> 1. Problem statement: Although there's lots of definitions of=20
> "early media" in existing documents, I think there's broad=20
> agreement that the stuff that causes problems is media that=20
> arrives before an SDP offer/answer exchange has completed. =20
> The problem is therefore to enable the SDP exchange(s) in the=20
> INVITE transaction to complete more quickly, and make it more=20
> difficult (if not impossible) to send media before it has=20
> completed.  We are explicitly NOT trying to prevent the=20
> sending of media before a 200/OK or other closing of an=20
> INVITE transaction.
>=20
> 2. Requirements: Here, we try to create a consolidated list=20
> of requirements for an early media fix.
>=20
> 3. Analysis of Solutions: Lots of folks have put forward=20
> proposals to fix the early media problem.  This section tries=20
> to enumerate most of them and evaluate them against the=20
> requirements in section 2.
>=20
> It hasn't shown up yet on the I-D tracker, so I've attached a copy.
>=20
> --Richard
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 12 10:09:33 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGcmo-0003Cf-Q7; Mon, 12 Feb 2007 10:08:10 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGcmn-0003BB-5w; Mon, 12 Feb 2007 10:08:09 -0500
Received: from mx12.bbn.com ([128.33.0.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HGcmk-00010V-S5; Mon, 12 Feb 2007 10:08:09 -0500
Received: from po2.bbn.com ([128.33.0.56])
	by mx12.bbn.com with esmtp (Exim 4.60)
	(envelope-from <rbarnes@bbn.com>)
	id 1HGcmb-0006ir-5J; Mon, 12 Feb 2007 10:07:57 -0500
Received: from [127.0.0.1] (col-dhcp33-244-152.bbn.com [128.33.244.152])
	by po2.bbn.com (8.11.6+Sun/8.10.2) with ESMTP id l1CF7uw15227;
	Mon, 12 Feb 2007 10:07:57 -0500 (EST)
Message-ID: <45D082D1.3000203@bbn.com>
Date: Mon, 12 Feb 2007 10:08:01 -0500
From: Richard Barnes <rbarnes@bbn.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Christer Holmberg (JO/LMF)" <christer.holmberg@ericsson.com>
References: <7374777208BDC7449D5620EF9423256701FF1078@esealmw113.eemea.ericsson.se>
In-Reply-To: <7374777208BDC7449D5620EF9423256701FF1078@esealmw113.eemea.ericsson.se>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 10d3e4e3c32e363f129e380e644649be
Cc: sip@ietf.org, sipping@ietf.org
Subject: [Sipping] Re: [Sip] New draft on early media
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Christer,

Thanks for your comments.  Responses inline.

Cheers,
--Richard

> I haven't had time to read the document yet, but I am not sure I agree
> with the statement saying that the biggest problem is media arriving
> before the SDP answer. In some environments such media will not even
> reach the terminal, due to gating etc.

Perhaps a better term would be "media sent before an SDP answer has been 
received".  As you point out, sending media before an SDP offer/answer 
exchange has completed poses the risk that that media will not arrive. 
To put it in the terms used in the document, media clipping is a problem 
caused by early (pre-SDP-answer) media, and thus one which this document 
seeks to solve.

> Many implementations DO send an SDP answer (or, at least an unreliable
> "preview" of the SDP answer) before (or at the same time as) early media
> is sent.

This is the behavior we would like to standardize -- that an SDP answer 
is *always* sent before media of any type.  What we're trying to do with 
this document is describe the requirements for how that answer is sent, 
and consolidate the many ways people have proposed to send it.

> And, in the cases where implementations do NOT send an SDP answer before
> they start sending early media - what behavior would you be able to
> standardize?

We would prefer that standards not allow this behavior -- that seems 
like the right thing to me.  What's a little trickier is that you'd like 
to not have the answerer sending media before the answer has *arrived*. 
  When the answer is carried in a reliable message, you could just 
forbid sending media before receipt has been confirmed.  When the 
transport is unreliable, the situation is a little less clear, but I 
think there's still possibilities for standardization, e.g., by 
forbidding answerers to send media until they've received media.  This 
is discussed in more length in the document.

> 
> Regards,
> 
> Christer 
> 
> 
>> -----Original Message-----
>> From: Richard Barnes [mailto:rbarnes@bbn.com] 
>> Sent: 8. helmikuuta 2007 21:47
>> To: sip@ietf.org; sipping@ietf.org
>> Subject: [Sip] New draft on early media
>>
>> After the big discussions on SIPPING last summer about how to 
>> fix early media, and the reaction at the last couple of IETFs 
>> to Brian Stucker's draft on how to cope with it, I've put 
>> together a draft to try to consolidate all of the discussions 
>> and move toward a solution.
>>
>> draft-barnes-sip-em-ps-req-sol includes three important parts:
>>
>> 1. Problem statement: Although there's lots of definitions of 
>> "early media" in existing documents, I think there's broad 
>> agreement that the stuff that causes problems is media that 
>> arrives before an SDP offer/answer exchange has completed.  
>> The problem is therefore to enable the SDP exchange(s) in the 
>> INVITE transaction to complete more quickly, and make it more 
>> difficult (if not impossible) to send media before it has 
>> completed.  We are explicitly NOT trying to prevent the 
>> sending of media before a 200/OK or other closing of an 
>> INVITE transaction.
>>
>> 2. Requirements: Here, we try to create a consolidated list 
>> of requirements for an early media fix.
>>
>> 3. Analysis of Solutions: Lots of folks have put forward 
>> proposals to fix the early media problem.  This section tries 
>> to enumerate most of them and evaluate them against the 
>> requirements in section 2.
>>
>> It hasn't shown up yet on the I-D tracker, so I've attached a copy.
>>
>> --Richard
>>
> 
> 


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 12 11:02:31 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGdd3-0007Ce-GL; Mon, 12 Feb 2007 11:02:09 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGdd1-0007CR-Vs; Mon, 12 Feb 2007 11:02:08 -0500
Received: from dsl001-129-069.dfw1.dsl.speakeasy.net ([72.1.129.69]
	helo=vicuna.estacado.net) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HGdcx-0005YI-La; Mon, 12 Feb 2007 11:02:07 -0500
Received: from [172.17.1.65] ([172.17.1.65]) (authenticated bits=0)
	by vicuna.estacado.net (8.13.8/8.13.8) with ESMTP id l1CG1nBQ080777
	(version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO);
	Mon, 12 Feb 2007 10:01:50 -0600 (CST)
	(envelope-from rjsparks@estacado.net)
Mime-Version: 1.0 (Apple Message framework v752.3)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <D7DED047-5FD4-4E9F-9446-9425305A3645@estacado.net>
Content-Transfer-Encoding: 7bit
From: Robert Sparks <rjsparks@estacado.net>
Date: Mon, 12 Feb 2007 10:01:45 -0600
To: sip-implementors@cs.columbia.edu, SIP LIST <sip@ietf.org>, simple@ietf.org,
	sipping LIST <sipping@ietf.org>
X-Mailer: Apple Mail (2.752.3)
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 7aefe408d50e9c7c47615841cb314bed
Cc: 
Subject: [Sipping] Registration for SIPit 20 is open
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

SIPit 20 will be hosted by Alcatel-Lucent in Antwerp, Belgium April  
16-20,2007.

Registration is now open - see http://www.sipit.net for more  
information.

RjS

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 12 15:25:55 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGhj8-0002TG-7X; Mon, 12 Feb 2007 15:24:42 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGhj6-0002T7-TY; Mon, 12 Feb 2007 15:24:40 -0500
Received: from zcars04f.nortel.com ([47.129.242.57])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HGhj4-0003K3-HK; Mon, 12 Feb 2007 15:24:40 -0500
Received: from zrc2hxm0.corp.nortel.com (zrc2hxm0.corp.nortel.com
	[47.103.123.71])
	by zcars04f.nortel.com (Switch-2.2.6/Switch-2.2.0) with ESMTP id
	l1CKOR518035; Mon, 12 Feb 2007 15:24:27 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Re: [Sip] New draft on early media
Date: Mon, 12 Feb 2007 14:24:05 -0600
Message-ID: <1ECE0EB50388174790F9694F77522CCF0F21D58F@zrc2hxm0.corp.nortel.com>
In-Reply-To: <45D082D1.3000203@bbn.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Re: [Sip] New draft on early media
Thread-Index: AcdOuBzwzIh5r0ljTA6EYziSrjehxgAKoryw
From: "Francois Audet" <audet@nortel.com>
To: "Richard Barnes" <rbarnes@bbn.com>,
	"Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 87a3f533bb300b99e2a18357f3c1563d
Cc: sip@ietf.org, sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi Richard,

Forcing media cut-through to occur only after an Answer has been
received
does solve SOME of the early media problems. As others have mentioned,
there
are environments that operate like this already. I also know many
products that
do not cut-through media before the answer is received.

But there are other problems related with early media. Brian Stucker's
draft
listed a bunch of them.=20

Forking for example comes to mind: even if we did mandate cut-through
after
answer, we would still have problems related with forking (i.e., how to
pick
the stream that needs to be played, and how to correlate the stream with
the=20
signaling). How to deal with SRTP key exchange in that scenario is an
even=20
bigger problem.=20

Clipping is a possible side-effect of your proposal. We'd need some
mechanism to=20
mitigate it, in particular for auto-answer scenarios.


> -----Original Message-----
> From: Richard Barnes [mailto:rbarnes@bbn.com]=20
> Sent: Monday, February 12, 2007 07:08
> To: Christer Holmberg (JO/LMF)
> Cc: sip@ietf.org; sipping@ietf.org
> Subject: [Sipping] Re: [Sip] New draft on early media
>=20
> Christer,
>=20
> Thanks for your comments.  Responses inline.
>=20
> Cheers,
> --Richard
>=20
> > I haven't had time to read the document yet, but I am not=20
> sure I agree=20
> > with the statement saying that the biggest problem is media=20
> arriving=20
> > before the SDP answer. In some environments such media will=20
> not even=20
> > reach the terminal, due to gating etc.
>=20
> Perhaps a better term would be "media sent before an SDP=20
> answer has been received".  As you point out, sending media=20
> before an SDP offer/answer exchange has completed poses the=20
> risk that that media will not arrive.=20
> To put it in the terms used in the document, media clipping=20
> is a problem caused by early (pre-SDP-answer) media, and thus=20
> one which this document seeks to solve.
>=20
> > Many implementations DO send an SDP answer (or, at least an=20
> unreliable=20
> > "preview" of the SDP answer) before (or at the same time as) early=20
> > media is sent.
>=20
> This is the behavior we would like to standardize -- that an=20
> SDP answer is *always* sent before media of any type.  What=20
> we're trying to do with this document is describe the=20
> requirements for how that answer is sent, and consolidate the=20
> many ways people have proposed to send it.
>=20
> > And, in the cases where implementations do NOT send an SDP answer=20
> > before they start sending early media - what behavior would you be=20
> > able to standardize?
>=20
> We would prefer that standards not allow this behavior --=20
> that seems like the right thing to me.  What's a little=20
> trickier is that you'd like to not have the answerer sending=20
> media before the answer has *arrived*.=20
>   When the answer is carried in a reliable message, you could=20
> just forbid sending media before receipt has been confirmed. =20
> When the transport is unreliable, the situation is a little=20
> less clear, but I think there's still possibilities for=20
> standardization, e.g., by forbidding answerers to send media=20
> until they've received media.  This is discussed in more=20
> length in the document.
>=20
> >=20
> > Regards,
> >=20
> > Christer
> >=20
> >=20
> >> -----Original Message-----
> >> From: Richard Barnes [mailto:rbarnes@bbn.com]
> >> Sent: 8. helmikuuta 2007 21:47
> >> To: sip@ietf.org; sipping@ietf.org
> >> Subject: [Sip] New draft on early media
> >>
> >> After the big discussions on SIPPING last summer about how to fix=20
> >> early media, and the reaction at the last couple of IETFs to Brian=20
> >> Stucker's draft on how to cope with it, I've put together=20
> a draft to=20
> >> try to consolidate all of the discussions and move toward=20
> a solution.
> >>
> >> draft-barnes-sip-em-ps-req-sol includes three important parts:
> >>
> >> 1. Problem statement: Although there's lots of definitions=20
> of "early=20
> >> media" in existing documents, I think there's broad agreement that=20
> >> the stuff that causes problems is media that arrives before an SDP=20
> >> offer/answer exchange has completed.
> >> The problem is therefore to enable the SDP exchange(s) in=20
> the INVITE=20
> >> transaction to complete more quickly, and make it more=20
> difficult (if=20
> >> not impossible) to send media before it has completed.  We are=20
> >> explicitly NOT trying to prevent the sending of media=20
> before a 200/OK=20
> >> or other closing of an INVITE transaction.
> >>
> >> 2. Requirements: Here, we try to create a consolidated list of=20
> >> requirements for an early media fix.
> >>
> >> 3. Analysis of Solutions: Lots of folks have put forward=20
> proposals to=20
> >> fix the early media problem.  This section tries to=20
> enumerate most of=20
> >> them and evaluate them against the requirements in section 2.
> >>
> >> It hasn't shown up yet on the I-D tracker, so I've attached a copy.
> >>
> >> --Richard
> >>
> >=20
> >=20
>=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 12 17:58:29 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGk6L-0001g5-Uy; Mon, 12 Feb 2007 17:56:49 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGk6K-0001fs-G0
	for sipping@ietf.org; Mon, 12 Feb 2007 17:56:48 -0500
Received: from alnrmhc16.comcast.net ([204.127.225.96])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGk6J-00012R-6W
	for sipping@ietf.org; Mon, 12 Feb 2007 17:56:48 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (alnrmhc16) with ESMTP
	id <20070212225644b1600iah0he>; Mon, 12 Feb 2007 22:56:44 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1CMueAM020625
	for <sipping@ietf.org>; Mon, 12 Feb 2007 17:56:40 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1CMueqK020621;
	Mon, 12 Feb 2007 17:56:40 -0500
Date: Mon, 12 Feb 2007 17:56:40 -0500
Message-Id: <200702122256.l1CMueqK020621@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <24CCCC428EFEA2469BF046DB3C7A8D224538F2@namail5.corp.adobe.com>
	(hsinnrei@adobe.com)
Subject: Re: [Sipping] Service Identification in SIP
References: <45C0AEF2.8010101@ericsson.com><532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com><1170835642.29763.54.camel@n95.nomadiclab.com><45CA08B2.4050603@cisco.com><532B18E13CF9E64380EF5FDDE265E071B6192E@blr-m2-msg.wipro.com><45CC7697.1060204@cisco.com><1171031930.3464.92.camel@localhost.localdomain>
	<200702110449.l1B4nfk2020329@dragon.ariadne.com>
	<24CCCC428EFEA2469BF046DB3C7A8D224538F2@namail5.corp.adobe.com>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 3002fc2e661cd7f114cb6bae92fe88f1
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   From: "Henry Sinnreich" <hsinnrei@adobe.com>

   >Could somebody please lay out an example showing the utility of service
   >identification?

   Sure: The service identification is most likely a new charging
   contraption invented in a marketing department. As such, any protocol
   significance cannot be found nor explained.

I think that it's more complicated than that.  If it was only a
matter of billing based on the services used, a fairly simple scheme
for monitoring the communication stream and generating billing entries
would be simpler and more efficient.

Let me throw out an idea (which I've probably stolen from someone
else), one which I recall based on the BT announcement.

AFAICT, for many service providers, an incoming call (regardless of
the destination E.164 number) will get routed to one of a set of
"application servers" (AS), each one of which handles a particular
"service".  The application server is responsible for managing the
final link to the phone.  From an Internet perspective, what is really
happening is that the UA is partitioned into two parts, part on the AS
and part on the real phone.

This has some advantages, in particular that a new service can be
implemented by buying and installing an AS, as long as the AS-to-phone
link provides adequate facilities for the lowest-level UI functions.

There are disadvantages, too, particularly that both ends of a call
have to agree on the "service" being used, and that each service in
effect defines a subset network, consisting of the ASs for that
service, and all phones which are serviced by an AS for that service.
Worse, services cannot be mix-and-matched in any way, even if both
users would be willing to pay the fees to support all of the needed
infrastructure.

In some cases, this sort of architecture may be difficult to avoid.
For instance, the push-to-talk (PTT) feature.  The way it's
implemented is rather a cheat -- The call is first sent to an AS.  If
the AS thinks that the callee's phone is registered with the network,
and the last status update from the callee's phone says that it will
answer PTT calls immediately, then the AS sends a "call connected"
indication to the caller before it has achieved a voice channel
connection with the callee phone.  It then starts buffering the
caller's audio for the several seconds it needs to gain a voice
connection to the callee phone.  Once it gets a connection, it plays
the buffered audio, using various tricks to make playback faster than
real-time.

But given the limitation that the last link to the callee's phone does
not impelment "immediate connection", there is no way provide PTT
without an AS.

I think the architectural challenge is how to support this sort of
network but without the nasty effects of "siloing":

- If a caller requests to use a particular feature, and the callee is
  unable or unwilling to support that feature, and the caller is
  willing to proceed without using the feature, the call should
  proceed without the feature.

- A call can request multiple features, and it will proceed using
  whatever subset of them are supported by the callee.

I don't see this as intrinsically impossible to achieve, even if every
feature is implemented by ASs.  Assuming the features are reasonably
well-behaved, it should be possible to chain a call through several
ASs before reaching the callee phone.

(This may require a separate charge for the use of each AS, but that's
not unfair -- by hypothesis, the ASs implement features that the
callee phones do not, and someone has to pay for the ASs.)

It would be helpful if each INVITE declared all of the features that
would ever be used during the call.  In some cases, the INVITE already
contains the needed information (particularly in regard to the media
features).  But it seems that we should avoid this requirement if
possible.  OTOH, it seems plausible that an AS could be inserted into
the service chain during a dialog -- I've plotted out call flows by
which a B2BUA can insert itself into an existing dialog using
INVITE/Join, and a cluster of ASs in a closed zone of a network can
perform more dexterous rerouting tricks.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 12 23:07:10 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGouU-0006AL-PG; Mon, 12 Feb 2007 23:04:54 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGouT-0006AC-3N
	for sipping@ietf.org; Mon, 12 Feb 2007 23:04:53 -0500
Received: from wip-ec-wd.wipro.com ([203.91.193.32])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGouN-0003OM-RN
	for sipping@ietf.org; Mon, 12 Feb 2007 23:04:53 -0500
Received: from wip-ec-wd.wipro.com (localhost.wipro.com [127.0.0.1])
	by localhost (Postfix) with ESMTP id 284E1205E3
	for <sipping@ietf.org>; Tue, 13 Feb 2007 09:26:54 +0530 (IST)
Received: from blr-ec-bh02.wipro.com (blr-ec-bh02.wipro.com [10.201.50.92])
	by wip-ec-wd.wipro.com (Postfix) with ESMTP id 12C8B205DC
	for <sipping@ietf.org>; Tue, 13 Feb 2007 09:26:54 +0530 (IST)
Received: from blr-m2-msg.wipro.com ([10.116.50.99]) by blr-ec-bh02.wipro.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 09:34:41 +0530
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Tue, 13 Feb 2007 09:39:35 +0530
Message-ID: <532B18E13CF9E64380EF5FDDE265E071B61B9E@blr-m2-msg.wipro.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdMTgqajKN0nfG7QfGdu+gL50dmgQC1gUJA
References: <45C0AEF2.8010101@ericsson.com>	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
	<1170835642.29763.54.camel@n95.nomadiclab.com>
	<45CA08B2.4050603@cisco.com>
	<532B18E13CF9E64380EF5FDDE265E071B6192E@blr-m2-msg.wipro.com>
	<45CC7697.1060204@cisco.com>
From: <ravishankar.shiroor@wipro.com>
To: <pkyzivat@cisco.com>
X-OriginalArrivalTime: 13 Feb 2007 04:04:41.0652 (UTC)
	FILETIME=[16658F40:01C74F24]
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 2b2ad76aced9b1d558e34a970a85c027
Cc: sipping@ietf.org, Gonzalo.Camarillo@ericsson.com, mary.barnes@nortel.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Paul Kyzivat wrote:
> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
> Sent: Friday, February 09, 2007 6:57 PM
> To: RaviShankar Shiroor (WT01 - Product Engineering Solutions)
> Cc: salvatore.loreto@ericsson.com; sipping@ietf.org;=20
> Gonzalo.Camarillo@ericsson.com; mary.barnes@nortel.com
> Subject: Re: [Sipping] Service Identification in SIP
>=20
>=20
>=20
> ravishankar.shiroor@wipro.com wrote:
> > Hi Paul,
> >=20
> > IMO, service identification is different from identifying the tool=20
> > that should be used for executing that service on the end=20
> device. In=20
> > the analogy that you have given, e-mail is the service.=20
> > Eudora/gmail/... Are the tools that are used to provide the service.
>=20
> The comment below was about identifying the application in=20
> the end system, not the protocol.
>=20
> In the case of email, the "service" is determined by the=20
> protocol and the address/port of the destination. Nobody has=20
> suggested adding headers to mail messages to identify the=20
> service/application in the end system that should process the=20
> mail message. But here we see proposals to add a header to a=20
> sip message to identify the service/application in the UAS.

I would not mind if the e-mail that is delivering meeting notification
automatically gets delivered to my calendar application. If I have three
different tools handling it, I would set one of them to preferred
calendar application. Similarly if some one sends me a mail with
contacts, I would like it to be picked up by my Contacts application.
I see gmail/Eudora as just tools to execute certain applications. But
the e-mail being delivered should be able to say if it is a calendar
notification/plain mail/.... Application.

Regards,
Ravi.

>=20
> 	Paul
>=20
> > Regards,
> > Ravi.
> >=20
> >=20
> > --
> > =20
> > Ravishankar. G. Shiroor
> > Wipro Technologies, Bangalore.
> > =20
> > ravishankar.shiroor@wipro.com
> > --
> > =20
> >=20
> >> -----Original Message-----
> >> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
> >> Sent: Wednesday, February 07, 2007 10:43 PM
> >> To: Salvatore Loreto (JO/LMF)
> >> Cc: RaviShankar Shiroor (WT01 - Product Engineering Solutions);=20
> >> sipping@ietf.org; Gonzalo.Camarillo@ericsson.com;=20
> >> mary.barnes@nortel.com
> >> Subject: Re: [Sipping] Service Identification in SIP
> >>
> >> I still don't get the intent of identifying the application.=20
> >> It seems to me this is evil - akin to sending an email message and=20
> >> identifying that the recipient must use Outlook to receive it, not=20
> >> Thunderbird, Eudora, gmail, or whatever.
> >>
> >> I also wish somebody could give better examples of a communication=20
> >> service. Chess is too artificial to be believable, and IMO PoC is=20
> >> only a service because of the way the implementation has been done.
> >> Conceptually PoC could be nothing but a special floor control=20
> >> protocol accompanying a voice session. In that case none of the=20
> >> Service mechanism would be needed.
> >>
> >> 	Paul
> >>
> >> Salvatore Loreto (JO/LMF) wrote:
> >>> hi Ravi,
> >>>
> >>> in IMS a Service is a "type of communication defined by a service=20
> >>> definition that specifies the rules and procedures and
> >> allowed media,
> >>> if any, for a specific type of communication and that
> >> utilises the IMS
> >>> enabler. An IMS communication service defines what is
> >> possible within
> >>> a single SIP session or standalone transaction."
> >>> So a Service implies some combination of capabilities and
> >> ways to use
> >>> those capabilities.
> >>> If you want the Service specifies the context for a communication.
> >>>
> >>> Instead the Application define the specific communication
> >> (or if you
> >>> want application) that you want run in the specific=20
> context. So you=20
> >>> should have several different applications designed to run on a=20
> >>> specific environment or that have different performance or
> >> behavior in
> >>> different context (running on top of different service).
> >>>
> >>> Lets go to Jonathan's example of the Chess.
> >>> In his example the Chess is a service, for IMS instead is an=20
> >>> application that you can run on top of different service.
> >>> For example you can run Chess application on top of PoC
> >> service or on
> >>> top of the "standard" service. So let say you run Chess on
> >> top of PoC
> >>> service, then you can use only voice in the way that PoC=20
> specifies.
> >>>
> >>> /sal
> >>>
> >>>
> >>>
> >>>
> >>> On Thu, 2007-02-01 at 16:39 +0530,
> >> ravishankar.shiroor@wipro.com wrote:
> >>>> Hello,
> >>>>
> >>>> IMO, Service Identification is definitely needed and
> >> should be taken
> >>>> up as a WG as proposed below.
> >>>>
> >>>> Further, I think Service Identification scope should be=20
> as per the=20
> >>>> requirement in Jonathan's draft, as I still do not=20
> understand the=20
> >>>> need for having a Communication Service Identifier and another=20
> >>>> Application Identifier as given in the
> >> 3gpp-ics-requirements document.
> >>>> Regards,
> >>>> Ravi.
> >>>>
> >>>> --
> >>>> =20
> >>>> Ravishankar. G. Shiroor
> >>>> Wipro Technologies, Bangalore.
> >>>> =20
> >>>> ravishankar.shiroor@wipro.com
> >>>> --
> >>>> =20
> >>>>
> >>>>> -----Original Message-----
> >>>>> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
> >>>>> Sent: Wednesday, January 31, 2007 8:30 PM
> >>>>> To: sipping
> >>>>> Cc: Mary Barnes
> >>>>> Subject: [Sipping] Service Identification in SIP
> >>>>>
> >>>>> Folks,
> >>>>>
> >>>>> during the last year, there have been a number of=20
> threads on the=20
> >>>>> mailing list discussing service identification in SIP. In=20
> >>>>> particular, the following thread discussed 3GPP
> >> requirements on this
> >>>>> area:
> >>>>>
> >>>>>
> >> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
> >>>>> At a later point, those 3GPP requirements were=20
> documented in the=20
> >>>>> following (expired by now) Internet Draft, which was
> >> presented in a
> >>>>> face-to-face IETF meeting.
> >>>>>
> >>>>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
> >>>>> cs-requirements-00.txt
> >>>>>
> >>>>> Lately, the following draft has been submitted to the archives:
> >>>>>
> >>>>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
> >>>>> rvice-identification-00.txt
> >>>>>
> >>>>> This draft proposes a solution that intends to meet known=20
> >>>>> requirements in this area.
> >>>>>
> >>>>> Given the amount of discussions on issues related to service=20
> >>>>> identification we have seen during the last year, we
> >> believe this is
> >>>>> a fairly important topic for the SIPPING WG to tackle. At some=20
> >>>>> point, we would like to have a milestone with its
> >> associated WG item
> >>>>> dealing with this issue.
> >>>>>
> >>>>> Therefore, we would like to encourage people to read this
> >> draft and
> >>>>> send comments to the list. Additionally, if you believe=20
> that this=20
> >>>>> work is (or is not) important, we would like to hear=20
> that as well.
> >>>>>
> >>>>> Cheers,
> >>>>>
> >>>>> Gonzalo
> >>>>> SIPPING co-chair
> >>>>>
> >>>>>
> >>>>> _______________________________________________
> >>>>> Sipping mailing list
> >> https://www1.ietf.org/mailman/listinfo/sipping
> >>>>> This list is for NEW development of the application of SIP Use=20
> >>>>> sip-implementors@cs.columbia.edu for questions on=20
> current sip Use=20
> >>>>> sip@ietf.org for new developments of core SIP
> >>>>>
> >>>> _______________________________________________
> >>>> Sipping mailing list
> >> https://www1.ietf.org/mailman/listinfo/sipping
> >>>> This list is for NEW development of the application of SIP Use=20
> >>>> sip-implementors@cs.columbia.edu for questions on=20
> current sip Use=20
> >>>> sip@ietf.org for new developments of core SIP
> >>>
> >>> _______________________________________________
> >>> Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> >>> This list is for NEW development of the application of SIP Use=20
> >>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> >>> sip@ietf.org for new developments of core SIP
> >>>
> >=20
> >=20
> > The information contained in this electronic message and=20
> any attachments to this message are intended for the=20
> exclusive use of the addressee(s) and may contain=20
> proprietary, confidential or privileged information. If you=20
> are not the intended recipient, you should not disseminate,=20
> distribute or copy this e-mail. Please notify the sender=20
> immediately and destroy all copies of this message and any=20
> attachments.=20
> >=20
> > WARNING: Computer viruses can be transmitted via email. The=20
> recipient should check this email and any attachments for the=20
> presence of viruses. The company accepts no liability for any=20
> damage caused by any virus transmitted by this email.
> > =20
> > www.wipro.com
> >=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 12 23:11:52 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGp0v-0000Zr-F0; Mon, 12 Feb 2007 23:11:33 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGp0u-0000Zd-7k
	for sipping@ietf.org; Mon, 12 Feb 2007 23:11:32 -0500
Received: from wip-ec-wd.wipro.com ([203.91.193.32])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGp0s-0004YH-5u
	for sipping@ietf.org; Mon, 12 Feb 2007 23:11:32 -0500
Received: from wip-ec-wd.wipro.com (localhost.wipro.com [127.0.0.1])
	by localhost (Postfix) with ESMTP id 360E7205DF
	for <sipping@ietf.org>; Tue, 13 Feb 2007 09:33:36 +0530 (IST)
Received: from blr-ec-bh02.wipro.com (blr-ec-bh02.wipro.com [10.201.50.92])
	by wip-ec-wd.wipro.com (Postfix) with ESMTP id 18E1E205D9
	for <sipping@ietf.org>; Tue, 13 Feb 2007 09:33:36 +0530 (IST)
Received: from blr-m2-msg.wipro.com ([10.116.50.99]) by blr-ec-bh02.wipro.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 09:41:23 +0530
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Subject: RE: [Sipping] Service Identification in SIP
Date: Tue, 13 Feb 2007 09:46:17 +0530
Message-ID: <532B18E13CF9E64380EF5FDDE265E071B61BAB@blr-m2-msg.wipro.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdMW5TvBUF8cSnkQl6kqnCgfOZ63QCyct+g
References: <45C0AEF2.8010101@ericsson.com>
	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
	<1170835642.29763.54.camel@n95.nomadiclab.com>
	<532B18E13CF9E64380EF5FDDE265E071B61978@blr-m2-msg.wipro.com>
	<1171033470.3464.115.camel@localhost.localdomain>
From: <ravishankar.shiroor@wipro.com>
To: <salvatore.loreto@ericsson.com>
X-OriginalArrivalTime: 13 Feb 2007 04:11:23.0597 (UTC)
	FILETIME=[05F97FD0:01C74F25]
X-Spam-Score: 0.3 (/)
X-Scan-Signature: 79668c4a0135b05f663aafe71bc273fd
Cc: sipping@ietf.org, Gonzalo.Camarillo@ericsson.com, mary.barnes@nortel.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1740718433=="
Errors-To: sipping-bounces@ietf.org

This is a multi-part message in MIME format.

--===============1740718433==
Content-class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C74F25.05BF531B"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C74F25.05BF531B
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

Savatore Loreto wrote:

________________________________

	From: Salvatore Loreto (JO/LMF)
[mailto:salvatore.loreto@ericsson.com]=20
	Sent: Friday, February 09, 2007 8:35 PM
	To: RaviShankar Shiroor (WT01 - Product Engineering Solutions)
	Cc: Gonzalo.Camarillo@ericsson.com; sipping@ietf.org;
mary.barnes@nortel.com
	Subject: RE: [Sipping] Service Identification in SIP
=09
=09
	Hi Ravi,
=09
	if you consider the Jonathan draft, he says that=20
=09

		"...chess game that=20
		utilizes voice commands to move pieces.  When one player
says "Queen
		to D7", the software on their phone recognizes this and
moves the
		piece.  At the same time, the user's voice is sent to
the other
		player, where it is both rendered to the user, and
interpreted
		locally in order to move the piece."


	and he suggests that the following urn
=09
	   urn:service:games:voice-chess=20
=09
	could be used to identify the voice chess application.
=09
=09
	I guess that with voice he supposes audio, but nowadays we can
also have voice
	using PoC. Then we need a way to specify which kind of voice we
want: a standard
	full-duplex or a PoC.=20
	Maybe the chess game works well with, or it is designed for a
standard audio but=20
	not for a PoC audio.
	So we need to address also the voice preference we want.
=09
	I like your proposal, having something hierarchical was
something I was thinking too
	I don't know if it is your idea, but with the service urn you
are proposing:
=09
	Urn:service:voice.poc.chess
=09
	I imaging you are asking to play a Chess game, that work
utilizing voice commands,=20
	but specifically voice over PoC.
	[RAVI>> ] Yes. That is what I am proposing. Thus, I do not see
the need for specifying the communication service identifier separately.
Only the service URN (whatever it is called - service
identifier/application identifier) is sufficient to meet the
requirements of application identification on the end device.
	=20
	This is also compliant to the draft-ietf-ecrit-service-urn (see
section 3)
=09

		"Declaration of syntactic structure: The URN consists of
a
		hierarchical service identifier, with a sequence of
labels
		separated by periods.  The left-most label is the most
significant
		one and is called 'top-level service', while names to
the right
		are called 'sub-services'.  The set of allowable
characters is the
		same as that for domain names [1] and a subset of the
labels
		allowed in [5].  Labels are case-insensitive and MUST be
specified
		in all lower-case.  For any given service URN,
service-identifiers
		can be removed right-to-left and the resulting URN is
still valid,
		referring a more generic service.  In other words, if a
service
		'x.y.z' exists, the URNs 'x' and 'x.y' are also valid
service
		URNs."

=09
	Regards
	sal
	=20

	On Fri, 2007-02-09 at 13:44 +0530, ravishankar.shiroor@wipro.com
wrote:=20

		Hi Salvatore,
	=09
		As an end user, poc is also an application that I use
and chess is also
		an application that I can use. The way I see it, in your
example, it
		will be running chess application on top of PoC
application.
	=09
		So the service urn for this would be something like:
	=09
		Urn:service:voice.poc.chess
	=09
		Regards,
		Ravi.
	=09
		--
		=20
		Ravishankar. G. Shiroor
		Wipro Technologies, Bangalore.
		=20
		ravishankar.shiroor@wipro.com
		--
		=20
	=09
		> -----Original Message-----
		> From: Salvatore Loreto (JO/LMF)=20
		> [mailto:salvatore.loreto@ericsson.com]=20
		> Sent: Wednesday, February 07, 2007 1:37 PM
		> To: RaviShankar Shiroor (WT01 - Product Engineering
Solutions)
		> Cc: Gonzalo.Camarillo@ericsson.com; sipping@ietf.org;=20
		> mary.barnes@nortel.com
		> Subject: RE: [Sipping] Service Identification in SIP
		>=20
		> hi Ravi,
		>=20
		> in IMS a Service is a "type of communication defined
by a=20
		> service definition that specifies the rules and
procedures=20
		> and allowed media, if any, for a specific type of=20
		> communication and that utilises the IMS enabler. An
IMS=20
		> communication service defines what is possible within
a=20
		> single SIP session or standalone transaction."
		> So a Service implies some combination of capabilities
and=20
		> ways to use those capabilities.
		> If you want the Service specifies the context for a
communication.
		>=20
		> Instead the Application define the specific
communication (or=20
		> if you want application) that you want run in the
specific=20
		> context. So you should have several different
applications=20
		> designed to run on a specific environment or that have

		> different performance or behavior in different context

		> (running on top of different service).
		>=20
		> Lets go to Jonathan's example of the Chess.
		> In his example the Chess is a service, for IMS instead
is an=20
		> application that you can run on top of different
service.
		> For example you can run Chess application on top of
PoC=20
		> service or on top of the "standard" service. So let
say you=20
		> run Chess on top of PoC service, then you can use only
voice=20
		> in the way that PoC specifies.
		>=20
		> /sal
		>=20
		>=20
		>=20
		>=20
		> On Thu, 2007-02-01 at 16:39 +0530,=20
		> ravishankar.shiroor@wipro.com wrote:
		> > Hello,
		> >=20
		> > IMO, Service Identification is definitely needed and
should=20
		> be taken=20
		> > up as a WG as proposed below.
		> >=20
		> > Further, I think Service Identification scope should
be as per the=20
		> > requirement in Jonathan's draft, as I still do not=20
		> understand the need=20
		> > for having a Communication Service Identifier and
another=20
		> Application=20
		> > Identifier as given in the 3gpp-ics-requirements
document.
		> >=20
		> > Regards,
		> > Ravi.
		> >=20
		> > --
		> > =20
		> > Ravishankar. G. Shiroor
		> > Wipro Technologies, Bangalore.
		> > =20
		> > ravishankar.shiroor@wipro.com
		> > --
		> > =20
		> >=20
		> > > -----Original Message-----
		> > > From: Gonzalo Camarillo
[mailto:Gonzalo.Camarillo@ericsson.com]
		> > > Sent: Wednesday, January 31, 2007 8:30 PM
		> > > To: sipping
		> > > Cc: Mary Barnes
		> > > Subject: [Sipping] Service Identification in SIP
		> > >=20
		> > > Folks,
		> > >=20
		> > > during the last year, there have been a number of
threads on the=20
		> > > mailing list discussing service identification in
SIP. In=20
		> > > particular, the following thread discussed 3GPP=20
		> requirements on this=20
		> > > area:
		> > >=20
		> > >=20
		>
http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
		> > >=20
		> > > At a later point, those 3GPP requirements were
documented in the=20
		> > > following (expired by now) Internet Draft, which
was=20
		> presented in a=20
		> > > face-to-face IETF meeting.
		> > >=20
		> > >
http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
		> > > cs-requirements-00.txt
		> > >=20
		> > > Lately, the following draft has been submitted to
the archives:
		> > >=20
		> > >
http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
		> > > rvice-identification-00.txt
		> > >=20
		> > > This draft proposes a solution that intends to
meet known=20
		> > > requirements in this area.
		> > >=20
		> > > Given the amount of discussions on issues related
to service=20
		> > > identification we have seen during the last year,
we=20
		> believe this is=20
		> > > a fairly important topic for the SIPPING WG to
tackle. At some=20
		> > > point, we would like to have a milestone with its=20
		> associated WG item=20
		> > > dealing with this issue.
		> > >=20
		> > > Therefore, we would like to encourage people to
read this=20
		> draft and=20
		> > > send comments to the list. Additionally, if you
believe that this=20
		> > > work is (or is not) important, we would like to
hear that as well.
		> > >=20
		> > > Cheers,
		> > >=20
		> > > Gonzalo
		> > > SIPPING co-chair
		> > >=20
		> > >=20
		> > > _______________________________________________
		> > > Sipping mailing list =20
		> https://www1.ietf.org/mailman/listinfo/sipping
		> > > This list is for NEW development of the
application of SIP Use=20
		> > > sip-implementors@cs.columbia.edu for questions on
current sip Use=20
		> > > sip@ietf.org for new developments of core SIP
		> > >=20
		> >=20
		> > _______________________________________________
		> > Sipping mailing list
https://www1.ietf.org/mailman/listinfo/sipping
		> > This list is for NEW development of the application
of SIP Use=20
		> > sip-implementors@cs.columbia.edu for questions on
current sip Use=20
		> > sip@ietf.org for new developments of core SIP
		>=20
		>=20
	=09
	=09
		The information contained in this electronic message and
any attachments to this message are intended for the exclusive use of
the addressee(s) and may contain proprietary, confidential or privileged
information. If you are not the intended recipient, you should not
disseminate, distribute or copy this e-mail. Please notify the sender
immediately and destroy all copies of this message and any attachments.=20
	=09
		WARNING: Computer viruses can be transmitted via email.
The recipient should check this email and any attachments for the
presence of viruses. The company accepts no liability for any damage
caused by any virus transmitted by this email.
		=20
		www.wipro.com


------_=_NextPart_001_01C74F25.05BF531B
Content-Type: text/html;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; =
charset=3Dus-ascii">
<META content=3D"MSHTML 6.00.2900.3020" name=3DGENERATOR></HEAD>
<BODY>
<DIV dir=3Dltr align=3Dleft><FONT face=3DArial><FONT=20
size=3D2>Savatore&nbsp;Loreto&nbsp;wrote:</FONT></FONT></DIV>
<BLOCKQUOTE dir=3Dltr=20
style=3D"PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #0000ff 2px =
solid; MARGIN-RIGHT: 0px">
  <DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr align=3Dleft>
  <HR tabIndex=3D-1>
  <FONT face=3DTahoma size=3D2><B>From:</B> Salvatore Loreto (JO/LMF)=20
  [mailto:salvatore.loreto@ericsson.com] <BR><B>Sent:</B> Friday, =
February 09,=20
  2007 8:35 PM<BR><B>To:</B> RaviShankar Shiroor (WT01 - Product =
Engineering=20
  Solutions)<BR><B>Cc:</B> Gonzalo.Camarillo@ericsson.com; =
sipping@ietf.org;=20
  mary.barnes@nortel.com<BR><B>Subject:</B> RE: [Sipping] Service =
Identification=20
  in SIP<BR></FONT><BR></DIV>
  <DIV></DIV>Hi Ravi,<BR><BR>if you consider the Jonathan draft, he says =
that=20
  <BR>
  <BLOCKQUOTE>"<I>...chess game that</I> <PRE><I>utilizes voice commands =
to move pieces.&nbsp; When one player says "Queen</I>
<I>to D7", the software on their phone recognizes this and moves the</I>
<I>piece.&nbsp; At the same time, the user's voice is sent to the =
other</I>
<I>player, where it is both rendered to the user, and interpreted</I>
<I>locally in order to move the piece.</I>"
</PRE></BLOCKQUOTE><BR>and he suggests that the following=20
  urn<BR><BR>&nbsp;&nbsp; urn:service:games:voice-chess <BR><BR>could be =
used to=20
  identify the voice chess application.<BR><BR><PRE>I guess that with =
voice he supposes audio, but nowadays we can also have voice
using PoC. Then we need a way to specify which kind of voice we want: a =
standard
full-duplex or a PoC.=20
Maybe the chess game works well with, or it is designed for a standard =
audio but=20
not for a PoC audio.
So we need to address also the voice preference we want.

I like your proposal, having something hierarchical was something I was =
thinking too
I don't know if it is your idea, but with the service urn you are =
proposing:

Urn:service:voice.poc.chess

I imaging you are asking to play a Chess game, that work utilizing voice =
commands,=20
but specifically voice over PoC.<BR><SPAN =
class=3D343551304-13022007><FONT face=3DArial color=3D#0000ff =
size=3D2>[RAVI&gt;&gt; ]&nbsp;Yes. That is what I am proposing. Thus, I =
do not see the need for specifying the communication service identifier =
separately. Only the service URN (whatever it is called - service =
identifier/application identifier) is sufficient to meet the =
requirements of application identification on the end =
device.</FONT></SPAN></PRE><PRE><SPAN =
class=3D343551304-13022007>&nbsp;</SPAN>
This is also compliant to the draft-ietf-ecrit-service-urn (see section =
3)

</PRE>
  <BLOCKQUOTE><PRE>"<I>Declaration of syntactic structure: The URN =
consists of a</I>
<I>hierarchical service identifier, with a sequence of labels</I>
<I>separated by periods.&nbsp; The left-most label is the most =
significant</I>
<I>one and is called 'top-level service', while names to the right</I>
<I>are called 'sub-services'.&nbsp; The set of allowable characters is =
the</I>
<I>same as that for domain names [1] and a subset of the labels</I>
<I>allowed in [5].&nbsp; Labels are case-insensitive and MUST be =
specified</I>
<I>in all lower-case.&nbsp; For any given service URN, =
service-identifiers</I>
<I>can be removed right-to-left and the resulting URN is still =
valid,</I>
<I>referring a more generic service.&nbsp; In other words, if a =
service</I>
<I>'x.y.z' exists, the URNs 'x' and 'x.y' are also valid service</I>
<I>URNs.</I>"
</PRE></BLOCKQUOTE><PRE>
Regards
sal
=20
</PRE><BR>On Fri, 2007-02-09 at 13:44 +0530, =
ravishankar.shiroor@wipro.com=20
  wrote:=20
  <BLOCKQUOTE TYPE=3D"CITE"><PRE>Hi Salvatore,

As an end user, poc is also an application that I use and chess is also
an application that I can use. The way I see it, in your example, it
will be running chess application on top of PoC application.

So the service urn for this would be something like:

Urn:service:voice.poc.chess

Regards,
Ravi.

--
=20
Ravishankar. G. Shiroor
Wipro Technologies, Bangalore.
=20
<FONT color=3D#000000><A =
href=3D"mailto:ravishankar.shiroor@wipro.com">ravishankar.shiroor@wipro.c=
om</A></FONT>
<FONT color=3D#000000>--</FONT>
<FONT color=3D#000000> </FONT>

<FONT color=3D#000000>&gt; -----Original Message-----</FONT>
<FONT color=3D#000000>&gt; From: Salvatore Loreto (JO/LMF) </FONT>
<FONT color=3D#000000>&gt; [mailto:<A =
href=3D"mailto:salvatore.loreto@ericsson.com">salvatore.loreto@ericsson.c=
om</A>] </FONT>
<FONT color=3D#000000>&gt; Sent: Wednesday, February 07, 2007 1:37 =
PM</FONT>
<FONT color=3D#000000>&gt; To: RaviShankar Shiroor (WT01 - Product =
Engineering Solutions)</FONT>
<FONT color=3D#000000>&gt; Cc: <A =
href=3D"mailto:Gonzalo.Camarillo@ericsson.com">Gonzalo.Camarillo@ericsson=
.com</A>; <A href=3D"mailto:sipping@ietf.org">sipping@ietf.org</A>; =
</FONT>
<FONT color=3D#000000>&gt; <A =
href=3D"mailto:mary.barnes@nortel.com">mary.barnes@nortel.com</A></FONT>
<FONT color=3D#000000>&gt; Subject: RE: [Sipping] Service Identification =
in SIP</FONT>
<FONT color=3D#000000>&gt; </FONT>
<FONT color=3D#000000>&gt; hi Ravi,</FONT>
<FONT color=3D#000000>&gt; </FONT>
<FONT color=3D#000000>&gt; in IMS a Service is a "type of communication =
defined by a </FONT>
<FONT color=3D#000000>&gt; service definition that specifies the rules =
and procedures </FONT>
<FONT color=3D#000000>&gt; and allowed media, if any, for a specific =
type of </FONT>
<FONT color=3D#000000>&gt; communication and that utilises the IMS =
enabler. An IMS </FONT>
<FONT color=3D#000000>&gt; communication service defines what is =
possible within a </FONT>
<FONT color=3D#000000>&gt; single SIP session or standalone =
transaction."</FONT>
<FONT color=3D#000000>&gt; So a Service implies some combination of =
capabilities and </FONT>
<FONT color=3D#000000>&gt; ways to use those capabilities.</FONT>
<FONT color=3D#000000>&gt; If you want the Service specifies the context =
for a communication.</FONT>
<FONT color=3D#000000>&gt; </FONT>
<FONT color=3D#000000>&gt; Instead the Application define the specific =
communication (or </FONT>
<FONT color=3D#000000>&gt; if you want application) that you want run in =
the specific </FONT>
<FONT color=3D#000000>&gt; context. So you should have several different =
applications </FONT>
<FONT color=3D#000000>&gt; designed to run on a specific environment or =
that have </FONT>
<FONT color=3D#000000>&gt; different performance or behavior in =
different context </FONT>
<FONT color=3D#000000>&gt; (running on top of different service).</FONT>
<FONT color=3D#000000>&gt; </FONT>
<FONT color=3D#000000>&gt; Lets go to Jonathan's example of the =
Chess.</FONT>
<FONT color=3D#000000>&gt; In his example the Chess is a service, for =
IMS instead is an </FONT>
<FONT color=3D#000000>&gt; application that you can run on top of =
different service.</FONT>
<FONT color=3D#000000>&gt; For example you can run Chess application on =
top of PoC </FONT>
<FONT color=3D#000000>&gt; service or on top of the "standard" service. =
So let say you </FONT>
<FONT color=3D#000000>&gt; run Chess on top of PoC service, then you can =
use only voice </FONT>
<FONT color=3D#000000>&gt; in the way that PoC specifies.</FONT>
<FONT color=3D#000000>&gt; </FONT>
<FONT color=3D#000000>&gt; /sal</FONT>
<FONT color=3D#000000>&gt; </FONT>
<FONT color=3D#000000>&gt; </FONT>
<FONT color=3D#000000>&gt; </FONT>
<FONT color=3D#000000>&gt; </FONT>
<FONT color=3D#000000>&gt; On Thu, 2007-02-01 at 16:39 +0530, </FONT>
<FONT color=3D#000000>&gt; <A =
href=3D"mailto:ravishankar.shiroor@wipro.com">ravishankar.shiroor@wipro.c=
om</A> wrote:</FONT>
<FONT color=3D#000000>&gt; &gt; Hello,</FONT>
<FONT color=3D#000000>&gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; IMO, Service Identification is =
definitely needed and should </FONT>
<FONT color=3D#000000>&gt; be taken </FONT>
<FONT color=3D#000000>&gt; &gt; up as a WG as proposed below.</FONT>
<FONT color=3D#000000>&gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; Further, I think Service Identification =
scope should be as per the </FONT>
<FONT color=3D#000000>&gt; &gt; requirement in Jonathan's draft, as I =
still do not </FONT>
<FONT color=3D#000000>&gt; understand the need </FONT>
<FONT color=3D#000000>&gt; &gt; for having a Communication Service =
Identifier and another </FONT>
<FONT color=3D#000000>&gt; Application </FONT>
<FONT color=3D#000000>&gt; &gt; Identifier as given in the =
3gpp-ics-requirements document.</FONT>
<FONT color=3D#000000>&gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; Regards,</FONT>
<FONT color=3D#000000>&gt; &gt; Ravi.</FONT>
<FONT color=3D#000000>&gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; --</FONT>
<FONT color=3D#000000>&gt; &gt;  </FONT>
<FONT color=3D#000000>&gt; &gt; Ravishankar. G. Shiroor</FONT>
<FONT color=3D#000000>&gt; &gt; Wipro Technologies, Bangalore.</FONT>
<FONT color=3D#000000>&gt; &gt;  </FONT>
<FONT color=3D#000000>&gt; &gt; <A =
href=3D"mailto:ravishankar.shiroor@wipro.com">ravishankar.shiroor@wipro.c=
om</A></FONT>
<FONT color=3D#000000>&gt; &gt; --</FONT>
<FONT color=3D#000000>&gt; &gt;  </FONT>
<FONT color=3D#000000>&gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; -----Original Message-----</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; From: Gonzalo Camarillo [mailto:<A =
href=3D"mailto:Gonzalo.Camarillo@ericsson.com">Gonzalo.Camarillo@ericsson=
.com</A>]</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; Sent: Wednesday, January 31, 2007 =
8:30 PM</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; To: sipping</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; Cc: Mary Barnes</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; Subject: [Sipping] Service =
Identification in SIP</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; Folks,</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; during the last year, there have =
been a number of threads on the </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; mailing list discussing service =
identification in SIP. In </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; particular, the following thread =
discussed 3GPP </FONT>
<FONT color=3D#000000>&gt; requirements on this </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; area:</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; </FONT>
<FONT color=3D#000000>&gt; <A =
href=3D"http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.ht=
ml">http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html</=
A></FONT>
<FONT color=3D#000000>&gt; &gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; At a later point, those 3GPP =
requirements were documented in the </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; following (expired by now) Internet =
Draft, which was </FONT>
<FONT color=3D#000000>&gt; presented in a </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; face-to-face IETF meeting.</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; <A =
href=3D"http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i">h=
ttp://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i</A></FONT>
<FONT color=3D#000000>&gt; &gt; &gt; cs-requirements-00.txt</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; Lately, the following draft has =
been submitted to the archives:</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; <A =
href=3D"http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se">h=
ttp://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se</A></FONT>
<FONT color=3D#000000>&gt; &gt; &gt; rvice-identification-00.txt</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; This draft proposes a solution that =
intends to meet known </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; requirements in this area.</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; Given the amount of discussions on =
issues related to service </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; identification we have seen during =
the last year, we </FONT>
<FONT color=3D#000000>&gt; believe this is </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; a fairly important topic for the =
SIPPING WG to tackle. At some </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; point, we would like to have a =
milestone with its </FONT>
<FONT color=3D#000000>&gt; associated WG item </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; dealing with this issue.</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; Therefore, we would like to =
encourage people to read this </FONT>
<FONT color=3D#000000>&gt; draft and </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; send comments to the list. =
Additionally, if you believe that this </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; work is (or is not) important, we =
would like to hear that as well.</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; Cheers,</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; Gonzalo</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; SIPPING co-chair</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; =
_______________________________________________</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; Sipping mailing list  </FONT>
<FONT color=3D#000000>&gt; <A =
href=3D"https://www1.ietf.org/mailman/listinfo/sipping">https://www1.ietf=
.org/mailman/listinfo/sipping</A></FONT>
<FONT color=3D#000000>&gt; &gt; &gt; This list is for NEW development of =
the application of SIP Use </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; <A =
href=3D"mailto:sip-implementors@cs.columbia.edu">sip-implementors@cs.colu=
mbia.edu</A> for questions on current sip Use </FONT>
<FONT color=3D#000000>&gt; &gt; &gt; <A =
href=3D"mailto:sip@ietf.org">sip@ietf.org</A> for new developments of =
core SIP</FONT>
<FONT color=3D#000000>&gt; &gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; </FONT>
<FONT color=3D#000000>&gt; &gt; =
_______________________________________________</FONT>
<FONT color=3D#000000>&gt; &gt; Sipping mailing list  <A =
href=3D"https://www1.ietf.org/mailman/listinfo/sipping">https://www1.ietf=
.org/mailman/listinfo/sipping</A></FONT>
<FONT color=3D#000000>&gt; &gt; This list is for NEW development of the =
application of SIP Use </FONT>
<FONT color=3D#000000>&gt; &gt; <A =
href=3D"mailto:sip-implementors@cs.columbia.edu">sip-implementors@cs.colu=
mbia.edu</A> for questions on current sip Use </FONT>
<FONT color=3D#000000>&gt; &gt; <A =
href=3D"mailto:sip@ietf.org">sip@ietf.org</A> for new developments of =
core SIP</FONT>
<FONT color=3D#000000>&gt; </FONT>
<FONT color=3D#000000>&gt; </FONT>


<FONT color=3D#000000>The information contained in this electronic =
message and any attachments to this message are intended for the =
exclusive use of the addressee(s) and may contain proprietary, =
confidential or privileged information. If you are not the intended =
recipient, you should not disseminate, distribute or copy this e-mail. =
Please notify the sender immediately and destroy all copies of this =
message and any attachments. </FONT>

<FONT color=3D#000000>WARNING: Computer viruses can be transmitted via =
email. The recipient should check this email and any attachments for the =
presence of viruses. The company accepts no liability for any damage =
caused by any virus transmitted by this email.</FONT>
<FONT color=3D#000000> </FONT>
<FONT color=3D#000000><A =
href=3D"http://www.wipro.com">www.wipro.com</A></FONT>
</PRE></BLOCKQUOTE></BLOCKQUOTE></BODY></HTML>

------_=_NextPart_001_01C74F25.05BF531B--


--===============1740718433==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1740718433==--




From sipping-bounces@ietf.org Mon Feb 12 23:37:19 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGpPP-0005KQ-Ml; Mon, 12 Feb 2007 23:36:51 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGpPN-0005KL-GS
	for sipping@ietf.org; Mon, 12 Feb 2007 23:36:49 -0500
Received: from rtp-iport-1.cisco.com ([64.102.122.148])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGpPM-0008U6-2e
	for sipping@ietf.org; Mon, 12 Feb 2007 23:36:49 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by rtp-iport-1.cisco.com with ESMTP; 12 Feb 2007 23:36:48 -0500
X-IronPort-AV: i="4.14,161,1170651600"; 
	d="scan'208"; a="52872980:sNHT55558784"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1D4alQD026243; 
	Mon, 12 Feb 2007 23:36:47 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1D4alOA014803; 
	Mon, 12 Feb 2007 23:36:47 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 12 Feb 2007 23:36:47 -0500
Received: from [10.86.240.17] ([10.86.240.17]) by xfe-rtp-201.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 12 Feb 2007 23:36:46 -0500
Message-ID: <45D1405E.30300@cisco.com>
Date: Mon, 12 Feb 2007 23:36:46 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Dale.Worley@comcast.net
Subject: Re: [Sipping] Service Identification in SIP
References: <45C0AEF2.8010101@ericsson.com><532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com><1170835642.29763.54.camel@n95.nomadiclab.com><45CA08B2.4050603@cisco.com><532B18E13CF9E64380EF5FDDE265E071B6192E@blr-m2-msg.wipro.com><45CC7697.1060204@cisco.com><1171031930.3464.92.camel@localhost.localdomain>	<200702110449.l1B4nfk2020329@dragon.ariadne.com>	<24CCCC428EFEA2469BF046DB3C7A8D224538F2@namail5.corp.adobe.com>
	<200702122256.l1CMueqK020621@dragon.ariadne.com>
In-Reply-To: <200702122256.l1CMueqK020621@dragon.ariadne.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 13 Feb 2007 04:36:46.0808 (UTC)
	FILETIME=[91E10D80:01C74F28]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=6307; t=1171341407;
	x=1172205407; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20 |To:=20Dale.Worley@comcast.net;
	bh=8QcKb0koIwUibyjg+kXAYvN+MAvF5AIb1OoiAzWsKAI=;
	b=IcttHZQhhj4XNtEqngAkDZwf21nQr0auvNwR70SUBzF8hSTGET8pOvwlbF0JZIseMWrjmEyX
	b16Rf1OUKWwcFcdPvdWwFeqOXLt4umfXq9cwipz+ybd/x1CmepL7d5CZ;
Authentication-Results: rtp-dkim-1; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: b22590c27682ace61775ee7b453b40d3
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Dale,

What you describe is largely consistent with the infrastruction of IMS, 
although there are a million ways that infrastructure could be applied.

Dale.Worley@comcast.net wrote:
>    From: "Henry Sinnreich" <hsinnrei@adobe.com>
> 
>    >Could somebody please lay out an example showing the utility of service
>    >identification?
> 
>    Sure: The service identification is most likely a new charging
>    contraption invented in a marketing department. As such, any protocol
>    significance cannot be found nor explained.
> 
> I think that it's more complicated than that.  If it was only a
> matter of billing based on the services used, a fairly simple scheme
> for monitoring the communication stream and generating billing entries
> would be simpler and more efficient.
> 
> Let me throw out an idea (which I've probably stolen from someone
> else), one which I recall based on the BT announcement.
> 
> AFAICT, for many service providers, an incoming call (regardless of
> the destination E.164 number) will get routed to one of a set of
> "application servers" (AS), each one of which handles a particular
> "service".  The application server is responsible for managing the
> final link to the phone.  From an Internet perspective, what is really
> happening is that the UA is partitioned into two parts, part on the AS
> and part on the real phone.

It is interesting to view the partitioning of the function between two 
servers as you suggest. I think it would be much more straightforward if 
such an AS was actually treated as a first class UAS that has registered 
and is selected in the same way that other registered UAs are selected 
for a call. Notably, that callerprefs could be used to make the choice.

Unfortunately, IMS has chosen to use a different algorithm to decide how 
to involve ASs.

> This has some advantages, in particular that a new service can be
> implemented by buying and installing an AS, as long as the AS-to-phone
> link provides adequate facilities for the lowest-level UI functions.
> 
> There are disadvantages, too, particularly that both ends of a call
> have to agree on the "service" being used, and that each service in
> effect defines a subset network, consisting of the ASs for that
> service, and all phones which are serviced by an AS for that service.
> Worse, services cannot be mix-and-matched in any way, even if both
> users would be willing to pay the fees to support all of the needed
> infrastructure.
> 
> In some cases, this sort of architecture may be difficult to avoid.
> For instance, the push-to-talk (PTT) feature.  The way it's
> implemented is rather a cheat -- The call is first sent to an AS.  If
> the AS thinks that the callee's phone is registered with the network,
> and the last status update from the callee's phone says that it will
> answer PTT calls immediately, then the AS sends a "call connected"
> indication to the caller before it has achieved a voice channel
> connection with the callee phone.  It then starts buffering the
> caller's audio for the several seconds it needs to gain a voice
> connection to the callee phone.  Once it gets a connection, it plays
> the buffered audio, using various tricks to make playback faster than
> real-time.
> 
> But given the limitation that the last link to the callee's phone does
> not impelment "immediate connection", there is no way provide PTT
> without an AS.

As I noted above, I think this can be handled better by treating the AS 
as the actual target. But I think this is a deviant case, and not a good 
one to generalize into a general purpose architecture.

> I think the architectural challenge is how to support this sort of
> network but without the nasty effects of "siloing":

Unless there are more applications that share this problem, it may be 
easier to simply treat it as a special case and wait for technology to 
obsolete it. I suspect that faster networks and more bandwidth will 
solve this, so that it can be handled e2e and viewed simply as special 
floor control and UI. If we could get there then it starts to look just 
like IM.

> - If a caller requests to use a particular feature, and the callee is
>   unable or unwilling to support that feature, and the caller is
>   willing to proceed without using the feature, the call should
>   proceed without the feature.
> 
> - A call can request multiple features, and it will proceed using
>   whatever subset of them are supported by the callee.
> 
> I don't see this as intrinsically impossible to achieve, even if every
> feature is implemented by ASs.  Assuming the features are reasonably
> well-behaved, it should be possible to chain a call through several
> ASs before reaching the callee phone.

Actually in IMSs case, the call can in fact be chained to multiple ASs. 
But of course they have to be applied in some order.

Its far from clear that services can in general be composed in this way 
and still work.

> (This may require a separate charge for the use of each AS, but that's
> not unfair -- by hypothesis, the ASs implement features that the
> callee phones do not, and someone has to pay for the ASs.)
> 
> It would be helpful if each INVITE declared all of the features that
> would ever be used during the call.  In some cases, the INVITE already
> contains the needed information (particularly in regard to the media
> features).  But it seems that we should avoid this requirement if
> possible.  OTOH, it seems plausible that an AS could be inserted into
> the service chain during a dialog -- I've plotted out call flows by
> which a B2BUA can insert itself into an existing dialog using
> INVITE/Join, and a cluster of ASs in a closed zone of a network can
> perform more dexterous rerouting tricks.

Lots of things are possible, but it will be a challenge to firgure out 
what to do in a particular circumstance.

> Dale
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 12 23:46:20 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGpYX-0002TF-Eu; Mon, 12 Feb 2007 23:46:17 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGpYV-0002TA-I7
	for sipping@ietf.org; Mon, 12 Feb 2007 23:46:15 -0500
Received: from rtp-iport-1.cisco.com ([64.102.122.148])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGpYT-0002Y1-Va
	for sipping@ietf.org; Mon, 12 Feb 2007 23:46:15 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by rtp-iport-1.cisco.com with ESMTP; 12 Feb 2007 23:46:13 -0500
X-IronPort-AV: i="4.14,161,1170651600"; 
	d="scan'208"; a="52873674:sNHT63881182"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1D4kDtx029618; 
	Mon, 12 Feb 2007 23:46:13 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1D4kDOA016076; 
	Mon, 12 Feb 2007 23:46:13 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 12 Feb 2007 23:46:13 -0500
Received: from [10.86.240.17] ([10.86.240.17]) by xfe-rtp-201.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 12 Feb 2007 23:46:12 -0500
Message-ID: <45D14293.30109@cisco.com>
Date: Mon, 12 Feb 2007 23:46:11 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: ravishankar.shiroor@wipro.com
Subject: Re: [Sipping] Service Identification in SIP
References: <45C0AEF2.8010101@ericsson.com>	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
	<1170835642.29763.54.camel@n95.nomadiclab.com>
	<45CA08B2.4050603@cisco.com>
	<532B18E13CF9E64380EF5FDDE265E071B6192E@blr-m2-msg.wipro.com>
	<45CC7697.1060204@cisco.com>
	<532B18E13CF9E64380EF5FDDE265E071B61B9E@blr-m2-msg.wipro.com>
In-Reply-To: <532B18E13CF9E64380EF5FDDE265E071B61B9E@blr-m2-msg.wipro.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 13 Feb 2007 04:46:12.0620 (UTC)
	FILETIME=[E3211CC0:01C74F29]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=10421; t=1171341973;
	x=1172205973; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20 |To:=20ravishankar.shiroor@wipro.com;
	bh=BDzJbrrxkWuyDzaF4kSxYExwM2PxT1qLLPlM1vkMzH8=;
	b=jhxZzVE1djkXCwwwqE3nNiD7hhwvGYqw2Kbg0bQMMuPvqEyZK6DHgkbMqQKUzh64d4770TPl
	OHjV5UJwpGj/kQT6Ugdyr1GQ4tcvQu6pzdFEl9m1GY5ZWCLpK+LIJsBc;
Authentication-Results: rtp-dkim-1; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: e178fd6cb61ffb6940cd878e7fea8606
Cc: sipping@ietf.org, Gonzalo.Camarillo@ericsson.com, mary.barnes@nortel.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



ravishankar.shiroor@wipro.com wrote:
> Paul Kyzivat wrote:
>> -----Original Message-----
>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
>> Sent: Friday, February 09, 2007 6:57 PM
>> To: RaviShankar Shiroor (WT01 - Product Engineering Solutions)
>> Cc: salvatore.loreto@ericsson.com; sipping@ietf.org; 
>> Gonzalo.Camarillo@ericsson.com; mary.barnes@nortel.com
>> Subject: Re: [Sipping] Service Identification in SIP
>>
>>
>>
>> ravishankar.shiroor@wipro.com wrote:
>>> Hi Paul,
>>>
>>> IMO, service identification is different from identifying the tool 
>>> that should be used for executing that service on the end 
>> device. In 
>>> the analogy that you have given, e-mail is the service. 
>>> Eudora/gmail/... Are the tools that are used to provide the service.
>> The comment below was about identifying the application in 
>> the end system, not the protocol.
>>
>> In the case of email, the "service" is determined by the 
>> protocol and the address/port of the destination. Nobody has 
>> suggested adding headers to mail messages to identify the 
>> service/application in the end system that should process the 
>> mail message. But here we see proposals to add a header to a 
>> sip message to identify the service/application in the UAS.
> 
> I would not mind if the e-mail that is delivering meeting notification
> automatically gets delivered to my calendar application. If I have three
> different tools handling it, I would set one of them to preferred
> calendar application. Similarly if some one sends me a mail with
> contacts, I would like it to be picked up by my Contacts application.
> I see gmail/Eudora as just tools to execute certain applications. But
> the e-mail being delivered should be able to say if it is a calendar
> notification/plain mail/.... Application.

Well, in this particular case,I would expect the meeting notification to 
be carried as one mime type in the body, and the contacts to be carried 
as another mime type. And its expected that these will be dispatched to 
a suitable application by the receiving email application *after* it is 
received.

Note that the mime type defines the information carried, not the 
application that will process it.

For SIP, the same thing is possible with body types. And further, the 
different types of media streams can be handed off to different 
applications.

	Paul

> Regards,
> Ravi.
> 
>> 	Paul
>>
>>> Regards,
>>> Ravi.
>>>
>>>
>>> --
>>>  
>>> Ravishankar. G. Shiroor
>>> Wipro Technologies, Bangalore.
>>>  
>>> ravishankar.shiroor@wipro.com
>>> --
>>>  
>>>
>>>> -----Original Message-----
>>>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
>>>> Sent: Wednesday, February 07, 2007 10:43 PM
>>>> To: Salvatore Loreto (JO/LMF)
>>>> Cc: RaviShankar Shiroor (WT01 - Product Engineering Solutions); 
>>>> sipping@ietf.org; Gonzalo.Camarillo@ericsson.com; 
>>>> mary.barnes@nortel.com
>>>> Subject: Re: [Sipping] Service Identification in SIP
>>>>
>>>> I still don't get the intent of identifying the application. 
>>>> It seems to me this is evil - akin to sending an email message and 
>>>> identifying that the recipient must use Outlook to receive it, not 
>>>> Thunderbird, Eudora, gmail, or whatever.
>>>>
>>>> I also wish somebody could give better examples of a communication 
>>>> service. Chess is too artificial to be believable, and IMO PoC is 
>>>> only a service because of the way the implementation has been done.
>>>> Conceptually PoC could be nothing but a special floor control 
>>>> protocol accompanying a voice session. In that case none of the 
>>>> Service mechanism would be needed.
>>>>
>>>> 	Paul
>>>>
>>>> Salvatore Loreto (JO/LMF) wrote:
>>>>> hi Ravi,
>>>>>
>>>>> in IMS a Service is a "type of communication defined by a service 
>>>>> definition that specifies the rules and procedures and
>>>> allowed media,
>>>>> if any, for a specific type of communication and that
>>>> utilises the IMS
>>>>> enabler. An IMS communication service defines what is
>>>> possible within
>>>>> a single SIP session or standalone transaction."
>>>>> So a Service implies some combination of capabilities and
>>>> ways to use
>>>>> those capabilities.
>>>>> If you want the Service specifies the context for a communication.
>>>>>
>>>>> Instead the Application define the specific communication
>>>> (or if you
>>>>> want application) that you want run in the specific 
>> context. So you 
>>>>> should have several different applications designed to run on a 
>>>>> specific environment or that have different performance or
>>>> behavior in
>>>>> different context (running on top of different service).
>>>>>
>>>>> Lets go to Jonathan's example of the Chess.
>>>>> In his example the Chess is a service, for IMS instead is an 
>>>>> application that you can run on top of different service.
>>>>> For example you can run Chess application on top of PoC
>>>> service or on
>>>>> top of the "standard" service. So let say you run Chess on
>>>> top of PoC
>>>>> service, then you can use only voice in the way that PoC 
>> specifies.
>>>>> /sal
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> On Thu, 2007-02-01 at 16:39 +0530,
>>>> ravishankar.shiroor@wipro.com wrote:
>>>>>> Hello,
>>>>>>
>>>>>> IMO, Service Identification is definitely needed and
>>>> should be taken
>>>>>> up as a WG as proposed below.
>>>>>>
>>>>>> Further, I think Service Identification scope should be 
>> as per the 
>>>>>> requirement in Jonathan's draft, as I still do not 
>> understand the 
>>>>>> need for having a Communication Service Identifier and another 
>>>>>> Application Identifier as given in the
>>>> 3gpp-ics-requirements document.
>>>>>> Regards,
>>>>>> Ravi.
>>>>>>
>>>>>> --
>>>>>>  
>>>>>> Ravishankar. G. Shiroor
>>>>>> Wipro Technologies, Bangalore.
>>>>>>  
>>>>>> ravishankar.shiroor@wipro.com
>>>>>> --
>>>>>>  
>>>>>>
>>>>>>> -----Original Message-----
>>>>>>> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
>>>>>>> Sent: Wednesday, January 31, 2007 8:30 PM
>>>>>>> To: sipping
>>>>>>> Cc: Mary Barnes
>>>>>>> Subject: [Sipping] Service Identification in SIP
>>>>>>>
>>>>>>> Folks,
>>>>>>>
>>>>>>> during the last year, there have been a number of 
>> threads on the 
>>>>>>> mailing list discussing service identification in SIP. In 
>>>>>>> particular, the following thread discussed 3GPP
>>>> requirements on this
>>>>>>> area:
>>>>>>>
>>>>>>>
>>>> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
>>>>>>> At a later point, those 3GPP requirements were 
>> documented in the 
>>>>>>> following (expired by now) Internet Draft, which was
>>>> presented in a
>>>>>>> face-to-face IETF meeting.
>>>>>>>
>>>>>>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
>>>>>>> cs-requirements-00.txt
>>>>>>>
>>>>>>> Lately, the following draft has been submitted to the archives:
>>>>>>>
>>>>>>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
>>>>>>> rvice-identification-00.txt
>>>>>>>
>>>>>>> This draft proposes a solution that intends to meet known 
>>>>>>> requirements in this area.
>>>>>>>
>>>>>>> Given the amount of discussions on issues related to service 
>>>>>>> identification we have seen during the last year, we
>>>> believe this is
>>>>>>> a fairly important topic for the SIPPING WG to tackle. At some 
>>>>>>> point, we would like to have a milestone with its
>>>> associated WG item
>>>>>>> dealing with this issue.
>>>>>>>
>>>>>>> Therefore, we would like to encourage people to read this
>>>> draft and
>>>>>>> send comments to the list. Additionally, if you believe 
>> that this 
>>>>>>> work is (or is not) important, we would like to hear 
>> that as well.
>>>>>>> Cheers,
>>>>>>>
>>>>>>> Gonzalo
>>>>>>> SIPPING co-chair
>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> Sipping mailing list
>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>>>> This list is for NEW development of the application of SIP Use 
>>>>>>> sip-implementors@cs.columbia.edu for questions on 
>> current sip Use 
>>>>>>> sip@ietf.org for new developments of core SIP
>>>>>>>
>>>>>> _______________________________________________
>>>>>> Sipping mailing list
>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>>> This list is for NEW development of the application of SIP Use 
>>>>>> sip-implementors@cs.columbia.edu for questions on 
>> current sip Use 
>>>>>> sip@ietf.org for new developments of core SIP
>>>>> _______________________________________________
>>>>> Sipping mailing list  
>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>> This list is for NEW development of the application of SIP Use 
>>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>>> sip@ietf.org for new developments of core SIP
>>>>>
>>>
>>> The information contained in this electronic message and 
>> any attachments to this message are intended for the 
>> exclusive use of the addressee(s) and may contain 
>> proprietary, confidential or privileged information. If you 
>> are not the intended recipient, you should not disseminate, 
>> distribute or copy this e-mail. Please notify the sender 
>> immediately and destroy all copies of this message and any 
>> attachments. 
>>> WARNING: Computer viruses can be transmitted via email. The 
>> recipient should check this email and any attachments for the 
>> presence of viruses. The company accepts no liability for any 
>> damage caused by any virus transmitted by this email.
>>>  
>>> www.wipro.com
>>>
> 
> 
> The information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments. 
> 
> WARNING: Computer viruses can be transmitted via email. The recipient should check this email and any attachments for the presence of viruses. The company accepts no liability for any damage caused by any virus transmitted by this email.
>  
> www.wipro.com
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 01:51:18 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGrUJ-0004WL-0D; Tue, 13 Feb 2007 01:50:03 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGrUH-0004RW-6H
	for sipping@ietf.org; Tue, 13 Feb 2007 01:50:01 -0500
Received: from wip-ec-wd.wipro.com ([203.91.193.32])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGrUB-0007ym-VQ
	for sipping@ietf.org; Tue, 13 Feb 2007 01:50:01 -0500
Received: from wip-ec-wd.wipro.com (localhost.wipro.com [127.0.0.1])
	by localhost (Postfix) with ESMTP id 26FF8205EE
	for <sipping@ietf.org>; Tue, 13 Feb 2007 12:12:02 +0530 (IST)
Received: from blr-ec-bh02.wipro.com (blr-ec-bh02.wipro.com [10.201.50.92])
	by wip-ec-wd.wipro.com (Postfix) with ESMTP id EA830205D9
	for <sipping@ietf.org>; Tue, 13 Feb 2007 12:12:01 +0530 (IST)
Received: from blr-m2-msg.wipro.com ([10.116.50.99]) by blr-ec-bh02.wipro.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 12:19:49 +0530
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Tue, 13 Feb 2007 12:24:43 +0530
Message-ID: <532B18E13CF9E64380EF5FDDE265E071B9E933@blr-m2-msg.wipro.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdPKf+3s2QQxsDtRcOvQCBqfI7OPgAB7/KQ
References: <45C0AEF2.8010101@ericsson.com>	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
	<1170835642.29763.54.camel@n95.nomadiclab.com>
	<45CA08B2.4050603@cisco.com>
	<532B18E13CF9E64380EF5FDDE265E071B6192E@blr-m2-msg.wipro.com>
	<45CC7697.1060204@cisco.com>
	<532B18E13CF9E64380EF5FDDE265E071B61B9E@blr-m2-msg.wipro.com>
	<45D14293.30109@cisco.com>
From: <ravishankar.shiroor@wipro.com>
To: <pkyzivat@cisco.com>
X-OriginalArrivalTime: 13 Feb 2007 06:49:49.0571 (UTC)
	FILETIME=[27FA0530:01C74F3B]
X-Spam-Score: 0.2 (/)
X-Scan-Signature: c0aa019322dfce838bd8604f5a841b57
Cc: sipping@ietf.org, Gonzalo.Camarillo@ericsson.com, mary.barnes@nortel.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Paul Kyzivat wrote:
> >>> IMO, service identification is different from identifying=20
> the tool=20
> >>> that should be used for executing that service on the end
> >> device. In
> >>> the analogy that you have given, e-mail is the service.=20
> >>> Eudora/gmail/... Are the tools that are used to provide=20
> the service.
> >> The comment below was about identifying the application in the end=20
> >> system, not the protocol.
> >>
> >> In the case of email, the "service" is determined by the=20
> protocol and=20
> >> the address/port of the destination. Nobody has suggested adding=20
> >> headers to mail messages to identify the=20
> service/application in the=20
> >> end system that should process the mail message. But here we see=20
> >> proposals to add a header to a sip message to identify the=20
> >> service/application in the UAS.
> >=20
> > I would not mind if the e-mail that is delivering meeting=20
> notification=20
> > automatically gets delivered to my calendar application. If I have=20
> > three different tools handling it, I would set one of them to=20
> > preferred calendar application. Similarly if some one sends=20
> me a mail=20
> > with contacts, I would like it to be picked up by my=20
> Contacts application.
> > I see gmail/Eudora as just tools to execute certain=20
> applications. But=20
> > the e-mail being delivered should be able to say if it is a=20
> calendar=20
> > notification/plain mail/.... Application.
>=20
> Well, in this particular case,I would expect the meeting=20
> notification to be carried as one mime type in the body, and=20
> the contacts to be carried as another mime type. And its=20
> expected that these will be dispatched to a suitable=20
> application by the receiving email application *after* it is received.
>=20
> Note that the mime type defines the information carried, not=20
> the application that will process it.
>=20
> For SIP, the same thing is possible with body types. And=20
> further, the different types of media streams can be handed=20
> off to different applications.
>=20
> 	Paul

Mime type does not just say it is ascii text; it says it contains
meeting notification - and that's how the application/service is
identified.=20
In this example, the payload is being identified as being used for the
calendar service.
In SIP, if the payload was being delivered as an attachment to SIP
message, then the mime type would be sufficient. If SIP is only setting
up the media path, how is the service to be identified? I think the
service identifier would be useful in that case.

>=20
> > Regards,
> > Ravi.
> >=20
> >> 	Paul
> >>
> >>> Regards,
> >>> Ravi.
> >>>
> >>>
> >>> --
> >>> =20
> >>> Ravishankar. G. Shiroor
> >>> Wipro Technologies, Bangalore.
> >>> =20
> >>> ravishankar.shiroor@wipro.com
> >>> --
> >>> =20
> >>>
> >>>> -----Original Message-----
> >>>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
> >>>> Sent: Wednesday, February 07, 2007 10:43 PM
> >>>> To: Salvatore Loreto (JO/LMF)
> >>>> Cc: RaviShankar Shiroor (WT01 - Product Engineering Solutions);=20
> >>>> sipping@ietf.org; Gonzalo.Camarillo@ericsson.com;=20
> >>>> mary.barnes@nortel.com
> >>>> Subject: Re: [Sipping] Service Identification in SIP
> >>>>
> >>>> I still don't get the intent of identifying the application.=20
> >>>> It seems to me this is evil - akin to sending an email=20
> message and=20
> >>>> identifying that the recipient must use Outlook to=20
> receive it, not=20
> >>>> Thunderbird, Eudora, gmail, or whatever.
> >>>>
> >>>> I also wish somebody could give better examples of a=20
> communication=20
> >>>> service. Chess is too artificial to be believable, and=20
> IMO PoC is=20
> >>>> only a service because of the way the implementation has=20
> been done.
> >>>> Conceptually PoC could be nothing but a special floor control=20
> >>>> protocol accompanying a voice session. In that case none of the=20
> >>>> Service mechanism would be needed.
> >>>>
> >>>> 	Paul
> >>>>
> >>>> Salvatore Loreto (JO/LMF) wrote:
> >>>>> hi Ravi,
> >>>>>
> >>>>> in IMS a Service is a "type of communication defined by=20
> a service=20
> >>>>> definition that specifies the rules and procedures and
> >>>> allowed media,
> >>>>> if any, for a specific type of communication and that
> >>>> utilises the IMS
> >>>>> enabler. An IMS communication service defines what is
> >>>> possible within
> >>>>> a single SIP session or standalone transaction."
> >>>>> So a Service implies some combination of capabilities and
> >>>> ways to use
> >>>>> those capabilities.
> >>>>> If you want the Service specifies the context for a=20
> communication.
> >>>>>
> >>>>> Instead the Application define the specific communication
> >>>> (or if you
> >>>>> want application) that you want run in the specific
> >> context. So you
> >>>>> should have several different applications designed to run on a=20
> >>>>> specific environment or that have different performance or
> >>>> behavior in
> >>>>> different context (running on top of different service).
> >>>>>
> >>>>> Lets go to Jonathan's example of the Chess.
> >>>>> In his example the Chess is a service, for IMS instead is an=20
> >>>>> application that you can run on top of different service.
> >>>>> For example you can run Chess application on top of PoC
> >>>> service or on
> >>>>> top of the "standard" service. So let say you run Chess on
> >>>> top of PoC
> >>>>> service, then you can use only voice in the way that PoC
> >> specifies.
> >>>>> /sal
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> On Thu, 2007-02-01 at 16:39 +0530,
> >>>> ravishankar.shiroor@wipro.com wrote:
> >>>>>> Hello,
> >>>>>>
> >>>>>> IMO, Service Identification is definitely needed and
> >>>> should be taken
> >>>>>> up as a WG as proposed below.
> >>>>>>
> >>>>>> Further, I think Service Identification scope should be
> >> as per the
> >>>>>> requirement in Jonathan's draft, as I still do not
> >> understand the
> >>>>>> need for having a Communication Service Identifier and another=20
> >>>>>> Application Identifier as given in the
> >>>> 3gpp-ics-requirements document.
> >>>>>> Regards,
> >>>>>> Ravi.
> >>>>>>
> >>>>>> --
> >>>>>> =20
> >>>>>> Ravishankar. G. Shiroor
> >>>>>> Wipro Technologies, Bangalore.
> >>>>>> =20
> >>>>>> ravishankar.shiroor@wipro.com
> >>>>>> --
> >>>>>> =20
> >>>>>>
> >>>>>>> -----Original Message-----
> >>>>>>> From: Gonzalo Camarillo=20
> [mailto:Gonzalo.Camarillo@ericsson.com]
> >>>>>>> Sent: Wednesday, January 31, 2007 8:30 PM
> >>>>>>> To: sipping
> >>>>>>> Cc: Mary Barnes
> >>>>>>> Subject: [Sipping] Service Identification in SIP
> >>>>>>>
> >>>>>>> Folks,
> >>>>>>>
> >>>>>>> during the last year, there have been a number of
> >> threads on the
> >>>>>>> mailing list discussing service identification in SIP. In=20
> >>>>>>> particular, the following thread discussed 3GPP
> >>>> requirements on this
> >>>>>>> area:
> >>>>>>>
> >>>>>>>
> >>>>=20
> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
> >>>>>>> At a later point, those 3GPP requirements were
> >> documented in the
> >>>>>>> following (expired by now) Internet Draft, which was
> >>>> presented in a
> >>>>>>> face-to-face IETF meeting.
> >>>>>>>
> >>>>>>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
> >>>>>>> cs-requirements-00.txt
> >>>>>>>
> >>>>>>> Lately, the following draft has been submitted to the=20
> archives:
> >>>>>>>
> >>>>>>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
> >>>>>>> rvice-identification-00.txt
> >>>>>>>
> >>>>>>> This draft proposes a solution that intends to meet known=20
> >>>>>>> requirements in this area.
> >>>>>>>
> >>>>>>> Given the amount of discussions on issues related to service=20
> >>>>>>> identification we have seen during the last year, we
> >>>> believe this is
> >>>>>>> a fairly important topic for the SIPPING WG to=20
> tackle. At some=20
> >>>>>>> point, we would like to have a milestone with its
> >>>> associated WG item
> >>>>>>> dealing with this issue.
> >>>>>>>
> >>>>>>> Therefore, we would like to encourage people to read this
> >>>> draft and
> >>>>>>> send comments to the list. Additionally, if you believe
> >> that this
> >>>>>>> work is (or is not) important, we would like to hear
> >> that as well.
> >>>>>>> Cheers,
> >>>>>>>
> >>>>>>> Gonzalo
> >>>>>>> SIPPING co-chair
> >>>>>>>
> >>>>>>>
> >>>>>>> _______________________________________________
> >>>>>>> Sipping mailing list
> >>>> https://www1.ietf.org/mailman/listinfo/sipping
> >>>>>>> This list is for NEW development of the application=20
> of SIP Use=20
> >>>>>>> sip-implementors@cs.columbia.edu for questions on
> >> current sip Use
> >>>>>>> sip@ietf.org for new developments of core SIP
> >>>>>>>
> >>>>>> _______________________________________________
> >>>>>> Sipping mailing list
> >>>> https://www1.ietf.org/mailman/listinfo/sipping
> >>>>>> This list is for NEW development of the application of SIP Use=20
> >>>>>> sip-implementors@cs.columbia.edu for questions on
> >> current sip Use
> >>>>>> sip@ietf.org for new developments of core SIP
> >>>>> _______________________________________________
> >>>>> Sipping mailing list
> >> https://www1.ietf.org/mailman/listinfo/sipping
> >>>>> This list is for NEW development of the application of SIP Use=20
> >>>>> sip-implementors@cs.columbia.edu for questions on=20
> current sip Use=20
> >>>>> sip@ietf.org for new developments of core SIP
> >>>>>
> >>>
> >>> The information contained in this electronic message and
> >> any attachments to this message are intended for the=20
> exclusive use of=20
> >> the addressee(s) and may contain proprietary, confidential or=20
> >> privileged information. If you are not the intended recipient, you=20
> >> should not disseminate, distribute or copy this e-mail.=20
> Please notify=20
> >> the sender immediately and destroy all copies of this=20
> message and any=20
> >> attachments.
> >>> WARNING: Computer viruses can be transmitted via email. The
> >> recipient should check this email and any attachments for the=20
> >> presence of viruses. The company accepts no liability for=20
> any damage=20
> >> caused by any virus transmitted by this email.
> >>> =20
> >>> www.wipro.com
> >>>
> >=20
> >=20
> > The information contained in this electronic message and=20
> any attachments to this message are intended for the=20
> exclusive use of the addressee(s) and may contain=20
> proprietary, confidential or privileged information. If you=20
> are not the intended recipient, you should not disseminate,=20
> distribute or copy this e-mail. Please notify the sender=20
> immediately and destroy all copies of this message and any=20
> attachments.=20
> >=20
> > WARNING: Computer viruses can be transmitted via email. The=20
> recipient should check this email and any attachments for the=20
> presence of viruses. The company accepts no liability for any=20
> damage caused by any virus transmitted by this email.
> > =20
> > www.wipro.com
> >=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 05:25:47 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGuoy-0006XP-5Q; Tue, 13 Feb 2007 05:23:36 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGuox-0006XK-Fd
	for sipping@ietf.org; Tue, 13 Feb 2007 05:23:35 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGuoq-0001v1-1z
	for sipping@ietf.org; Tue, 13 Feb 2007 05:23:35 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	71DAE2059A
	for <sipping@ietf.org>; Tue, 13 Feb 2007 11:23:19 +0100 (CET)
X-AuditID: c1b4fb3c-b17cbbb0000007de-21-45d191970c5a 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	661B32049A
	for <sipping@ietf.org>; Tue, 13 Feb 2007 11:23:19 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 11:23:18 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Date: Tue, 13 Feb 2007 11:23:18 +0100
Message-ID: <7374777208BDC7449D5620EF9423256702047C66@esealmw113.eemea.ericsson.se>
In-Reply-To: <1170835642.29763.54.camel@n95.nomadiclab.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Service Identification in SIP - proxy types
Thread-Index: AcdKpsbFDrW3FiTgTaqoCBJtt1RnoQEDfmXQ
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 13 Feb 2007 10:23:18.0827 (UTC)
	FILETIME=[FAE35FB0:01C74F58]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 08170828343bcf1325e4a0fb4584481c
Subject: [Sipping] Service Identification in SIP - proxy types
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,

First, the document talks about outbound proxy, home proxy, originating
proxy and terminating proxy, but there is no real
definition/explanation/reference to what those proxies really are.

Second, the draft also uses Jonathan's loose route draft. Have we
decided to work on that draft? I remember there was quite much
discussion at one point, but what happened after that? In any case, a
few words WHY the mechanism is needed would also be useful.

Regards,

Christer

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 05:57:43 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGvLH-0006zd-NA; Tue, 13 Feb 2007 05:56:59 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGvLF-0006zQ-W1
	for sipping@ietf.org; Tue, 13 Feb 2007 05:56:57 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGvLE-0005Ns-7a
	for sipping@ietf.org; Tue, 13 Feb 2007 05:56:57 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	C59F021119; Tue, 13 Feb 2007 11:56:55 +0100 (CET)
X-AuditID: c1b4fb3e-af6d2bb0000007e1-eb-45d199774ff5 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	A3991200DC; Tue, 13 Feb 2007 11:56:55 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 11:56:55 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [SIPPING] SDP offer/answer
	inearly	media:draft-sawada-sipping-sip-offeranswer-01.txt
Date: Tue, 13 Feb 2007 11:56:54 +0100
Message-ID: <7374777208BDC7449D5620EF9423256702047D9A@esealmw113.eemea.ericsson.se>
In-Reply-To: <455CC2D1.9020406@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [SIPPING] SDP offer/answer
	inearly	media:draft-sawada-sipping-sip-offeranswer-01.txt
Thread-Index: AccJwhyYskAXl4YdQ7q9Uk6feCN7lRDseyaQ
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>,
	"Siddhartha Bhakta" <Siddhartha.Bhakta@newport-networks.com>
X-OriginalArrivalTime: 13 Feb 2007 10:56:55.0231 (UTC)
	FILETIME=[ACC234F0:01C74F5D]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: e367d58950869b6582535ddf5a673488
Cc: sipping@ietf.org, Siddhartha Bhakta <sbhakta007@yahoo.co.in>,
	Brian Stucker <bstucker@nortel.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,

I once proposed that it should be possible to terminate an early dialog
with an 19x provisional response.

Then, if "fake forking" is used, the dialog not used anymore could be
terminted and the UAC knows he doesn't have to care about it (and
possible media coming on it).

Regards,

Christer
=20

> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
> Sent: 16. marraskuuta 2006 21:58
> To: Siddhartha Bhakta
> Cc: sipping@ietf.org; 'Siddhartha Bhakta'; 'Brian Stucker'
> Subject: Re: [SIPPING] SDP offer/answer inearly=20
> media:draft-sawada-sipping-sip-offeranswer-01.txt
>=20
>=20
>=20
> Siddhartha Bhakta wrote:
> > I agree with you. No matter whether different provisional responses=20
> > are creating different dialogs or not, case is same as far as media=20
> > rendering/committing is concerned. For media committing, there are=20
> > multiple answers for a given offer.
> >=20
> > Like you, I also can not understand what we are gaining by creating=20
> > multiple dialogs here except the chance to manage=20
> offer/answer state=20
> > as per RFC3261/RFC3264. It seems that multiple SDP answers=20
> for a given=20
> > SDP offer is the reality. We can not avoid that by applying=20
> any tricky=20
> > logic in dialog management (like Fake Forking).
>=20
> If you ignore dialogs, it becomes impossible to impose any=20
> offer/answer rules.
>=20
> By paying attention to the separate dialogs, it is possible=20
> to have offer/answer rules that make sense. If you just mix=20
> together offers or answers from multiple destinations it=20
> becomes impossible to make any sense of what you have once it=20
> is determined which dialog will be completed.
>=20
> For the period when there are multiple dialogs, you have two=20
> different things going on:
> - an offer/answer state machine is proceeding in each dialog, and
>    defining the current media parameters for that dialog
> - you are deciding which dialog(s) you want to process and send
>    media to.
>=20
> As Brian notes, it is complex to decide which dialog to honor=20
> during the early stage. But once you settle on one dialog at=20
> least you will then know what to do.
>=20
> 	Paul
>=20
> > -----Original Message-----
> > From: Brian Stucker [mailto:bstucker@nortel.com]
> > Sent: 16 November 2006 17:37
> > To: Hadriel Kaplan; Siddhartha Bhakta; sipping@ietf.org
> > Cc: Siddhartha Bhakta
> > Subject: RE: [SIPPING] SDP offer/answer in early=20
> > media:draft-sawada-sipping-sip-offeranswer-01.txt
> >=20
> > Fake forking...
> >=20
> > Please note, if there's SDP in the other provisional=20
> responses, you're=20
> > going to run into problems potentially getting themedia rendered=20
> > (assuming that's your goal).
> >=20
> > Regards,
> > Brian
> >=20
> >> -----Original Message-----
> >> From: Hadriel Kaplan [mailto:HKaplan@acmepacket.com]
> >> Sent: Sunday, November 12, 2006 8:40 PM
> >> To: 'Siddhartha Bhakta'; sipping@ietf.org
> >> Cc: 'Siddhartha Bhakta'
> >> Subject: RE: [SIPPING] SDP offer/answer in early=20
> >> media:draft-sawada-sipping-sip-offeranswer-01.txt
> >>
> >> If the Proxy is sending back SDP by itself it's not a proxy, but a=20
> >> b2bua, as is the PBX.
> >> As such, it should be answering in a different dialog, so its SDP=20
> >> answer should be in a 183 with a different To-tag from the=20
> 180, and=20
> >> the PBX's 180 SDP should have a different To-tag than the 200ok. =20
> >> That would make them look like a forked call and follow the RFCs.
> >>
> >> Of course the reality is often different, but being a=20
> b2bua yourself=20
> >> it's in your power to fix it.  How you do that is not up=20
> to the IETF=20
> >> to define, as the very idea of it goes against the IETF's=20
> view of the=20
> >> world. (and I say that in a positive way, as I think it would be=20
> >> practically impossible for the IETF to do otherwise)
> >>
> >> -hadriel
> >>
> >>
> >>> -----Original Message-----
> >>> From: Siddhartha Bhakta [mailto:sbhakta007@yahoo.co.in]
> >>> Sent: Sunday, November 12, 2006 2:11 PM
> >>> To: sipping@ietf.org
> >>> Cc: sbhakta007@yahoo.co.in; Siddhartha Bhakta
> >>> Subject: [SIPPING] SDP offer/answer in early
> >>> media:draft-sawada-sipping- sip-offeranswer-01.txt
> >>>
> >>> I am re-sending it as in last email, the figure was distorted.
> >>>
> >>> Though, post RFC3261 drafts indicate that one valid=20
> answer would be=20
> >>> there for one SDP offer, the our old friend early-media
> >> flow is still
> >>> present in practice.
> >>>
> >>> B2BUA/SBC     Proxy         PABX        Bob
> >>>   |            |            |            |
> >>>   | INVITE F1  |            |            |
> >>>   |----------->| INVITE F2  |            |
> >>>   |   183 F3   |----------->| INVITE F4  |
> >>>   |<-----------|            |----------->|
> >>>   |            |            |            |
> >>>   |            |            |   180 without SDP F5
> >>>   |            |            |<-----------|
> >>>   |            |            |            |
> >>>   |            |  180 with SDP F6        |
> >>>   |   180 F7   |<-----------|            |
> >>>   |<-----------|            |  200 F8    |
> >>>   |            |  200 F9    |<-----------|
> >>>   |   200 F10  |<-----------|            |
> >>>   |<-----------|            |            |
> >>>   |   ACK F11  |            |            |
> >>>   |----------->|  ACK F12   |            |
From sipping-bounces@ietf.org Tue Feb 13 05:57:43 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGvLH-0006zd-NA; Tue, 13 Feb 2007 05:56:59 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGvLF-0006zQ-W1
	for sipping@ietf.org; Tue, 13 Feb 2007 05:56:57 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGvLE-0005Ns-7a
	for sipping@ietf.org; Tue, 13 Feb 2007 05:56:57 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	C59F021119; Tue, 13 Feb 2007 11:56:55 +0100 (CET)
X-AuditID: c1b4fb3e-af6d2bb0000007e1-eb-45d199774ff5 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	A3991200DC; Tue, 13 Feb 2007 11:56:55 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 11:56:55 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [SIPPING] SDP offer/answer
	inearly	media:draft-sawada-sipping-sip-offeranswer-01.txt
Date: Tue, 13 Feb 2007 11:56:54 +0100
Message-ID: <7374777208BDC7449D5620EF9423256702047D9A@esealmw113.eemea.ericsson.se>
In-Reply-To: <455CC2D1.9020406@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [SIPPING] SDP offer/answer
	inearly	media:draft-sawada-sipping-sip-offeranswer-01.txt
Thread-Index: AccJwhyYskAXl4YdQ7q9Uk6feCN7lRDseyaQ
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>,
	"Siddhartha Bhakta" <Siddhartha.Bhakta@newport-networks.com>
X-OriginalArrivalTime: 13 Feb 2007 10:56:55.0231 (UTC)
	FILETIME=[ACC234F0:01C74F5D]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: e367d58950869b6582535ddf5a673488
Cc: sipping@ietf.org, Siddhartha Bhakta <sbhakta007@yahoo.co.in>,
	Brian Stucker <bstucker@nortel.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,

I once proposed that it should be possible to terminate an early dialog
with an 19x provisional response.

Then, if "fake forking" is used, the dialog not used anymore could be
terminted and the UAC knows he doesn't have to care about it (and
possible media coming on it).

Regards,

Christer
=20

> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
> Sent: 16. marraskuuta 2006 21:58
> To: Siddhartha Bhakta
> Cc: sipping@ietf.org; 'Siddhartha Bhakta'; 'Brian Stucker'
> Subject: Re: [SIPPING] SDP offer/answer inearly=20
> media:draft-sawada-sipping-sip-offeranswer-01.txt
>=20
>=20
>=20
> Siddhartha Bhakta wrote:
> > I agree with you. No matter whether different provisional responses=20
> > are creating different dialogs or not, case is same as far as media=20
> > rendering/committing is concerned. For media committing, there are=20
> > multiple answers for a given offer.
> >=20
> > Like you, I also can not understand what we are gaining by creating=20
> > multiple dialogs here except the chance to manage=20
> offer/answer state=20
> > as per RFC3261/RFC3264. It seems that multiple SDP answers=20
> for a given=20
> > SDP offer is the reality. We can not avoid that by applying=20
> any tricky=20
> > logic in dialog management (like Fake Forking).
>=20
> If you ignore> >>>   |            |----------->|  ACK F13   |
> >>>   |            |            |----------->|
> >>>
> >>> Our B2BUA is facing the problem as specified above.
> >>> The first answer is carried
> >>> by F3. This SDP is specified by Proxy. The 2nd SDP answer
> >> is carried
> >>> by F6,F7.
> >>> In fact, this SDP is specified by PABX. The third SDP answer is=20
> >>> carried by F8,F9,F10.
> >>> This SDP is specified by called party (Bob).
> >>>
> >>> As per the RFC3261 & RFC3264, the SDP answer carried by
> >> F6,F7 should
> >>> match with F3 as far as B2BUA is concerned. Therefore,=20
> B2BUA shall=20
> >>> ignore it. This is leading to the fact that SIP UA behind=20
> B2BUA can=20
> >>> not listen to ringtone fed by PABX.
> >>> Similarly, SDP answer carried by F8,F9,F10 shall be ignored. This=20
> >>> shall lead to the fact that SIP UA behind B2BUA can not listen to=20
> >>> Bob's voice. Too bad.
> >>> This problem is due to the fact that RFC3261 & RFC3264 are not=20
> >>> backward compatible.
> >>>
> >>>
> >>> The sec 2.2. of
> >>> draft-sawada-sipping-sip-offeranswer-01.txt indicates that UPDATE=20
> >>> should be used to update the media in early media. But in
> >> practice old
> >>> early-media flow (i.e., sending different SDPs over different 1xx=20
> >>> responses of
> >>> INVITE) is quite common.
> >>> Can we somehow make new SDP offer/answer specified in
> >>> RFC3261 & RFC3264 backward
> >>> compatible?
> >>>
> >>> The old standard (early-media), allows multiple SDP answers
> >> of single
> >>> SDP offer. can we somehow induce this thing in SDP
> >> offer/answer model?=20
> >>> If many of you think the need of it then I may comeup with
> >> some thumb
> >>> rule for the same.
> >>>
> >>> Thanks & Regards,
> >>> Siddhartha
> >>
> >> _______________________________________________
> >> Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> >> This list is for NEW development of the application of SIP Use=20
> >> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> >> sip@ietf.org for new developments of core SIP
> >>
> >=20
> >=20
> >=20
> > ---------------
> > This e-mail may contain confidential and/or privileged=20
> information. If you are not the intended recipient (or have=20
> received this e-mail in error) please notify the sender=20
> immediately and delete this e-mail. Any unauthorized copying,=20
> disclosure or distribution of the contents in this e-mail is=20
> strictly forbidden.
> > ---------------
> >=20
> >=20
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP Use=20
> > sip-implementors@cs.columbia.edu for questions on current sip Use=20
> > sip@ietf.org for new developments of core SIP
> >=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP

From sipping-bounces@ietf.org Tue Feb 13 05:57:43 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGvLI-00070z-Ag; Tue, 13 Feb 2007 05:57:00 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGvLF-0006zR-W9
	for sipping@ietf.org; Tue, 13 Feb 2007 05:56:58 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGvL9-0005NR-6u
	for sipping@ietf.org; Tue, 13 Feb 2007 05:56:57 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson dialogs, it becomes impossible to impose any=20
> offer/answer rules.
>=20
> By paying attention to the separate dialogs, it is possible=20
> to have offer/answer rules that make sense. If you just mix=20
> together offers or answers from multiple destinations it=20
> becomes impossible to make any sense of what you have once it=20
> is determined which dialog will be completed.
>=20
> For the period when there are multiple dialogs, you have two=20
> different things going on:
> - an offer/answer state machine is proceeding in each dialog, and
>    defining the current media parameters for that dialog
> - you are deciding which dialog(s) you want to process and send
>    media to.
>=20
> As Brian notes, it is complex to decide which dialog to honor=20
> during the early stage. But once you settle on one dialog at=20
> least you will then know what to do.
>=20
> 	Paul
>=20
> > -----Original Message-----
> > From: Brian Stucker [mailto:bstucker@nortel.com]
> > Sent: 16 November 2006 17:37
> > To: Hadriel Kaplan; Siddhartha Bhakta; sipping@ietf.org
> > Cc: Siddhartha Bhakta
> > Subject: RE: [SIPPING] SDP offer/answer in early=20
> > media:draft-sawada-sipping-sip-offeranswer-01.txt
> >=20
> > Fake forking...
> >=20
> > Please note, if there's SDP in the other provisional=20
> responses, you're=20
> > going to run into problems potentially getting themedia rendered=20
> > (assuming that's your goal).
> >=20
> > Regards,
> > Brian
> >=20
> >> -----Original Message-----
> >> From: Hadriel Kaplan [mailto:HKaplan@acmepacket.com]
> >> Sent: Sunday, November 12, 2006 8:40 PM
> >> To: 'Siddhartha Bhakta'; sipping@ietf.org
> >> Cc: 'Siddhartha Bhakta'
> >> Subject: RE: [SIPPING] SDP offer/answer in early=20
> >> media:draft-sawada-sipping-sip-offeranswer-01.txt
> >>
> >> If the Proxy is sending back SDP by itself it's not a proxy, but a=20
> >> b2bua, as is the PBX.
> >> As such, it should be answering in a different dialog, so its SDP=20
> >> answer should be in a 183 with a different To-tag from the=20
> 180, and=20
> >> the PBX's 180 SDP should have a different To-tag than the 200ok. =20
> >> That would make them look like a forked call and follow the RFCs.
> >>
> >> Of course the reality is often different, but being a=20
> b2bua yourself=20
> >> it's in your power to fix it.  How you do that is not up=20
> to the IETF=20
> >> to define, as the very idea of it goes against the IETF's=20
> view of the=20
> >> world. (and I say that in a positive way, as I think it would be=20
> >> practically impossible for the IETF to do otherwise)
> >>
> >> -hadriel
> >>
> >>
> >>> -----Original Message-----
> >>> From: Siddhartha Bhakta [mailto:sbhakta007@yahoo.co.in]
> >>> Sent: Sunday, November 12, 2006 2:11 PM
> >>> To: sipping@ietf.org
> >>> Cc: sbhakta007@yahoo.co.in; Siddhartha Bhakta
> >>> Subject: [SIPPING] SDP offer/answer in early
> >>> media:draft-sawada-sipping- sip-offeranswer-01.txt
> >>>
> >>> I am re-sending it as in last email, the figure was distorted.
> >>>
> >>> Though, post RFC3261 drafts indicate that one valid=20
> answer would be=20
> >>> there for one SDP offer, the our old friend early-media
> >> flow is still
> >>> present in practice.
> >>>
> >>> B2BUA/SBC     Proxy         PABX        Bob
> >>>   |            |            |            |
> >>>   | INVITE F1  |            |            |
> >>>   |----------->| INVITE F2  |            |
> >>>   |   183 F3   |----------->| INVITE F4  |
> >>>   |<-----------|            |----------->|
> >>>   |            |            |            |
> >>>   |            |            |   180 without SDP F5
> >>>   |            |            |<-----------|
> >>>   |            |            |            |
> >>>   |            |  180 with SDP F6        |
> >>>   |   180 F7   |<-----------|            |
> >>>   |<-----------|            |  200 F8    |
> >>>   |            |  200 F9    |<-----------|
> >>>   |   200 F10  |<-----------|            |
> >>>   |<-----------|            |            |
> >>>   |   ACK F11  |            |            |
> >>>   |----------->|  ACK F12   |            |
.se (Symantec Mail Security) with ESMTP id
	B0F8821112; Tue, 13 Feb 2007 11:56:48 +0100 (CET)
X-AuditID: c1b4fb3e-af6d2bb0000007e1-e2-45d199707f96 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	8E2EF200DC; Tue, 13 Feb 2007 11:56:48 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 11:56:48 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] SDP
	Rollbackindraft-sawada-sipping-sip-offeranswer-01.txt
Date: Tue, 13 Feb 2007 11:56:47 +0100
Message-ID: <7374777208BDC7449D5620EF9423256702047D98@esealmw113.eemea.ericsson.se>
In-Reply-To: <455E2386.4080900@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] SDP
	Rollbackindraft-sawada-sipping-sip-offeranswer-01.txt
Thread-Index: AccKjG+x9v7fniDTS7miiuo4yK7SERC5gZ2Q
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>,
	"Sanjay Sinha \(sanjsinh\)" <sanjsinh@cisco.com>
X-OriginalArrivalTime: 13 Feb 2007 10:56:48.0122 (UTC)
	FILETIME=[A88575A0:01C74F5D]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: f8ee348dcc4be4a59bc395f7cd6343ad
Cc: tu-sawada@kddi.com, sipping@ietf.org,
	Siddhartha Bhakta <Siddhartha.Bhakta@newport-networks.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,

One approach would of course be to not allow reliable provisional
responses for re-INVITEs in the first place.

But, then we MAY have a problem with preconditions - but not
necessarily.

The SDP answerer could simply choose not to send 200 OK until he has
reserved the needed resources. Ie he does not send any provisional
responses.

The SDP offerer would then send a second re-INVITE/UPDATE when he has
reserved the needed resources (unfortunatly he can't use the ACK to
indicate it).

Regards,

Christer

=20

> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
> Sent: 17. marraskuuta 2006 23:03
> To: Sanjay Sinha (sanjsinh)
> Cc: tu-sawada@kddi.com; sipping@ietf.org; Siddhartha Bhakta
> Subject: Re: [Sipping] SDP=20
> Rollbackindraft-sawada-sipping-sip-offeranswer-01.txt
>=20
> (To those who haven't been paying attention, this is about=20
> what to do about SDP when a reINVITE fails. RFC 3261, section=20
> 14.1, says that if a non-2xx final response is received the=20
> session parameters must remain unchanged, as if no reINVITE=20
> had been issued. Herein that is being described as "rollback"=20
> of the SDP.)
>=20
> There are several interesting questions relative to this topic:
>=20
> - are the existing RFCs clear about what should be done?
>    (Do all informed and reasonable people agree what this is?)
>=20
> - what do currently deployed implementations do?
>=20
> - what is the *most reasonable* thing to do?
>=20
> There has been quite a bit of discussion in the past on this,=20
> though we could probably some more because I think only a few=20
> people participated.
>=20
> IMO the existing RFCs are *not* clear, because we are getting=20
> multiple interpretations that can be justified to some=20
> extent. The rollback rule only appears in 3261, and doesn't=20
> really address the additional issues raised by 3262, 3311, and 3312.
>=20
> There are also good arguments for either of two approaches=20
> being the most reasonable thing> >>>   |            |----------->|  ACK F13   |
> >>>   |            |            |----------->|
> >>>
> >>> Our B2BUA is facing the problem as specified above.
> >>> The first answer is carried
> >>> by F3. This SDP is specified by Proxy. The 2nd SDP answer
> >> is carried
> >>> by F6,F7.
> >>> In fact, this SDP is specified by PABX. The third SDP answer is=20
> >>> carried by F8,F9,F10.
> >>> This SDP is specified by called party (Bob).
> >>>
> >>> As per the RFC3261 & RFC3264, the SDP answer carried by
> >> F6,F7 should
> >>> match with F3 as far as B2BUA is concerned. Therefore,=20
> B2BUA shall=20
> >>> ignore it. This is leading to the fact that SIP UA behind=20
> B2BUA can=20
> >>> not listen to ringtone fed by PABX.
> >>> Similarly, SDP answer carried by F8,F9,F10 shall be ignored. This=20
> >>> shall lead to the fact that SIP UA behind B2BUA can not listen to=20
> >>> Bob's voice. Too bad.
> >>> This problem is due to the fact that RFC3261 & RFC3264 are not=20
> >>> backward compatible.
> >>>
> >>>
> >>> The sec 2.2. of
> >>> draft-sawada-sipping-sip-offeranswer-01.txt indicates that UPDATE=20
> >>> should be used to update the media in early media. But in
> >> practice old
> >>> early-media flow (i.e., sending different SDPs over different 1xx=20
> >>> responses of
> >>> INVITE) is quite common.
> >>> Can we somehow make new SDP offer/answer specified in
> >>> RFC3261 & RFC3264 backward
> >>> compatible?
> >>>
> >>> The old standard (early-media), allows multiple SDP answers
> >> of single
> >>> SDP offer. can we somehow induce this thing in SDP
> >> offer/answer model?=20
> >>> If many of you think the need of it then I may comeup with
> >> some thumb
> >>> rule for the same.
> >>>
> >>> Thanks & Regards,
> >>> Siddhartha
> >>
> >> _______________________________________________
> >> Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> >> This list is for NEW development of the application of SIP Use=20
> >> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> >> sip@ietf.org for new developments of core SIP
> >>
> >=20
> >=20
> >=20
> > ---------------
> > This e-mail may contain confidential and/or privileged=20
> information. If you are not the intended recipient (or have=20
> received this e-mail in error) please notify the sender=20
> immediately and delete this e-mail. Any unauthorized copying,=20
> disclosure or distribution of the contents in this e-mail is=20
> strictly forbidden.
> > ---------------
> >=20
> >=20
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP Use=20
> > sip-implementors@cs.columbia.edu for questions on current sip Use=20
> > sip@ietf.org for new developments of core SIP
> >=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP

From sipping-bounces@ietf.org Tue Feb 13 05:57:43 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGvLI-00070z-Ag; Tue, 13 Feb 2007 05:57:00 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGvLF-0006zR-W9
	for sipping@ietf.org; Tue, 13 Feb 2007 05:56:58 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGvL9-0005NR-6u
	for sipping@ietf.org; Tue, 13 Feb 2007 05:56:57 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson to do.
>=20
> The two competing interpretations are:
>=20
> 1) once an answer has been transmitted reliably it is committed and
>     will not be rolled back, even if a reinvite fails.
>=20
> 2) an entire sequence of offers and answers (there could be many)
>     that is initiated by a reinvite, is rolled back if the reinvite
>     fails.
>=20
> Of these, (1) is somewhat easier to implement. It also=20
> eliminates some race conditions that seem unresolvable,=20
> though obscure, that can arise with (2). The downside is that=20
> when multiple offers and answers are used to resolve=20
> preconditions, and then the reinvite fails, you end up in an=20
> interim state with unresolved preconditions. The proper=20
> behavior in that state is not clear.
>=20
> With (2) the precondition problem goes away. A few ambiguous=20
> race conditions have been identified, but they are obscure=20
> cases that can probably be resolved by just just writing some=20
> best practices to avoid them. Its not entirely clear to me if=20
> this approach requires retention of any more state than (1),=20
> but if so it is only slightly more.
>=20
> It would be helpful if people that have opinions on this=20
> subject identify which approach they advocate and why. I'm=20
> not expecting to see a new approach different from either of=20
> the above, but if somebody has one, please spell out how it=20
> differs from these, and why it is better.
>=20
> 	Thanks,
> 	Paul
>=20
> Sanjay Sinha (sanjsinh) wrote:
> > SDP rollback applies only if previously there has been an=20
> offer-answer=20
> > exchange and an early dialog or a confirmed dialog has been=20
> established.
> > In that case, if SDP in re-Invite/UPDATE is rejected, then session=20
> > continues with previously negotiated characteristics. This is=20
> > mentioned in RFC 3261. And this draft does mention that if offer in=20
> > Prack or in a sip response is rejected, then answerer has=20
> to send an=20
> > updated offer. So I am not sure how is the new proposal=20
> different than=20
> > what has been already stated.
> > =20
> > Sanjay
> >=20
> >    =20
> --------------------------------------------------------------
> ----------
> >     *From:* Siddhartha Bhakta
> >     [mailto:Siddhartha.Bhakta@newport-networks.com]
> >     *Sent:* Tuesday, November 14, 2006 12:37 PM
> >     *To:* tu-sawada@kddi.com; Paul Kyzivat (pkyzivat)
> >     *Cc:* sipping@ietf.org
> >     *Subject:* [Sipping] SDP Rollback
> >     indraft-sawada-sipping-sip-offeranswer-01.txt
> >=20
> >     Hi,
> >=20
> >     =20
> >=20
> >     I am not sure (as I am a very new user to SIPPING
> >=20
> >     group) whether SDP rollback has ever been discussed
> >=20
> >     regarding this draft or not. I personally feel that
> >=20
> >     'SDP rollback' is very relevant topics regarding this
> >=20
> >     draft.
> >=20
> >     =20
> >=20
> >     The sec 1.2 indicates the response(488) to reject any
> >=20
> >     SDP offer. But the missing part is that if any SIP
> >=20
> >     request carrying SDP offer is rejected by 3xx-6xx then
> >=20
> >     SDP offer should be rolled-back. The rollback means,
> >=20
> >     applying last session description.
> >=20
> >     =20
> >=20
> >     Rather than specifying different SIP messages
> >=20
> >     separately, it would be better to have some thumb rule
> >=20
> >     that shall be quite easy to follow as far as SDP
> >=20
> >     rollback is concerned. I am proposing following thumb
> >=20
> >     rule. Please indicate whether that makes sense or not.
> >=20
> >     =20
> >=20
> >     The SDP rollback shall be associated with transaction
> >=20
> >     rollback/rejection.
> >=20
> >     =20
> >=20
> >     [1] If the transaction request (carrying SDP offer) is
> >=20
> >     rejected then SDP offer shall be rolled-back. The
> >=20
> >     exception is the ACK request. The ACK (of 2xx) can not
> >=20
> >     be rejected. (The ACK of 3xx-6xx is not the
> >=20
> >     transaction initiating request).
> >=20
> >     =20
> >=20
> >     [2] If any transaction request carries SDP answer
> >=20
> >  .se (Symantec Mail Security) with ESMTP id
	B0F8821112; Tue, 13 Feb 2007 11:56:48 +0100 (CET)
X-AuditID: c1b4fb3e-af6d2bb0000007e1-e2-45d199707f96 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	8E2EF200DC; Tue, 13 Feb 2007 11:56:48 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 11:56:48 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] SDP
	Rollbackindraft-sawada-sipping-sip-offeranswer-01.txt
Date: Tue, 13 Feb 2007 11:56:47 +0100
Message-ID: <7374777208BDC7449D5620EF9423256702047D98@esealmw113.eemea.ericsson.se>
In-Reply-To: <455E2386.4080900@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] SDP
	Rollbackindraft-sawada-sipping-sip-offeranswer-01.txt
Thread-Index: AccKjG+x9v7fniDTS7miiuo4yK7SERC5gZ2Q
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>,
	"Sanjay Sinha \(sanjsinh\)" <sanjsinh@cisco.com>
X-OriginalArrivalTime: 13 Feb 2007 10:56:48.0122 (UTC)
	FILETIME=[A88575A0:01C74F5D]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: f8ee348dcc4be4a59bc395f7cd6343ad
Cc: tu-sawada@kddi.com, sipping@ietf.org,
	Siddhartha Bhakta <Siddhartha.Bhakta@newport-networks.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,

One approach would of course be to not allow reliable provisional
responses for re-INVITEs in the first place.

But, then we MAY have a problem with preconditions - but not
necessarily.

The SDP answerer could simply choose not to send 200 OK until he has
reserved the needed resources. Ie he does not send any provisional
responses.

The SDP offerer would then send a second re-INVITE/UPDATE when he has
reserved the needed resources (unfortunatly he can't use the ACK to
indicate it).

Regards,

Christer

=20

> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
> Sent: 17. marraskuuta 2006 23:03
> To: Sanjay Sinha (sanjsinh)
> Cc: tu-sawada@kddi.com; sipping@ietf.org; Siddhartha Bhakta
> Subject: Re: [Sipping] SDP=20
> Rollbackindraft-sawada-sipping-sip-offeranswer-01.txt
>=20
> (To those who haven't been paying attention, this is about=20
> what to do about SDP when a reINVITE fails. RFC 3261, section=20
> 14.1, says that if a non-2xx final response is received the=20
> session parameters must remain unchanged, as if no reINVITE=20
> had been issued. Herein that is being described as "rollback"=20
> of the SDP.)
>=20
> There are several interesting questions relative to this topic:
>=20
> - are the existing RFCs clear about what should be done?
>    (Do all informed and reasonable people agree what this is?)
>=20
> - what do currently deployed implementations do?
>=20
> - what is the *most reasonable* thing to do?
>=20
> There has been quite a bit of discussion in the past on this,=20
> though we could probably some more because I think only a few=20
> people participated.
>=20
> IMO the existing RFCs are *not* clear, because we are getting=20
> multiple interpretations that can be justified to some=20
> extent. The rollback rule only appears in 3261, and doesn't=20
> really address the additional issues raised by 3262, 3311, and 3312.
>=20
> There are also good arguments for either of two approaches=20
> being the most reasonable thing   (e.g., ACK or PRACK) then that transaction can not
> >=20
> >     rejected as there shall be the confusion over whether
> >=20
> >     SDP shall be rolled-back or not. I suppose, SDP answer
> >=20
> >     can not be rolled-back. There is no confusion for ACK,
> >=20
> >     as it can not rejected but for PRACK the restriction
> >=20
> >     should be there that it can not be rejected if it
> >=20
> >     carries SDP answer.
> >=20
> >     =20
> >=20
> >     This draft says that PRACK(irrespective of whether it
> >=20
> >     is carrying SDP offer/answer) can not be rejected. I
> >=20
> >     can not understand the rationale about this statement.
> >=20
> >     Though, I can appreciate why PRACK MUST not be
> >=20
> >     rejected if it carries SDP answer.
> >=20
> >     =20
> >=20
> >     [3] If any SIP response contains SDP offer then that
> >=20
> >     SDP can not be rolled-back. If any SIP entity wants to
> >=20
> >     rollback the SDP offer carried by SIP response, it
> >=20
> >     should initiate a SIP transaction carrying old SDP to
> >=20
> >     accomplish it.
> >=20
> >     =20
> >=20
> >     There is a special case for re-INVITE. If multiple SDP
> >=20
> >     offer/answer happens(using PRACK/UPDATE) within
> >=20
> >     re-INVITE transaction and re-INVITE is rejected by
> >=20
> >     4xx-6xx then whether all the SDP offer/answer happens
> >=20
> >     within re-INVITE transaction shall be rolled-back or
> >=20
> >     not. My suggestion would be that once SDP offer/answer
> >=20
> >     is completed, that can not rolled-back by means of
> >=20
> >     transaction rejection. That means, all the completed
> >=20
> >     SDP offer/answer shall not be rolled-back due to
> >=20
> >     4xx-6xx of re-INVITE. This decision shall save the
> >=20
> >     work of SIP UA, SIP SBC, B2BUA etc. that follows SDP
> >=20
> >     offer/answer state.
> >=20
> >     =20
> >=20
> >     Please comment.
> >=20
> >     =20
> >=20
> >     =20
> >=20
> >     Thanks and Regards,
> >=20
> >     Siddhartha
> >=20
> >     =20
> >=20
> >=20
> >=20
> >    =20
> --------------------------------------------------------------
> ----------
> >     ---------------
> >     This e-mail may contain confidential and/or privileged=20
> information.
> >     If you are not the intended recipient (or have received=20
> this e-mail
> >     in error) please notify the sender immediately and delete this
> >     e-mail. Any unauthorized copying, disclosure or=20
> distribution of the
> >     contents in this e-mail is strictly forbidden.
> >     ---------------
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP





 to do.
>=20
> The two competing interpretations are:
>=20
> 1) once an answer has been transmitted reliably it is committed and
>     will not be rolled back, even if a reinvite fails.
>=20
> 2) an entire sequence of offers and answers (there could be many)
>     that is initiated by a reinvite, is rolled back if the reinvite
>     fails.
>=20
> Of these, (1) is somewhat easier to implement. It also=20
> eliminates some race conditions that seem unresolvable,=20
> though obscure, that can arise with (2). The downside is that=20
> when multiple offers and answers are used to resolve=20
> preconditions, and then the reinvite fails, you end up in an=20
> interim state with unresolved preconditions. The proper=20
> behavior in that state is not clear.
>=20
> With (2) the precondition problem goes away. A few ambiguous=20
> race conditions have been identified, but they are obscure=20
> cases that can probably be resolved by just just writing some=20
> best practices to avoid them. Its not entirely clear to me if=20
> this approach requires retention of any more state than (1),=20
> but if so it is only slightly more.
>=20
> It would be helpful if people that have opinions on this=20
> subject identify which approach they advocate and why. I'm=20
> not expecting to see a new approach different from either of=20
> the above, but if somebody has one, please spell out how it=20
> differs from these, and why it is better.
>=20
> 	Thanks,
> 	Paul
>=20
> Sanjay Sinha (sanjsinh) wrote:
> > SDP rollback applies only if previously there has been an=20
> offer-answer=20
> > exchange and an early dialog or a confirmed dialog has been=20
> established.
> > In that case, if SDP in re-Invite/UPDATE is rejected, then session=20
> > continues with previously negotiated characteristics. This is=20
> > mentioned in RFC 3261. And this draft does mention that if offer in=20
> > Prack or in a sip response is rejected, then answerer has=20
> to send an=20
> > updated offer. So I am not sure how is the new proposal=20
> different than=20
> > what has been already stated.
> > =20
> > Sanjay
> >=20
> >    =20
> --------------------------------------------------------------
> ----------
> >     *From:* Siddhartha Bhakta
> >     [mailto:Siddhartha.Bhakta@newport-networks.com]
> >     *Sent:* Tuesday, November 14, 2006 12:37 PM
> >     *To:* tu-sawada@kddi.com; Paul Kyzivat (pkyzivat)
> >     *Cc:* sipping@ietf.org
> >     *Subject:* [Sipping] SDP Rollback
> >     indraft-sawada-sipping-sip-offeranswer-01.txt
> >=20
> >     Hi,
> >=20
> >     =20
> >=20
> >     I am not sure (as I am a very new user to SIPPING
> >=20
> >     group) whether SDP rollback has ever been discussed
> >=20
> >     regarding this draft or not. I personally feel that
> >=20
> >     'SDP rollback' is very relevant topics regarding this
> >=20
> >     draft.
> >=20
> >     =20
> >=20
> >     The sec 1.2 indicates the response(488) to reject any
> >=20
> >     SDP offer. But the missing part is that if any SIP
> >=20
> >     request carrying SDP offer is rejected by 3xx-6xx then
> >=20
> >     SDP offer should be rolled-back. The rollback means,
> >=20
> >     applying last session description.
> >=20
> >     =20
> >=20
> >     Rather than specifying different SIP messages
> >=20
> >     separately, it would be better to have some thumb rule
> >=20
> >     that shall be quite easy to follow as far as SDP
> >=20
> >     rollback is concerned. I am proposing following thumb
> >=20
> >     rule. Please indicate whether that makes sense or not.
> >=20
> >     =20
> >=20
> >     The SDP rollback shall be associated with transaction
> >=20
> >     rollback/rejection.
> >=20
> >     =20
> >=20
> >     [1] If the transaction request (carrying SDP offer) is
> >=20
> >     rejected then SDP offer shall be rolled-back. The
> >=20
> >     exception is the ACK request. The ACK (of 2xx) can not
> >=20
> >     be rejected. (The ACK of 3xx-6xx is not the
> >=20
> >     transaction initiating request).
> >=20
> >     =20
> >=20
> >     [2] If any transaction request carries SDP answer
> >=20
> >     (e.g., ACK or PRACK) then that transaction can not
> >=20
> >     rejected as there shall be the confusion over whether
> >=20
> >     SDP shall be rolled-back or not. I suppose, SDP answer
> >=20
> >     can not be rolled-back. There is no confusion for ACK,
> >=20
> >     as it can not rejected but for PRACK the restriction
> >=20
> >     should be there that it can not be rejected if it
> >=20
> >     carries SDP answer.
> >=20
> >     =20
> >=20
> >     This draft says that PRACK(irrespective of whether it
> >=20
> >     is carrying SDP offer/answer) can not be rejected. I
> >=20
> >     can not understand the rationale about this statement.
> >=20
> >     Though, I can appreciate why PRACK MUST not be
> >=20
> >     rejected if it carries SDP answer.
> >=20
> >     =20
> >=20
> >     [3] If any SIP response contains SDP offer then that
> >=20
> >     SDP can not be rolled-back. If any SIP entity wants to
> >=20
> >     rollback the SDP offer carried by SIP response, it
> >=20
> >     should initiate a SIP transaction carrying old SDP to
> >=20
> >     accomplish it.
> >=20
> >     =20
> >=20
> >     There is a special case for re-INVITE. If multiple SDP
> >=20
> >     offer/answer happens(using PRACK/UPDATE) within
> >=20
> >     re-INVITE transaction and re-INVITE is rejected by
> >=20
> >     4xx-6xx then whether all the SDP offer/answer happens
> >=20
> >     within re-INVITE transaction shall be rolled-back or
> >=20
> >     not. My suggestion would be that once SDP offer/answer
> >=20
> >     is completed, that can not rolled-back by means of
> >=20
> >     transaction rejection. That means, all the completed
> >=20
> >     SDP offer/answer shall not be rolled-back due to
> >=20
> >     4xx-6xx of re-INVITE. This decision shall save the
> >=20
> >     work of SIP UA, SIP SBC, B2BUA etc. that follows SDP
> >=20
> >     offer/answer state.
> >=20
> >     =20
> >=20
> >     Please comment.
> >=20
> >     =20
> >=20
> >     =20
> >=20
> >     Thanks and Regards,
> >=20
> >     Siddhartha
> >=20
> >     =20
> >=20
> >=20
> >=20
> >    =20
> --------------------------------------------------------------
> ----------
> >     ---------------
> >     This e-mail may contain confidential and/or privileged=20
> information.
> >     If you are not the intended recipient (or have received=20
> this e-mail
> >     in error) please notify the sender immediately and delete this
> >     e-mail. Any unauthorized copying, disclosure or=20
> distribution of the
> >     contents in this e-mail is strictly forbidden.
> >     ---------------
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP





From sipping-bounces@ietf.org Tue Feb 13 06:27:52 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGvoK-00061s-VI; Tue, 13 Feb 2007 06:27:00 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGvoJ-00061J-Hp
	for sipping@ietf.org; Tue, 13 Feb 2007 06:26:59 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGvoC-000117-1T
	for sipping@ietf.org; Tue, 13 Feb 2007 06:26:59 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	7715820F98
	for <sipping@ietf.org>; Tue, 13 Feb 2007 12:26:47 +0100 (CET)
X-AuditID: c1b4fb3e-b06d4bb0000007e1-6f-45d1a077fb56 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	67BEC20483
	for <sipping@ietf.org>; Tue, 13 Feb 2007 12:26:47 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 12:26:46 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP - proxy types
Date: Tue, 13 Feb 2007 12:26:46 +0100
Message-ID: <7374777208BDC7449D5620EF9423256702047EAF@esealmw113.eemea.ericsson.se>
In-Reply-To: <7374777208BDC7449D5620EF9423256702047C66@esealmw113.eemea.ericsson.se>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP - proxy types
Thread-Index: AcdKpsbFDrW3FiTgTaqoCBJtt1RnoQEDfmXQACojU7A=
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 13 Feb 2007 11:26:46.0767 (UTC)
	FILETIME=[D898E7F0:01C74F61]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8b30eb7682a596edff707698f4a80f7d
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

=20
Third, why is GRUU mandated?

Regards,

Christer



> -----Original Message-----
> From: Christer Holmberg (JO/LMF)=20
> [mailto:christer.holmberg@ericsson.com]=20
> Sent: 13. helmikuuta 2007 12:23
> To: sipping@ietf.org
> Subject: [Sipping] Service Identification in SIP - proxy types
>=20
>=20
> Hi,
>=20
> First, the document talks about outbound proxy, home proxy,=20
> originating proxy and terminating proxy, but there is no real=20
> definition/explanation/reference to what those proxies really are.
>=20
> Second, the draft also uses Jonathan's loose route draft.=20
> Have we decided to work on that draft? I remember there was=20
> quite much discussion at one point, but what happened after=20
> that? In any case, a few words WHY the mechanism is needed=20
> would also be useful.
>=20
> Regards,
>=20
> Christer
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 06:37:35 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGvyB-0004r2-UY; Tue, 13 Feb 2007 06:37:11 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGvyA-0004qw-Co
	for sipping@ietf.org; Tue, 13 Feb 2007 06:37:10 -0500
Received: from mailgate.siemenscomms.co.uk ([195.171.110.225]
	helo=bemg01.siemenscomms.co.uk)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGvy9-0002FA-Mm
	for sipping@ietf.org; Tue, 13 Feb 2007 06:37:10 -0500
Received: from ntht207e.uksgcs.siemenscomms.co.uk ([137.223.247.82])
	by siemenscomms.co.uk (PMDF V6.0-24 #40642)
	with ESMTP id <0JDE00M7TG7W0U@siemenscomms.co.uk> for sipping@ietf.org;
	Tue, 13 Feb 2007 11:35:56 +0000 (GMT)
Received: by ntht207e.uksgcs.siemenscomms.co.uk with Internet Mail Service
	(5.5.2657.72)	id <D2K3XDWM>; Tue, 13 Feb 2007 11:35:57 +0000
Content-return: allowed
Date: Tue, 13 Feb 2007 11:35:54 +0000
From: "Elwell, John" <john.elwell@siemens.com>
Subject: RE: [SIPPING] SDP offer/answer	inearly media:draft-sawada-sipping
	-sip-offeranswer-01.txt
To: "Christer Holmberg (JO/LMF)" <christer.holmberg@ericsson.com>,
	Paul Kyzivat <pkyzivat@cisco.com>,
	Siddhartha Bhakta <Siddhartha.Bhakta@newport-networks.com>
Message-id: <50B1CBA96870A34799A506B2313F26670B05FD2C@ntht201e.siemenscomms.co.uk>
MIME-version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
Content-type: text/plain
Content-transfer-encoding: 7BIT
X-Spam-Score: 0.0 (/)
X-Scan-Signature: fb93e867a11a29ac1dc5018706b412ac
Cc: sipping@ietf.org, Siddhartha Bhakta <sbhakta007@yahoo.co.in>,
	Brian Stucker <bstucker@nortel.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Christer,

I don't know what you mean by "fake forking", but I think there is an issue
with serial forking, e.g.:
- INVITE request to user@example.com
- proxy for example.com forwards to PSTN gateway
- PSTN gateway provides early media
- proxy is scripted to cancel the PSTN call after a certain time and forward
the request elsewhere (e.g., to a mobile or voicemail)
- the new destination sends early media or answers quickly and sends regular
media
- the UAC will probably still be rendering early media from the PSTN and
will not know that this is in the process of stopping (even if packets stop
arriving, this could be due to silence)

In this case your proposed 19x provisional response could tell the UAC to
stop listening to the corresponding RTP stream immediately.

John

> -----Original Message-----
> From: Christer Holmberg (JO/LMF) 
> [mailto:christer.holmberg@ericsson.com] 
> Sent: 13 February 2007 10:57
> To: Paul Kyzivat; Siddhartha Bhakta
> Cc: sipping@ietf.org; Siddhartha Bhakta; Brian Stucker
> Subject: RE: [SIPPING] SDP offer/answer inearly 
> media:draft-sawada-sipping-sip-offeranswer-01.txt
> 
> 
> Hi,
> 
> I once proposed that it should be possible to terminate an 
> early dialog
> with an 19x provisional response.
> 
> Then, if "fake forking" is used, the dialog not used anymore could be
> terminted and the UAC knows he doesn't have to care about it (and
> possible media coming on it).
> 
> Regards,
> 
> Christer
>  
> 
> > -----Original Message-----
> > From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
> > Sent: 16. marraskuuta 2006 21:58
> > To: Siddhartha Bhakta
> > Cc: sipping@ietf.org; 'Siddhartha Bhakta'; 'Brian Stucker'
> > Subject: Re: [SIPPING] SDP offer/answer inearly 
> > media:draft-sawada-sipping-sip-offeranswer-01.txt
> > 
> > 
> > 
> > Siddhartha Bhakta wrote:
> > > I agree with you. No matter whether different provisional 
> responses 
> > > are creating different dialogs or not, case is same as 
> far as media 
> > > rendering/committing is concerned. For media committing, 
> there are 
> > > multiple answers for a given offer.
> > > 
> > > Like you, I also can not understand what we are gaining 
> by creating 
> > > multiple dialogs here except the chance to manage 
> > offer/answer state 
> > > as per RFC3261/RFC3264. It seems that multiple SDP answers 
> > for a given 
> > > SDP offer is the reality. We can not avoid that by applying 
> > any tricky 
> > > logic in dialog management (like Fake Forking).
> > 
> > If you ignore dialogs, it becomes impossible to impose any 
> > offer/answer rules.
> > 
> > By paying attention to the separate dialogs, it is possible 
> > to have offer/answer rules that make sense. If you just mix 
> > together offers or answers from multiple destinations it 
> > becomes impossible to make any sense of what you have once it 
> > is determined which dialog will be completed.
> > 
> > For the period when there are multiple dialogs, you have two 
> > different things going on:
> > - an offer/answer state machine is proceeding in each dialog, and
> >    defining the current media parameters for that dialog
> > - you are deciding which dialog(s) you want to process and send
> >    media to.
> > 
> > As Brian notes, it is complex to decide which dialog to honor 
> > during the early stage. But once you settle on one dialog at 
> > least you will then know what to do.
> > 
> > 	Paul
> > 
> > > -----Original Message-----
> > > From: Brian Stucker [mailto:bstucker@nortel.com]
> > > Sent: 16 November 2006 17:37
> > > To: Hadriel Kaplan; Siddhartha Bhakta; sipping@ietf.org
> > > Cc: Siddhartha Bhakta
> > > Subject: RE: [SIPPING] SDP offer/answer in early 
> > > media:draft-sawada-sipping-sip-offeranswer-01.txt
> > > 
> > > Fake forking...
> > > 
> > > Please note, if there's SDP in the other provisional 
> > responses, you're 
> > > going to run into problems potentially getting themedia rendered 
> > > (assuming that's your goal).
> > > 
> > > Regards,
> > > Brian
> > > 
> > >> -----Original Message-----
> > >> From: Hadriel Kaplan [mailto:HKaplan@acmepacket.com]
> > >> Sent: Sunday, November 12, 2006 8:40 PM
> > >> To: 'Siddhartha Bhakta'; sipping@ietf.org
> > >> Cc: 'Siddhartha Bhakta'
> > >> Subject: RE: [SIPPING] SDP offer/answer in early 
> > >> media:draft-sawada-sipping-sip-offeranswer-01.txt
> > >>
> > >> If the Proxy is sending back SDP by itself it's not a 
> proxy, but a 
> > >> b2bua, as is the PBX.
> > >> As such, it should be answering in a different dialog, 
> so its SDP 
> > >> answer should be in a 183 with a different To-tag from the 
> > 180, and 
> > >> the PBX's 180 SDP should have a different To-tag than 
> the 200ok.  
> > >> That would make them look like a forked call and follow the RFCs.
> > >>
> > >> Of course the reality is often different, but being a 
> > b2bua yourself 
> > >> it's in your power to fix it.  How you do that is not up 
> > to the IETF 
> > >> to define, as the very idea of it goes against the IETF's 
> > view of the 
> > >> world. (and I say that in a positive way, as I think it would be 
> > >> practically impossible for the IETF to do otherwise)
> > >>
> > >> -hadriel
> > >>
> > >>
> > >>> -----Original Message-----
> > >>> From: Siddhartha Bhakta [mailto:sbhakta007@yahoo.co.in]
> > >>> Sent: Sunday, November 12, 2006 2:11 PM
> > >>> To: sipping@ietf.org
> > >>> Cc: sbhakta007@yahoo.co.in; Siddhartha Bhakta
> > >>> Subject: [SIPPING] SDP offer/answer in early
> > >>> media:draft-sawada-sipping- sip-offeranswer-01.txt
> > >>>
> > >>> I am re-sending it as in last email, the figure was distorted.
> > >>>
> > >>> Though, post RFC3261 drafts indicate that one valid 
> > answer would be 
> > >>> there for one SDP offer, the our old friend early-media
> > >> flow is still
> > >>> present in practice.
> > >>>
> > >>> B2BUA/SBC     Proxy         PABX        Bob
> > >>>   |            |            |            |
> > >>>   | INVITE F1  |            |            |
> > >>>   |----------->| INVITE F2  |            |
> > >>>   |   183 F3   |----------->| INVITE F4  |
> > >>>   |<-----------|            |----------->|
> > >>>   |            |            |            |
> > >>>   |            |            |   180 without SDP F5
> > >>>   |            |            |<-----------|
> > >>>   |            |            |            |
> > >>>   |            |  180 with SDP F6        |
> > >>>   |   180 F7   |<-----------|            |
> > >>>   |<-----------|            |  200 F8    |
> > >>>   |            |  200 F9    |<-----------|
> > >>>   |   200 F10  |<-----------|            |
> > >>>   |<-----------|            |            |
> > >>>   |   ACK F11  |            |            |
> > >>>   |----------->|  ACK F12   |            |
> > >>>   |            |----------->|  ACK F13   |
> > >>>   |            |            |----------->|
> > >>>
> > >>> Our B2BUA is facing the problem as specified above.
> > >>> The first answer is carried
> > >>> by F3. This SDP is specified by Proxy. The 2nd SDP answer
> > >> is carried
> > >>> by F6,F7.
> > >>> In fact, this SDP is specified by PABX. The third SDP answer is 
> > >>> carried by F8,F9,F10.
> > >>> This SDP is specified by called party (Bob).
> > >>>
> > >>> As per the RFC3261 & RFC3264, the SDP answer carried by
> > >> F6,F7 should
> > >>> match with F3 as far as B2BUA is concerned. Therefore, 
> > B2BUA shall 
> > >>> ignore it. This is leading to the fact that SIP UA behind 
> > B2BUA can 
> > >>> not listen to ringtone fed by PABX.
> > >>> Similarly, SDP answer carried by F8,F9,F10 shall be 
> ignored. This 
> > >>> shall lead to the fact that SIP UA behind B2BUA can not 
> listen to 
> > >>> Bob's voice. Too bad.
> > >>> This problem is due to the fact that RFC3261 & RFC3264 are not 
> > >>> backward compatible.
> > >>>
> > >>>
> > >>> The sec 2.2. of
> > >>> draft-sawada-sipping-sip-offeranswer-01.txt indicates 
> that UPDATE 
> > >>> should be used to update the media in early media. But in
> > >> practice old
> > >>> early-media flow (i.e., sending different SDPs over 
> different 1xx 
> > >>> responses of
> > >>> INVITE) is quite common.
> > >>> Can we somehow make new SDP offer/answer specified in
> > >>> RFC3261 & RFC3264 backward
> > >>> compatible?
> > >>>
> > >>> The old standard (early-media), allows multiple SDP answers
> > >> of single
> > >>> SDP offer. can we somehow induce this thing in SDP
> > >> offer/answer model? 
> > >>> If many of you think the need of it then I may comeup with
> > >> some thumb
> > >>> rule for the same.
> > >>>
> > >>> Thanks & Regards,
> > >>> Siddhartha
> > >>
> > >> _______________________________________________
> > >> Sipping mailing list  
> > https://www1.ietf.org/mailman/listinfo/sipping
> > >> This list is for NEW development of the application of SIP Use 
> > >> sip-implementors@cs.columbia.edu for questions on 
> current sip Use 
> > >> sip@ietf.org for new developments of core SIP
> > >>
> > > 
> > > 
> > > 
> > > ---------------
> > > This e-mail may contain confidential and/or privileged 
> > information. If you are not the intended recipient (or have 
> > received this e-mail in error) please notify the sender 
> > immediately and delete this e-mail. Any unauthorized copying, 
> > disclosure or distribution of the contents in this e-mail is 
> > strictly forbidden.
> > > ---------------
> > > 
> > > 
> > > _______________________________________________
> > > Sipping mailing list  
> https://www1.ietf.org/mailman/listinfo/sipping
> > > This list is for NEW development of the application of SIP Use 
> > > sip-implementors@cs.columbia.edu for questions on current sip Use 
> > > sip@ietf.org for new developments of core SIP
> > > 
> > 
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP 
> > Use sip-implementors@cs.columbia.edu for questions on current 
> > sip Use sip@ietf.org for new developments of core SIP
> > 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 07:12:51 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGwVt-0007XA-NQ; Tue, 13 Feb 2007 07:12:01 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGwVr-0007Qw-Df
	for sipping@ietf.org; Tue, 13 Feb 2007 07:11:59 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGwVj-000826-C0
	for sipping@ietf.org; Tue, 13 Feb 2007 07:11:59 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	B26D720F2A; Tue, 13 Feb 2007 13:11:46 +0100 (CET)
X-AuditID: c1b4fb3e-b16d6bb0000007e1-4f-45d1ab0275c7 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	8803B204A5; Tue, 13 Feb 2007 13:11:46 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 13:11:46 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [SIPPING] SDP offer/answer	inearly
	media:draft-sawada-sipping-sip-offeranswer-01.txt
Date: Tue, 13 Feb 2007 13:11:45 +0100
Message-ID: <7374777208BDC7449D5620EF94232567020480BE@esealmw113.eemea.ericsson.se>
In-Reply-To: <50B1CBA96870A34799A506B2313F26670B05FD2C@ntht201e.siemenscomms.co.uk>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [SIPPING] SDP offer/answer	inearly
	media:draft-sawada-sipping-sip-offeranswer-01.txt
Thread-Index: AcdPY06RIufrIME6TceHiIKUn8q1qAABBKyA
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Elwell, John" <john.elwell@siemens.com>,
	"Paul Kyzivat" <pkyzivat@cisco.com>,
	"Siddhartha Bhakta" <Siddhartha.Bhakta@newport-networks.com>
X-OriginalArrivalTime: 13 Feb 2007 12:11:46.0035 (UTC)
	FILETIME=[217C8430:01C74F68]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: abda3837e791065a13ac6f11cf8e625a
Cc: sipping@ietf.org, Siddhartha Bhakta <sbhakta007@yahoo.co.in>,
	Brian Stucker <bstucker@nortel.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,=20

>I don't know what you mean by "fake forking",

I think your example shows pretty well what I mean :) The only
difference is that I refer to a scenario where the "forking proxy"
actually is a B2BUA that generates early media ITSELF.=20

So, it creates a dialog between itself and the UAC. Then it e.g. play
some announcment, and forwards the INVITE to the UAS. The UAS also
creates a dialog between itself and the UAC - which is separate
(different To tag) from the dialog between the UAC and B2BUA. Now, the
UAC sees two dialogs, so it thinks that forking has occured.

>but I think=20
>there is an issue with serial forking, e.g.:
>- INVITE request to user@example.com
>- proxy for example.com forwards to PSTN gateway
>- PSTN gateway provides early media
>- proxy is scripted to cancel the PSTN call after a certain=20
>time and forward the request elsewhere (e.g., to a mobile or=20
>voicemail)
>- the new destination sends early media or answers quickly=20
>and sends regular media
>- the UAC will probably still be rendering early media from=20
>the PSTN and will not know that this is in the process of=20
>stopping (even if packets stop arriving, this could be due to silence)
>=20
>In this case your proposed 19x provisional response could=20
>tell the UAC to stop listening to the corresponding RTP=20
>stream immediately.

Yes, that is the idea behind the 19x response.

Regadrs,

Christer





> > -----Original Message-----
> > From: Christer Holmberg (JO/LMF)
> > [mailto:christer.holmberg@ericsson.com]
> > Sent: 13 February 2007 10:57
> > To: Paul Kyzivat; Siddhartha Bhakta
> > Cc: sipping@ietf.org; Siddhartha Bhakta; Brian Stucker
> > Subject: RE: [SIPPING] SDP offer/answer inearly=20
> > media:draft-sawada-sipping-sip-offeranswer-01.txt
> >=20
> >=20
> > Hi,
> >=20
> > I once proposed that it should be possible to terminate an early=20
> > dialog with an 19x provisional response.
> >=20
> > Then, if "fake forking" is used, the dialog not used=20
> anymore could be=20
> > terminted and the UAC knows he doesn't have to care about it (and=20
> > possible media coming on it).
> >=20
> > Regards,
> >=20
> > Christer
> > =20
> >=20
> > > -----Original Message-----
> > > From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
> > > Sent: 16. marraskuuta 2006 21:58
> > > To: Siddhartha Bhakta
> > > Cc: sipping@ietf.org; 'Siddhartha Bhakta'; 'Brian Stucker'
> > > Subject: Re: [SIPPING] SDP offer/answer inearly=20
> > > media:draft-sawada-sipping-sip-offeranswer-01.txt
> > >=20
> > >=20
> > >=20
> > > Siddhartha Bhakta wrote:
> > > > I agree with you. No matter whether different provisional
> > responses
> > > > are creating different dialogs or not, case is same as
> > far as media
> > > > rendering/committing is concerned. For media committing,
> > there are
> > > > multiple answers for a given offer.
> > > >=20
> > > > Like you, I also can not understand what we are gaining
> > by creating
> > > > multiple dialogs here except the chance to manage
> > > offer/answer state
> > > > as per RFC3261/RFC3264. It seems that multiple SDP answers
> > > for a given
> > > > SDP offer is the reality. We can not avoid that by applying
> > > any tricky
> > > > logic in dialog management (like Fake Forking).
> > >=20
> > > If you ignore dialogs, it becomes impossible to impose any=20
> > > offer/answer rules.
> > >=20
> > > By paying attention to the separate dialogs, it is=20
> possible to have=20
> > > offer/answer rules that make sense. If you just mix=20
> together offers=20
> > > or answers from multiple destinations it becomes=20
> impossible to make=20
> > > any sense of what you have once it is determined which=20
> dialog will=20
> > > be completed.
> > >=20
> > > For the period when there are multiple dialogs, you have two=20
> > > different things going on:
> > > - an offer/answer state machine is proceeding in each dialog, and
> > >    defining the current media parameters for that dialog
> > > - you are deciding which dialog(s) you want to process and send
> > >    media to.
> > >=20
> > > As Brian notes, it is complex to decide which dialog to=20
> honor during=20
> > > the early stage. But once you settle on one dialog at=20
> least you will=20
> > > then know what to do.
> > >=20
> > > 	Paul
> > >=20
> > > > -----Original Message-----
> > > > From: Brian Stucker [mailto:bstucker@nortel.com]
> > > > Sent: 16 November 2006 17:37
> > > > To: Hadriel Kaplan; Siddhartha Bhakta; sipping@ietf.org
> > > > Cc: Siddhartha Bhakta
> > > > Subject: RE: [SIPPING] SDP offer/answer in early=20
> > > > media:draft-sawada-sipping-sip-offeranswer-01.txt
> > > >=20
> > > > Fake forking...
> > > >=20
> > > > Please note, if there's SDP in the other provisional
> > > responses, you're
> > > > going to run into problems potentially getting themedia=20
> rendered=20
> > > > (assuming that's your goal).
> > > >=20
> > > > Regards,
> > > > Brian
> > > >=20
> > > >> -----Original Message-----
> > > >> From: Hadriel Kaplan [mailto:HKaplan@acmepacket.com]
> > > >> Sent: Sunday, November 12, 2006 8:40 PM
> > > >> To: 'Siddhartha Bhakta'; sipping@ietf.org
> > > >> Cc: 'Siddhartha Bhakta'
> > > >> Subject: RE: [SIPPING] SDP offer/answer in early=20
> > > >> media:draft-sawada-sipping-sip-offeranswer-01.txt
> > > >>
> > > >> If the Proxy is sending back SDP by itself it's not a
> > proxy, but a
> > > >> b2bua, as is the PBX.
> > > >> As such, it should be answering in a different dialog,
> > so its SDP
> > > >> answer should be in a 183 with a different To-tag from the
> > > 180, and
> > > >> the PBX's 180 SDP should have a different To-tag than
> > the 200ok. =20
> > > >> That would make them look like a forked call and=20
> follow the RFCs.
> > > >>
> > > >> Of course the reality is often different, but being a
> > > b2bua yourself
> > > >> it's in your power to fix it.  How you do that is not up
> > > to the IETF
> > > >> to define, as the very idea of it goes against the IETF's
> > > view of the
> > > >> world. (and I say that in a positive way, as I think=20
> it would be=20
> > > >> practically impossible for the IETF to do otherwise)
> > > >>
> > > >> -hadriel
> > > >>
> > > >>
> > > >>> -----Original Message-----
> > > >>> From: Siddhartha Bhakta [mailto:sbhakta007@yahoo.co.in]
> > > >>> Sent: Sunday, November 12, 2006 2:11 PM
> > > >>> To: sipping@ietf.org
> > > >>> Cc: sbhakta007@yahoo.co.in; Siddhartha Bhakta
> > > >>> Subject: [SIPPING] SDP offer/answer in early
> > > >>> media:draft-sawada-sipping- sip-offeranswer-01.txt
> > > >>>
> > > >>> I am re-sending it as in last email, the figure was distorted.
> > > >>>
> > > >>> Though, post RFC3261 drafts indicate that one valid
> > > answer would be
> > > >>> there for one SDP offer, the our old friend early-media
> > > >> flow is still
> > > >>> present in practice.
> > > >>>
> > > >>> B2BUA/SBC     Proxy         PABX        Bob
> > > >>>   |            |            |            |
> > > >>>   | INVITE F1  |            |            |
> > > >>>   |----------->| INVITE F2  |            |
> > > >>>   |   183 F3   |----------->| INVITE F4  |
> > > >>>   |<-----------|            |----------->|
> > > >>>   |            |            |            |
> > > >>>   |            |            |   180 without SDP F5
> > > >>>   |            |            |<-----------|
> > > >>>   |            |            |            |
> > > >>>   |            |  180 with SDP F6        |
> > > >>>   |   180 F7   |<-----------|            |
> > > >>>   |<-----------|            |  200 F8    |
> > > >>>   |            |  200 F9    |<-----------|
> > > >>>   |   200 F10  |<-----------|            |
> > > >>>   |<-----------|            |            |
> > > >>>   |   ACK F11  |            |            |
> > > >>>   |----------->|  ACK F12   |            |
> > > >>>   |            |----------->|  ACK F13   |
> > > >>>   |            |            |----------->|
> > > >>>
> > > >>> Our B2BUA is facing the problem as specified above.
> > > >>> The first answer is carried
> > > >>> by F3. This SDP is specified by Proxy. The 2nd SDP answer
> > > >> is carried
> > > >>> by F6,F7.
> > > >>> In fact, this SDP is specified by PABX. The third SDP=20
> answer is=20
> > > >>> carried by F8,F9,F10.
> > > >>> This SDP is specified by called party (Bob).
> > > >>>
> > > >>> As per the RFC3261 & RFC3264, the SDP answer carried by
> > > >> F6,F7 should
> > > >>> match with F3 as far as B2BUA is concerned. Therefore,
> > > B2BUA shall
> > > >>> ignore it. This is leading to the fact that SIP UA behind
> > > B2BUA can
> > > >>> not listen to ringtone fed by PABX.
> > > >>> Similarly, SDP answer carried by F8,F9,F10 shall be
> > ignored. This
> > > >>> shall lead to the fact that SIP UA behind B2BUA can not
> > listen to
> > > >>> Bob's voice. Too bad.
> > > >>> This problem is due to the fact that RFC3261 &=20
> RFC3264 are not=20
> > > >>> backward compatible.
> > > >>>
> > > >>>
> > > >>> The sec 2.2. of
> > > >>> draft-sawada-sipping-sip-offeranswer-01.txt indicates
> > that UPDATE
> > > >>> should be used to update the media in early media. But in
> > > >> practice old
> > > >>> early-media flow (i.e., sending different SDPs over
> > different 1xx
> > > >>> responses of
> > > >>> INVITE) is quite common.
> > > >>> Can we somehow make new SDP offer/answer specified in
> > > >>> RFC3261 & RFC3264 backward
> > > >>> compatible?
> > > >>>
> > > >>> The old standard (early-media), allows multiple SDP answers
> > > >> of single
> > > >>> SDP offer. can we somehow induce this thing in SDP
> > > >> offer/answer model?=20
> > > >>> If many of you think the need of it then I may comeup with
> > > >> some thumb
> > > >>> rule for the same.
> > > >>>
> > > >>> Thanks & Regards,
> > > >>> Siddhartha
> > > >>
> > > >> _______________________________________________
> > > >> Sipping mailing list
> > > https://www1.ietf.org/mailman/listinfo/sipping
> > > >> This list is for NEW development of the application of SIP Use=20
> > > >> sip-implementors@cs.columbia.edu for questions on
> > current sip Use
> > > >> sip@ietf.org for new developments of core SIP
> > > >>
> > > >=20
> > > >=20
> > > >=20
> > > > ---------------
> > > > This e-mail may contain confidential and/or privileged
> > > information. If you are not the intended recipient (or=20
> have received=20
> > > this e-mail in error) please notify the sender immediately and=20
> > > delete this e-mail. Any unauthorized copying, disclosure or=20
> > > distribution of the contents in this e-mail is strictly forbidden.
> > > > ---------------
> > > >=20
> > > >=20
> > > > _______________________________________________
> > > > Sipping mailing list
> > https://www1.ietf.org/mailman/listinfo/sipping
> > > > This list is for NEW development of the application of SIP Use=20
> > > > sip-implementors@cs.columbia.edu for questions on=20
> current sip Use=20
> > > > sip@ietf.org for new developments of core SIP
> > > >=20
> > >=20
> > > _______________________________________________
> > > Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> > > This list is for NEW development of the application of SIP Use=20
> > > sip-implementors@cs.columbia.edu for questions on current sip Use=20
> > > sip@ietf.org for new developments of core SIP
> > >=20
> >=20
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP Use=20
> > sip-implementors@cs.columbia.edu for questions on current sip Use=20
> > sip@ietf.org for new developments of core SIP
> >=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 08:10:00 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGxOi-00036x-5o; Tue, 13 Feb 2007 08:08:40 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGxOg-000358-Os
	for sipping@ietf.org; Tue, 13 Feb 2007 08:08:38 -0500
Received: from zrtps0kp.nortel.com ([47.140.192.56])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGxOe-0004C4-5Q
	for sipping@ietf.org; Tue, 13 Feb 2007 08:08:38 -0500
Received: from zrc2hxm2.corp.nortel.com (zrc2hxm2.corp.nortel.com
	[47.103.123.73])
	by zrtps0kp.nortel.com (Switch-2.2.6/Switch-2.2.0) with ESMTP id
	l1DD8R828136; Tue, 13 Feb 2007 08:08:27 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [SIPPING] SDP offer/answer	inearly
	media:draft-sawada-sipping-sip-offeranswer-01.txt
Date: Tue, 13 Feb 2007 07:08:26 -0600
Message-ID: <6FC4416DDE56C44DA0AEE67BC7CA437112B05283@zrc2hxm2.corp.nortel.com>
In-Reply-To: <7374777208BDC7449D5620EF94232567020480BE@esealmw113.eemea.ericsson.se>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [SIPPING] SDP offer/answer	inearly
	media:draft-sawada-sipping-sip-offeranswer-01.txt
Thread-index: AcdPY06RIufrIME6TceHiIKUn8q1qAABBKyAAAHFdtA=
From: "Brian Stucker" <bstucker@nortel.com>
To: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>,
	"Elwell, John" <john.elwell@siemens.com>,
	"Paul Kyzivat" <pkyzivat@cisco.com>,
	"Siddhartha Bhakta" <Siddhartha.Bhakta@newport-networks.com>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 5b943e80df8c8cad631fd60298783617
Cc: sipping@ietf.org, Siddhartha Bhakta <sbhakta007@yahoo.co.in>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Fake forking is a classic example of the middle screwing up a process
that should only be between final endpoints.

A box in the middle forks the call off to an endpoint that the
originator never intended to call that is acting as if it is one of the
called party's devices, when it is not, and has no intention of ever
answering the call.

Although a 19x response might make knowing when early media from the
interloper to the SDP exchange is no longer necessary, the same thing
can be handled by (most) clients today by simply sending a 180 back to
the originator to switch back to local ringing when that endpoint no
longer has any media it wishes the originator to render (not that the
originator had to render anything in the first place, but that's a
separate discussion).

What would be nice is if each dialog was handled explicitly, rather than
implicitly at the originator for early media (i.e. the originator
explicitly contacts the device that wishes to generate early media
rather than having the network resort to tricking the originator into
rendering the desired media).

I have a draft sitting in the queue that proposes such a mechanism. The
XML2RFC template still used (C) 2006 when I generated the draft text, so
it got bounced during the I-D nits check. Until it appears in the
archives, anyone interested can email me and I'll send you a copy of the
draft.

Regards,
Brian

> -----Original Message-----
> From: Christer Holmberg (JO/LMF)=20
> [mailto:christer.holmberg@ericsson.com]=20
> Sent: Tuesday, February 13, 2007 9:12 PM
> To: Elwell, John; Paul Kyzivat; Siddhartha Bhakta
> Cc: sipping@ietf.org; Siddhartha Bhakta; Stucker, Brian (RICH1:AR00)
> Subject: RE: [SIPPING] SDP offer/answer inearly=20
> media:draft-sawada-sipping-sip-offeranswer-01.txt
>=20
>=20
> Hi,=20
>=20
> >I don't know what you mean by "fake forking",
>=20
> I think your example shows pretty well what I mean :) The=20
> only difference is that I refer to a scenario where the=20
> "forking proxy"
> actually is a B2BUA that generates early media ITSELF.=20
>=20
> So, it creates a dialog between itself and the UAC. Then it=20
> e.g. play some announcment, and forwards the INVITE to the=20
> UAS. The UAS also creates a dialog between itself and the UAC=20
> - which is separate (different To tag) from the dialog=20
> between the UAC and B2BUA. Now, the UAC sees two dialogs, so=20
> it thinks that forking has occured.
>=20
> >but I think
> >there is an issue with serial forking, e.g.:
> >- INVITE request to user@example.com
> >- proxy for example.com forwards to PSTN gateway
> >- PSTN gateway provides early media
> >- proxy is scripted to cancel the PSTN call after a certain time and=20
> >forward the request elsewhere (e.g., to a mobile or
> >voicemail)
> >- the new destination sends early media or answers quickly and sends=20
> >regular media
> >- the UAC will probably still be rendering early media from the PSTN=20
> >and will not know that this is in the process of stopping (even if=20
> >packets stop arriving, this could be due to silence)
> >=20
> >In this case your proposed 19x provisional response could=20
> tell the UAC=20
> >to stop listening to the corresponding RTP stream immediately.
>=20
> Yes, that is the idea behind the 19x response.
>=20
> Regadrs,
>=20
> Christer
>=20
>=20
>=20
>=20
>=20
> > > -----Original Message-----
> > > From: Christer Holmberg (JO/LMF)
> > > [mailto:christer.holmberg@ericsson.com]
> > > Sent: 13 February 2007 10:57
> > > To: Paul Kyzivat; Siddhartha Bhakta
> > > Cc: sipping@ietf.org; Siddhartha Bhakta; Brian Stucker
> > > Subject: RE: [SIPPING] SDP offer/answer inearly=20
> > > media:draft-sawada-sipping-sip-offeranswer-01.txt
> > >=20
> > >=20
> > > Hi,
> > >=20
> > > I once proposed that it should be possible to terminate an early=20
> > > dialog with an 19x provisional response.
> > >=20
> > > Then, if "fake forking" is used, the dialog not used
> > anymore could be
> > > terminted and the UAC knows he doesn't have to care about it (and=20
> > > possible media coming on it).
> > >=20
> > > Regards,
> > >=20
> > > Christer
> > > =20
> > >=20
> > > > -----Original Message-----
> > > > From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
> > > > Sent: 16. marraskuuta 2006 21:58
> > > > To: Siddhartha Bhakta
> > > > Cc: sipping@ietf.org; 'Siddhartha Bhakta'; 'Brian Stucker'
> > > > Subject: Re: [SIPPING] SDP offer/answer inearly=20
> > > > media:draft-sawada-sipping-sip-offeranswer-01.txt
> > > >=20
> > > >=20
> > > >=20
> > > > Siddhartha Bhakta wrote:
> > > > > I agree with you. No matter whether different provisional
> > > responses
> > > > > are creating different dialogs or not, case is same as
> > > far as media
> > > > > rendering/committing is concerned. For media committing,
> > > there are
> > > > > multiple answers for a given offer.
> > > > >=20
> > > > > Like you, I also can not understand what we are gaining
> > > by creating
> > > > > multiple dialogs here except the chance to manage
> > > > offer/answer state
> > > > > as per RFC3261/RFC3264. It seems that multiple SDP answers
> > > > for a given
> > > > > SDP offer is the reality. We can not avoid that by applying
> > > > any tricky
> > > > > logic in dialog management (like Fake Forking).
> > > >=20
> > > > If you ignore dialogs, it becomes impossible to impose any=20
> > > > offer/answer rules.
> > > >=20
> > > > By paying attention to the separate dialogs, it is
> > possible to have
> > > > offer/answer rules that make sense. If you just mix
> > together offers
> > > > or answers from multiple destinations it becomes
> > impossible to make
> > > > any sense of what you have once it is determined which
> > dialog will
> > > > be completed.
> > > >=20
> > > > For the period when there are multiple dialogs, you have two=20
> > > > different things going on:
> > > > - an offer/answer state machine is proceeding in each=20
> dialog, and
> > > >    defining the current media parameters for that dialog
> > > > - you are deciding which dialog(s) you want to process and send
> > > >    media to.
> > > >=20
> > > > As Brian notes, it is complex to decide which dialog to
> > honor during
> > > > the early stage. But once you settle on one dialog at
> > least you will
> > > > then know what to do.
> > > >=20
> > > > 	Paul
> > > >=20
> > > > > -----Original Message-----
> > > > > From: Brian Stucker [mailto:bstucker@nortel.com]
> > > > > Sent: 16 November 2006 17:37
> > > > > To: Hadriel Kaplan; Siddhartha Bhakta; sipping@ietf.org
> > > > > Cc: Siddhartha Bhakta
> > > > > Subject: RE: [SIPPING] SDP offer/answer in early=20
> > > > > media:draft-sawada-sipping-sip-offeranswer-01.txt
> > > > >=20
> > > > > Fake forking...
> > > > >=20
> > > > > Please note, if there's SDP in the other provisional
> > > > responses, you're
> > > > > going to run into problems potentially getting themedia
> > rendered
> > > > > (assuming that's your goal).
> > > > >=20
> > > > > Regards,
> > > > > Brian
> > > > >=20
> > > > >> -----Original Message-----
> > > > >> From: Hadriel Kaplan [mailto:HKaplan@acmepacket.com]
> > > > >> Sent: Sunday, November 12, 2006 8:40 PM
> > > > >> To: 'Siddhartha Bhakta'; sipping@ietf.org
> > > > >> Cc: 'Siddhartha Bhakta'
> > > > >> Subject: RE: [SIPPING] SDP offer/answer in early=20
> > > > >> media:draft-sawada-sipping-sip-offeranswer-01.txt
> > > > >>
> > > > >> If the Proxy is sending back SDP by itself it's not a
> > > proxy, but a
> > > > >> b2bua, as is the PBX.
> > > > >> As such, it should be answering in a different dialog,
> > > so its SDP
> > > > >> answer should be in a 183 with a different To-tag from the
> > > > 180, and
> > > > >> the PBX's 180 SDP should have a different To-tag than
> > > the 200ok. =20
> > > > >> That would make them look like a forked call and
> > follow the RFCs.
> > > > >>
> > > > >> Of course the reality is often different, but being a
> > > > b2bua yourself
> > > > >> it's in your power to fix it.  How you do that is not up
> > > > to the IETF
> > > > >> to define, as the very idea of it goes against the IETF's
> > > > view of the
> > > > >> world. (and I say that in a positive way, as I think
> > it would be
> > > > >> practically impossible for the IETF to do otherwise)
> > > > >>
> > > > >> -hadriel
> > > > >>
> > > > >>
> > > > >>> -----Original Message-----
> > > > >>> From: Siddhartha Bhakta [mailto:sbhakta007@yahoo.co.in]
> > > > >>> Sent: Sunday, November 12, 2006 2:11 PM
> > > > >>> To: sipping@ietf.org
> > > > >>> Cc: sbhakta007@yahoo.co.in; Siddhartha Bhakta
> > > > >>> Subject: [SIPPING] SDP offer/answer in early
> > > > >>> media:draft-sawada-sipping- sip-offeranswer-01.txt
> > > > >>>
> > > > >>> I am re-sending it as in last email, the figure was=20
> distorted.
> > > > >>>
> > > > >>> Though, post RFC3261 drafts indicate that one valid
> > > > answer would be
> > > > >>> there for one SDP offer, the our old friend early-media
> > > > >> flow is still
> > > > >>> present in practice.
> > > > >>>
> > > > >>> B2BUA/SBC     Proxy         PABX        Bob
> > > > >>>   |            |            |            |
> > > > >>>   | INVITE F1  |            |            |
> > > > >>>   |----------->| INVITE F2  |            |
> > > > >>>   |   183 F3   |----------->| INVITE F4  |
> > > > >>>   |<-----------|            |----------->|
> > > > >>>   |            |            |            |
> > > > >>>   |            |            |   180 without SDP F5
> > > > >>>   |            |            |<-----------|
> > > > >>>   |            |            |            |
> > > > >>>   |            |  180 with SDP F6        |
> > > > >>>   |   180 F7   |<-----------|            |
> > > > >>>   |<-----------|            |  200 F8    |
> > > > >>>   |            |  200 F9    |<-----------|
> > > > >>>   |   200 F10  |<-----------|            |
> > > > >>>   |<-----------|            |            |
> > > > >>>   |   ACK F11  |            |            |
> > > > >>>   |----------->|  ACK F12   |            |
> > > > >>>   |            |----------->|  ACK F13   |
> > > > >>>   |            |            |----------->|
> > > > >>>
> > > > >>> Our B2BUA is facing the problem as specified above.
> > > > >>> The first answer is carried
> > > > >>> by F3. This SDP is specified by Proxy. The 2nd SDP answer
> > > > >> is carried
> > > > >>> by F6,F7.
> > > > >>> In fact, this SDP is specified by PABX. The third SDP
> > answer is
> > > > >>> carried by F8,F9,F10.
> > > > >>> This SDP is specified by called party (Bob).
> > > > >>>
> > > > >>> As per the RFC3261 & RFC3264, the SDP answer carried by
> > > > >> F6,F7 should
> > > > >>> match with F3 as far as B2BUA is concerned. Therefore,
> > > > B2BUA shall
> > > > >>> ignore it. This is leading to the fact that SIP UA behind
> > > > B2BUA can
> > > > >>> not listen to ringtone fed by PABX.
> > > > >>> Similarly, SDP answer carried by F8,F9,F10 shall be
> > > ignored. This
> > > > >>> shall lead to the fact that SIP UA behind B2BUA can not
> > > listen to
> > > > >>> Bob's voice. Too bad.
> > > > >>> This problem is due to the fact that RFC3261 &
> > RFC3264 are not
> > > > >>> backward compatible.
> > > > >>>
> > > > >>>
> > > > >>> The sec 2.2. of
> > > > >>> draft-sawada-sipping-sip-offeranswer-01.txt indicates
> > > that UPDATE
> > > > >>> should be used to update the media in early media. But in
> > > > >> practice old
> > > > >>> early-media flow (i.e., sending different SDPs over
> > > different 1xx
> > > > >>> responses of
> > > > >>> INVITE) is quite common.
> > > > >>> Can we somehow make new SDP offer/answer specified in
> > > > >>> RFC3261 & RFC3264 backward
> > > > >>> compatible?
> > > > >>>
> > > > >>> The old standard (early-media), allows multiple SDP answers
> > > > >> of single
> > > > >>> SDP offer. can we somehow induce this thing in SDP
> > > > >> offer/answer model?=20
> > > > >>> If many of you think the need of it then I may comeup with
> > > > >> some thumb
> > > > >>> rule for the same.
> > > > >>>
> > > > >>> Thanks & Regards,
> > > > >>> Siddhartha
> > > > >>
> > > > >> _______________________________________________
> > > > >> Sipping mailing list
> > > > https://www1.ietf.org/mailman/listinfo/sipping
> > > > >> This list is for NEW development of the application=20
> of SIP Use=20
> > > > >> sip-implementors@cs.columbia.edu for questions on
> > > current sip Use
> > > > >> sip@ietf.org for new developments of core SIP
> > > > >>
> > > > >=20
> > > > >=20
> > > > >=20
> > > > > ---------------
> > > > > This e-mail may contain confidential and/or privileged
> > > > information. If you are not the intended recipient (or
> > have received
> > > > this e-mail in error) please notify the sender immediately and=20
> > > > delete this e-mail. Any unauthorized copying, disclosure or=20
> > > > distribution of the contents in this e-mail is strictly=20
> forbidden.
> > > > > ---------------
> > > > >=20
> > > > >=20
> > > > > _______________________________________________
> > > > > Sipping mailing list
> > > https://www1.ietf.org/mailman/listinfo/sipping
> > > > > This list is for NEW development of the application=20
> of SIP Use=20
> > > > > sip-implementors@cs.columbia.edu for questions on
> > current sip Use
> > > > > sip@ietf.org for new developments of core SIP
> > > > >=20
> > > >=20
> > > > _______________________________________________
> > > > Sipping mailing list
> > https://www1.ietf.org/mailman/listinfo/sipping
> > > > This list is for NEW development of the application of SIP Use=20
> > > > sip-implementors@cs.columbia.edu for questions on=20
> current sip Use=20
> > > > sip@ietf.org for new developments of core SIP
> > > >=20
> > >=20
> > > _______________________________________________
> > > Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> > > This list is for NEW development of the application of SIP Use=20
> > > sip-implementors@cs.columbia.edu for questions on current sip Use=20
> > > sip@ietf.org for new developments of core SIP
> > >=20
> >=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 08:32:33 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGxlO-0007Y3-EM; Tue, 13 Feb 2007 08:32:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGxlN-0007XV-CN
	for sipping@ietf.org; Tue, 13 Feb 2007 08:32:05 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGxlD-00085W-MX
	for sipping@ietf.org; Tue, 13 Feb 2007 08:32:05 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	232A92117B; Tue, 13 Feb 2007 14:31:53 +0100 (CET)
X-AuditID: c1b4fb3e-b1ed7bb0000007e1-f9-45d1bdc94cf6 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	0445F21173; Tue, 13 Feb 2007 14:31:53 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 14:31:52 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [SIPPING] SDP offer/answer	inearly
	media:draft-sawada-sipping-sip-offeranswer-01.txt
Date: Tue, 13 Feb 2007 14:31:51 +0100
Message-ID: <7374777208BDC7449D5620EF942325670206A53E@esealmw113.eemea.ericsson.se>
In-Reply-To: <6FC4416DDE56C44DA0AEE67BC7CA437112B05283@zrc2hxm2.corp.nortel.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [SIPPING] SDP offer/answer	inearly
	media:draft-sawada-sipping-sip-offeranswer-01.txt
Thread-Index: AcdPY06RIufrIME6TceHiIKUn8q1qAABBKyAAAHFdtAAAOJ2YA==
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Brian Stucker" <bstucker@nortel.com>,
	"Elwell, John" <john.elwell@siemens.com>,
	"Paul Kyzivat" <pkyzivat@cisco.com>,
	"Siddhartha Bhakta" <Siddhartha.Bhakta@newport-networks.com>
X-OriginalArrivalTime: 13 Feb 2007 13:31:52.0744 (UTC)
	FILETIME=[52821A80:01C74F73]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 6fc5b1c74c5bed09a3a9da2884900dec
Cc: sipping@ietf.org, Siddhartha Bhakta <sbhakta007@yahoo.co.in>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,=20

>Fake forking is a classic example of the middle screwing up a=20
>process that should only be between final endpoints.

If the B2BUA is providing early media it IS an endpoint.
=20
>A box in the middle forks the call off to an endpoint that=20
>the originator never intended to call that is acting as if it=20
>is one of the called party's devices, when it is not, and has=20
>no intention of ever answering the call.

So what? It could as well be an answering machine, or something else
which is never going to answer the call.

Anyway, I think this is more an architecture issue, and has nothing to
do with early media such. From a UAC perspective the call has been
forked, so from that perspective it really doesn't matter if the media
comes from a "real" UAS or from a node in the network.

>Although a 19x response might make knowing when early media=20
>from the interloper to the SDP exchange is no longer=20
>necessary, the same thing can be handled by (most) clients=20
>today by simply sending a 180 back to the originator to=20
>switch back to local ringing when that endpoint no longer has=20
>any media it wishes the originator to render (not that the=20
>originator had to render anything in the first place, but=20
>that's a separate discussion).

The simply-sending-back-a-180 is not a standardized mechanism.

And, 19x would help also with other things than early media. For
example, you can release QoS related (or other) resources reserved for
that specific dialog, and you can terminate pre-condition procedures.
19x would not only say "you will not get any media on this dialog" - it
would say "you will not get anything on this dialog, because it doesn't
exist anymore".
=20
>What would be nice is if each dialog was handled explicitly,=20
>rather than implicitly at the originator for early media=20
>(i.e. the originator explicitly contacts the device that=20
>wishes to generate early media rather than having the network=20
>resort to tricking the originator into rendering the desired media).

I don't understand how that would work. Some of the early media is
generated because I am trying to call YOU.

>I have a draft sitting in the queue that proposes such a=20
>mechanism. The XML2RFC template still used (C) 2006 when I=20
>generated the draft text, so it got bounced during the I-D=20
>nits check. Until it appears in the archives, anyone=20
>interested can email me and I'll send you a copy of the draft.

Please do.

Regards,

Christer






> > -----Original Message-----
> > From: Christer Holmberg (JO/LMF)
> > [mailto:christer.holmberg@ericsson.com]
> > Sent: Tuesday, February 13, 2007 9:12 PM
> > To: Elwell, John; Paul Kyzivat; Siddhartha Bhakta
> > Cc: sipping@ietf.org; Siddhartha Bhakta; Stucker, Brian (RICH1:AR00)
> > Subject: RE: [SIPPING] SDP offer/answer inearly=20
> > media:draft-sawada-sipping-sip-offeranswer-01.txt
> >=20
> >=20
> > Hi,
> >=20
> > >I don't know what you mean by "fake forking",
> >=20
> > I think your example shows pretty well what I mean :) The only=20
> > difference is that I refer to a scenario where the "forking proxy"
> > actually is a B2BUA that generates early media ITSELF.=20
> >=20
> > So, it creates a dialog between itself and the UAC. Then it=20
> e.g. play=20
> > some announcment, and forwards the INVITE to the UAS. The UAS also=20
> > creates a dialog between itself and the UAC
> > - which is separate (different To tag) from the dialog=20
> between the UAC=20
> > and B2BUA. Now, the UAC sees two dialogs, so it thinks that forking=20
> > has occured.
> >=20
> > >but I think
> > >there is an issue with serial forking, e.g.:
> > >- INVITE request to user@example.com
> > >- proxy for example.com forwards to PSTN gateway
> > >- PSTN gateway provides early media
> > >- proxy is scripted to cancel the PSTN call after a=20
> certain time and=20
> > >forward the request elsewhere (e.g., to a mobile or
> > >voicemail)
> > >- the new destination sends early media or answers quickly=20
> and sends=20
> > >regular media
> > >- the UAC will probably still be rendering early media=20
> from the PSTN=20
> > >and will not know that this is in the process of stopping (even if=20
> > >packets stop arriving, this could be due to silence)
> > >=20
> > >In this case your proposed 19x provisional response could
> > tell the UAC
> > >to stop listening to the corresponding RTP stream immediately.
> >=20
> > Yes, that is the idea behind the 19x response.
> >=20
> > Regadrs,
> >=20
> > Christer
> >=20
> >=20
> >=20
> >=20
> >=20
> > > > -----Original Message-----
> > > > From: Christer Holmberg (JO/LMF)
> > > > [mailto:christer.holmberg@ericsson.com]
> > > > Sent: 13 February 2007 10:57
> > > > To: Paul Kyzivat; Siddhartha Bhakta
> > > > Cc: sipping@ietf.org; Siddhartha Bhakta; Brian Stucker
> > > > Subject: RE: [SIPPING] SDP offer/answer inearly=20
> > > > media:draft-sawada-sipping-sip-offeranswer-01.txt
> > > >=20
> > > >=20
> > > > Hi,
> > > >=20
> > > > I once proposed that it should be possible to terminate=20
> an early=20
> > > > dialog with an 19x provisional response.
> > > >=20
> > > > Then, if "fake forking" is used, the dialog not used
> > > anymore could be
> > > > terminted and the UAC knows he doesn't have to care=20
> about it (and=20
> > > > possible media coming on it).
> > > >=20
> > > > Regards,
> > > >=20
> > > > Christer
> > > > =20
> > > >=20
> > > > > -----Original Message-----
> > > > > From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
> > > > > Sent: 16. marraskuuta 2006 21:58
> > > > > To: Siddhartha Bhakta
> > > > > Cc: sipping@ietf.org; 'Siddhartha Bhakta'; 'Brian Stucker'
> > > > > Subject: Re: [SIPPING] SDP offer/answer inearly=20
> > > > > media:draft-sawada-sipping-sip-offeranswer-01.txt
> > > > >=20
> > > > >=20
> > > > >=20
> > > > > Siddhartha Bhakta wrote:
> > > > > > I agree with you. No matter whether different provisional
> > > > responses
> > > > > > are creating different dialogs or not, case is same as
> > > > far as media
> > > > > > rendering/committing is concerned. For media committing,
> > > > there are
> > > > > > multiple answers for a given offer.
> > > > > >=20
> > > > > > Like you, I also can not understand what we are gaining
> > > > by creating
> > > > > > multiple dialogs here except the chance to manage
> > > > > offer/answer state
> > > > > > as per RFC3261/RFC3264. It seems that multiple SDP answers
> > > > > for a given
> > > > > > SDP offer is the reality. We can not avoid that by applying
> > > > > any tricky
> > > > > > logic in dialog management (like Fake Forking).
> > > > >=20
> > > > > If you ignore dialogs, it becomes impossible to impose any=20
> > > > > offer/answer rules.
> > > > >=20
> > > > > By paying attention to the separate dialogs, it is
> > > possible to have
> > > > > offer/answer rules that make sense. If you just mix
> > > together offers
> > > > > or answers from multiple destinations it becomes
> > > impossible to make
> > > > > any sense of what you have once it is determined which
> > > dialog will
> > > > > be completed.
> > > > >=20
> > > > > For the period when there are multiple dialogs, you have two=20
> > > > > different things going on:
> > > > > - an offer/answer state machine is proceeding in each
> > dialog, and
> > > > >    defining the current media parameters for that dialog
> > > > > - you are deciding which dialog(s) you want to=20
> process and send
> > > > >    media to.
> > > > >=20
> > > > > As Brian notes, it is complex to decide which dialog to
> > > honor during
> > > > > the early stage. But once you settle on one dialog at
> > > least you will
> > > > > then know what to do.
> > > > >=20
> > > > > 	Paul
> > > > >=20
> > > > > > -----Original Message-----
> > > > > > From: Brian Stucker [mailto:bstucker@nortel.com]
> > > > > > Sent: 16 November 2006 17:37
> > > > > > To: Hadriel Kaplan; Siddhartha Bhakta; sipping@ietf.org
> > > > > > Cc: Siddhartha Bhakta
> > > > > > Subject: RE: [SIPPING] SDP offer/answer in early=20
> > > > > > media:draft-sawada-sipping-sip-offeranswer-01.txt
> > > > > >=20
> > > > > > Fake forking...
> > > > > >=20
> > > > > > Please note, if there's SDP in the other provisional
> > > > > responses, you're
> > > > > > going to run into problems potentially getting themedia
> > > rendered
> > > > > > (assuming that's your goal).
> > > > > >=20
> > > > > > Regards,
> > > > > > Brian
> > > > > >=20
> > > > > >> -----Original Message-----
> > > > > >> From: Hadriel Kaplan [mailto:HKaplan@acmepacket.com]
> > > > > >> Sent: Sunday, November 12, 2006 8:40 PM
> > > > > >> To: 'Siddhartha Bhakta'; sipping@ietf.org
> > > > > >> Cc: 'Siddhartha Bhakta'
> > > > > >> Subject: RE: [SIPPING] SDP offer/answer in early=20
> > > > > >> media:draft-sawada-sipping-sip-offeranswer-01.txt
> > > > > >>
> > > > > >> If the Proxy is sending back SDP by itself it's not a
> > > > proxy, but a
> > > > > >> b2bua, as is the PBX.
> > > > > >> As such, it should be answering in a different dialog,
> > > > so its SDP
> > > > > >> answer should be in a 183 with a different To-tag from the
> > > > > 180, and
> > > > > >> the PBX's 180 SDP should have a different To-tag than
> > > > the 200ok. =20
> > > > > >> That would make them look like a forked call and
> > > follow the RFCs.
> > > > > >>
> > > > > >> Of course the reality is often different, but being a
> > > > > b2bua yourself
> > > > > >> it's in your power to fix it.  How you do that is not up
> > > > > to the IETF
> > > > > >> to define, as the very idea of it goes against the IETF's
> > > > > view of the
> > > > > >> world. (and I say that in a positive way, as I think
> > > it would be
> > > > > >> practically impossible for the IETF to do otherwise)
> > > > > >>
> > > > > >> -hadriel
> > > > > >>
> > > > > >>
> > > > > >>> -----Original Message-----
> > > > > >>> From: Siddhartha Bhakta [mailto:sbhakta007@yahoo.co.in]
> > > > > >>> Sent: Sunday, November 12, 2006 2:11 PM
> > > > > >>> To: sipping@ietf.org
> > > > > >>> Cc: sbhakta007@yahoo.co.in; Siddhartha Bhakta
> > > > > >>> Subject: [SIPPING] SDP offer/answer in early
> > > > > >>> media:draft-sawada-sipping- sip-offeranswer-01.txt
> > > > > >>>
> > > > > >>> I am re-sending it as in last email, the figure was
> > distorted.
> > > > > >>>
> > > > > >>> Though, post RFC3261 drafts indicate that one valid
> > > > > answer would be
> > > > > >>> there for one SDP offer, the our old friend early-media
> > > > > >> flow is still
> > > > > >>> present in practice.
> > > > > >>>
> > > > > >>> B2BUA/SBC     Proxy         PABX        Bob
> > > > > >>>   |            |            |            |
> > > > > >>>   | INVITE F1  |            |            |
> > > > > >>>   |----------->| INVITE F2  |            |
> > > > > >>>   |   183 F3   |----------->| INVITE F4  |
> > > > > >>>   |<-----------|            |----------->|
> > > > > >>>   |            |            |            |
> > > > > >>>   |            |            |   180 without SDP F5
> > > > > >>>   |            |            |<-----------|
> > > > > >>>   |            |            |            |
> > > > > >>>   |            |  180 with SDP F6        |
> > > > > >>>   |   180 F7   |<-----------|            |
> > > > > >>>   |<-----------|            |  200 F8    |
> > > > > >>>   |            |  200 F9    |<-----------|
> > > > > >>>   |   200 F10  |<-----------|            |
> > > > > >>>   |<-----------|            |            |
> > > > > >>>   |   ACK F11  |            |            |
> > > > > >>>   |----------->|  ACK F12   |            |
> > > > > >>>   |            |----------->|  ACK F13   |
> > > > > >>>   |            |            |----------->|
> > > > > >>>
> > > > > >>> Our B2BUA is facing the problem as specified above.
> > > > > >>> The first answer is carried
> > > > > >>> by F3. This SDP is specified by Proxy. The 2nd SDP answer
> > > > > >> is carried
> > > > > >>> by F6,F7.
> > > > > >>> In fact, this SDP is specified by PABX. The third SDP
> > > answer is
> > > > > >>> carried by F8,F9,F10.
> > > > > >>> This SDP is specified by called party (Bob).
> > > > > >>>
> > > > > >>> As per the RFC3261 & RFC3264, the SDP answer carried by
> > > > > >> F6,F7 should
> > > > > >>> match with F3 as far as B2BUA is concerned. Therefore,
> > > > > B2BUA shall
> > > > > >>> ignore it. This is leading to the fact that SIP UA behind
> > > > > B2BUA can
> > > > > >>> not listen to ringtone fed by PABX.
> > > > > >>> Similarly, SDP answer carried by F8,F9,F10 shall be
> > > > ignored. This
> > > > > >>> shall lead to the fact that SIP UA behind B2BUA can not
> > > > listen to
> > > > > >>> Bob's voice. Too bad.
> > > > > >>> This problem is due to the fact that RFC3261 &
> > > RFC3264 are not
> > > > > >>> backward compatible.
> > > > > >>>
> > > > > >>>
> > > > > >>> The sec 2.2. of
> > > > > >>> draft-sawada-sipping-sip-offeranswer-01.txt indicates
> > > > that UPDATE
> > > > > >>> should be used to update the media in early media. But in
> > > > > >> practice old
> > > > > >>> early-media flow (i.e., sending different SDPs over
> > > > different 1xx
> > > > > >>> responses of
> > > > > >>> INVITE) is quite common.
> > > > > >>> Can we somehow make new SDP offer/answer specified in
> > > > > >>> RFC3261 & RFC3264 backward
> > > > > >>> compatible?
> > > > > >>>
> > > > > >>> The old standard (early-media), allows multiple=20
> SDP answers
> > > > > >> of single
> > > > > >>> SDP offer. can we somehow induce this thing in SDP
> > > > > >> offer/answer model?=20
> > > > > >>> If many of you think the need of it then I may comeup with
> > > > > >> some thumb
> > > > > >>> rule for the same.
> > > > > >>>
> > > > > >>> Thanks & Regards,
> > > > > >>> Siddhartha
> > > > > >>
> > > > > >> _______________________________________________
> > > > > >> Sipping mailing list
> > > > > https://www1.ietf.org/mailman/listinfo/sipping
> > > > > >> This list is for NEW development of the application
> > of SIP Use
> > > > > >> sip-implementors@cs.columbia.edu for questions on
> > > > current sip Use
> > > > > >> sip@ietf.org for new developments of core SIP
> > > > > >>
> > > > > >=20
> > > > > >=20
> > > > > >=20
> > > > > > ---------------
> > > > > > This e-mail may contain confidential and/or privileged
> > > > > information. If you are not the intended recipient (or
> > > have received
> > > > > this e-mail in error) please notify the sender=20
> immediately and=20
> > > > > delete this e-mail. Any unauthorized copying, disclosure or=20
> > > > > distribution of the contents in this e-mail is strictly
> > forbidden.
> > > > > > ---------------
> > > > > >=20
> > > > > >=20
> > > > > > _______________________________________________
> > > > > > Sipping mailing list
> > > > https://www1.ietf.org/mailman/listinfo/sipping
> > > > > > This list is for NEW development of the application
> > of SIP Use
> > > > > > sip-implementors@cs.columbia.edu for questions on
> > > current sip Use
> > > > > > sip@ietf.org for new developments of core SIP
> > > > > >=20
> > > > >=20
> > > > > _______________________________________________
> > > > > Sipping mailing list
> > > https://www1.ietf.org/mailman/listinfo/sipping
> > > > > This list is for NEW development of the application=20
> of SIP Use=20
> > > > > sip-implementors@cs.columbia.edu for questions on
> > current sip Use
> > > > > sip@ietf.org for new developments of core SIP
> > > > >=20
> > > >=20
> > > > _______________________________________________
> > > > Sipping mailing list
> > https://www1.ietf.org/mailman/listinfo/sipping
> > > > This list is for NEW development of the application of SIP Use=20
> > > > sip-implementors@cs.columbia.edu for questions on=20
> current sip Use=20
> > > > sip@ietf.org for new developments of core SIP
> > > >=20
> > >=20
> >=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 08:35:25 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGxoQ-0000LR-OF; Tue, 13 Feb 2007 08:35:14 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGxoP-0000Jw-Qm
	for sipping@ietf.org; Tue, 13 Feb 2007 08:35:13 -0500
Received: from mailgate.siemenscomms.co.uk ([195.171.110.225]
	helo=bemg01.siemenscomms.co.uk)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGxoO-00007X-T5
	for sipping@ietf.org; Tue, 13 Feb 2007 08:35:13 -0500
Received: from ntht207e.uksgcs.siemenscomms.co.uk ([137.223.247.82])
	by siemenscomms.co.uk (PMDF V6.0-24 #40642)
	with ESMTP id <0JDE002C2LONZW@siemenscomms.co.uk> for sipping@ietf.org;
	Tue, 13 Feb 2007 13:33:59 +0000 (GMT)
Received: by ntht207e.uksgcs.siemenscomms.co.uk with Internet Mail Service
	(5.5.2657.72)	id <D2K3XHAH>; Tue, 13 Feb 2007 13:33:59 +0000
Content-return: allowed
Date: Tue, 13 Feb 2007 13:33:58 +0000
From: "Elwell, John" <john.elwell@siemens.com>
Subject: RE: [SIPPING] SDP offer/answer	inearly media:draft-sawada-sipping
	-sip-offeranswer-01.txt
To: Brian Stucker <bstucker@nortel.com>,
	"Christer Holmberg (JO/LMF)" <christer.holmberg@ericsson.com>,
	Paul Kyzivat <pkyzivat@cisco.com>,
	Siddhartha Bhakta <Siddhartha.Bhakta@newport-networks.com>
Message-id: <50B1CBA96870A34799A506B2313F26670B05FEF4@ntht201e.siemenscomms.co.uk>
MIME-version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
Content-type: text/plain
Content-transfer-encoding: 7BIT
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 919b3965bd46f7460d234f848680b238
Cc: sipping@ietf.org, Siddhartha Bhakta <sbhakta007@yahoo.co.in>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Brian,

See below,

John 

> -----Original Message-----
> From: Brian Stucker [mailto:bstucker@nortel.com] 
> Sent: 13 February 2007 13:08
> To: Christer Holmberg (JO/LMF); Elwell, John; Paul Kyzivat; 
> Siddhartha Bhakta
> Cc: sipping@ietf.org; Siddhartha Bhakta
> Subject: RE: [SIPPING] SDP offer/answer inearly 
> media:draft-sawada-sipping-sip-offeranswer-01.txt
> 
> Fake forking is a classic example of the middle screwing up a process
> that should only be between final endpoints.
> 
> A box in the middle forks the call off to an endpoint that the
> originator never intended to call that is acting as if it is 
> one of the
> called party's devices, when it is not, and has no intention of ever
> answering the call.
> 
> Although a 19x response might make knowing when early media from the
> interloper to the SDP exchange is no longer necessary, the same thing
> can be handled by (most) clients today by simply sending a 180 back to
> the originator to switch back to local ringing when that endpoint no
> longer has any media it wishes the originator to render (not that the
> originator had to render anything in the first place, but that's a
> separate discussion).
[JRE] I don't think 180 means that. It means play ringback tone until media
starts to arrive. If media is already arriving, I don't think 180 has the
semantics of stop rendering media and revert to ringback tone. I don't even
think it has the semantics of wait until media appears to have stopped and
then revert to ringback tone. I am not saying that one of these would not be
an unreasonable thing to do, and also I am not denying that some
implementations might do one of these already. But I don't know where it is
specified in an RFC. Perhaps some clarification is needed.

> 
> What would be nice is if each dialog was handled explicitly, 
> rather than
> implicitly at the originator for early media (i.e. the originator
> explicitly contacts the device that wishes to generate early media
> rather than having the network resort to tricking the originator into
> rendering the desired media).
> 
> I have a draft sitting in the queue that proposes such a 
> mechanism. The
> XML2RFC template still used (C) 2006 when I generated the 
> draft text, so
> it got bounced during the I-D nits check. Until it appears in the
> archives, anyone interested can email me and I'll send you a 
> copy of the
> draft.
[JRE] Yes please

> 
> Regards,
> Brian
> 
> > -----Original Message-----
> > From: Christer Holmberg (JO/LMF) 
> > [mailto:christer.holmberg@ericsson.com] 
> > Sent: Tuesday, February 13, 2007 9:12 PM
> > To: Elwell, John; Paul Kyzivat; Siddhartha Bhakta
> > Cc: sipping@ietf.org; Siddhartha Bhakta; Stucker, Brian (RICH1:AR00)
> > Subject: RE: [SIPPING] SDP offer/answer inearly 
> > media:draft-sawada-sipping-sip-offeranswer-01.txt
> > 
> > 
> > Hi, 
> > 
> > >I don't know what you mean by "fake forking",
> > 
> > I think your example shows pretty well what I mean :) The 
> > only difference is that I refer to a scenario where the 
> > "forking proxy"
> > actually is a B2BUA that generates early media ITSELF. 
> > 
> > So, it creates a dialog between itself and the UAC. Then it 
> > e.g. play some announcment, and forwards the INVITE to the 
> > UAS. The UAS also creates a dialog between itself and the UAC 
> > - which is separate (different To tag) from the dialog 
> > between the UAC and B2BUA. Now, the UAC sees two dialogs, so 
> > it thinks that forking has occured.
> > 
> > >but I think
> > >there is an issue with serial forking, e.g.:
> > >- INVITE request to user@example.com
> > >- proxy for example.com forwards to PSTN gateway
> > >- PSTN gateway provides early media
> > >- proxy is scripted to cancel the PSTN call after a 
> certain time and 
> > >forward the request elsewhere (e.g., to a mobile or
> > >voicemail)
> > >- the new destination sends early media or answers quickly 
> and sends 
> > >regular media
> > >- the UAC will probably still be rendering early media 
> from the PSTN 
> > >and will not know that this is in the process of stopping (even if 
> > >packets stop arriving, this could be due to silence)
> > > 
> > >In this case your proposed 19x provisional response could 
> > tell the UAC 
> > >to stop listening to the corresponding RTP stream immediately.
> > 
> > Yes, that is the idea behind the 19x response.
> > 
> > Regadrs,
> > 
> > Christer
> > 
> > 
> > 
> > 
> > 
> > > > -----Original Message-----
> > > > From: Christer Holmberg (JO/LMF)
> > > > [mailto:christer.holmberg@ericsson.com]
> > > > Sent: 13 February 2007 10:57
> > > > To: Paul Kyzivat; Siddhartha Bhakta
> > > > Cc: sipping@ietf.org; Siddhartha Bhakta; Brian Stucker
> > > > Subject: RE: [SIPPING] SDP offer/answer inearly 
> > > > media:draft-sawada-sipping-sip-offeranswer-01.txt
> > > > 
> > > > 
> > > > Hi,
> > > > 
> > > > I once proposed that it should be possible to terminate 
> an early 
> > > > dialog with an 19x provisional response.
> > > > 
> > > > Then, if "fake forking" is used, the dialog not used
> > > anymore could be
> > > > terminted and the UAC knows he doesn't have to care 
> about it (and 
> > > > possible media coming on it).
> > > > 
> > > > Regards,
> > > > 
> > > > Christer
> > > >  
> > > > 
> > > > > -----Original Message-----
> > > > > From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
> > > > > Sent: 16. marraskuuta 2006 21:58
> > > > > To: Siddhartha Bhakta
> > > > > Cc: sipping@ietf.org; 'Siddhartha Bhakta'; 'Brian Stucker'
> > > > > Subject: Re: [SIPPING] SDP offer/answer inearly 
> > > > > media:draft-sawada-sipping-sip-offeranswer-01.txt
> > > > > 
> > > > > 
> > > > > 
> > > > > Siddhartha Bhakta wrote:
> > > > > > I agree with you. No matter whether different provisional
> > > > responses
> > > > > > are creating different dialogs or not, case is same as
> > > > far as media
> > > > > > rendering/committing is concerned. For media committing,
> > > > there are
> > > > > > multiple answers for a given offer.
> > > > > > 
> > > > > > Like you, I also can not understand what we are gaining
> > > > by creating
> > > > > > multiple dialogs here except the chance to manage
> > > > > offer/answer state
> > > > > > as per RFC3261/RFC3264. It seems that multiple SDP answers
> > > > > for a given
> > > > > > SDP offer is the reality. We can not avoid that by applying
> > > > > any tricky
> > > > > > logic in dialog management (like Fake Forking).
> > > > > 
> > > > > If you ignore dialogs, it becomes impossible to impose any 
> > > > > offer/answer rules.
> > > > > 
> > > > > By paying attention to the separate dialogs, it is
> > > possible to have
> > > > > offer/answer rules that make sense. If you just mix
> > > together offers
> > > > > or answers from multiple destinations it becomes
> > > impossible to make
> > > > > any sense of what you have once it is determined which
> > > dialog will
> > > > > be completed.
> > > > > 
> > > > > For the period when there are multiple dialogs, you have two 
> > > > > different things going on:
> > > > > - an offer/answer state machine is proceeding in each 
> > dialog, and
> > > > >    defining the current media parameters for that dialog
> > > > > - you are deciding which dialog(s) you want to 
> process and send
> > > > >    media to.
> > > > > 
> > > > > As Brian notes, it is complex to decide which dialog to
> > > honor during
> > > > > the early stage. But once you settle on one dialog at
> > > least you will
> > > > > then know what to do.
> > > > > 
> > > > > 	Paul
> > > > > 
> > > > > > -----Original Message-----
> > > > > > From: Brian Stucker [mailto:bstucker@nortel.com]
> > > > > > Sent: 16 November 2006 17:37
> > > > > > To: Hadriel Kaplan; Siddhartha Bhakta; sipping@ietf.org
> > > > > > Cc: Siddhartha Bhakta
> > > > > > Subject: RE: [SIPPING] SDP offer/answer in early 
> > > > > > media:draft-sawada-sipping-sip-offeranswer-01.txt
> > > > > > 
> > > > > > Fake forking...
> > > > > > 
> > > > > > Please note, if there's SDP in the other provisional
> > > > > responses, you're
> > > > > > going to run into problems potentially getting themedia
> > > rendered
> > > > > > (assuming that's your goal).
> > > > > > 
> > > > > > Regards,
> > > > > > Brian
> > > > > > 
> > > > > >> -----Original Message-----
> > > > > >> From: Hadriel Kaplan [mailto:HKaplan@acmepacket.com]
> > > > > >> Sent: Sunday, November 12, 2006 8:40 PM
> > > > > >> To: 'Siddhartha Bhakta'; sipping@ietf.org
> > > > > >> Cc: 'Siddhartha Bhakta'
> > > > > >> Subject: RE: [SIPPING] SDP offer/answer in early 
> > > > > >> media:draft-sawada-sipping-sip-offeranswer-01.txt
> > > > > >>
> > > > > >> If the Proxy is sending back SDP by itself it's not a
> > > > proxy, but a
> > > > > >> b2bua, as is the PBX.
> > > > > >> As such, it should be answering in a different dialog,
> > > > so its SDP
> > > > > >> answer should be in a 183 with a different To-tag from the
> > > > > 180, and
> > > > > >> the PBX's 180 SDP should have a different To-tag than
> > > > the 200ok.  
> > > > > >> That would make them look like a forked call and
> > > follow the RFCs.
> > > > > >>
> > > > > >> Of course the reality is often different, but being a
> > > > > b2bua yourself
> > > > > >> it's in your power to fix it.  How you do that is not up
> > > > > to the IETF
> > > > > >> to define, as the very idea of it goes against the IETF's
> > > > > view of the
> > > > > >> world. (and I say that in a positive way, as I think
> > > it would be
> > > > > >> practically impossible for the IETF to do otherwise)
> > > > > >>
> > > > > >> -hadriel
> > > > > >>
> > > > > >>
> > > > > >>> -----Original Message-----
> > > > > >>> From: Siddhartha Bhakta [mailto:sbhakta007@yahoo.co.in]
> > > > > >>> Sent: Sunday, November 12, 2006 2:11 PM
> > > > > >>> To: sipping@ietf.org
> > > > > >>> Cc: sbhakta007@yahoo.co.in; Siddhartha Bhakta
> > > > > >>> Subject: [SIPPING] SDP offer/answer in early
> > > > > >>> media:draft-sawada-sipping- sip-offeranswer-01.txt
> > > > > >>>
> > > > > >>> I am re-sending it as in last email, the figure was 
> > distorted.
> > > > > >>>
> > > > > >>> Though, post RFC3261 drafts indicate that one valid
> > > > > answer would be
> > > > > >>> there for one SDP offer, the our old friend early-media
> > > > > >> flow is still
> > > > > >>> present in practice.
> > > > > >>>
> > > > > >>> B2BUA/SBC     Proxy         PABX        Bob
> > > > > >>>   |            |            |            |
> > > > > >>>   | INVITE F1  |            |            |
> > > > > >>>   |----------->| INVITE F2  |            |
> > > > > >>>   |   183 F3   |----------->| INVITE F4  |
> > > > > >>>   |<-----------|            |----------->|
> > > > > >>>   |            |            |            |
> > > > > >>>   |            |            |   180 without SDP F5
> > > > > >>>   |            |            |<-----------|
> > > > > >>>   |            |            |            |
> > > > > >>>   |            |  180 with SDP F6        |
> > > > > >>>   |   180 F7   |<-----------|            |
> > > > > >>>   |<-----------|            |  200 F8    |
> > > > > >>>   |            |  200 F9    |<-----------|
> > > > > >>>   |   200 F10  |<-----------|            |
> > > > > >>>   |<-----------|            |            |
> > > > > >>>   |   ACK F11  |            |            |
> > > > > >>>   |----------->|  ACK F12   |            |
> > > > > >>>   |            |----------->|  ACK F13   |
> > > > > >>>   |            |            |----------->|
> > > > > >>>
> > > > > >>> Our B2BUA is facing the problem as specified above.
> > > > > >>> The first answer is carried
> > > > > >>> by F3. This SDP is specified by Proxy. The 2nd SDP answer
> > > > > >> is carried
> > > > > >>> by F6,F7.
> > > > > >>> In fact, this SDP is specified by PABX. The third SDP
> > > answer is
> > > > > >>> carried by F8,F9,F10.
> > > > > >>> This SDP is specified by called party (Bob).
> > > > > >>>
> > > > > >>> As per the RFC3261 & RFC3264, the SDP answer carried by
> > > > > >> F6,F7 should
> > > > > >>> match with F3 as far as B2BUA is concerned. Therefore,
> > > > > B2BUA shall
> > > > > >>> ignore it. This is leading to the fact that SIP UA behind
> > > > > B2BUA can
> > > > > >>> not listen to ringtone fed by PABX.
> > > > > >>> Similarly, SDP answer carried by F8,F9,F10 shall be
> > > > ignored. This
> > > > > >>> shall lead to the fact that SIP UA behind B2BUA can not
> > > > listen to
> > > > > >>> Bob's voice. Too bad.
> > > > > >>> This problem is due to the fact that RFC3261 &
> > > RFC3264 are not
> > > > > >>> backward compatible.
> > > > > >>>
> > > > > >>>
> > > > > >>> The sec 2.2. of
> > > > > >>> draft-sawada-sipping-sip-offeranswer-01.txt indicates
> > > > that UPDATE
> > > > > >>> should be used to update the media in early media. But in
> > > > > >> practice old
> > > > > >>> early-media flow (i.e., sending different SDPs over
> > > > different 1xx
> > > > > >>> responses of
> > > > > >>> INVITE) is quite common.
> > > > > >>> Can we somehow make new SDP offer/answer specified in
> > > > > >>> RFC3261 & RFC3264 backward
> > > > > >>> compatible?
> > > > > >>>
> > > > > >>> The old standard (early-media), allows multiple 
> SDP answers
> > > > > >> of single
> > > > > >>> SDP offer. can we somehow induce this thing in SDP
> > > > > >> offer/answer model? 
> > > > > >>> If many of you think the need of it then I may comeup with
> > > > > >> some thumb
> > > > > >>> rule for the same.
> > > > > >>>
> > > > > >>> Thanks & Regards,
> > > > > >>> Siddhartha
> > > > > >>
> > > > > >> _______________________________________________
> > > > > >> Sipping mailing list
> > > > > https://www1.ietf.org/mailman/listinfo/sipping
> > > > > >> This list is for NEW development of the application 
> > of SIP Use 
> > > > > >> sip-implementors@cs.columbia.edu for questions on
> > > > current sip Use
> > > > > >> sip@ietf.org for new developments of core SIP
> > > > > >>
> > > > > > 
> > > > > > 
> > > > > > 
> > > > > > ---------------
> > > > > > This e-mail may contain confidential and/or privileged
> > > > > information. If you are not the intended recipient (or
> > > have received
> > > > > this e-mail in error) please notify the sender 
> immediately and 
> > > > > delete this e-mail. Any unauthorized copying, disclosure or 
> > > > > distribution of the contents in this e-mail is strictly 
> > forbidden.
> > > > > > ---------------
> > > > > > 
> > > > > > 
> > > > > > _______________________________________________
> > > > > > Sipping mailing list
> > > > https://www1.ietf.org/mailman/listinfo/sipping
> > > > > > This list is for NEW development of the application 
> > of SIP Use 
> > > > > > sip-implementors@cs.columbia.edu for questions on
> > > current sip Use
> > > > > > sip@ietf.org for new developments of core SIP
> > > > > > 
> > > > > 
> > > > > _______________________________________________
> > > > > Sipping mailing list
> > > https://www1.ietf.org/mailman/listinfo/sipping
> > > > > This list is for NEW development of the application 
> of SIP Use 
> > > > > sip-implementors@cs.columbia.edu for questions on 
> > current sip Use 
> > > > > sip@ietf.org for new developments of core SIP
> > > > > 
> > > > 
> > > > _______________________________________________
> > > > Sipping mailing list  
> > https://www1.ietf.org/mailman/listinfo/sipping
> > > > This list is for NEW development of the application of SIP Use 
> > > > sip-implementors@cs.columbia.edu for questions on 
> current sip Use 
> > > > sip@ietf.org for new developments of core SIP
> > > > 
> > > 
> > 
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 08:39:14 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGxsB-0002iF-CJ; Tue, 13 Feb 2007 08:39:07 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGxsA-0002gZ-PG
	for sipping@ietf.org; Tue, 13 Feb 2007 08:39:06 -0500
Received: from rtp-iport-1.cisco.com ([64.102.122.148])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGxs8-00010J-6n
	for sipping@ietf.org; Tue, 13 Feb 2007 08:39:06 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by rtp-iport-1.cisco.com with ESMTP; 13 Feb 2007 08:39:04 -0500
X-IronPort-AV: i="4.14,163,1170651600"; 
	d="scan'208"; a="52906692:sNHT67849040"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1DDd4wd022844; 
	Tue, 13 Feb 2007 08:39:04 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1DDd2VV019185; 
	Tue, 13 Feb 2007 08:39:02 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 08:39:02 -0500
Received: from [161.44.182.226] ([161.44.182.226]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 08:39:02 -0500
Message-ID: <45D1BF75.5000801@cisco.com>
Date: Tue, 13 Feb 2007 08:39:01 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Christer Holmberg (JO/LMF)" <christer.holmberg@ericsson.com>
Subject: Re: [Sipping] SDP
	Rollbackindraft-sawada-sipping-sip-offeranswer-01.txt
References: <7374777208BDC7449D5620EF9423256702047D98@esealmw113.eemea.ericsson.se>
In-Reply-To: <7374777208BDC7449D5620EF9423256702047D98@esealmw113.eemea.ericsson.se>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 13 Feb 2007 13:39:02.0438 (UTC)
	FILETIME=[52A03460:01C74F74]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=9924; t=1171373944;
	x=1172237944; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20SDP=20Rollbackindraft-sawada-sipping-sip-
	offeranswer-01.txt |Sender:=20
	|To:=20=22Christer=20Holmberg=20(JO/LMF)=22=20<christer.holmberg@ericsson
	.com>; bh=3K8xzIuyL72SHP+f0ZbHdcl0yfdcIEPLwBN32JQC9BU=;
	b=fkJA61MoPNQY7NgoCNv2221JJnk1keZtvPJUoWg+WVgwVGAeXCTOIYtK6w+DIhykt+fC8w3M
	5ZVHLbFQKj9p2eADFHYgIFGqo8nQWaSwge5zteZcB3vIxhG90eqY3Wwn;
Authentication-Results: rtp-dkim-2; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8a4bcf8f67063cac573319207fe3db35
Cc: "Sanjay Sinha \(sanjsinh\)" <sanjsinh@cisco.com>, tu-sawada@kddi.com,
	sipping@ietf.org,
	Siddhartha Bhakta <Siddhartha.Bhakta@newport-networks.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Christer Holmberg (JO/LMF) wrote:
> Hi,
> 
> One approach would of course be to not allow reliable provisional
> responses for re-INVITEs in the first place.
> 
> But, then we MAY have a problem with preconditions - but not
> necessarily.
> 
> The SDP answerer could simply choose not to send 200 OK until he has
> reserved the needed resources. Ie he does not send any provisional
> responses.
> 
> The SDP offerer would then send a second re-INVITE/UPDATE when he has
> reserved the needed resources (unfortunatly he can't use the ACK to
> indicate it).

This would present some serious problems in the case of offerless 
reinvites. In that case, delaying the offer to the 200 means that the 
UAC has no way to reject the offer.

Nevertheless, restricting the offer/answer patterns that are allowed in 
reinvite (more restrictive than what is permitted in the initial invite) 
could be effective.

The question is whether it is a more attractive alternative than either 
of the others mentioned. To me it comes down to trading off simplicity, 
flexibility, and backward compatibility. I don't see simplicity or 
flexibility as being markedly different among these either. So it may be 
that this should be judged on the basis of backwards compatibility. But 
that may be hard to get good info on. Ideally we should take some sort 
of survey, but in the end I suppose that is roughly equivalent to voting.

	Paul

> Regards,
> 
> Christer
> 
>  
> 
>> -----Original Message-----
>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
>> Sent: 17. marraskuuta 2006 23:03
>> To: Sanjay Sinha (sanjsinh)
>> Cc: tu-sawada@kddi.com; sipping@ietf.org; Siddhartha Bhakta
>> Subject: Re: [Sipping] SDP 
>> Rollbackindraft-sawada-sipping-sip-offeranswer-01.txt
>>
>> (To those who haven't been paying attention, this is about 
>> what to do about SDP when a reINVITE fails. RFC 3261, section 
>> 14.1, says that if a non-2xx final response is received the 
>> session parameters must remain unchanged, as if no reINVITE 
>> had been issued. Herein that is being described as "rollback" 
>> of the SDP.)
>>
>> There are several interesting questions relative to this topic:
>>
>> - are the existing RFCs clear about what should be done?
>>    (Do all informed and reasonable people agree what this is?)
>>
>> - what do currently deployed implementations do?
>>
>> - what is the *most reasonable* thing to do?
>>
>> There has been quite a bit of discussion in the past on this, 
>> though we could probably some more because I think only a few 
>> people participated.
>>
>> IMO the existing RFCs are *not* clear, because we are getting 
>> multiple interpretations that can be justified to some 
>> extent. The rollback rule only appears in 3261, and doesn't 
>> really address the additional issues raised by 3262, 3311, and 3312.
>>
>> There are also good arguments for either of two approaches 
>> being the most reasonable thing to do.
>>
>> The two competing interpretations are:
>>
>> 1) once an answer has been transmitted reliably it is committed and
>>     will not be rolled back, even if a reinvite fails.
>>
>> 2) an entire sequence of offers and answers (there could be many)
>>     that is initiated by a reinvite, is rolled back if the reinvite
>>     fails.
>>
>> Of these, (1) is somewhat easier to implement. It also 
>> eliminates some race conditions that seem unresolvable, 
>> though obscure, that can arise with (2). The downside is that 
>> when multiple offers and answers are used to resolve 
>> preconditions, and then the reinvite fails, you end up in an 
>> interim state with unresolved preconditions. The proper 
>> behavior in that state is not clear.
>>
>> With (2) the precondition problem goes away. A few ambiguous 
>> race conditions have been identified, but they are obscure 
>> cases that can probably be resolved by just just writing some 
>> best practices to avoid them. Its not entirely clear to me if 
>> this approach requires retention of any more state than (1), 
>> but if so it is only slightly more.
>>
>> It would be helpful if people that have opinions on this 
>> subject identify which approach they advocate and why. I'm 
>> not expecting to see a new approach different from either of 
>> the above, but if somebody has one, please spell out how it 
>> differs from these, and why it is better.
>>
>> 	Thanks,
>> 	Paul
>>
>> Sanjay Sinha (sanjsinh) wrote:
>>> SDP rollback applies only if previously there has been an 
>> offer-answer 
>>> exchange and an early dialog or a confirmed dialog has been 
>> established.
>>> In that case, if SDP in re-Invite/UPDATE is rejected, then session 
>>> continues with previously negotiated characteristics. This is 
>>> mentioned in RFC 3261. And this draft does mention that if offer in 
>>> Prack or in a sip response is rejected, then answerer has 
>> to send an 
>>> updated offer. So I am not sure how is the new proposal 
>> different than 
>>> what has been already stated.
>>>  
>>> Sanjay
>>>
>>>     
>> --------------------------------------------------------------
>> ----------
>>>     *From:* Siddhartha Bhakta
>>>     [mailto:Siddhartha.Bhakta@newport-networks.com]
>>>     *Sent:* Tuesday, November 14, 2006 12:37 PM
>>>     *To:* tu-sawada@kddi.com; Paul Kyzivat (pkyzivat)
>>>     *Cc:* sipping@ietf.org
>>>     *Subject:* [Sipping] SDP Rollback
>>>     indraft-sawada-sipping-sip-offeranswer-01.txt
>>>
>>>     Hi,
>>>
>>>      
>>>
>>>     I am not sure (as I am a very new user to SIPPING
>>>
>>>     group) whether SDP rollback has ever been discussed
>>>
>>>     regarding this draft or not. I personally feel that
>>>
>>>     'SDP rollback' is very relevant topics regarding this
>>>
>>>     draft.
>>>
>>>      
>>>
>>>     The sec 1.2 indicates the response(488) to reject any
>>>
>>>     SDP offer. But the missing part is that if any SIP
>>>
>>>     request carrying SDP offer is rejected by 3xx-6xx then
>>>
>>>     SDP offer should be rolled-back. The rollback means,
>>>
>>>     applying last session description.
>>>
>>>      
>>>
>>>     Rather than specifying different SIP messages
>>>
>>>     separately, it would be better to have some thumb rule
>>>
>>>     that shall be quite easy to follow as far as SDP
>>>
>>>     rollback is concerned. I am proposing following thumb
>>>
>>>     rule. Please indicate whether that makes sense or not.
>>>
>>>      
>>>
>>>     The SDP rollback shall be associated with transaction
>>>
>>>     rollback/rejection.
>>>
>>>      
>>>
>>>     [1] If the transaction request (carrying SDP offer) is
>>>
>>>     rejected then SDP offer shall be rolled-back. The
>>>
>>>     exception is the ACK request. The ACK (of 2xx) can not
>>>
>>>     be rejected. (The ACK of 3xx-6xx is not the
>>>
>>>     transaction initiating request).
>>>
>>>      
>>>
>>>     [2] If any transaction request carries SDP answer
>>>
>>>     (e.g., ACK or PRACK) then that transaction can not
>>>
>>>     rejected as there shall be the confusion over whether
>>>
>>>     SDP shall be rolled-back or not. I suppose, SDP answer
>>>
>>>     can not be rolled-back. There is no confusion for ACK,
>>>
>>>     as it can not rejected but for PRACK the restriction
>>>
>>>     should be there that it can not be rejected if it
>>>
>>>     carries SDP answer.
>>>
>>>      
>>>
>>>     This draft says that PRACK(irrespective of whether it
>>>
>>>     is carrying SDP offer/answer) can not be rejected. I
>>>
>>>     can not understand the rationale about this statement.
>>>
>>>     Though, I can appreciate why PRACK MUST not be
>>>
>>>     rejected if it carries SDP answer.
>>>
>>>      
>>>
>>>     [3] If any SIP response contains SDP offer then that
>>>
>>>     SDP can not be rolled-back. If any SIP entity wants to
>>>
>>>     rollback the SDP offer carried by SIP response, it
>>>
>>>     should initiate a SIP transaction carrying old SDP to
>>>
>>>     accomplish it.
>>>
>>>      
>>>
>>>     There is a special case for re-INVITE. If multiple SDP
>>>
>>>     offer/answer happens(using PRACK/UPDATE) within
>>>
>>>     re-INVITE transaction and re-INVITE is rejected by
>>>
>>>     4xx-6xx then whether all the SDP offer/answer happens
>>>
>>>     within re-INVITE transaction shall be rolled-back or
>>>
>>>     not. My suggestion would be that once SDP offer/answer
>>>
>>>     is completed, that can not rolled-back by means of
>>>
>>>     transaction rejection. That means, all the completed
>>>
>>>     SDP offer/answer shall not be rolled-back due to
>>>
>>>     4xx-6xx of re-INVITE. This decision shall save the
>>>
>>>     work of SIP UA, SIP SBC, B2BUA etc. that follows SDP
>>>
>>>     offer/answer state.
>>>
>>>      
>>>
>>>     Please comment.
>>>
>>>      
>>>
>>>      
>>>
>>>     Thanks and Regards,
>>>
>>>     Siddhartha
>>>
>>>      
>>>
>>>
>>>
>>>     
>> --------------------------------------------------------------
>> ----------
>>>     ---------------
>>>     This e-mail may contain confidential and/or privileged 
>> information.
>>>     If you are not the intended recipient (or have received 
>> this e-mail
>>>     in error) please notify the sender immediately and delete this
>>>     e-mail. Any unauthorized copying, disclosure or 
>> distribution of the
>>>     contents in this e-mail is strictly forbidden.
>>>     ---------------
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP 
>> Use sip-implementors@cs.columbia.edu for questions on current 
>> sip Use sip@ietf.org for new developments of core SIP
>>
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 08:45:30 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGxxy-0006dg-Kp; Tue, 13 Feb 2007 08:45:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGxxx-0006dN-7c
	for sipping@ietf.org; Tue, 13 Feb 2007 08:45:05 -0500
Received: from [217.205.209.100] (helo=vegastream.com)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGxxv-0002UF-St
	for sipping@ietf.org; Tue, 13 Feb 2007 08:45:05 -0500
content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
X-MimeOLE: Produced By Microsoft Exchange V6.0.6487.1
Subject: RE: [SIPPING] SDP
	offer/answer	inearlymedia:draft-sawada-sipping-sip-offeranswer-01.txt
Date: Tue, 13 Feb 2007 13:44:59 -0000
Message-ID: <8F1F2062CA7D754A838F0E66097D23A0C24180@veg-brk-svr-pri.vegastream>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [SIPPING] SDP
	offer/answer	inearlymedia:draft-sawada-sipping-sip-offeranswer-01.txt
Thread-Index: AcdPY06RIufrIME6TceHiIKUn8q1qAABBKyAAAHFdtAAAUG08A==
From: "Attila Sipos" <Attila.Sipos@vegastream.com>
To: "Brian Stucker" <bstucker@nortel.com>,
	"Christer Holmberg (JO/LMF)" <christer.holmberg@ericsson.com>,
	"Elwell, John" <john.elwell@siemens.com>,
	"Paul Kyzivat" <pkyzivat@cisco.com>,
	"Siddhartha Bhakta" <Siddhartha.Bhakta@newport-networks.com>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: c0bedb65cce30976f0bf60a0a39edea4
Cc: sipping@ietf.org, Siddhartha Bhakta <sbhakta007@yahoo.co.in>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


>>Although a 19x response might make knowing when early media from the
>>interloper to the SDP exchange is no longer necessary, the same thing
>>can be handled by (most) clients today by simply sending a 180 back to
>>the originator to switch back to local ringing when that endpoint no
>>longer has any media it wishes the originator to render (not that the

Is this really going to work?
If it works, it is just luck I think.
In RFC 3261 it says...

             The UAC MUST treat the first session
         description it receives as the answer, and MUST ignore any
         session descriptions in subsequent responses to the initial
         INVITE.

So should lack of SDP get ignored?
Intuitively I would say yes.

Isn't UPDATE the only legal way for the fake fork
to say "don't talk or listen to me"?

Regards,

Attila


-----Original Message-----
From: Brian Stucker [mailto:bstucker@nortel.com]
Sent: 13 February 2007 13:08
To: Christer Holmberg (JO/LMF); Elwell, John; Paul Kyzivat; Siddhartha
Bhakta
Cc: sipping@ietf.org; Siddhartha Bhakta
Subject: RE: [SIPPING] SDP offer/answer
inearlymedia:draft-sawada-sipping-sip-offeranswer-01.txt


Fake forking is a classic example of the middle screwing up a process
that should only be between final endpoints.

A box in the middle forks the call off to an endpoint that the
originator never intended to call that is acting as if it is one of the
called party's devices, when it is not, and has no intention of ever
answering the call.

Although a 19x response might make knowing when early media from the
interloper to the SDP exchange is no longer necessary, the same thing
can be handled by (most) clients today by simply sending a 180 back to
the originator to switch back to local ringing when that endpoint no
longer has any media it wishes the originator to render (not that the
originator had to render anything in the first place, but that's a
separate discussion).

What would be nice is if each dialog was handled explicitly, rather than
implicitly at the originator for early media (i.e. the originator
explicitly contacts the device that wishes to generate early media
rather than having the network resort to tricking the originator into
rendering the desired media).

I have a draft sitting in the queue that proposes such a mechanism. The
XML2RFC template still used (C) 2006 when I generated the draft text, so
it got bounced during the I-D nits check. Until it appears in the
archives, anyone interested can email me and I'll send you a copy of the
draft.

Regards,
Brian

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 08:45:30 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGxxl-0006WM-I8; Tue, 13 Feb 2007 08:44:53 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGxxj-0006WF-Jc
	for sipping@ietf.org; Tue, 13 Feb 2007 08:44:51 -0500
Received: from rtp-iport-2.cisco.com ([64.102.122.149])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGxxi-0002SR-2y
	for sipping@ietf.org; Tue, 13 Feb 2007 08:44:51 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by rtp-iport-2.cisco.com with ESMTP; 13 Feb 2007 08:44:49 -0500
X-IronPort-AV: i="4.14,163,1170651600"; 
	d="scan'208"; a="113783615:sNHT63597018"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1DDin5r025380; 
	Tue, 13 Feb 2007 08:44:49 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1DDinOA019532; 
	Tue, 13 Feb 2007 08:44:49 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 08:44:49 -0500
Received: from [161.44.182.226] ([161.44.182.226]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 08:44:49 -0500
Message-ID: <45D1C0D0.4070401@cisco.com>
Date: Tue, 13 Feb 2007 08:44:48 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Christer Holmberg (JO/LMF)" <christer.holmberg@ericsson.com>
Subject: Re: [SIPPING] SDP offer/answer
	inearly	media:draft-sawada-sipping-sip-offeranswer-01.txt
References: <7374777208BDC7449D5620EF9423256702047D9A@esealmw113.eemea.ericsson.se>
In-Reply-To: <7374777208BDC7449D5620EF9423256702047D9A@esealmw113.eemea.ericsson.se>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 13 Feb 2007 13:44:49.0125 (UTC)
	FILETIME=[21447150:01C74F75]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=9075; t=1171374289;
	x=1172238289; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[SIPPING]=20SDP=20offer/answer=20inearly=09media=3Adr
	aft-sawada-sipping-sip-offeranswer-01.txt |Sender:=20
	|To:=20=22Christer=20Holmberg=20(JO/LMF)=22=20<christer.holmberg@ericsson
	.com>; bh=3x+qcpyP6xXvHwI26ZBr+MTBOIp6aHPAAY8nppIGH2Y=;
	b=Nf0WN/tH6OVj+U2NtdS6SY4t9M64AfIyMCf99pdYsaTzXGEzSmr3zfPcRICwFPLh+ITen/QN
	FB3k90iWKHzyGnwW2YrUD9X+P5O8mcJPAM38mJ3EKV/R6v2kKgkiHeiT;
Authentication-Results: rtp-dkim-2; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: f2728948111f2edaaf8980b5b9de55af
Cc: Brian Stucker <bstucker@nortel.com>, sipping@ietf.org,
	Siddhartha Bhakta <sbhakta007@yahoo.co.in>,
	Siddhartha Bhakta <Siddhartha.Bhakta@newport-networks.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Christer Holmberg (JO/LMF) wrote:
> Hi,
> 
> I once proposed that it should be possible to terminate an early dialog
> with an 19x provisional response.

How do you see this being used with "real" forking? Would the proxy that 
is doing the forking send the 190 response on behalf of early dialogs 
that it is canceling?

> Then, if "fake forking" is used, the dialog not used anymore could be
> terminted and the UAC knows he doesn't have to care about it (and
> possible media coming on it).

What is the prevalence of "fake" forking relative to "real" forking with 
early media?

I think there are merits to this approach, though it needs to be worked 
out in detail.

	Paul

> Regards,
> 
> Christer
>  
> 
>> -----Original Message-----
>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
>> Sent: 16. marraskuuta 2006 21:58
>> To: Siddhartha Bhakta
>> Cc: sipping@ietf.org; 'Siddhartha Bhakta'; 'Brian Stucker'
>> Subject: Re: [SIPPING] SDP offer/answer inearly 
>> media:draft-sawada-sipping-sip-offeranswer-01.txt
>>
>>
>>
>> Siddhartha Bhakta wrote:
>>> I agree with you. No matter whether different provisional responses 
>>> are creating different dialogs or not, case is same as far as media 
>>> rendering/committing is concerned. For media committing, there are 
>>> multiple answers for a given offer.
>>>
>>> Like you, I also can not understand what we are gaining by creating 
>>> multiple dialogs here except the chance to manage 
>> offer/answer state 
>>> as per RFC3261/RFC3264. It seems that multiple SDP answers 
>> for a given 
>>> SDP offer is the reality. We can not avoid that by applying 
>> any tricky 
>>> logic in dialog management (like Fake Forking).
>> If you ignore dialogs, it becomes impossible to impose any 
>> offer/answer rules.
>>
>> By paying attention to the separate dialogs, it is possible 
>> to have offer/answer rules that make sense. If you just mix 
>> together offers or answers from multiple destinations it 
>> becomes impossible to make any sense of what you have once it 
>> is determined which dialog will be completed.
>>
>> For the period when there are multiple dialogs, you have two 
>> different things going on:
>> - an offer/answer state machine is proceeding in each dialog, and
>>    defining the current media parameters for that dialog
>> - you are deciding which dialog(s) you want to process and send
>>    media to.
>>
>> As Brian notes, it is complex to decide which dialog to honor 
>> during the early stage. But once you settle on one dialog at 
>> least you will then know what to do.
>>
>> 	Paul
>>
>>> -----Original Message-----
>>> From: Brian Stucker [mailto:bstucker@nortel.com]
>>> Sent: 16 November 2006 17:37
>>> To: Hadriel Kaplan; Siddhartha Bhakta; sipping@ietf.org
>>> Cc: Siddhartha Bhakta
>>> Subject: RE: [SIPPING] SDP offer/answer in early 
>>> media:draft-sawada-sipping-sip-offeranswer-01.txt
>>>
>>> Fake forking...
>>>
>>> Please note, if there's SDP in the other provisional 
>> responses, you're 
>>> going to run into problems potentially getting themedia rendered 
>>> (assuming that's your goal).
>>>
>>> Regards,
>>> Brian
>>>
>>>> -----Original Message-----
>>>> From: Hadriel Kaplan [mailto:HKaplan@acmepacket.com]
>>>> Sent: Sunday, November 12, 2006 8:40 PM
>>>> To: 'Siddhartha Bhakta'; sipping@ietf.org
>>>> Cc: 'Siddhartha Bhakta'
>>>> Subject: RE: [SIPPING] SDP offer/answer in early 
>>>> media:draft-sawada-sipping-sip-offeranswer-01.txt
>>>>
>>>> If the Proxy is sending back SDP by itself it's not a proxy, but a 
>>>> b2bua, as is the PBX.
>>>> As such, it should be answering in a different dialog, so its SDP 
>>>> answer should be in a 183 with a different To-tag from the 
>> 180, and 
>>>> the PBX's 180 SDP should have a different To-tag than the 200ok.  
>>>> That would make them look like a forked call and follow the RFCs.
>>>>
>>>> Of course the reality is often different, but being a 
>> b2bua yourself 
>>>> it's in your power to fix it.  How you do that is not up 
>> to the IETF 
>>>> to define, as the very idea of it goes against the IETF's 
>> view of the 
>>>> world. (and I say that in a positive way, as I think it would be 
>>>> practically impossible for the IETF to do otherwise)
>>>>
>>>> -hadriel
>>>>
>>>>
>>>>> -----Original Message-----
>>>>> From: Siddhartha Bhakta [mailto:sbhakta007@yahoo.co.in]
>>>>> Sent: Sunday, November 12, 2006 2:11 PM
>>>>> To: sipping@ietf.org
>>>>> Cc: sbhakta007@yahoo.co.in; Siddhartha Bhakta
>>>>> Subject: [SIPPING] SDP offer/answer in early
>>>>> media:draft-sawada-sipping- sip-offeranswer-01.txt
>>>>>
>>>>> I am re-sending it as in last email, the figure was distorted.
>>>>>
>>>>> Though, post RFC3261 drafts indicate that one valid 
>> answer would be 
>>>>> there for one SDP offer, the our old friend early-media
>>>> flow is still
>>>>> present in practice.
>>>>>
>>>>> B2BUA/SBC     Proxy         PABX        Bob
>>>>>   |            |            |            |
>>>>>   | INVITE F1  |            |            |
>>>>>   |----------->| INVITE F2  |            |
>>>>>   |   183 F3   |----------->| INVITE F4  |
>>>>>   |<-----------|            |----------->|
>>>>>   |            |            |            |
>>>>>   |            |            |   180 without SDP F5
>>>>>   |            |            |<-----------|
>>>>>   |            |            |            |
>>>>>   |            |  180 with SDP F6        |
>>>>>   |   180 F7   |<-----------|            |
>>>>>   |<-----------|            |  200 F8    |
>>>>>   |            |  200 F9    |<-----------|
>>>>>   |   200 F10  |<-----------|            |
>>>>>   |<-----------|            |            |
>>>>>   |   ACK F11  |            |            |
>>>>>   |----------->|  ACK F12   |            |
>>>>>   |            |----------->|  ACK F13   |
>>>>>   |            |            |----------->|
>>>>>
>>>>> Our B2BUA is facing the problem as specified above.
>>>>> The first answer is carried
>>>>> by F3. This SDP is specified by Proxy. The 2nd SDP answer
>>>> is carried
>>>>> by F6,F7.
>>>>> In fact, this SDP is specified by PABX. The third SDP answer is 
>>>>> carried by F8,F9,F10.
>>>>> This SDP is specified by called party (Bob).
>>>>>
>>>>> As per the RFC3261 & RFC3264, the SDP answer carried by
>>>> F6,F7 should
>>>>> match with F3 as far as B2BUA is concerned. Therefore, 
>> B2BUA shall 
>>>>> ignore it. This is leading to the fact that SIP UA behind 
>> B2BUA can 
>>>>> not listen to ringtone fed by PABX.
>>>>> Similarly, SDP answer carried by F8,F9,F10 shall be ignored. This 
>>>>> shall lead to the fact that SIP UA behind B2BUA can not listen to 
>>>>> Bob's voice. Too bad.
>>>>> This problem is due to the fact that RFC3261 & RFC3264 are not 
>>>>> backward compatible.
>>>>>
>>>>>
>>>>> The sec 2.2. of
>>>>> draft-sawada-sipping-sip-offeranswer-01.txt indicates that UPDATE 
>>>>> should be used to update the media in early media. But in
>>>> practice old
>>>>> early-media flow (i.e., sending different SDPs over different 1xx 
>>>>> responses of
>>>>> INVITE) is quite common.
>>>>> Can we somehow make new SDP offer/answer specified in
>>>>> RFC3261 & RFC3264 backward
>>>>> compatible?
>>>>>
>>>>> The old standard (early-media), allows multiple SDP answers
>>>> of single
>>>>> SDP offer. can we somehow induce this thing in SDP
>>>> offer/answer model? 
>>>>> If many of you think the need of it then I may comeup with
>>>> some thumb
>>>>> rule for the same.
>>>>>
>>>>> Thanks & Regards,
>>>>> Siddhartha
>>>> _______________________________________________
>>>> Sipping mailing list  
>> https://www1.ietf.org/mailman/listinfo/sipping
>>>> This list is for NEW development of the application of SIP Use 
>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>> sip@ietf.org for new developments of core SIP
>>>>
>>>
>>>
>>> ---------------
>>> This e-mail may contain confidential and/or privileged 
>> information. If you are not the intended recipient (or have 
>> received this e-mail in error) please notify the sender 
>> immediately and delete this e-mail. Any unauthorized copying, 
>> disclosure or distribution of the contents in this e-mail is 
>> strictly forbidden.
>>> ---------------
>>>
>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP Use 
>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>> sip@ietf.org for new developments of core SIP
>>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP 
>> Use sip-implementors@cs.columbia.edu for questions on current 
>> sip Use sip@ietf.org for new developments of core SIP
>>
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 08:54:01 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGy6U-0001rS-LM; Tue, 13 Feb 2007 08:53:54 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGy6T-0001rI-Ot
	for sipping@ietf.org; Tue, 13 Feb 2007 08:53:53 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGy6S-0004Wu-0J
	for sipping@ietf.org; Tue, 13 Feb 2007 08:53:53 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	663C7205F6; Tue, 13 Feb 2007 14:53:47 +0100 (CET)
X-AuditID: c1b4fb3c-ae7c5bb0000007de-b8-45d1c2eb3cf4 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	3660A2045C; Tue, 13 Feb 2007 14:53:47 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 14:53:46 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [SIPPING] SDP offer/answer
	inearly	media:draft-sawada-sipping-sip-offeranswer-01.txt
Date: Tue, 13 Feb 2007 14:53:46 +0100
Message-ID: <7374777208BDC7449D5620EF942325670206A64F@esealmw113.eemea.ericsson.se>
In-Reply-To: <45D1C0D0.4070401@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [SIPPING] SDP offer/answer
	inearly	media:draft-sawada-sipping-sip-offeranswer-01.txt
Thread-Index: AcdPdSQAXoAgP5mPTcCbIRk307JoGgAAKQKg
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>
X-OriginalArrivalTime: 13 Feb 2007 13:53:46.0968 (UTC)
	FILETIME=[61D8C580:01C74F76]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 249cd1efd3d5e0d09114abe826a41235
Cc: Brian Stucker <bstucker@nortel.com>, sipping@ietf.org,
	Siddhartha Bhakta <sbhakta007@yahoo.co.in>,
	Siddhartha Bhakta <Siddhartha.Bhakta@newport-networks.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,=20

>>I once proposed that it should be possible to terminate an early=20
>>dialog with an 19x provisional response.
>=20
>How do you see this being used with "real" forking? Would the=20
>proxy that is doing the forking send the 190 response on=20
>behalf of early dialogs that it is canceling?

Yes.=20

Also, if the forking proxy receives a 456xx from a UAS it can forward a
190 towards the UAC (forwarding the 456xxx would terminate the whole
call setup).

>>Then, if "fake forking" is used, the dialog not used anymore could be
terminted and the UAC knows he doesn't have to care about it (and=20
>>possible media coming on it).
>=20
>What is the prevalence of "fake" forking relative to "real" forking
with early media?

I don't, but at least in TISPAN "fake forking" is a specified mechanism
for announcement sending by an AS.

Regards,

Christer




>I think there are merits to this approach, though it needs to be worked
out in detail.



>=20
> 	Paul
>=20
> > Regards,
> >=20
> > Christer
> > =20
> >=20
> >> -----Original Message-----
> >> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
> >> Sent: 16. marraskuuta 2006 21:58
> >> To: Siddhartha Bhakta
> >> Cc: sipping@ietf.org; 'Siddhartha Bhakta'; 'Brian Stucker'
> >> Subject: Re: [SIPPING] SDP offer/answer inearly=20
> >> media:draft-sawada-sipping-sip-offeranswer-01.txt
> >>
> >>
> >>
> >> Siddhartha Bhakta wrote:
> >>> I agree with you. No matter whether different provisional=20
> responses=20
> >>> are creating different dialogs or not, case is same as=20
> far as media=20
> >>> rendering/committing is concerned. For media committing,=20
> there are=20
> >>> multiple answers for a given offer.
> >>>
> >>> Like you, I also can not understand what we are gaining=20
> by creating=20
> >>> multiple dialogs here except the chance to manage
> >> offer/answer state
> >>> as per RFC3261/RFC3264. It seems that multiple SDP answers
> >> for a given
> >>> SDP offer is the reality. We can not avoid that by applying
> >> any tricky
> >>> logic in dialog management (like Fake Forking).
> >> If you ignore dialogs, it becomes impossible to impose any=20
> >> offer/answer rules.
> >>
> >> By paying attention to the separate dialogs, it is=20
> possible to have=20
> >> offer/answer rules that make sense. If you just mix=20
> together offers=20
> >> or answers from multiple destinations it becomes=20
> impossible to make=20
> >> any sense of what you have once it is determined which=20
> dialog will be=20
> >> completed.
> >>
> >> For the period when there are multiple dialogs, you have two=20
> >> different things going on:
> >> - an offer/answer state machine is proceeding in each dialog, and
> >>    defining the current media parameters for that dialog
> >> - you are deciding which dialog(s) you want to process and send
> >>    media to.
> >>
> >> As Brian notes, it is complex to decide which dialog to=20
> honor during=20
> >> the early stage. But once you settle on one dialog at=20
> least you will=20
> >> then know what to do.
> >>
> >> 	Paul
> >>
> >>> -----Original Message-----
> >>> From: Brian Stucker [mailto:bstucker@nortel.com]
> >>> Sent: 16 November 2006 17:37
> >>> To: Hadriel Kaplan; Siddhartha Bhakta; sipping@ietf.org
> >>> Cc: Siddhartha Bhakta
> >>> Subject: RE: [SIPPING] SDP offer/answer in early=20
> >>> media:draft-sawada-sipping-sip-offeranswer-01.txt
> >>>
> >>> Fake forking...
> >>>
> >>> Please note, if there's SDP in the other provisional
> >> responses, you're
> >>> going to run into problems potentially getting themedia rendered=20
> >>> (assuming that's your goal).
> >>>
> >>> Regards,
> >>> Brian
> >>>
> >>>> -----Original Message-----
> >>>> From: Hadriel Kaplan [mailto:HKaplan@acmepacket.com]
> >>>> Sent: Sunday, November 12, 2006 8:40 PM
> >>>> To: 'Siddhartha Bhakta'; sipping@ietf.org
> >>>> Cc: 'Siddhartha Bhakta'
> >>>> Subject: RE: [SIPPING] SDP offer/answer in early=20
> >>>> media:draft-sawada-sipping-sip-offeranswer-01.txt
> >>>>
> >>>> If the Proxy is sending back SDP by itself it's not a=20
> proxy, but a=20
> >>>> b2bua, as is the PBX.
> >>>> As such, it should be answering in a different dialog,=20
> so its SDP=20
> >>>> answer should be in a 183 with a different To-tag from the
> >> 180, and
> >>>> the PBX's 180 SDP should have a different To-tag than=20
> the 200ok. =20
> >>>> That would make them look like a forked call and follow the RFCs.
> >>>>
> >>>> Of course the reality is often different, but being a
> >> b2bua yourself
> >>>> it's in your power to fix it.  How you do that is not up
> >> to the IETF
> >>>> to define, as the very idea of it goes against the IETF's
> >> view of the
> >>>> world. (and I say that in a positive way, as I think it would be=20
> >>>> practically impossible for the IETF to do otherwise)
> >>>>
> >>>> -hadriel
> >>>>
> >>>>
> >>>>> -----Original Message-----
> >>>>> From: Siddhartha Bhakta [mailto:sbhakta007@yahoo.co.in]
> >>>>> Sent: Sunday, November 12, 2006 2:11 PM
> >>>>> To: sipping@ietf.org
> >>>>> Cc: sbhakta007@yahoo.co.in; Siddhartha Bhakta
> >>>>> Subject: [SIPPING] SDP offer/answer in early
> >>>>> media:draft-sawada-sipping- sip-offeranswer-01.txt
> >>>>>
> >>>>> I am re-sending it as in last email, the figure was distorted.
> >>>>>
> >>>>> Though, post RFC3261 drafts indicate that one valid
> >> answer would be
> >>>>> there for one SDP offer, the our old friend early-media
> >>>> flow is still
> >>>>> present in practice.
> >>>>>
> >>>>> B2BUA/SBC     Proxy         PABX        Bob
> >>>>>   |            |            |            |
> >>>>>   | INVITE F1  |            |            |
> >>>>>   |----------->| INVITE F2  |            |
> >>>>>   |   183 F3   |----------->| INVITE F4  |
> >>>>>   |<-----------|            |----------->|
> >>>>>   |            |            |            |
> >>>>>   |            |            |   180 without SDP F5
> >>>>>   |            |            |<-----------|
> >>>>>   |            |            |            |
> >>>>>   |            |  180 with SDP F6        |
> >>>>>   |   180 F7   |<-----------|            |
> >>>>>   |<-----------|            |  200 F8    |
> >>>>>   |            |  200 F9    |<-----------|
> >>>>>   |   200 F10  |<-----------|            |
> >>>>>   |<-----------|            |            |
> >>>>>   |   ACK F11  |            |            |
> >>>>>   |----------->|  ACK F12   |            |
> >>>>>   |            |----------->|  ACK F13   |
> >>>>>   |            |            |----------->|
> >>>>>
> >>>>> Our B2BUA is facing the problem as specified above.
> >>>>> The first answer is carried
> >>>>> by F3. This SDP is specified by Proxy. The 2nd SDP answer
> >>>> is carried
> >>>>> by F6,F7.
> >>>>> In fact, this SDP is specified by PABX. The third SDP answer is=20
> >>>>> carried by F8,F9,F10.
> >>>>> This SDP is specified by called party (Bob).
> >>>>>
> >>>>> As per the RFC3261 & RFC3264, the SDP answer carried by
> >>>> F6,F7 should
> >>>>> match with F3 as far as B2BUA is concerned. Therefore,
> >> B2BUA shall
> >>>>> ignore it. This is leading to the fact that SIP UA behind
> >> B2BUA can
> >>>>> not listen to ringtone fed by PABX.
> >>>>> Similarly, SDP answer carried by F8,F9,F10 shall be=20
> ignored. This=20
> >>>>> shall lead to the fact that SIP UA behind B2BUA can not=20
> listen to=20
> >>>>> Bob's voice. Too bad.
> >>>>> This problem is due to the fact that RFC3261 & RFC3264 are not=20
> >>>>> backward compatible.
> >>>>>
> >>>>>
> >>>>> The sec 2.2. of
> >>>>> draft-sawada-sipping-sip-offeranswer-01.txt indicates=20
> that UPDATE=20
> >>>>> should be used to update the media in early media. But in
> >>>> practice old
> >>>>> early-media flow (i.e., sending different SDPs over=20
> different 1xx=20
> >>>>> responses of
> >>>>> INVITE) is quite common.
> >>>>> Can we somehow make new SDP offer/answer specified in
> >>>>> RFC3261 & RFC3264 backward
> >>>>> compatible?
> >>>>>
> >>>>> The old standard (early-media), allows multiple SDP answers
> >>>> of single
> >>>>> SDP offer. can we somehow induce this thing in SDP
> >>>> offer/answer model?=20
> >>>>> If many of you think the need of it then I may comeup with
> >>>> some thumb
> >>>>> rule for the same.
> >>>>>
> >>>>> Thanks & Regards,
> >>>>> Siddhartha
> >>>> _______________________________________________
> >>>> Sipping mailing list
> >> https://www1.ietf.org/mailman/listinfo/sipping
> >>>> This list is for NEW development of the application of SIP Use=20
> >>>> sip-implementors@cs.columbia.edu for questions on=20
> current sip Use=20
> >>>> sip@ietf.org for new developments of core SIP
> >>>>
> >>>
> >>>
> >>> ---------------
> >>> This e-mail may contain confidential and/or privileged
> >> information. If you are not the intended recipient (or=20
> have received=20
> >> this e-mail in error) please notify the sender immediately=20
> and delete=20
> >> this e-mail. Any unauthorized copying, disclosure or=20
> distribution of=20
> >> the contents in this e-mail is strictly forbidden.
> >>> ---------------
> >>>
> >>>
> >>> _______________________________________________
> >>> Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> >>> This list is for NEW development of the application of SIP Use=20
> >>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> >>> sip@ietf.org for new developments of core SIP
> >>>
> >> _______________________________________________
> >> Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> >> This list is for NEW development of the application of SIP=20
> >> Use sip-implementors@cs.columbia.edu for questions on current=20
> >> sip Use sip@ietf.org for new developments of core SIP
> >>
> >=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 08:58:38 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGyAx-0005Ru-VN; Tue, 13 Feb 2007 08:58:32 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGyAw-0005R7-Dk
	for sipping@ietf.org; Tue, 13 Feb 2007 08:58:30 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGyAo-0005my-KQ
	for sipping@ietf.org; Tue, 13 Feb 2007 08:58:30 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	2A23B208CA; Tue, 13 Feb 2007 14:58:22 +0100 (CET)
X-AuditID: c1b4fb3c-af7c7bb0000007de-f1-45d1c3fe07b6 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	0104820401; Tue, 13 Feb 2007 14:58:22 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 14:58:21 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [SIPPING] SDP
	offer/answer	inearlymedia:draft-sawada-sipping-sip-offeranswer-01.txt
Date: Tue, 13 Feb 2007 14:58:20 +0100
Message-ID: <7374777208BDC7449D5620EF942325670206A68D@esealmw113.eemea.ericsson.se>
In-Reply-To: <8F1F2062CA7D754A838F0E66097D23A0C24180@veg-brk-svr-pri.vegastream>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [SIPPING] SDP
	offer/answer	inearlymedia:draft-sawada-sipping-sip-offeranswer-01.txt
Thread-Index: AcdPY06RIufrIME6TceHiIKUn8q1qAABBKyAAAHFdtAAAUG08AAAxB6Q
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Attila Sipos" <Attila.Sipos@vegastream.com>,
	"Brian Stucker" <bstucker@nortel.com>,
	"Elwell, John" <john.elwell@siemens.com>,
	"Paul Kyzivat" <pkyzivat@cisco.com>,
	"Siddhartha Bhakta" <Siddhartha.Bhakta@newport-networks.com>
X-OriginalArrivalTime: 13 Feb 2007 13:58:21.0791 (UTC)
	FILETIME=[05A76EF0:01C74F77]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 3002fc2e661cd7f114cb6bae92fe88f1
Cc: sipping@ietf.org, Siddhartha Bhakta <sbhakta007@yahoo.co.in>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

=20

>>>Although a 19x response might make knowing when early media from the=20
>>>interloper to the SDP exchange is no longer necessary, the same thing

>>>can be handled by (most) clients today by simply sending a 180 back
to=20
>>>the originator to switch back to local ringing when that=20
>>>endpoint no longer has any media it wishes the originator to render=20
>(not that the
>=20
>Is this really going to work?
>If it works, it is just luck I think.
>In RFC 3261 it says...
>=20
>              The UAC MUST treat the first session
>          description it receives as the answer, and MUST ignore any
>          session descriptions in subsequent responses to the initial
>          INVITE.
>=20
>So should lack of SDP get ignored?
>Intuitively I would say yes.
>=20
>Isn't UPDATE the only legal way for the fake fork to say=20
>"don't talk or listen to me"?

An UPDATE could work in the case where the B2BUA establishes a dialog,
because the B2BUA acts as the UAS.

But, in real forking cases you can't expect the end-terminal UAS to send
an UPDATE (e.g. if they don't accept a call and will send an error
response), so the forking proxy would have to do it - which would mean
it becomes a B2BUA. I think it would be much easier for the forking
proxy to simply send a 19x.

Regards,

Christer





> -----Original Message-----
> From: Brian Stucker [mailto:bstucker@nortel.com]
> Sent: 13 February 2007 13:08
> To: Christer Holmberg (JO/LMF); Elwell, John; Paul Kyzivat;=20
> Siddhartha Bhakta
> Cc: sipping@ietf.org; Siddhartha Bhakta
> Subject: RE: [SIPPING] SDP offer/answer
> inearlymedia:draft-sawada-sipping-sip-offeranswer-01.txt
>=20
>=20
> Fake forking is a classic example of the middle screwing up a=20
> process that should only be between final endpoints.
>=20
> A box in the middle forks the call off to an endpoint that=20
> the originator never intended to call that is acting as if it=20
> is one of the called party's devices, when it is not, and has=20
> no intention of ever answering the call.
>=20
> Although a 19x response might make knowing when early media=20
> from the interloper to the SDP exchange is no longer=20
> necessary, the same thing can be handled by (most) clients=20
> today by simply sending a 180 back to the originator to=20
> switch back to local ringing when that endpoint no longer has=20
> any media it wishes the originator to render (not that the=20
> originator had to render anything in the first place, but=20
> that's a separate discussion).
>=20
> What would be nice is if each dialog was handled explicitly,=20
> rather than implicitly at the originator for early media=20
> (i.e. the originator explicitly contacts the device that=20
> wishes to generate early media rather than having the network=20
> resort to tricking the originator into rendering the desired media).
>=20
> I have a draft sitting in the queue that proposes such a=20
> mechanism. The XML2RFC template still used (C) 2006 when I=20
> generated the draft text, so it got bounced during the I-D=20
> nits check. Until it appears in the archives, anyone=20
> interested can email me and I'll send you a copy of the draft.
>=20
> Regards,
> Brian
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 09:11:31 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGyMX-00022a-0z; Tue, 13 Feb 2007 09:10:29 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGyMU-00022V-OP
	for sipping@ietf.org; Tue, 13 Feb 2007 09:10:26 -0500
Received: from rtp-iport-2.cisco.com ([64.102.122.149])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGyMT-0008SJ-53
	for sipping@ietf.org; Tue, 13 Feb 2007 09:10:26 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by rtp-iport-2.cisco.com with ESMTP; 13 Feb 2007 09:10:25 -0500
X-IronPort-AV: i="4.14,163,1170651600"; 
	d="scan'208"; a="113788790:sNHT73655808"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1DEAOwf014867; 
	Tue, 13 Feb 2007 09:10:24 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1DEAOOA026547; 
	Tue, 13 Feb 2007 09:10:24 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 09:10:24 -0500
Received: from [161.44.182.226] ([161.44.182.226]) by
	xfe-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 09:10:23 -0500
Message-ID: <45D1C6CF.30801@cisco.com>
Date: Tue, 13 Feb 2007 09:10:23 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Elwell, John" <john.elwell@siemens.com>
Subject: Re: [SIPPING] SDP offer/answer	inearly media:draft-sawada-sipping
	-sip-offeranswer-01.txt
References: <50B1CBA96870A34799A506B2313F26670B05FD2C@ntht201e.siemenscomms.co.uk>
In-Reply-To: <50B1CBA96870A34799A506B2313F26670B05FD2C@ntht201e.siemenscomms.co.uk>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 13 Feb 2007 14:10:23.0562 (UTC)
	FILETIME=[B3DCF2A0:01C74F78]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=10880; t=1171375824;
	x=1172239824; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[SIPPING]=20SDP=20offer/answer=09inearly=20media=3Adr
	aft-sawada-sipping=0A=20-sip-offeranswer-01.txt |Sender:=20
	|To:=20=22Elwell,=20John=22=20<john.elwell@siemens.com>;
	bh=3mXjfkIOh2p5rzb+SYB5GtZgo4RQi5VsC9qqYxDDxwc=;
	b=CvlOlbEYnOGQ8YaCMqZgCRw8VWvIrD0br1gbqURT+LqH94XzqphRnlnTE/RSRr6jZCV63Vei
	bfbBaE5m4+tizKpI53FJZySlcVD2hc2zwebua1iz1OruggruCwpeOrpG;
Authentication-Results: rtp-dkim-1; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 40161b1d86420e0807d771943d981d25
Cc: Siddhartha Bhakta <Siddhartha.Bhakta@newport-networks.com>,
	sipping@ietf.org, Brian Stucker <bstucker@nortel.com>,
	Siddhartha Bhakta <sbhakta007@yahoo.co.in>,
	"Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Elwell, John wrote:
> Christer,
> 
> I don't know what you mean by "fake forking", but I think there is an issue
> with serial forking, e.g.:
> - INVITE request to user@example.com
> - proxy for example.com forwards to PSTN gateway
> - PSTN gateway provides early media
> - proxy is scripted to cancel the PSTN call after a certain time and forward
> the request elsewhere (e.g., to a mobile or voicemail)
> - the new destination sends early media or answers quickly and sends regular
> media
> - the UAC will probably still be rendering early media from the PSTN and
> will not know that this is in the process of stopping (even if packets stop
> arriving, this could be due to silence)
> 
> In this case your proposed 19x provisional response could tell the UAC to
> stop listening to the corresponding RTP stream immediately.

I don't see how that would work. To work, the UAC must correlate the 190 
response with the particular media stream being received, so it knows to 
abandon it. But there isn't any well defined way to do that. *If* the 
UAS uses symmetric RTP, and if there is no NAT in the middle, then it 
might be able to do it. I guess this might work in conjunction with ICE.

	Paul

> John
> 
>> -----Original Message-----
>> From: Christer Holmberg (JO/LMF) 
>> [mailto:christer.holmberg@ericsson.com] 
>> Sent: 13 February 2007 10:57
>> To: Paul Kyzivat; Siddhartha Bhakta
>> Cc: sipping@ietf.org; Siddhartha Bhakta; Brian Stucker
>> Subject: RE: [SIPPING] SDP offer/answer inearly 
>> media:draft-sawada-sipping-sip-offeranswer-01.txt
>>
>>
>> Hi,
>>
>> I once proposed that it should be possible to terminate an 
>> early dialog
>> with an 19x provisional response.
>>
>> Then, if "fake forking" is used, the dialog not used anymore could be
>> terminted and the UAC knows he doesn't have to care about it (and
>> possible media coming on it).
>>
>> Regards,
>>
>> Christer
>>  
>>
>>> -----Original Message-----
>>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
>>> Sent: 16. marraskuuta 2006 21:58
>>> To: Siddhartha Bhakta
>>> Cc: sipping@ietf.org; 'Siddhartha Bhakta'; 'Brian Stucker'
>>> Subject: Re: [SIPPING] SDP offer/answer inearly 
>>> media:draft-sawada-sipping-sip-offeranswer-01.txt
>>>
>>>
>>>
>>> Siddhartha Bhakta wrote:
>>>> I agree with you. No matter whether different provisional 
>> responses 
>>>> are creating different dialogs or not, case is same as 
>> far as media 
>>>> rendering/committing is concerned. For media committing, 
>> there are 
>>>> multiple answers for a given offer.
>>>>
>>>> Like you, I also can not understand what we are gaining 
>> by creating 
>>>> multiple dialogs here except the chance to manage 
>>> offer/answer state 
>>>> as per RFC3261/RFC3264. It seems that multiple SDP answers 
>>> for a given 
>>>> SDP offer is the reality. We can not avoid that by applying 
>>> any tricky 
>>>> logic in dialog management (like Fake Forking).
>>> If you ignore dialogs, it becomes impossible to impose any 
>>> offer/answer rules.
>>>
>>> By paying attention to the separate dialogs, it is possible 
>>> to have offer/answer rules that make sense. If you just mix 
>>> together offers or answers from multiple destinations it 
>>> becomes impossible to make any sense of what you have once it 
>>> is determined which dialog will be completed.
>>>
>>> For the period when there are multiple dialogs, you have two 
>>> different things going on:
>>> - an offer/answer state machine is proceeding in each dialog, and
>>>    defining the current media parameters for that dialog
>>> - you are deciding which dialog(s) you want to process and send
>>>    media to.
>>>
>>> As Brian notes, it is complex to decide which dialog to honor 
>>> during the early stage. But once you settle on one dialog at 
>>> least you will then know what to do.
>>>
>>> 	Paul
>>>
>>>> -----Original Message-----
>>>> From: Brian Stucker [mailto:bstucker@nortel.com]
>>>> Sent: 16 November 2006 17:37
>>>> To: Hadriel Kaplan; Siddhartha Bhakta; sipping@ietf.org
>>>> Cc: Siddhartha Bhakta
>>>> Subject: RE: [SIPPING] SDP offer/answer in early 
>>>> media:draft-sawada-sipping-sip-offeranswer-01.txt
>>>>
>>>> Fake forking...
>>>>
>>>> Please note, if there's SDP in the other provisional 
>>> responses, you're 
>>>> going to run into problems potentially getting themedia rendered 
>>>> (assuming that's your goal).
>>>>
>>>> Regards,
>>>> Brian
>>>>
>>>>> -----Original Message-----
>>>>> From: Hadriel Kaplan [mailto:HKaplan@acmepacket.com]
>>>>> Sent: Sunday, November 12, 2006 8:40 PM
>>>>> To: 'Siddhartha Bhakta'; sipping@ietf.org
>>>>> Cc: 'Siddhartha Bhakta'
>>>>> Subject: RE: [SIPPING] SDP offer/answer in early 
>>>>> media:draft-sawada-sipping-sip-offeranswer-01.txt
>>>>>
>>>>> If the Proxy is sending back SDP by itself it's not a 
>> proxy, but a 
>>>>> b2bua, as is the PBX.
>>>>> As such, it should be answering in a different dialog, 
>> so its SDP 
>>>>> answer should be in a 183 with a different To-tag from the 
>>> 180, and 
>>>>> the PBX's 180 SDP should have a different To-tag than 
>> the 200ok.  
>>>>> That would make them look like a forked call and follow the RFCs.
>>>>>
>>>>> Of course the reality is often different, but being a 
>>> b2bua yourself 
>>>>> it's in your power to fix it.  How you do that is not up 
>>> to the IETF 
>>>>> to define, as the very idea of it goes against the IETF's 
>>> view of the 
>>>>> world. (and I say that in a positive way, as I think it would be 
>>>>> practically impossible for the IETF to do otherwise)
>>>>>
>>>>> -hadriel
>>>>>
>>>>>
>>>>>> -----Original Message-----
>>>>>> From: Siddhartha Bhakta [mailto:sbhakta007@yahoo.co.in]
>>>>>> Sent: Sunday, November 12, 2006 2:11 PM
>>>>>> To: sipping@ietf.org
>>>>>> Cc: sbhakta007@yahoo.co.in; Siddhartha Bhakta
>>>>>> Subject: [SIPPING] SDP offer/answer in early
>>>>>> media:draft-sawada-sipping- sip-offeranswer-01.txt
>>>>>>
>>>>>> I am re-sending it as in last email, the figure was distorted.
>>>>>>
>>>>>> Though, post RFC3261 drafts indicate that one valid 
>>> answer would be 
>>>>>> there for one SDP offer, the our old friend early-media
>>>>> flow is still
>>>>>> present in practice.
>>>>>>
>>>>>> B2BUA/SBC     Proxy         PABX        Bob
>>>>>>   |            |            |            |
>>>>>>   | INVITE F1  |            |            |
>>>>>>   |----------->| INVITE F2  |            |
>>>>>>   |   183 F3   |----------->| INVITE F4  |
>>>>>>   |<-----------|            |----------->|
>>>>>>   |            |            |            |
>>>>>>   |            |            |   180 without SDP F5
>>>>>>   |            |            |<-----------|
>>>>>>   |            |            |            |
>>>>>>   |            |  180 with SDP F6        |
>>>>>>   |   180 F7   |<-----------|            |
>>>>>>   |<-----------|            |  200 F8    |
>>>>>>   |            |  200 F9    |<-----------|
>>>>>>   |   200 F10  |<-----------|            |
>>>>>>   |<-----------|            |            |
>>>>>>   |   ACK F11  |            |            |
>>>>>>   |----------->|  ACK F12   |            |
>>>>>>   |            |----------->|  ACK F13   |
>>>>>>   |            |            |----------->|
>>>>>>
>>>>>> Our B2BUA is facing the problem as specified above.
>>>>>> The first answer is carried
>>>>>> by F3. This SDP is specified by Proxy. The 2nd SDP answer
>>>>> is carried
>>>>>> by F6,F7.
>>>>>> In fact, this SDP is specified by PABX. The third SDP answer is 
>>>>>> carried by F8,F9,F10.
>>>>>> This SDP is specified by called party (Bob).
>>>>>>
>>>>>> As per the RFC3261 & RFC3264, the SDP answer carried by
>>>>> F6,F7 should
>>>>>> match with F3 as far as B2BUA is concerned. Therefore, 
>>> B2BUA shall 
>>>>>> ignore it. This is leading to the fact that SIP UA behind 
>>> B2BUA can 
>>>>>> not listen to ringtone fed by PABX.
>>>>>> Similarly, SDP answer carried by F8,F9,F10 shall be 
>> ignored. This 
>>>>>> shall lead to the fact that SIP UA behind B2BUA can not 
>> listen to 
>>>>>> Bob's voice. Too bad.
>>>>>> This problem is due to the fact that RFC3261 & RFC3264 are not 
>>>>>> backward compatible.
>>>>>>
>>>>>>
>>>>>> The sec 2.2. of
>>>>>> draft-sawada-sipping-sip-offeranswer-01.txt indicates 
>> that UPDATE 
>>>>>> should be used to update the media in early media. But in
>>>>> practice old
>>>>>> early-media flow (i.e., sending different SDPs over 
>> different 1xx 
>>>>>> responses of
>>>>>> INVITE) is quite common.
>>>>>> Can we somehow make new SDP offer/answer specified in
>>>>>> RFC3261 & RFC3264 backward
>>>>>> compatible?
>>>>>>
>>>>>> The old standard (early-media), allows multiple SDP answers
>>>>> of single
>>>>>> SDP offer. can we somehow induce this thing in SDP
>>>>> offer/answer model? 
>>>>>> If many of you think the need of it then I may comeup with
>>>>> some thumb
>>>>>> rule for the same.
>>>>>>
>>>>>> Thanks & Regards,
>>>>>> Siddhartha
>>>>> _______________________________________________
>>>>> Sipping mailing list  
>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>> This list is for NEW development of the application of SIP Use 
>>>>> sip-implementors@cs.columbia.edu for questions on 
>> current sip Use 
>>>>> sip@ietf.org for new developments of core SIP
>>>>>
>>>>
>>>>
>>>> ---------------
>>>> This e-mail may contain confidential and/or privileged 
>>> information. If you are not the intended recipient (or have 
>>> received this e-mail in error) please notify the sender 
>>> immediately and delete this e-mail. Any unauthorized copying, 
>>> disclosure or distribution of the contents in this e-mail is 
>>> strictly forbidden.
>>>> ---------------
>>>>
>>>>
>>>> _______________________________________________
>>>> Sipping mailing list  
>> https://www1.ietf.org/mailman/listinfo/sipping
>>>> This list is for NEW development of the application of SIP Use 
>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>> sip@ietf.org for new developments of core SIP
>>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP 
>>> Use sip-implementors@cs.columbia.edu for questions on current 
>>> sip Use sip@ietf.org for new developments of core SIP
>>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP
>> Use sip-implementors@cs.columbia.edu for questions on current sip
>> Use sip@ietf.org for new developments of core SIP
>>
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 09:26:44 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGybd-0004I7-1L; Tue, 13 Feb 2007 09:26:05 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGybb-0004Gz-JQ
	for sipping@ietf.org; Tue, 13 Feb 2007 09:26:03 -0500
Received: from rtp-iport-1.cisco.com ([64.102.122.148])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGyba-0004O9-99
	for sipping@ietf.org; Tue, 13 Feb 2007 09:26:03 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by rtp-iport-1.cisco.com with ESMTP; 13 Feb 2007 09:26:02 -0500
X-IronPort-AV: i="4.14,163,1170651600"; 
	d="scan'208"; a="52913499:sNHT55172320"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1DEQ2Pa014466; 
	Tue, 13 Feb 2007 09:26:02 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1DEQ1VV001608; 
	Tue, 13 Feb 2007 09:26:01 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 09:26:01 -0500
Received: from [161.44.182.226] ([161.44.182.226]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 09:26:01 -0500
Message-ID: <45D1CA78.5010601@cisco.com>
Date: Tue, 13 Feb 2007 09:26:00 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Attila Sipos <Attila.Sipos@vegastream.com>
Subject: Re: [SIPPING] SDP
	offer/answer	inearlymedia:draft-sawada-sipping-sip-offeranswer-01.txt
References: <8F1F2062CA7D754A838F0E66097D23A0C24180@veg-brk-svr-pri.vegastream>
In-Reply-To: <8F1F2062CA7D754A838F0E66097D23A0C24180@veg-brk-svr-pri.vegastream>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 13 Feb 2007 14:26:01.0141 (UTC)
	FILETIME=[E2B42650:01C74F7A]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=3299; t=1171376762;
	x=1172240762; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[SIPPING]=20SDP=20offer/answer=09inearlymedia=3Adraft
	-sawada-sipping-sip-offeranswer-01.txt |Sender:=20
	|To:=20Attila=20Sipos=20<Attila.Sipos@vegastream.com>;
	bh=tRtyJrDuU7xGEe5BVfgpIsSorilrahGW8RL2oZ3/qhA=;
	b=J8Ga1vJ0MCOdLMQyFcEmRqCxjXVShkq7E9pCy8KACNGmc0UYyJJN2wAgmUIYvjyBR3DBJo/Z
	CvQMbPdQ1SuxP8ehsgkDeDsiM3/NQhsggwUgNvbzEt2dGikeUp1TBZpO;
Authentication-Results: rtp-dkim-2; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 31247fb3be228bb596db9127becad0bc
Cc: Siddhartha Bhakta <sbhakta007@yahoo.co.in>,
	Siddhartha Bhakta <Siddhartha.Bhakta@newport-networks.com>,
	Brian Stucker <bstucker@nortel.com>, "Elwell,
	John" <john.elwell@siemens.com>,
	"Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>,
	sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Attila Sipos wrote:
>>> Although a 19x response might make knowing when early media from the
>>> interloper to the SDP exchange is no longer necessary, the same thing
>>> can be handled by (most) clients today by simply sending a 180 back to
>>> the originator to switch back to local ringing when that endpoint no
>>> longer has any media it wishes the originator to render (not that the
> 
> Is this really going to work?
> If it works, it is just luck I think.
> In RFC 3261 it says...
> 
>              The UAC MUST treat the first session
>          description it receives as the answer, and MUST ignore any
>          session descriptions in subsequent responses to the initial
>          INVITE.

Whether it says it or not, the above is intended to apply *per dialog*. 
Each early dialog is handled independently.

It would be applicable if the 190 carried a session description. I 
haven't seen, or don't recall, the specifics of what had been proposed 
for the 190, so I don't know if it depended on a session description or not.

> So should lack of SDP get ignored?
> Intuitively I would say yes.
> 
> Isn't UPDATE the only legal way for the fake fork
> to say "don't talk or listen to me"?

My impression is that the intent is to say: "this early dialog is 
ended". It has to be something <200 so that it can be returned while 
allowing the other forks to proceed.

	Paul

> Regards,
> 
> Attila
> 
> 
> -----Original Message-----
> From: Brian Stucker [mailto:bstucker@nortel.com]
> Sent: 13 February 2007 13:08
> To: Christer Holmberg (JO/LMF); Elwell, John; Paul Kyzivat; Siddhartha
> Bhakta
> Cc: sipping@ietf.org; Siddhartha Bhakta
> Subject: RE: [SIPPING] SDP offer/answer
> inearlymedia:draft-sawada-sipping-sip-offeranswer-01.txt
> 
> 
> Fake forking is a classic example of the middle screwing up a process
> that should only be between final endpoints.
> 
> A box in the middle forks the call off to an endpoint that the
> originator never intended to call that is acting as if it is one of the
> called party's devices, when it is not, and has no intention of ever
> answering the call.
> 
> Although a 19x response might make knowing when early media from the
> interloper to the SDP exchange is no longer necessary, the same thing
> can be handled by (most) clients today by simply sending a 180 back to
> the originator to switch back to local ringing when that endpoint no
> longer has any media it wishes the originator to render (not that the
> originator had to render anything in the first place, but that's a
> separate discussion).
> 
> What would be nice is if each dialog was handled explicitly, rather than
> implicitly at the originator for early media (i.e. the originator
> explicitly contacts the device that wishes to generate early media
> rather than having the network resort to tricking the originator into
> rendering the desired media).
> 
> I have a draft sitting in the queue that proposes such a mechanism. The
> XML2RFC template still used (C) 2006 when I generated the draft text, so
> it got bounced during the I-D nits check. Until it appears in the
> archives, anyone interested can email me and I'll send you a copy of the
> draft.
> 
> Regards,
> Brian
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 09:46:53 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGyvG-0007En-BJ; Tue, 13 Feb 2007 09:46:22 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGyvD-0007Df-5b
	for sipping@ietf.org; Tue, 13 Feb 2007 09:46:19 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGytm-0001mm-Hh
	for sipping@ietf.org; Tue, 13 Feb 2007 09:44:57 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	C172520702; Tue, 13 Feb 2007 15:44:39 +0100 (CET)
X-AuditID: c1b4fb3e-ae6d0bb0000007e1-6c-45d1ced72d65 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	A468C204B7; Tue, 13 Feb 2007 15:44:39 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 15:44:39 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [SIPPING] SDP
	offer/answer	inearlymedia:draft-sawada-sipping-sip-offeranswer-01.txt
Date: Tue, 13 Feb 2007 15:44:38 +0100
Message-ID: <7374777208BDC7449D5620EF942325670206A8B6@esealmw113.eemea.ericsson.se>
In-Reply-To: <45D1CA78.5010601@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [SIPPING] SDP
	offer/answer	inearlymedia:draft-sawada-sipping-sip-offeranswer-01.txt
Thread-Index: AcdPeuZocSj8O0BzTvu6rm25+PFrTQAAlEWQ
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>,
	"Attila Sipos" <Attila.Sipos@vegastream.com>
X-OriginalArrivalTime: 13 Feb 2007 14:44:39.0370 (UTC)
	FILETIME=[7D385AA0:01C74F7D]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 32b73d73e8047ed17386f9799119ce43
Cc: sipping@ietf.org,
	Siddhartha Bhakta <Siddhartha.Bhakta@newport-networks.com>,
	"Elwell, John" <john.elwell@siemens.com>,
	Siddhartha Bhakta <sbhakta007@yahoo.co.in>,
	Brian Stucker <bstucker@nortel.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,=20

>It would be applicable if the 190 carried a session description. I
haven't seen, or don't recall, the specifics=20
>of what had been proposed for the 190, so I don't know if it depended
on a session description or not.

Well, the first time I brought it up it was shot down before we even
started the discussions on the details :)

Regards,

Christer





>=20
> > So should lack of SDP get ignored?
> > Intuitively I would say yes.
> >=20
> > Isn't UPDATE the only legal way for the fake fork to say=20
> "don't talk=20
> > or listen to me"?
>=20
> My impression is that the intent is to say: "this early=20
> dialog is ended". It has to be something <200 so that it can=20
> be returned while allowing the other forks to proceed.
>=20
> 	Paul
>=20
> > Regards,
> >=20
> > Attila
> >=20
> >=20
> > -----Original Message-----
> > From: Brian Stucker [mailto:bstucker@nortel.com]
> > Sent: 13 February 2007 13:08
> > To: Christer Holmberg (JO/LMF); Elwell, John; Paul Kyzivat;=20
> Siddhartha=20
> > Bhakta
> > Cc: sipping@ietf.org; Siddhartha Bhakta
> > Subject: RE: [SIPPING] SDP offer/answer=20
> > inearlymedia:draft-sawada-sipping-sip-offeranswer-01.txt
> >=20
> >=20
> > Fake forking is a classic example of the middle screwing up=20
> a process=20
> > that should only be between final endpoints.
> >=20
> > A box in the middle forks the call off to an endpoint that the=20
> > originator never intended to call that is acting as if it is one of=20
> > the called party's devices, when it is not, and has no intention of=20
> > ever answering the call.
> >=20
> > Although a 19x response might make knowing when early media=20
> from the=20
> > interloper to the SDP exchange is no longer necessary, the=20
> same thing=20
> > can be handled by (most) clients today by simply sending a=20
> 180 back to=20
> > the originator to switch back to local ringing when that=20
> endpoint no=20
> > longer has any media it wishes the originator to render=20
> (not that the=20
> > originator had to render anything in the first place, but that's a=20
> > separate discussion).
> >=20
> > What would be nice is if each dialog was handled explicitly, rather=20
> > than implicitly at the originator for early media (i.e. the=20
> originator=20
> > explicitly contacts the device that wishes to generate early media=20
> > rather than having the network resort to tricking the=20
> originator into=20
> > rendering the desired media).
> >=20
> > I have a draft sitting in the queue that proposes such a mechanism.=20
> > The XML2RFC template still used (C) 2006 when I generated the draft=20
> > text, so it got bounced during the I-D nits check. Until it=20
> appears in=20
> > the archives, anyone interested can email me and I'll send=20
> you a copy=20
> > of the draft.
> >=20
> > Regards,
> > Brian
> >=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 09:50:34 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGyzK-0001pA-04; Tue, 13 Feb 2007 09:50:34 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGyzJ-0001ov-2y
	for sipping@ietf.org; Tue, 13 Feb 2007 09:50:33 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGyzI-0004ES-69
	for sipping@ietf.org; Tue, 13 Feb 2007 09:50:33 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	797F5205F9; Tue, 13 Feb 2007 15:50:27 +0100 (CET)
X-AuditID: c1b4fb3c-aefc6bb0000007de-ef-45d1d033c0a2 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	4D15C205DD; Tue, 13 Feb 2007 15:50:27 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 15:50:26 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [SIPPING] SDP offer/answer	inearly media:draft-sawada-sipping
	-sip-offeranswer-01.txt
Date: Tue, 13 Feb 2007 15:50:25 +0100
Message-ID: <7374777208BDC7449D5620EF942325670206A8EE@esealmw113.eemea.ericsson.se>
In-Reply-To: <45D1C6CF.30801@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [SIPPING] SDP offer/answer	inearly media:draft-sawada-sipping
	-sip-offeranswer-01.txt
Thread-Index: AcdPeLyr3yXLlp79QRKVPCjnBBcP9QABRSmw
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>,
	"Elwell, John" <john.elwell@siemens.com>
X-OriginalArrivalTime: 13 Feb 2007 14:50:26.0953 (UTC)
	FILETIME=[4C654F90:01C74F7E]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 9e5c23589e6cce06555030c0194c9e2b
Cc: Brian Stucker <bstucker@nortel.com>, sipping@ietf.org,
	Siddhartha Bhakta <sbhakta007@yahoo.co.in>,
	Siddhartha Bhakta <Siddhartha.Bhakta@newport-networks.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,=20

> > I don't know what you mean by "fake forking", but I think=20
> there is an=20
> > issue with serial forking, e.g.:
> > - INVITE request to user@example.com
> > - proxy for example.com forwards to PSTN gateway
> > - PSTN gateway provides early media
> > - proxy is scripted to cancel the PSTN call after a certain=20
> time and=20
> > forward the request elsewhere (e.g., to a mobile or voicemail)
> > - the new destination sends early media or answers quickly=20
> and sends=20
> > regular media
> > - the UAC will probably still be rendering early media from=20
> the PSTN=20
> > and will not know that this is in the process of stopping (even if=20
> > packets stop arriving, this could be due to silence)
> >=20
> > In this case your proposed 19x provisional response could=20
> tell the UAC=20
> > to stop listening to the corresponding RTP stream immediately.
>=20
>I don't see how that would work. To work, the UAC must=20
>correlate the 190 response with the particular media stream=20
>being received, so it knows to abandon it. But there isn't=20
>any well defined way to do that. *If* the UAS uses symmetric=20
>RTP, and if there is no NAT in the middle, then it might be=20
>able to do it. I guess this might work in conjunction with ICE.

True, it MAY not always be possible to correlate the 190 response with
the media stream. On the other hand, if 190 is sent on a dialog, the
media for that dialog will most likely also terminate, since the dialog
with the node sending the media has been terminated.

Regards,

Christer








>=20
> 	Paul
>=20
> > John
> >=20
> >> -----Original Message-----
> >> From: Christer Holmberg (JO/LMF)
> >> [mailto:christer.holmberg@ericsson.com]
> >> Sent: 13 February 2007 10:57
> >> To: Paul Kyzivat; Siddhartha Bhakta
> >> Cc: sipping@ietf.org; Siddhartha Bhakta; Brian Stucker
> >> Subject: RE: [SIPPING] SDP offer/answer inearly=20
> >> media:draft-sawada-sipping-sip-offeranswer-01.txt
> >>
> >>
> >> Hi,
> >>
> >> I once proposed that it should be possible to terminate an early=20
> >> dialog with an 19x provisional response.
> >>
> >> Then, if "fake forking" is used, the dialog not used=20
> anymore could be=20
> >> terminted and the UAC knows he doesn't have to care about it (and=20
> >> possible media coming on it).
> >>
> >> Regards,
> >>
> >> Christer
> >> =20
> >>
> >>> -----Original Message-----
> >>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
> >>> Sent: 16. marraskuuta 2006 21:58
> >>> To: Siddhartha Bhakta
> >>> Cc: sipping@ietf.org; 'Siddhartha Bhakta'; 'Brian Stucker'
> >>> Subject: Re: [SIPPING] SDP offer/answer inearly=20
> >>> media:draft-sawada-sipping-sip-offeranswer-01.txt
> >>>
> >>>
> >>>
> >>> Siddhartha Bhakta wrote:
> >>>> I agree with you. No matter whether different provisional
> >> responses
> >>>> are creating different dialogs or not, case is same as
> >> far as media
> >>>> rendering/committing is concerned. For media committing,
> >> there are
> >>>> multiple answers for a given offer.
> >>>>
> >>>> Like you, I also can not understand what we are gaining
> >> by creating
> >>>> multiple dialogs here except the chance to manage
> >>> offer/answer state
> >>>> as per RFC3261/RFC3264. It seems that multiple SDP answers
> >>> for a given
> >>>> SDP offer is the reality. We can not avoid that by applying
> >>> any tricky
> >>>> logic in dialog management (like Fake Forking).
> >>> If you ignore dialogs, it becomes impossible to impose any=20
> >>> offer/answer rules.
> >>>
> >>> By paying attention to the separate dialogs, it is=20
> possible to have=20
> >>> offer/answer rules that make sense. If you just mix=20
> together offers=20
> >>> or answers from multiple destinations it becomes=20
> impossible to make=20
> >>> any sense of what you have once it is determined which=20
> dialog will=20
> >>> be completed.
> >>>
> >>> For the period when there are multiple dialogs, you have two=20
> >>> different things going on:
> >>> - an offer/answer state machine is proceeding in each dialog, and
> >>>    defining the current media parameters for that dialog
> >>> - you are deciding which dialog(s) you want to process and send
> >>>    media to.
> >>>
> >>> As Brian notes, it is complex to decide which dialog to=20
> honor during=20
> >>> the early stage. But once you settle on one dialog at=20
> least you will=20
> >>> then know what to do.
> >>>
> >>> 	Paul
> >>>
> >>>> -----Original Message-----
> >>>> From: Brian Stucker [mailto:bstucker@nortel.com]
> >>>> Sent: 16 November 2006 17:37
> >>>> To: Hadriel Kaplan; Siddhartha Bhakta; sipping@ietf.org
> >>>> Cc: Siddhartha Bhakta
> >>>> Subject: RE: [SIPPING] SDP offer/answer in early=20
> >>>> media:draft-sawada-sipping-sip-offeranswer-01.txt
> >>>>
> >>>> Fake forking...
> >>>>
> >>>> Please note, if there's SDP in the other provisional
> >>> responses, you're
> >>>> going to run into problems potentially getting themedia rendered=20
> >>>> (assuming that's your goal).
> >>>>
> >>>> Regards,
> >>>> Brian
> >>>>
> >>>>> -----Original Message-----
> >>>>> From: Hadriel Kaplan [mailto:HKaplan@acmepacket.com]
> >>>>> Sent: Sunday, November 12, 2006 8:40 PM
> >>>>> To: 'Siddhartha Bhakta'; sipping@ietf.org
> >>>>> Cc: 'Siddhartha Bhakta'
> >>>>> Subject: RE: [SIPPING] SDP offer/answer in early=20
> >>>>> media:draft-sawada-sipping-sip-offeranswer-01.txt
> >>>>>
> >>>>> If the Proxy is sending back SDP by itself it's not a
> >> proxy, but a
> >>>>> b2bua, as is the PBX.
> >>>>> As such, it should be answering in a different dialog,
> >> so its SDP
> >>>>> answer should be in a 183 with a different To-tag from the
> >>> 180, and
> >>>>> the PBX's 180 SDP should have a different To-tag than
> >> the 200ok. =20
> >>>>> That would make them look like a forked call and follow=20
> the RFCs.
> >>>>>
> >>>>> Of course the reality is often different, but being a
> >>> b2bua yourself
> >>>>> it's in your power to fix it.  How you do that is not up
> >>> to the IETF
> >>>>> to define, as the very idea of it goes against the IETF's
> >>> view of the
> >>>>> world. (and I say that in a positive way, as I think it=20
> would be=20
> >>>>> practically impossible for the IETF to do otherwise)
> >>>>>
> >>>>> -hadriel
> >>>>>
> >>>>>
> >>>>>> -----Original Message-----
> >>>>>> From: Siddhartha Bhakta [mailto:sbhakta007@yahoo.co.in]
> >>>>>> Sent: Sunday, November 12, 2006 2:11 PM
> >>>>>> To: sipping@ietf.org
> >>>>>> Cc: sbhakta007@yahoo.co.in; Siddhartha Bhakta
> >>>>>> Subject: [SIPPING] SDP offer/answer in early
> >>>>>> media:draft-sawada-sipping- sip-offeranswer-01.txt
> >>>>>>
> >>>>>> I am re-sending it as in last email, the figure was distorted.
> >>>>>>
> >>>>>> Though, post RFC3261 drafts indicate that one valid
> >>> answer would be
> >>>>>> there for one SDP offer, the our old friend early-media
> >>>>> flow is still
> >>>>>> present in practice.
> >>>>>>
> >>>>>> B2BUA/SBC     Proxy         PABX        Bob
> >>>>>>   |            |            |            |
> >>>>>>   | INVITE F1  |            |            |
> >>>>>>   |----------->| INVITE F2  |            |
> >>>>>>   |   183 F3   |----------->| INVITE F4  |
> >>>>>>   |<-----------|            |----------->|
> >>>>>>   |            |            |            |
> >>>>>>   |            |            |   180 without SDP F5
> >>>>>>   |            |            |<-----------|
> >>>>>>   |            |            |            |
> >>>>>>   |            |  180 with SDP F6        |
> >>>>>>   |   180 F7   |<-----------|            |
> >>>>>>   |<-----------|            |  200 F8    |
> >>>>>>   |            |  200 F9    |<-----------|
> >>>>>>   |   200 F10  |<-----------|            |
> >>>>>>   |<-----------|            |            |
> >>>>>>   |   ACK F11  |            |            |
> >>>>>>   |----------->|  ACK F12   |            |
> >>>>>>   |            |----------->|  ACK F13   |
> >>>>>>   |            |            |----------->|
> >>>>>>
> >>>>>> Our B2BUA is facing the problem as specified above.
> >>>>>> The first answer is carried
> >>>>>> by F3. This SDP is specified by Proxy. The 2nd SDP answer
> >>>>> is carried
> >>>>>> by F6,F7.
> >>>>>> In fact, this SDP is specified by PABX. The third SDP=20
> answer is=20
> >>>>>> carried by F8,F9,F10.
> >>>>>> This SDP is specified by called party (Bob).
> >>>>>>
> >>>>>> As per the RFC3261 & RFC3264, the SDP answer carried by
> >>>>> F6,F7 should
> >>>>>> match with F3 as far as B2BUA is concerned. Therefore,
> >>> B2BUA shall
> >>>>>> ignore it. This is leading to the fact that SIP UA behind
> >>> B2BUA can
> >>>>>> not listen to ringtone fed by PABX.
> >>>>>> Similarly, SDP answer carried by F8,F9,F10 shall be
> >> ignored. This
> >>>>>> shall lead to the fact that SIP UA behind B2BUA can not
> >> listen to
> >>>>>> Bob's voice. Too bad.
> >>>>>> This problem is due to the fact that RFC3261 & RFC3264 are not=20
> >>>>>> backward compatible.
> >>>>>>
> >>>>>>
> >>>>>> The sec 2.2. of
> >>>>>> draft-sawada-sipping-sip-offeranswer-01.txt indicates
> >> that UPDATE
> >>>>>> should be used to update the media in early media. But in
> >>>>> practice old
> >>>>>> early-media flow (i.e., sending different SDPs over
> >> different 1xx
> >>>>>> responses of
> >>>>>> INVITE) is quite common.
> >>>>>> Can we somehow make new SDP offer/answer specified in
> >>>>>> RFC3261 & RFC3264 backward
> >>>>>> compatible?
> >>>>>>
> >>>>>> The old standard (early-media), allows multiple SDP answers
> >>>>> of single
> >>>>>> SDP offer. can we somehow induce this thing in SDP
> >>>>> offer/answer model?=20
> >>>>>> If many of you think the need of it then I may comeup with
> >>>>> some thumb
> >>>>>> rule for the same.
> >>>>>>
> >>>>>> Thanks & Regards,
> >>>>>> Siddhartha
> >>>>> _______________________________________________
> >>>>> Sipping mailing list
> >>> https://www1.ietf.org/mailman/listinfo/sipping
> >>>>> This list is for NEW development of the application of SIP Use=20
> >>>>> sip-implementors@cs.columbia.edu for questions on
> >> current sip Use
> >>>>> sip@ietf.org for new developments of core SIP
> >>>>>
> >>>>
> >>>>
> >>>> ---------------
> >>>> This e-mail may contain confidential and/or privileged
> >>> information. If you are not the intended recipient (or=20
> have received=20
> >>> this e-mail in error) please notify the sender immediately and=20
> >>> delete this e-mail. Any unauthorized copying, disclosure or=20
> >>> distribution of the contents in this e-mail is strictly forbidden.
> >>>> ---------------
> >>>>
> >>>>
> >>>> _______________________________________________
> >>>> Sipping mailing list =20
> >> https://www1.ietf.org/mailman/listinfo/sipping
> >>>> This list is for NEW development of the application of SIP Use=20
> >>>> sip-implementors@cs.columbia.edu for questions on=20
> current sip Use=20
> >>>> sip@ietf.org for new developments of core SIP
> >>>>
> >>> _______________________________________________
> >>> Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> >>> This list is for NEW development of the application of SIP=20
> >>> Use sip-implementors@cs.columbia.edu for questions on current=20
> >>> sip Use sip@ietf.org for new developments of core SIP
> >>>
> >> _______________________________________________
> >> Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> >> This list is for NEW development of the application of SIP
> >> Use sip-implementors@cs.columbia.edu for questions on current sip
> >> Use sip@ietf.org for new developments of core SIP
> >>
> >=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 10:02:59 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HGzAw-0007k4-FQ; Tue, 13 Feb 2007 10:02:34 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HGzAv-0007eP-4G
	for sipping@ietf.org; Tue, 13 Feb 2007 10:02:33 -0500
Received: from mailgate.siemenscomms.co.uk ([195.171.110.225]
	helo=bemg01.siemenscomms.co.uk)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HGzAr-0000pC-QK
	for sipping@ietf.org; Tue, 13 Feb 2007 10:02:33 -0500
Received: from ntht207e.uksgcs.siemenscomms.co.uk ([137.223.247.82])
	by siemenscomms.co.uk (PMDF V6.0-24 #40642)
	with ESMTP id <0JDE001J5PJQVQ@siemenscomms.co.uk> for sipping@ietf.org;
	Tue, 13 Feb 2007 14:57:26 +0000 (GMT)
Received: by ntht207e.uksgcs.siemenscomms.co.uk with Internet Mail Service
	(5.5.2657.72)	id <D2K3XJMP>; Tue, 13 Feb 2007 14:57:26 +0000
Content-return: allowed
Date: Tue, 13 Feb 2007 14:57:24 +0000
From: "Elwell, John" <john.elwell@siemens.com>
Subject: RE: [SIPPING] SDP offer/answer	inearly media:draft-sawada-sipping
	-sip-offeranswer-01.txt
To: Paul Kyzivat <pkyzivat@cisco.com>
Message-id: <50B1CBA96870A34799A506B2313F26670B060053@ntht201e.siemenscomms.co.uk>
MIME-version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
Content-type: text/plain
Content-transfer-encoding: 7BIT
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 9f79b8e383fd3af2b1b5b1d0910f6094
Cc: Siddhartha Bhakta <Siddhartha.Bhakta@newport-networks.com>,
	sipping@ietf.org, Brian Stucker <bstucker@nortel.com>,
	Siddhartha Bhakta <sbhakta007@yahoo.co.in>,
	"Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Paul,

Yes, I failed to elaborate on that. Another means of correlation is if MIKEY
is used or if something along the lines of the Stucker early-media-coping
draft. These two mechanisms would provide the SSRC value, but unfortunately
that is still not guaranteed to be unique.

John

> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
> Sent: 13 February 2007 14:10
> To: Elwell, John
> Cc: Christer Holmberg (JO/LMF); Siddhartha Bhakta; 
> sipping@ietf.org; Siddhartha Bhakta; Brian Stucker
> Subject: Re: [SIPPING] SDP offer/answer inearly 
> media:draft-sawada-sipping -sip-offeranswer-01.txt
> 
> 
> 
> Elwell, John wrote:
> > Christer,
> > 
> > I don't know what you mean by "fake forking", but I think 
> there is an issue
> > with serial forking, e.g.:
> > - INVITE request to user@example.com
> > - proxy for example.com forwards to PSTN gateway
> > - PSTN gateway provides early media
> > - proxy is scripted to cancel the PSTN call after a certain 
> time and forward
> > the request elsewhere (e.g., to a mobile or voicemail)
> > - the new destination sends early media or answers quickly 
> and sends regular
> > media
> > - the UAC will probably still be rendering early media from 
> the PSTN and
> > will not know that this is in the process of stopping (even 
> if packets stop
> > arriving, this could be due to silence)
> > 
> > In this case your proposed 19x provisional response could 
> tell the UAC to
> > stop listening to the corresponding RTP stream immediately.
> 
> I don't see how that would work. To work, the UAC must 
> correlate the 190 
> response with the particular media stream being received, so 
> it knows to 
> abandon it. But there isn't any well defined way to do that. *If* the 
> UAS uses symmetric RTP, and if there is no NAT in the middle, then it 
> might be able to do it. I guess this might work in 
> conjunction with ICE.
> 
> 	Paul
> 
> > John
> > 
> >> -----Original Message-----
> >> From: Christer Holmberg (JO/LMF) 
> >> [mailto:christer.holmberg@ericsson.com] 
> >> Sent: 13 February 2007 10:57
> >> To: Paul Kyzivat; Siddhartha Bhakta
> >> Cc: sipping@ietf.org; Siddhartha Bhakta; Brian Stucker
> >> Subject: RE: [SIPPING] SDP offer/answer inearly 
> >> media:draft-sawada-sipping-sip-offeranswer-01.txt
> >>
> >>
> >> Hi,
> >>
> >> I once proposed that it should be possible to terminate an 
> >> early dialog
> >> with an 19x provisional response.
> >>
> >> Then, if "fake forking" is used, the dialog not used 
> anymore could be
> >> terminted and the UAC knows he doesn't have to care about it (and
> >> possible media coming on it).
> >>
> >> Regards,
> >>
> >> Christer
> >>  
> >>
> >>> -----Original Message-----
> >>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
> >>> Sent: 16. marraskuuta 2006 21:58
> >>> To: Siddhartha Bhakta
> >>> Cc: sipping@ietf.org; 'Siddhartha Bhakta'; 'Brian Stucker'
> >>> Subject: Re: [SIPPING] SDP offer/answer inearly 
> >>> media:draft-sawada-sipping-sip-offeranswer-01.txt
> >>>
> >>>
> >>>
> >>> Siddhartha Bhakta wrote:
> >>>> I agree with you. No matter whether different provisional 
> >> responses 
> >>>> are creating different dialogs or not, case is same as 
> >> far as media 
> >>>> rendering/committing is concerned. For media committing, 
> >> there are 
> >>>> multiple answers for a given offer.
> >>>>
> >>>> Like you, I also can not understand what we are gaining 
> >> by creating 
> >>>> multiple dialogs here except the chance to manage 
> >>> offer/answer state 
> >>>> as per RFC3261/RFC3264. It seems that multiple SDP answers 
> >>> for a given 
> >>>> SDP offer is the reality. We can not avoid that by applying 
> >>> any tricky 
> >>>> logic in dialog management (like Fake Forking).
> >>> If you ignore dialogs, it becomes impossible to impose any 
> >>> offer/answer rules.
> >>>
> >>> By paying attention to the separate dialogs, it is possible 
> >>> to have offer/answer rules that make sense. If you just mix 
> >>> together offers or answers from multiple destinations it 
> >>> becomes impossible to make any sense of what you have once it 
> >>> is determined which dialog will be completed.
> >>>
> >>> For the period when there are multiple dialogs, you have two 
> >>> different things going on:
> >>> - an offer/answer state machine is proceeding in each dialog, and
> >>>    defining the current media parameters for that dialog
> >>> - you are deciding which dialog(s) you want to process and send
> >>>    media to.
> >>>
> >>> As Brian notes, it is complex to decide which dialog to honor 
> >>> during the early stage. But once you settle on one dialog at 
> >>> least you will then know what to do.
> >>>
> >>> 	Paul
> >>>
> >>>> -----Original Message-----
> >>>> From: Brian Stucker [mailto:bstucker@nortel.com]
> >>>> Sent: 16 November 2006 17:37
> >>>> To: Hadriel Kaplan; Siddhartha Bhakta; sipping@ietf.org
> >>>> Cc: Siddhartha Bhakta
> >>>> Subject: RE: [SIPPING] SDP offer/answer in early 
> >>>> media:draft-sawada-sipping-sip-offeranswer-01.txt
> >>>>
> >>>> Fake forking...
> >>>>
> >>>> Please note, if there's SDP in the other provisional 
> >>> responses, you're 
> >>>> going to run into problems potentially getting themedia rendered 
> >>>> (assuming that's your goal).
> >>>>
> >>>> Regards,
> >>>> Brian
> >>>>
> >>>>> -----Original Message-----
> >>>>> From: Hadriel Kaplan [mailto:HKaplan@acmepacket.com]
> >>>>> Sent: Sunday, November 12, 2006 8:40 PM
> >>>>> To: 'Siddhartha Bhakta'; sipping@ietf.org
> >>>>> Cc: 'Siddhartha Bhakta'
> >>>>> Subject: RE: [SIPPING] SDP offer/answer in early 
> >>>>> media:draft-sawada-sipping-sip-offeranswer-01.txt
> >>>>>
> >>>>> If the Proxy is sending back SDP by itself it's not a 
> >> proxy, but a 
> >>>>> b2bua, as is the PBX.
> >>>>> As such, it should be answering in a different dialog, 
> >> so its SDP 
> >>>>> answer should be in a 183 with a different To-tag from the 
> >>> 180, and 
> >>>>> the PBX's 180 SDP should have a different To-tag than 
> >> the 200ok.  
> >>>>> That would make them look like a forked call and follow 
> the RFCs.
> >>>>>
> >>>>> Of course the reality is often different, but being a 
> >>> b2bua yourself 
> >>>>> it's in your power to fix it.  How you do that is not up 
> >>> to the IETF 
> >>>>> to define, as the very idea of it goes against the IETF's 
> >>> view of the 
> >>>>> world. (and I say that in a positive way, as I think it 
> would be 
> >>>>> practically impossible for the IETF to do otherwise)
> >>>>>
> >>>>> -hadriel
> >>>>>
> >>>>>
> >>>>>> -----Original Message-----
> >>>>>> From: Siddhartha Bhakta [mailto:sbhakta007@yahoo.co.in]
> >>>>>> Sent: Sunday, November 12, 2006 2:11 PM
> >>>>>> To: sipping@ietf.org
> >>>>>> Cc: sbhakta007@yahoo.co.in; Siddhartha Bhakta
> >>>>>> Subject: [SIPPING] SDP offer/answer in early
> >>>>>> media:draft-sawada-sipping- sip-offeranswer-01.txt
> >>>>>>
> >>>>>> I am re-sending it as in last email, the figure was distorted.
> >>>>>>
> >>>>>> Though, post RFC3261 drafts indicate that one valid 
> >>> answer would be 
> >>>>>> there for one SDP offer, the our old friend early-media
> >>>>> flow is still
> >>>>>> present in practice.
> >>>>>>
> >>>>>> B2BUA/SBC     Proxy         PABX        Bob
> >>>>>>   |            |            |            |
> >>>>>>   | INVITE F1  |            |            |
> >>>>>>   |----------->| INVITE F2  |            |
> >>>>>>   |   183 F3   |----------->| INVITE F4  |
> >>>>>>   |<-----------|            |----------->|
> >>>>>>   |            |            |            |
> >>>>>>   |            |            |   180 without SDP F5
> >>>>>>   |            |            |<-----------|
> >>>>>>   |            |            |            |
> >>>>>>   |            |  180 with SDP F6        |
> >>>>>>   |   180 F7   |<-----------|            |
> >>>>>>   |<-----------|            |  200 F8    |
> >>>>>>   |            |  200 F9    |<-----------|
> >>>>>>   |   200 F10  |<-----------|            |
> >>>>>>   |<-----------|            |            |
> >>>>>>   |   ACK F11  |            |            |
> >>>>>>   |----------->|  ACK F12   |            |
> >>>>>>   |            |----------->|  ACK F13   |
> >>>>>>   |            |            |----------->|
> >>>>>>
> >>>>>> Our B2BUA is facing the problem as specified above.
> >>>>>> The first answer is carried
> >>>>>> by F3. This SDP is specified by Proxy. The 2nd SDP answer
> >>>>> is carried
> >>>>>> by F6,F7.
> >>>>>> In fact, this SDP is specified by PABX. The third SDP 
> answer is 
> >>>>>> carried by F8,F9,F10.
> >>>>>> This SDP is specified by called party (Bob).
> >>>>>>
> >>>>>> As per the RFC3261 & RFC3264, the SDP answer carried by
> >>>>> F6,F7 should
> >>>>>> match with F3 as far as B2BUA is concerned. Therefore, 
> >>> B2BUA shall 
> >>>>>> ignore it. This is leading to the fact that SIP UA behind 
> >>> B2BUA can 
> >>>>>> not listen to ringtone fed by PABX.
> >>>>>> Similarly, SDP answer carried by F8,F9,F10 shall be 
> >> ignored. This 
> >>>>>> shall lead to the fact that SIP UA behind B2BUA can not 
> >> listen to 
> >>>>>> Bob's voice. Too bad.
> >>>>>> This problem is due to the fact that RFC3261 & RFC3264 are not 
> >>>>>> backward compatible.
> >>>>>>
> >>>>>>
> >>>>>> The sec 2.2. of
> >>>>>> draft-sawada-sipping-sip-offeranswer-01.txt indicates 
> >> that UPDATE 
> >>>>>> should be used to update the media in early media. But in
> >>>>> practice old
> >>>>>> early-media flow (i.e., sending different SDPs over 
> >> different 1xx 
> >>>>>> responses of
> >>>>>> INVITE) is quite common.
> >>>>>> Can we somehow make new SDP offer/answer specified in
> >>>>>> RFC3261 & RFC3264 backward
> >>>>>> compatible?
> >>>>>>
> >>>>>> The old standard (early-media), allows multiple SDP answers
> >>>>> of single
> >>>>>> SDP offer. can we somehow induce this thing in SDP
> >>>>> offer/answer model? 
> >>>>>> If many of you think the need of it then I may comeup with
> >>>>> some thumb
> >>>>>> rule for the same.
> >>>>>>
> >>>>>> Thanks & Regards,
> >>>>>> Siddhartha
> >>>>> _______________________________________________
> >>>>> Sipping mailing list  
> >>> https://www1.ietf.org/mailman/listinfo/sipping
> >>>>> This list is for NEW development of the application of SIP Use 
> >>>>> sip-implementors@cs.columbia.edu for questions on 
> >> current sip Use 
> >>>>> sip@ietf.org for new developments of core SIP
> >>>>>
> >>>>
> >>>>
> >>>> ---------------
> >>>> This e-mail may contain confidential and/or privileged 
> >>> information. If you are not the intended recipient (or have 
> >>> received this e-mail in error) please notify the sender 
> >>> immediately and delete this e-mail. Any unauthorized copying, 
> >>> disclosure or distribution of the contents in this e-mail is 
> >>> strictly forbidden.
> >>>> ---------------
> >>>>
> >>>>
> >>>> _______________________________________________
> >>>> Sipping mailing list  
> >> https://www1.ietf.org/mailman/listinfo/sipping
> >>>> This list is for NEW development of the application of SIP Use 
> >>>> sip-implementors@cs.columbia.edu for questions on 
> current sip Use 
> >>>> sip@ietf.org for new developments of core SIP
> >>>>
> >>> _______________________________________________
> >>> Sipping mailing list  
> https://www1.ietf.org/mailman/listinfo/sipping
> >>> This list is for NEW development of the application of SIP 
> >>> Use sip-implementors@cs.columbia.edu for questions on current 
> >>> sip Use sip@ietf.org for new developments of core SIP
> >>>
> >> _______________________________________________
> >> Sipping mailing list  
> https://www1.ietf.org/mailman/listinfo/sipping
> >> This list is for NEW development of the application of SIP
> >> Use sip-implementors@cs.columbia.edu for questions on current sip
> >> Use sip@ietf.org for new developments of core SIP
> >>
> > 
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 13:19:31 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HH2En-00081D-Fb; Tue, 13 Feb 2007 13:18:45 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HH2El-000812-EM
	for sipping@ietf.org; Tue, 13 Feb 2007 13:18:43 -0500
Received: from exprod6og54.obsmtp.com ([64.18.1.189])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1HH2Ek-0007bu-3R
	for sipping@ietf.org; Tue, 13 Feb 2007 13:18:43 -0500
Received: from source ([192.150.20.142]) by exprod6ob54.postini.com
	([64.18.5.12]) with SMTP; Tue, 13 Feb 2007 10:18:40 PST
Received: from inner-relay-3.eur.adobe.com (inner-relay-3b [10.128.4.236])
	by outbound-smtp-2.corp.adobe.com (8.12.10/8.12.10) with ESMTP id
	l1DIJ5Uh000652
	for <sipping@ietf.org>; Tue, 13 Feb 2007 10:19:05 -0800 (PST)
Received: from fe1.corp.adobe.com (fe1.corp.adobe.com [10.8.192.70])
	by inner-relay-3.eur.adobe.com (8.12.10/8.12.9) with ESMTP id
	l1DII9mc021981
	for <sipping@ietf.org>; Tue, 13 Feb 2007 10:18:38 -0800 (PST)
Received: from namail5.corp.adobe.com ([10.8.192.88]) by fe1.corp.adobe.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 10:18:30 -0800
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Tue, 13 Feb 2007 10:18:30 -0800
Message-ID: <24CCCC428EFEA2469BF046DB3C7A8D22453C24@namail5.corp.adobe.com>
In-Reply-To: <200702122256.l1CMueqK020621@dragon.ariadne.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdO+YvA09OnvI+MRL2ceCkkYINQ5AAnY7Nw
References: <45C0AEF2.8010101@ericsson.com><532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com><1170835642.29763.54.camel@n95.nomadiclab.com><45CA08B2.4050603@cisco.com><532B18E13CF9E64380EF5FDDE265E071B6192E@blr-m2-msg.wipro.com><45CC7697.1060204@cisco.com><1171031930.3464.92.camel@localhost.localdomain><200702110449.l1B4nfk2020329@dragon.ariadne.com><24CCCC428EFEA2469BF046DB3C7A8D224538F2@namail5.corp.adobe.com>
	<200702122256.l1CMueqK020621@dragon.ariadne.com>
From: "Henry Sinnreich" <hsinnrei@adobe.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 13 Feb 2007 18:18:30.0720 (UTC)
	FILETIME=[5D4CFC00:01C74F9B]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 2409bba43e9c8d580670fda8b695204a
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

It must be my lack of imagination for not being able to come up with a
credible usage scenario, given the following:

- Is playing chess by voice a credible usage scenario? :-)
Especially since voice recognition is now so often found in PCs and in
mobile phones as well. In other words, chess playing by voice may also
be an endpoint application, without any network involvement. Actually,
if placed in the endpoint it works just as well on poor bandwidth links
by contrast to chess in the network that may require adequate broadband.

- Is anyone's boss or spouse going to approve extra phone charges above
the flat rate for playing chess at a time when the gaming industry has
pioneered free VoIP?
Will such a service be taxed extra and if yes using by what criteria?
Gambling comes to my mind :-)

- VoIP is reported in the press as a give-away and I have seen no
reports of any VoIP provider recording a profit on VoIP. I assume the
losses are often hidden by bundling in triple play, etc.

Given my lack of imagination, can anyone share a more credible usage
scenario for Service Identification in SIP?

What's the use of a standard without a credible usage scenario?

Thanks, Henry


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 14:41:52 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HH3WH-0006cN-Hk; Tue, 13 Feb 2007 14:40:53 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HH3WF-0006aE-HY; Tue, 13 Feb 2007 14:40:51 -0500
Received: from mx11.bbn.com ([128.33.0.80])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HH3WE-0008A6-4v; Tue, 13 Feb 2007 14:40:51 -0500
Received: from dommiel.bbn.com ([192.1.122.15] helo=[127.0.0.1])
	by mx11.bbn.com with esmtp (Exim 4.60)
	(envelope-from <rbarnes@bbn.com>)
	id 1HH3W8-0003pa-5w; Tue, 13 Feb 2007 14:40:45 -0500
Message-ID: <45D2143D.9020907@bbn.com>
Date: Tue, 13 Feb 2007 14:40:45 -0500
From: Richard Barnes <rbarnes@bbn.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Francois Audet <audet@nortel.com>
Subject: Re: [Sipping] Re: [Sip] New draft on early media
References: <1ECE0EB50388174790F9694F77522CCF0F21D58F@zrc2hxm0.corp.nortel.com>
In-Reply-To: <1ECE0EB50388174790F9694F77522CCF0F21D58F@zrc2hxm0.corp.nortel.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 29dc808194f5fb921c09d0040806d6eb
Cc: sip@ietf.org, sipping@ietf.org,
	"Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Francois,

I think I have some responses, but I'd like to ask you first: What do 
you mean by "media cut-through"?

Thanks,
--Richard

Francois Audet wrote:
> Hi Richard,
> 
> Forcing media cut-through to occur only after an Answer has been
> received
> does solve SOME of the early media problems. As others have mentioned,
> there
> are environments that operate like this already. I also know many
> products that
> do not cut-through media before the answer is received.
> 
> But there are other problems related with early media. Brian Stucker's
> draft
> listed a bunch of them. 
> 
> Forking for example comes to mind: even if we did mandate cut-through
> after
> answer, we would still have problems related with forking (i.e., how to
> pick
> the stream that needs to be played, and how to correlate the stream with
> the 
> signaling). How to deal with SRTP key exchange in that scenario is an
> even 
> bigger problem. 
> 
> Clipping is a possible side-effect of your proposal. We'd need some
> mechanism to 
> mitigate it, in particular for auto-answer scenarios.
> 
> 
>> -----Original Message-----
>> From: Richard Barnes [mailto:rbarnes@bbn.com] 
>> Sent: Monday, February 12, 2007 07:08
>> To: Christer Holmberg (JO/LMF)
>> Cc: sip@ietf.org; sipping@ietf.org
>> Subject: [Sipping] Re: [Sip] New draft on early media
>>
>> Christer,
>>
>> Thanks for your comments.  Responses inline.
>>
>> Cheers,
>> --Richard
>>
>>> I haven't had time to read the document yet, but I am not 
>> sure I agree 
>>> with the statement saying that the biggest problem is media 
>> arriving 
>>> before the SDP answer. In some environments such media will 
>> not even 
>>> reach the terminal, due to gating etc.
>> Perhaps a better term would be "media sent before an SDP 
>> answer has been received".  As you point out, sending media 
>> before an SDP offer/answer exchange has completed poses the 
>> risk that that media will not arrive. 
>> To put it in the terms used in the document, media clipping 
>> is a problem caused by early (pre-SDP-answer) media, and thus 
>> one which this document seeks to solve.
>>
>>> Many implementations DO send an SDP answer (or, at least an 
>> unreliable 
>>> "preview" of the SDP answer) before (or at the same time as) early 
>>> media is sent.
>> This is the behavior we would like to standardize -- that an 
>> SDP answer is *always* sent before media of any type.  What 
>> we're trying to do with this document is describe the 
>> requirements for how that answer is sent, and consolidate the 
>> many ways people have proposed to send it.
>>
>>> And, in the cases where implementations do NOT send an SDP answer 
>>> before they start sending early media - what behavior would you be 
>>> able to standardize?
>> We would prefer that standards not allow this behavior -- 
>> that seems like the right thing to me.  What's a little 
>> trickier is that you'd like to not have the answerer sending 
>> media before the answer has *arrived*. 
>>   When the answer is carried in a reliable message, you could 
>> just forbid sending media before receipt has been confirmed.  
>> When the transport is unreliable, the situation is a little 
>> less clear, but I think there's still possibilities for 
>> standardization, e.g., by forbidding answerers to send media 
>> until they've received media.  This is discussed in more 
>> length in the document.
>>
>>> Regards,
>>>
>>> Christer
>>>
>>>
>>>> -----Original Message-----
>>>> From: Richard Barnes [mailto:rbarnes@bbn.com]
>>>> Sent: 8. helmikuuta 2007 21:47
>>>> To: sip@ietf.org; sipping@ietf.org
>>>> Subject: [Sip] New draft on early media
>>>>
>>>> After the big discussions on SIPPING last summer about how to fix 
>>>> early media, and the reaction at the last couple of IETFs to Brian 
>>>> Stucker's draft on how to cope with it, I've put together 
>> a draft to 
>>>> try to consolidate all of the discussions and move toward 
>> a solution.
>>>> draft-barnes-sip-em-ps-req-sol includes three important parts:
>>>>
>>>> 1. Problem statement: Although there's lots of definitions 
>> of "early 
>>>> media" in existing documents, I think there's broad agreement that 
>>>> the stuff that causes problems is media that arrives before an SDP 
>>>> offer/answer exchange has completed.
>>>> The problem is therefore to enable the SDP exchange(s) in 
>> the INVITE 
>>>> transaction to complete more quickly, and make it more 
>> difficult (if 
>>>> not impossible) to send media before it has completed.  We are 
>>>> explicitly NOT trying to prevent the sending of media 
>> before a 200/OK 
>>>> or other closing of an INVITE transaction.
>>>>
>>>> 2. Requirements: Here, we try to create a consolidated list of 
>>>> requirements for an early media fix.
>>>>
>>>> 3. Analysis of Solutions: Lots of folks have put forward 
>> proposals to 
>>>> fix the early media problem.  This section tries to 
>> enumerate most of 
>>>> them and evaluate them against the requirements in section 2.
>>>>
>>>> It hasn't shown up yet on the I-D tracker, so I've attached a copy.
>>>>
>>>> --Richard
>>>>
>>>
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP 
>> Use sip-implementors@cs.columbia.edu for questions on current 
>> sip Use sip@ietf.org for new developments of core SIP
>>
> 
> 


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 15:32:26 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HH4JO-0005dx-JZ; Tue, 13 Feb 2007 15:31:38 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HH4JM-0005dr-Si
	for sipping@ietf.org; Tue, 13 Feb 2007 15:31:36 -0500
Received: from sj-iport-4.cisco.com ([171.68.10.86])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HH4JI-0005HA-Up
	for sipping@ietf.org; Tue, 13 Feb 2007 15:31:36 -0500
Received: from sj-dkim-6.cisco.com ([171.68.10.81])
	by sj-iport-4.cisco.com with ESMTP; 13 Feb 2007 12:31:31 -0800
X-IronPort-AV: i="4.14,164,1170662400"; 
	d="scan'208"; a="39267552:sNHT752062653"
Received: from sj-core-3.cisco.com (sj-core-3.cisco.com [171.68.223.137])
	by sj-dkim-6.cisco.com (8.12.11/8.12.11) with ESMTP id l1DKVV2r029822; 
	Tue, 13 Feb 2007 12:31:31 -0800
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by sj-core-3.cisco.com (8.12.10/8.12.6) with ESMTP id l1DKUqiQ009111;
	Tue, 13 Feb 2007 12:31:26 -0800 (PST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 15:31:09 -0500
Received: from [161.44.182.226] ([161.44.182.226]) by
	xfe-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 15:31:08 -0500
Message-ID: <45D2200D.8020302@cisco.com>
Date: Tue, 13 Feb 2007 15:31:09 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: ravishankar.shiroor@wipro.com
Subject: Re: [Sipping] Service Identification in SIP
References: <45C0AEF2.8010101@ericsson.com>	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
	<1170835642.29763.54.camel@n95.nomadiclab.com>
	<45CA08B2.4050603@cisco.com>
	<532B18E13CF9E64380EF5FDDE265E071B6192E@blr-m2-msg.wipro.com>
	<45CC7697.1060204@cisco.com>
	<532B18E13CF9E64380EF5FDDE265E071B61B9E@blr-m2-msg.wipro.com>
	<45D14293.30109@cisco.com>
	<532B18E13CF9E64380EF5FDDE265E071B9E933@blr-m2-msg.wipro.com>
In-Reply-To: <532B18E13CF9E64380EF5FDDE265E071B9E933@blr-m2-msg.wipro.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 13 Feb 2007 20:31:08.0625 (UTC)
	FILETIME=[E494EC10:01C74FAD]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=13796; t=1171398691;
	x=1172262691; c=relaxed/simple; s=sjdkim6002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20; bh=F5gsmmKeyrA1v+zkRN70UbLI9r1r9CgOTHzX3oVoKic=;
	b=aT0Y7CbURZXqIDRsjL4/F0eHNztWHAiqzHODxZlZ7CnS7msQsiYilBDcoRFcxpNnqVs4zaPi
	YGvj6+94iafI0zQb72vT9PeWK7A57/AIe66uSWTLGTnzSiLWElvazh4i;
Authentication-Results: sj-dkim-6; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim6002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 32a65c0bf5eb4ec26489239c7cdd0636
Cc: sipping@ietf.org, Gonzalo.Camarillo@ericsson.com, mary.barnes@nortel.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



ravishankar.shiroor@wipro.com wrote:
> 
> Paul Kyzivat wrote:
>>>>> IMO, service identification is different from identifying 
>> the tool 
>>>>> that should be used for executing that service on the end
>>>> device. In
>>>>> the analogy that you have given, e-mail is the service. 
>>>>> Eudora/gmail/... Are the tools that are used to provide 
>> the service.
>>>> The comment below was about identifying the application in the end 
>>>> system, not the protocol.
>>>>
>>>> In the case of email, the "service" is determined by the 
>> protocol and 
>>>> the address/port of the destination. Nobody has suggested adding 
>>>> headers to mail messages to identify the 
>> service/application in the 
>>>> end system that should process the mail message. But here we see 
>>>> proposals to add a header to a sip message to identify the 
>>>> service/application in the UAS.
>>> I would not mind if the e-mail that is delivering meeting 
>> notification 
>>> automatically gets delivered to my calendar application. If I have 
>>> three different tools handling it, I would set one of them to 
>>> preferred calendar application. Similarly if some one sends 
>> me a mail 
>>> with contacts, I would like it to be picked up by my 
>> Contacts application.
>>> I see gmail/Eudora as just tools to execute certain 
>> applications. But 
>>> the e-mail being delivered should be able to say if it is a 
>> calendar 
>>> notification/plain mail/.... Application.
>> Well, in this particular case,I would expect the meeting 
>> notification to be carried as one mime type in the body, and 
>> the contacts to be carried as another mime type. And its 
>> expected that these will be dispatched to a suitable 
>> application by the receiving email application *after* it is received.
>>
>> Note that the mime type defines the information carried, not 
>> the application that will process it.
>>
>> For SIP, the same thing is possible with body types. And 
>> further, the different types of media streams can be handed 
>> off to different applications.
>>
>> 	Paul
> 
> Mime type does not just say it is ascii text; it says it contains
> meeting notification - and that's how the application/service is
> identified. 

Yep, I know. At least it does if the meeting notification has a well 
defined machine understandable syntax.

> In this example, the payload is being identified as being used for the
> calendar service.

Well, for you it might be a calendar service. For me it might be 
something else that processes meeting notifications and perhaps various 
other things as well.

> In SIP, if the payload was being delivered as an attachment to SIP
> message, then the mime type would be sufficient. If SIP is only setting
> up the media path, how is the service to be identified? I think the
> service identifier would be useful in that case.

Well, the obvious answer, though perhaps not sufficient, is that the 
media description is enough for the recipient to decide. Jonathan has 
used Chess as an example of where the media description is insufficient, 
but it is a very contrived example.

PoC has been offered as another example. But in that case I think it 
would not be hard to come up with an appropriate use of the media 
description to distinguish this case, though it may well require some 
new extension. PoC is clearly distinguished from regular voice by having 
a special floor control. It would be entirely appropriate use of SDP to 
indicate the need for that floor control. So I'm not yet convinced PoC 
is a great example either, though it is better than Chess.

An advantage to using the media description is that potentially each 
media stream can be treated separately, or they can be coupled if that 
is required. Note that in email I can have several attachments of 
differing mime types, and application selection is done separately for 
each. It should be possible to do the same for SIP. For instance, if an 
invite that includes audio and MSRP, it should be possible for those to 
be handled by different applications on the device. (Of course it should 
also be possible for a single application to handle both if it is capable.)

There are already drafts/rfcs that provide for labeling media streams 
for the role they are intended to play for the session. These could 
easily be used by the recipient to make special assignments of media to 
applications.

There may indeed still be cases when something else is needed. But, like 
Henry, I'm looking for some plausible use cases to consider.

	Paul

>>> Regards,
>>> Ravi.
>>>
>>>> 	Paul
>>>>
>>>>> Regards,
>>>>> Ravi.
>>>>>
>>>>>
>>>>> --
>>>>>  
>>>>> Ravishankar. G. Shiroor
>>>>> Wipro Technologies, Bangalore.
>>>>>  
>>>>> ravishankar.shiroor@wipro.com
>>>>> --
>>>>>  
>>>>>
>>>>>> -----Original Message-----
>>>>>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
>>>>>> Sent: Wednesday, February 07, 2007 10:43 PM
>>>>>> To: Salvatore Loreto (JO/LMF)
>>>>>> Cc: RaviShankar Shiroor (WT01 - Product Engineering Solutions); 
>>>>>> sipping@ietf.org; Gonzalo.Camarillo@ericsson.com; 
>>>>>> mary.barnes@nortel.com
>>>>>> Subject: Re: [Sipping] Service Identification in SIP
>>>>>>
>>>>>> I still don't get the intent of identifying the application. 
>>>>>> It seems to me this is evil - akin to sending an email 
>> message and 
>>>>>> identifying that the recipient must use Outlook to 
>> receive it, not 
>>>>>> Thunderbird, Eudora, gmail, or whatever.
>>>>>>
>>>>>> I also wish somebody could give better examples of a 
>> communication 
>>>>>> service. Chess is too artificial to be believable, and 
>> IMO PoC is 
>>>>>> only a service because of the way the implementation has 
>> been done.
>>>>>> Conceptually PoC could be nothing but a special floor control 
>>>>>> protocol accompanying a voice session. In that case none of the 
>>>>>> Service mechanism would be needed.
>>>>>>
>>>>>> 	Paul
>>>>>>
>>>>>> Salvatore Loreto (JO/LMF) wrote:
>>>>>>> hi Ravi,
>>>>>>>
>>>>>>> in IMS a Service is a "type of communication defined by 
>> a service 
>>>>>>> definition that specifies the rules and procedures and
>>>>>> allowed media,
>>>>>>> if any, for a specific type of communication and that
>>>>>> utilises the IMS
>>>>>>> enabler. An IMS communication service defines what is
>>>>>> possible within
>>>>>>> a single SIP session or standalone transaction."
>>>>>>> So a Service implies some combination of capabilities and
>>>>>> ways to use
>>>>>>> those capabilities.
>>>>>>> If you want the Service specifies the context for a 
>> communication.
>>>>>>> Instead the Application define the specific communication
>>>>>> (or if you
>>>>>>> want application) that you want run in the specific
>>>> context. So you
>>>>>>> should have several different applications designed to run on a 
>>>>>>> specific environment or that have different performance or
>>>>>> behavior in
>>>>>>> different context (running on top of different service).
>>>>>>>
>>>>>>> Lets go to Jonathan's example of the Chess.
>>>>>>> In his example the Chess is a service, for IMS instead is an 
>>>>>>> application that you can run on top of different service.
>>>>>>> For example you can run Chess application on top of PoC
>>>>>> service or on
>>>>>>> top of the "standard" service. So let say you run Chess on
>>>>>> top of PoC
>>>>>>> service, then you can use only voice in the way that PoC
>>>> specifies.
>>>>>>> /sal
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On Thu, 2007-02-01 at 16:39 +0530,
>>>>>> ravishankar.shiroor@wipro.com wrote:
>>>>>>>> Hello,
>>>>>>>>
>>>>>>>> IMO, Service Identification is definitely needed and
>>>>>> should be taken
>>>>>>>> up as a WG as proposed below.
>>>>>>>>
>>>>>>>> Further, I think Service Identification scope should be
>>>> as per the
>>>>>>>> requirement in Jonathan's draft, as I still do not
>>>> understand the
>>>>>>>> need for having a Communication Service Identifier and another 
>>>>>>>> Application Identifier as given in the
>>>>>> 3gpp-ics-requirements document.
>>>>>>>> Regards,
>>>>>>>> Ravi.
>>>>>>>>
>>>>>>>> --
>>>>>>>>  
>>>>>>>> Ravishankar. G. Shiroor
>>>>>>>> Wipro Technologies, Bangalore.
>>>>>>>>  
>>>>>>>> ravishankar.shiroor@wipro.com
>>>>>>>> --
>>>>>>>>  
>>>>>>>>
>>>>>>>>> -----Original Message-----
>>>>>>>>> From: Gonzalo Camarillo 
>> [mailto:Gonzalo.Camarillo@ericsson.com]
>>>>>>>>> Sent: Wednesday, January 31, 2007 8:30 PM
>>>>>>>>> To: sipping
>>>>>>>>> Cc: Mary Barnes
>>>>>>>>> Subject: [Sipping] Service Identification in SIP
>>>>>>>>>
>>>>>>>>> Folks,
>>>>>>>>>
>>>>>>>>> during the last year, there have been a number of
>>>> threads on the
>>>>>>>>> mailing list discussing service identification in SIP. In 
>>>>>>>>> particular, the following thread discussed 3GPP
>>>>>> requirements on this
>>>>>>>>> area:
>>>>>>>>>
>>>>>>>>>
>> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
>>>>>>>>> At a later point, those 3GPP requirements were
>>>> documented in the
>>>>>>>>> following (expired by now) Internet Draft, which was
>>>>>> presented in a
>>>>>>>>> face-to-face IETF meeting.
>>>>>>>>>
>>>>>>>>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
>>>>>>>>> cs-requirements-00.txt
>>>>>>>>>
>>>>>>>>> Lately, the following draft has been submitted to the 
>> archives:
>>>>>>>>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
>>>>>>>>> rvice-identification-00.txt
>>>>>>>>>
>>>>>>>>> This draft proposes a solution that intends to meet known 
>>>>>>>>> requirements in this area.
>>>>>>>>>
>>>>>>>>> Given the amount of discussions on issues related to service 
>>>>>>>>> identification we have seen during the last year, we
>>>>>> believe this is
>>>>>>>>> a fairly important topic for the SIPPING WG to 
>> tackle. At some 
>>>>>>>>> point, we would like to have a milestone with its
>>>>>> associated WG item
>>>>>>>>> dealing with this issue.
>>>>>>>>>
>>>>>>>>> Therefore, we would like to encourage people to read this
>>>>>> draft and
>>>>>>>>> send comments to the list. Additionally, if you believe
>>>> that this
>>>>>>>>> work is (or is not) important, we would like to hear
>>>> that as well.
>>>>>>>>> Cheers,
>>>>>>>>>
>>>>>>>>> Gonzalo
>>>>>>>>> SIPPING co-chair
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> _______________________________________________
>>>>>>>>> Sipping mailing list
>>>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>>>>>> This list is for NEW development of the application 
>> of SIP Use 
>>>>>>>>> sip-implementors@cs.columbia.edu for questions on
>>>> current sip Use
>>>>>>>>> sip@ietf.org for new developments of core SIP
>>>>>>>>>
>>>>>>>> _______________________________________________
>>>>>>>> Sipping mailing list
>>>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>>>>> This list is for NEW development of the application of SIP Use 
>>>>>>>> sip-implementors@cs.columbia.edu for questions on
>>>> current sip Use
>>>>>>>> sip@ietf.org for new developments of core SIP
>>>>>>> _______________________________________________
>>>>>>> Sipping mailing list
>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>>>> This list is for NEW development of the application of SIP Use 
>>>>>>> sip-implementors@cs.columbia.edu for questions on 
>> current sip Use 
>>>>>>> sip@ietf.org for new developments of core SIP
>>>>>>>
>>>>> The information contained in this electronic message and
>>>> any attachments to this message are intended for the 
>> exclusive use of 
>>>> the addressee(s) and may contain proprietary, confidential or 
>>>> privileged information. If you are not the intended recipient, you 
>>>> should not disseminate, distribute or copy this e-mail. 
>> Please notify 
>>>> the sender immediately and destroy all copies of this 
>> message and any 
>>>> attachments.
>>>>> WARNING: Computer viruses can be transmitted via email. The
>>>> recipient should check this email and any attachments for the 
>>>> presence of viruses. The company accepts no liability for 
>> any damage 
>>>> caused by any virus transmitted by this email.
>>>>>  
>>>>> www.wipro.com
>>>>>
>>>
>>> The information contained in this electronic message and 
>> any attachments to this message are intended for the 
>> exclusive use of the addressee(s) and may contain 
>> proprietary, confidential or privileged information. If you 
>> are not the intended recipient, you should not disseminate, 
>> distribute or copy this e-mail. Please notify the sender 
>> immediately and destroy all copies of this message and any 
>> attachments. 
>>> WARNING: Computer viruses can be transmitted via email. The 
>> recipient should check this email and any attachments for the 
>> presence of viruses. The company accepts no liability for any 
>> damage caused by any virus transmitted by this email.
>>>  
>>> www.wipro.com
>>>
> 
> 
> The information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments. 
> 
> WARNING: Computer viruses can be transmitted via email. The recipient should check this email and any attachments for the presence of viruses. The company accepts no liability for any damage caused by any virus transmitted by this email.
>  
> www.wipro.com
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 16:30:17 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HH5DL-0006Qu-J4; Tue, 13 Feb 2007 16:29:27 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HH5DK-0006QX-7S
	for sipping@ietf.org; Tue, 13 Feb 2007 16:29:26 -0500
Received: from sj-iport-5.cisco.com ([171.68.10.87])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HH5DH-0007CN-SY
	for sipping@ietf.org; Tue, 13 Feb 2007 16:29:26 -0500
Received: from sj-dkim-7.cisco.com ([171.68.10.88])
	by sj-iport-5.cisco.com with ESMTP; 13 Feb 2007 13:29:21 -0800
X-IronPort-AV: i="4.14,164,1170662400"; 
	d="scan'208"; a="388765677:sNHT108142698"
Received: from sj-core-3.cisco.com (sj-core-3.cisco.com [171.68.223.137])
	by sj-dkim-7.cisco.com (8.12.11/8.12.11) with ESMTP id l1DLTLeq024798
	for <sipping@ietf.org>; Tue, 13 Feb 2007 13:29:21 -0800
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by sj-core-3.cisco.com (8.12.10/8.12.6) with ESMTP id l1DLSeiG001610
	for <sipping@ietf.org>; Tue, 13 Feb 2007 13:29:21 -0800 (PST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 16:28:57 -0500
Received: from [161.44.182.226] ([161.44.182.226]) by
	xfe-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 16:28:57 -0500
Message-ID: <45D22D97.9090203@cisco.com>
Date: Tue, 13 Feb 2007 16:28:55 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: sipping <sipping@ietf.org>
References: <E1HH4bD-0002yh-Aq@stiedprstage1.ietf.org>
In-Reply-To: <E1HH4bD-0002yh-Aq@stiedprstage1.ietf.org>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 13 Feb 2007 21:28:57.0039 (UTC)
	FILETIME=[F7EAE5F0:01C74FB5]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=1164; t=1171402161;
	x=1172266161; c=relaxed/simple; s=sjdkim7002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20draft-worley-sipping-forking-02.txt
	|Sender:=20; bh=refBjppcszaPE5LwHVHVa5OMYKnKVdUUbRltBWAG9o8=;
	b=lA/3KbRRF38po2R77OZab0xU2FuslaPzpWnHBETGCO6C1gK0WgrZNlz4B4VJ6tdUJTO3Zoza
	u0QGyGiCV1R6o0kaMYeLgkCIfWCMIdUuXI5r4hwuE+LH6ejDfm2Ckw6e;
Authentication-Results: sj-dkim-7; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim7002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 9182cfff02fae4f1b6e9349e01d62f32
Subject: [Sipping] Re: draft-worley-sipping-forking-02.txt
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Dale,

Interesting proposal. I see a few issues that may or may not be resolvable:

A big one is the possibility of multiple failure responses. After 
sending such a request, when does one give up? This is already an issue 
for INVITE, but only after receiving a 2xx. And in that case you are 
probably keeping some state around for the session anyway. This would 
require some significant reworking of the transaction state machines.

Another issue is that this requires (I think) the UAC to alter its 
behavior when no-cancel is in effect. But 3841 permits a proxy to add 
the Request-Disposition header. If it does, the UAC won't know about it.

I think there may be some other things, but these are the big ones that 
leap out at me.

Note that the origins of callerprefs are ancient - before I became 
involved. And in all the time I was involved I don't recall the 
Request-Disposition header ever being discussed at all. It is clearly 
underspecified and perhaps just plain broken. It might be worth taking a 
survey to see if anyone has implemented it. Making some incompatible 
changes to fix it might be quite acceptable.

	Paul

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 16:34:26 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HH5I5-00086E-RT; Tue, 13 Feb 2007 16:34:21 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HH5I4-00085s-ON
	for sipping@ietf.org; Tue, 13 Feb 2007 16:34:20 -0500
Received: from ihemail1.lucent.com ([135.245.0.33])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HH5Hz-00086g-55
	for sipping@ietf.org; Tue, 13 Feb 2007 16:34:20 -0500
Received: from ilexp02.ndc.lucent.com (h135-3-39-2.lucent.com [135.3.39.2])
	by ihemail1.lucent.com (8.13.8/IER-o) with ESMTP id l1DLY86x022947
	for <sipping@ietf.org>; Tue, 13 Feb 2007 15:34:08 -0600 (CST)
Received: from ILEXC2U03.ndc.lucent.com ([135.3.39.10]) by
	ilexp02.ndc.lucent.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 13 Feb 2007 15:34:08 -0600
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Tue, 13 Feb 2007 15:34:08 -0600
Message-ID: <58F0634C629F3D4DB1D369F600D8A8531E9FBE@ILEXC2U03.ndc.lucent.com>
In-Reply-To: <45D2200D.8020302@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
thread-index: AcdPrjptBt+ci5A3TbOR1WSaUqy1DwABLsUA
References: <45C0AEF2.8010101@ericsson.com>	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com><1170835642.29763.54.camel@n95.nomadiclab.com><45CA08B2.4050603@cisco.com><532B18E13CF9E64380EF5FDDE265E071B6192E@blr-m2-msg.wipro.com><45CC7697.1060204@cisco.com><532B18E13CF9E64380EF5FDDE265E071B61B9E@blr-m2-msg.wipro.com><45D14293.30109@cisco.com><532B18E13CF9E64380EF5FDDE265E071B9E933@blr-m2-msg.wipro.com>
	<45D2200D.8020302@cisco.com>
From: "Esteban, Jairo O \(Jairo\)" <esteban@alcatel-lucent.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 13 Feb 2007 21:34:08.0221 (UTC)
	FILETIME=[B16580D0:01C74FB6]
X-Scanned-By: MIMEDefang 2.57 on 135.245.2.33
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 2409bba43e9c8d580670fda8b695204a
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

All,

After reading all the posts in this thread I am wondering if the
difficulty to find use cases has to do with the definition of the
problem and/or proposed solution. For example, one of the motivations
for this draft is the fact that media description is not always enough
to distinguish what the UE wants to do. But there are alternate
solutions besides filling this void using SIP. One possibility would be
to add this power to the SDP body, instead. At least in some of the
examples I have found here, that approach might work well. Do we need a
more powerful (and complex) SIP protocol or a more powerful SDP
body?...or both?

Just from the top of my head: If some involvement of the elements in the
path is required (e.g. dynamic routing), service/capability/feature data
shall be carried by the SIP message. If only the endpoints are involved,
it is up to the SDP body to convey this information and enable the
negotiation between all the participants of the session.=20

Just want to explore other possibilities...Would this be crazy?

An initial comment that comes to my mind is that there would be
different approaches to deal with services provided by Application
Servers and services provided by the UE. But, would it be really bad?
Aren't they both fundamentally different?

Jairo

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 17:09:47 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HH5pd-0004iQ-Ax; Tue, 13 Feb 2007 17:09:01 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HH5pb-0004iF-SU; Tue, 13 Feb 2007 17:08:59 -0500
Received: from zcars04f.nortel.com ([47.129.242.57])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HH5pa-000465-Dv; Tue, 13 Feb 2007 17:08:59 -0500
Received: from zrc2hxm0.corp.nortel.com (zrc2hxm0.corp.nortel.com
	[47.103.123.71])
	by zcars04f.nortel.com (Switch-2.2.6/Switch-2.2.0) with ESMTP id
	l1DM8rT09032; Tue, 13 Feb 2007 17:08:53 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Re: [Sip] New draft on early media
Date: Tue, 13 Feb 2007 16:08:50 -0600
Message-ID: <1ECE0EB50388174790F9694F77522CCF0F281DFC@zrc2hxm0.corp.nortel.com>
In-Reply-To: <45D2143D.9020907@bbn.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Re: [Sip] New draft on early media
Thread-Index: AcdPtq7FxGJRhOl2RYKBqE+agQuZ4gABNUkA
From: "Francois Audet" <audet@nortel.com>
To: "Richard Barnes" <rbarnes@bbn.com>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 156eddb66af16eef49a76ae923b15b92
Cc: sip@ietf.org, sipping@ietf.org,
	"Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I mean play incoming media.=20

> -----Original Message-----
> From: Richard Barnes [mailto:rbarnes@bbn.com]=20
> Sent: Tuesday, February 13, 2007 11:41
> To: Audet, Francois (SC100:3055)
> Cc: Christer Holmberg (JO/LMF); sip@ietf.org; sipping@ietf.org
> Subject: Re: [Sipping] Re: [Sip] New draft on early media
>=20
> Francois,
>=20
> I think I have some responses, but I'd like to ask you first:=20
> What do you mean by "media cut-through"?
>=20
> Thanks,
> --Richard
>=20
> Francois Audet wrote:
> > Hi Richard,
> >=20
> > Forcing media cut-through to occur only after an Answer has been=20
> > received does solve SOME of the early media problems. As=20
> others have=20
> > mentioned, there are environments that operate like this already. I=20
> > also know many products that do not cut-through media before the=20
> > answer is received.
> >=20
> > But there are other problems related with early media.=20
> Brian Stucker's=20
> > draft listed a bunch of them.
> >=20
> > Forking for example comes to mind: even if we did mandate=20
> cut-through=20
> > after answer, we would still have problems related with=20
> forking (i.e.,=20
> > how to pick the stream that needs to be played, and how to=20
> correlate=20
> > the stream with the signaling). How to deal with SRTP key=20
> exchange in=20
> > that scenario is an even bigger problem.
> >=20
> > Clipping is a possible side-effect of your proposal. We'd need some=20
> > mechanism to mitigate it, in particular for auto-answer scenarios.
> >=20
> >=20
> >> -----Original Message-----
> >> From: Richard Barnes [mailto:rbarnes@bbn.com]
> >> Sent: Monday, February 12, 2007 07:08
> >> To: Christer Holmberg (JO/LMF)
> >> Cc: sip@ietf.org; sipping@ietf.org
> >> Subject: [Sipping] Re: [Sip] New draft on early media
> >>
> >> Christer,
> >>
> >> Thanks for your comments.  Responses inline.
> >>
> >> Cheers,
> >> --Richard
> >>
> >>> I haven't had time to read the document yet, but I am not
> >> sure I agree
> >>> with the statement saying that the biggest problem is media
> >> arriving
> >>> before the SDP answer. In some environments such media will
> >> not even
> >>> reach the terminal, due to gating etc.
> >> Perhaps a better term would be "media sent before an SDP=20
> answer has=20
> >> been received".  As you point out, sending media before an SDP=20
> >> offer/answer exchange has completed poses the risk that that media=20
> >> will not arrive.
> >> To put it in the terms used in the document, media clipping is a=20
> >> problem caused by early (pre-SDP-answer) media, and thus one which=20
> >> this document seeks to solve.
> >>
> >>> Many implementations DO send an SDP answer (or, at least an
> >> unreliable
> >>> "preview" of the SDP answer) before (or at the same time=20
> as) early=20
> >>> media is sent.
> >> This is the behavior we would like to standardize -- that an SDP=20
> >> answer is *always* sent before media of any type.  What=20
> we're trying=20
> >> to do with this document is describe the requirements for how that=20
> >> answer is sent, and consolidate the many ways people have=20
> proposed to=20
> >> send it.
> >>
> >>> And, in the cases where implementations do NOT send an SDP answer=20
> >>> before they start sending early media - what behavior=20
> would you be=20
> >>> able to standardize?
> >> We would prefer that standards not allow this behavior --=20
> that seems=20
> >> like the right thing to me.  What's a little trickier is=20
> that you'd=20
> >> like to not have the answerer sending media before the answer has=20
> >> *arrived*.
> >>   When the answer is carried in a reliable message, you could just=20
> >> forbid sending media before receipt has been confirmed.
> >> When the transport is unreliable, the situation is a little less=20
> >> clear, but I think there's still possibilities for=20
> standardization,=20
> >> e.g., by forbidding answerers to send media until they've received=20
> >> media.  This is discussed in more length in the document.
> >>
> >>> Regards,
> >>>
> >>> Christer
> >>>
> >>>
> >>>> -----Original Message-----
> >>>> From: Richard Barnes [mailto:rbarnes@bbn.com]
> >>>> Sent: 8. helmikuuta 2007 21:47
> >>>> To: sip@ietf.org; sipping@ietf.org
> >>>> Subject: [Sip] New draft on early media
> >>>>
> >>>> After the big discussions on SIPPING last summer about=20
> how to fix=20
> >>>> early media, and the reaction at the last couple of=20
> IETFs to Brian=20
> >>>> Stucker's draft on how to cope with it, I've put together
> >> a draft to
> >>>> try to consolidate all of the discussions and move toward
> >> a solution.
> >>>> draft-barnes-sip-em-ps-req-sol includes three important parts:
> >>>>
> >>>> 1. Problem statement: Although there's lots of definitions
> >> of "early
> >>>> media" in existing documents, I think there's broad=20
> agreement that=20
> >>>> the stuff that causes problems is media that arrives=20
> before an SDP=20
> >>>> offer/answer exchange has completed.
> >>>> The problem is therefore to enable the SDP exchange(s) in
> >> the INVITE
> >>>> transaction to complete more quickly, and make it more
> >> difficult (if
> >>>> not impossible) to send media before it has completed.  We are=20
> >>>> explicitly NOT trying to prevent the sending of media
> >> before a 200/OK
> >>>> or other closing of an INVITE transaction.
> >>>>
> >>>> 2. Requirements: Here, we try to create a consolidated list of=20
> >>>> requirements for an early media fix.
> >>>>
> >>>> 3. Analysis of Solutions: Lots of folks have put forward
> >> proposals to
> >>>> fix the early media problem.  This section tries to
> >> enumerate most of
> >>>> them and evaluate them against the requirements in section 2.
> >>>>
> >>>> It hasn't shown up yet on the I-D tracker, so I've=20
> attached a copy.
> >>>>
> >>>> --Richard
> >>>>
> >>>
> >>
> >> _______________________________________________
> >> Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> >> This list is for NEW development of the application of SIP Use=20
> >> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> >> sip@ietf.org for new developments of core SIP
> >>
> >=20
> >=20
>=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 13 18:17:36 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HH6sy-0003TK-BR; Tue, 13 Feb 2007 18:16:32 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HH6sw-0003T7-N9
	for sipping@ietf.org; Tue, 13 Feb 2007 18:16:30 -0500
Received: from qfe1.f10207-20.atlanta2.level3.net ([67.72.93.25]
	helo=f10207-20.adc1.level3.com)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HH6sw-00071R-D5
	for sipping@ietf.org; Tue, 13 Feb 2007 18:16:30 -0500
Received: from machine77.Level3.com (qfe0.f4ff49-08.idc1.oss.level3.com
	[10.1.156.103])
	by f10207-20.adc1.level3.com (8.12.10/8.12.10) with ESMTP id
	l1DNGLea012505; Tue, 13 Feb 2007 23:16:22 GMT
Received: from machine77.Level3.com (localhost [127.0.0.1])
	by localhost.level3.com (Postfix) with ESMTP
	id B4A1E78B5BB; Tue, 13 Feb 2007 23:11:53 +0000 (GMT)
Received: from IDC1EXC0001B.corp.global.level3.com
	(idc1exc0001b.corp.global.level3.com [10.1.9.25])
	by scanner3.l3.com (Postfix) with SMTP
	id E7A5E78B7BB; Tue, 13 Feb 2007 23:02:04 +0000 (GMT)
Received: from idc1exc0006.corp.global.level3.com ([10.1.9.17]) by
	IDC1EXC0001B.corp.global.level3.com with Microsoft
	SMTPSVC(6.0.3790.1830); Tue, 13 Feb 2007 16:02:04 -0700
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: [Sipping] RE: draft-malas-performance-metrics-06
Date: Tue, 13 Feb 2007 16:02:03 -0700
Message-ID: <9A5D0F8924D0AA4BBEA4B7E0B0D450E2013AA56F@idc1exc0006.corp.global.level3.com>
In-Reply-To: <97F2AD7AA4EAD948827639B06CF2745F73C1F5@esealmw106.eemea.ericsson.se>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] RE: draft-malas-performance-metrics-06
thread-index: AcdJwqq2KVrk+P49QbivLfsE/PiKpAF/uY4g
From: "Malas, Daryl" <Daryl.Malas@Level3.com>
To: "Marian Delkinov (LN/EAB)" <marian.delkinov@ericsson.com>
X-OriginalArrivalTime: 13 Feb 2007 23:02:04.0747 (UTC)
	FILETIME=[FA7379B0:01C74FC2]
X-Spam-Score: 1.2 (+)
X-Scan-Signature: 03169bfe4792634a390035a01a6c6d2f
Cc: sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Marian...Comments inline...Also, I copied the Sipping mailing list,
because it is helpful to allow others within the Sipping community to
provide insight, thoughts into your comments as well.  I hope you do not
mind.

Thanks...--Daryl

> -----Original Message-----
> From: Marian Delkinov (LN/EAB) [mailto:marian.delkinov@ericsson.com]
> Sent: Tuesday, February 06, 2007 12:45 AM
> To: Malas, Daryl
> Subject: draft-malas-performance-metrics-06
>=20
> Hi,
>=20
> I'm working on implementing some IMS performance metrics into our SLA
> solutions. After performing a lot of searching through the different
> standardization bodies, the most useful document I found was your
draft.
>=20
>=20
> Would like to comment on some of the metrics and perhaps on the
generic
> approach in the document.
>=20
> Generic approach:
>=20
> It would be good to give some background information about what are
the
> major reasons for SIP calls being lost. A picture similar to the one
> used in ITU E411 recommendation for the interpretation of the network
> performance data would really help. Please, see example in attachment.
>=20
[DM] I could do this if the general mailing list agrees with the
request.  I would like to get their opinion as well.  In either case, I
will consider it for future revisions of the document.  I agree that
diagrams, in general, usually help provide clarity around concepts.
=20
> Metrics:
>=20
> Since dealing with SLAs, I would like to put the things more into a
> Customer perspective. Therefore, I'm more focused on how the service
> looks for the customer. First of all what is the service accessibility
> and secondly what is the service performance. Perhaps you can help me
> find out the most suitable metrics that you discuss in your document.
>=20
> 1) I try to define the SIP based telephony service effectiveness:
>=20
> The probability that an end user can access the IMS (SIP based)
> Telephony Service.
>=20
> In the PSTN/ISDN terms this is defined by ITU in Recommendation 411 as
> Network Effectiveness Rate:
>=20
>             Seizures Delivered to the far end terminal
> NER% =3D ---------------------------------------------------- x 100
> 		       Total Number of Seizures
>=20
> Note that in this case the end user behaviour is excluded, so seizures
> that reach answer signal, or user busy, or ring no answer are counted
as
> successful.
>=20
> When I look in your draft document, I don't clearly see similar
metric.
[DM]  If you look in the "Additional Metrics for Consideration" section
of the draft, I think you will find a similar metric to what you are
describing. Again, if you would like to see the metric included in the
document, please state that on the mailing list.  Thanks...

> However if used some of your definitions, it could perhaps be defined
as
> Session Effectiveness Rate (or any other suitable definition, e.g.
> Telephony Service Availability) as follows:
>=20
>          Total # of Session Requests - # of Failed Session Completions
> SEFR% =3D =
--------------------------------------------------------------
x
> 100
> 			Total # of Session Requests
>=20
> Or, if in the above the "# of Failed Session Completion" includes the
> end user behaviour, then perhaps a better definition would be:
>=20
>          Total # of Session Requests - # of ISA
> SEFR% =3D =
--------------------------------------------------------------
x
> 100
> 			Total # of Session Requests
>=20
> So, I would be really grateful if you share your opinion about that.
>=20
> 2) I also try to define the SIP based telephony call completion ratio.
> According to your document the Session Completion Rate (SCR) is the
> parameter that matches that. However I think of a more useful
> definition, that would really present the ability of the network to
> maintain successfully initiated session in the time. So, I would like
to
> hear what you think about the following definition proposal:
>=20
> Session Completion Rate is the probability that a successfully
> established session is maintained for a predetermined time until it is
> released intentionally by any of the UAs.
>=20
> According to your document it is defined as:
>          # of Successfully Completed Sessions
> SCR% =3D ---------------------------------------
>           Total # of Session Requests
>=20
> But if we follow my proposal it would be:
>=20
>          # of Successfully Completed Sessions
> SCR% =3D ------------------------------------------ x 100
>          # of Successfully Established Sessions
>=20
>=20
>=20
[DM] In concept I understand what you are trying to get at with this
change; however, this will require defining what an "Established
Session" means.  This will become very tricky.  Instead, I think you can
use the SCR as it is defined and then provide greater granularity by
calculating it for calls above and below a time certain duration.  What
do you think?

This is actually something we can perhaps easily measure, however it
> would be good if we can show this parameter in dependence with the
> Session Duration and that I haven't taken into account. In many cases,
> when jitter and packet loss increase, the sessions are terminated
> unintentionally. And the probability this to happen raises in time.
For
> example, for short sessions the SCR may look good, but for long
sessions
> it deteriorates.
>=20
> So, I wonder how to link this with the session duration time in order
to
> have a single parameter that really represents the network
performance.
> Will be happy to exchange ideas.
>=20
> Best regards!
> Mario.
>=20
> Marian Delkinov
> Senior Consultant
>=20
> Network Management and Service Assurance
>=20
> ERICSSON AB
> Lindholmspiren 11
> SE-417 56 Gothenburg
> Sweden
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 14 05:20:26 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHHDV-0004QT-KR; Wed, 14 Feb 2007 05:18:25 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHHDU-0004QO-4c
	for sipping@ietf.org; Wed, 14 Feb 2007 05:18:24 -0500
Received: from mailgate.siemenscomms.co.uk ([195.171.110.225]
	helo=bemg01.siemenscomms.co.uk)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHHDQ-0008Rl-K9
	for sipping@ietf.org; Wed, 14 Feb 2007 05:18:24 -0500
Received: from ntht207e.uksgcs.siemenscomms.co.uk ([137.223.247.82])
	by siemenscomms.co.uk (PMDF V6.0-24 #40642)
	with ESMTP id <0JDG00C8979XBS@siemenscomms.co.uk> for sipping@ietf.org;
	Wed, 14 Feb 2007 10:17:57 +0000 (GMT)
Received: by ntht207e.uksgcs.siemenscomms.co.uk with Internet Mail Service
	(5.5.2657.72)	id <D2K3XYFP>; Wed, 14 Feb 2007 10:17:58 +0000
Content-return: allowed
Date: Wed, 14 Feb 2007 10:17:57 +0000
From: "Elwell, John" <john.elwell@siemens.com>
To: bstucker@nortel.com, sipping@ietf.org
Message-id: <50B1CBA96870A34799A506B2313F26670B0A8CB9@ntht201e.siemenscomms.co.uk>
MIME-version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
Content-type: text/plain
Content-transfer-encoding: 7BIT
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 2086112c730e13d5955355df27e3074b
Cc: 
Subject: [Sipping] RE: I-D
	ACTION:draft-stucker-sipping-early-media-indirection-00.t	xt
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Brian,

Although I am interested in this general problem space and welcome
proposals, I had a hard time understanding the benefits this gives compared
to RFC 3960. For example, in section 6.2.2 it states:

"This dialog, known as an early-media dialog, allows the UAC to receive
early media from endpoints other than the called party's device(s) such as a
network media server, does not suffer from the problems of forking (since
each indirect early-media session creates a separate early media dialog),
allows for the UAC to specify a different RTP payload type (to facilitate
discrimination of early-media versus final-media), and gives both the UAC
and UAS the ability to explicitly start and stop the early media stream
without complicating the SDP negotiations for the final media."

It is not clear which of these properties of the new mechanism are not
available with RFC 3960. For example, when an INVITE request is forked, each
branch of the fork establishes a separate early dialog according to RFC
3960. Forking to a media server could either create a separate early dialog
or use a separate SDP offer-answer exchange (early session disposition)
within the context of an existing early dialog. The separate offer-answer
exchange permits things like use of a different RTP payload type or
different port. Explicitly stopping an early media stream can be done by a
further offer-answer exchange using early session disposition.

Section 6.2.3 fails to say how the proxy that instructed the UAC to play an
early media stream can subsequently tell the UAC to stop that stream
(assuming the dialog between the UAC and the media server does not pass
through that proxy).

The Early-Media header field definition does not appear to provide any
information about the nature of the early media, although the examples do
seem to show something of this nature, e.g., "service=ringback;user=userb",
but presumably this is information for the media server and not for
consumption by the UAC. The UAC will therefore not be able to determine
relative priorities of Early-Media header fields received on different early
dialogs from the same or different proxies.

John


> -----Original Message-----
> From: Internet-Drafts@ietf.org [mailto:Internet-Drafts@ietf.org] 
> Sent: 13 February 2007 20:50
> To: i-d-announce@ietf.org
> Subject: I-D 
> ACTION:draft-stucker-sipping-early-media-indirection-00.txt
> 
> A New Internet-Draft is available from the on-line Internet-Drafts 
> directories.
> 
> 
> 	Title		: Early Media INDirection (EMIND) in 
> the Session Initiation Protocol (SIP)
> 	Author(s)	: B. Stucker
> 	Filename	: 
> draft-stucker-sipping-early-media-indirection-00.txt
> 	Pages		: 18
> 	Date		: 2007-2-13
> 	
>    This document describes models that can be used to manage multiple
>    early media streams in the Session Initiation Protocol (SIP).  The
>    model seeks to allow calling party endpoints to be able to better
>    control the rendering of early media to the user, and distinguish
>    early media from final media.
> 
>    Although some of early media challenges are likely to never be
>    overcome, (e.g. when interworking with an ISUP PSTN 
> gateway that does
>    not take into account CPG or ACM messages), the potential 
> to improve
>    on what is already there does exist.  The recommendations in this
>    document are intended to be an update to RFC-3960 
> recommendations and
>    to avoid many of the complications outlined in
>    draft-stucker-sipping-early-media-coping.
> 
> 
> 
> A URL for this Internet-Draft is:
> http://www.ietf.org/internet-drafts/draft-stucker-sipping-earl
> y-media-indirection-00.txt
> 
> To remove yourself from the I-D Announcement list, send a message to 
> i-d-announce-request@ietf.org with the word unsubscribe in 
> the body of 
> the message. 
> You can also visit 
> https://www1.ietf.org/mailman/listinfo/I-D-announce 
> to change your subscription settings.
> 
> Internet-Drafts are also available by anonymous FTP. Login with the 
> username "anonymous" and a password of your e-mail address. After 
> logging in, type "cd internet-drafts" and then 
> "get draft-stucker-sipping-early-media-indirection-00.txt".
> 
> A list of Internet-Drafts directories can be found in
> http://www.ietf.org/shadow.html 
> or ftp://ftp.ietf.org/ietf/1shadow-sites.txt
> 
> Internet-Drafts can also be obtained by e-mail.
> 
> Send a message to:
> 	mailserv@ietf.org.
> In the body type:
> 	"FILE 
> /internet-drafts/draft-stucker-sipping-early-media-indirection
> -00.txt".
> 	
> NOTE:	The mail server at ietf.org can return the document in
> 	MIME-encoded form by using the "mpack" utility.  To use this
> 	feature, insert the command "ENCODING mime" before the "FILE"
> 	command.  To decode the response(s), you will need "munpack" or
> 	a MIME-compliant mail reader.  Different MIME-compliant 
> mail readers
> 	exhibit different behavior, especially when dealing with
> 	"multipart" MIME messages (i.e. documents which have been split
> 	up into multiple messages), so check your local documentation on
> 	how to manipulate these messages.
> 
> Below is the data which will enable a MIME compliant mail reader
> implementation to automatically retrieve the ASCII version of the
> Internet-Draft.
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 14 13:46:03 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHP6r-0003hR-Kd; Wed, 14 Feb 2007 13:44:05 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHP6q-0003aL-6t
	for sipping@ietf.org; Wed, 14 Feb 2007 13:44:04 -0500
Received: from alnrmhc13.comcast.net ([206.18.177.53])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHP6j-0004qd-Ht
	for sipping@ietf.org; Wed, 14 Feb 2007 13:44:04 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (alnrmhc13) with ESMTP
	id <20070214184352b13001k11ce>; Wed, 14 Feb 2007 18:43:52 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1EIhpAM017924
	for <sipping@ietf.org>; Wed, 14 Feb 2007 13:43:51 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1EIhpDq017920;
	Wed, 14 Feb 2007 13:43:51 -0500
Date: Wed, 14 Feb 2007 13:43:51 -0500
Message-Id: <200702141843.l1EIhpDq017920@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <45D22D97.9090203@cisco.com> (pkyzivat@cisco.com)
Subject: Re: [Sipping] Re: draft-worley-sipping-forking-02.txt
References: <E1HH4bD-0002yh-Aq@stiedprstage1.ietf.org>
	<45D22D97.9090203@cisco.com>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 39bd8f8cbb76cae18b7e23f7cf6b2b9f
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   From: Paul Kyzivat <pkyzivat@cisco.com>

   A big one is the possibility of multiple failure responses. After 
   sending such a request, when does one give up?

I assume you mean, "When does a UAC give up waiting for further
responses?"

In the cases I'm thinking of, the UAC is trying to accumulate a set of
responses, and expects that they will dribble in over time.  The
cannonical case is SUBSCRIBEs to obtain some sort of status info.  In
the call pick-up case, the agent trying to perform the pick-up has to
have such a time-out already, the point at which it decides what to
do.  In cases involving ongoing status monitoring, the information
from an additional subscription can be added to the information
already accumulated.

   This is already an issue for INVITE, but only after receiving a
   2xx. And in that case you are probably keeping some state around
   for the session anyway. This would require some significant
   reworking of the transaction state machines.

In the simplest implementation, the UAC can pick the first one, and
ignore the latter ones.  But if the UAC doesn't want to handle
multiple responses, it shouldn't ask for them.

   Another issue is that this requires (I think) the UAC to alter its 
   behavior when no-cancel is in effect. But 3841 permits a proxy to add 
   the Request-Disposition header. If it does, the UAC won't know about it.

I think that in various cases without Request-Disposition, the UAC can
be confronted with multiple responses anyway.  Mostly cases involving
stateless proxies or alternative routing of responses due to failure
of intermediate agents during the transaction.  But isn't ignoring all
latter responses an adequate solution?

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 14 13:59:12 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHPLC-00049K-Vh; Wed, 14 Feb 2007 13:58:54 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHPLB-00049E-GK
	for sipping@ietf.org; Wed, 14 Feb 2007 13:58:53 -0500
Received: from alnrmhc12.comcast.net ([206.18.177.52])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHPLA-0006Sd-8P
	for sipping@ietf.org; Wed, 14 Feb 2007 13:58:53 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (alnrmhc12) with ESMTP
	id <20070214185842b12001noude>; Wed, 14 Feb 2007 18:58:42 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1EIwgAM019015
	for <sipping@ietf.org>; Wed, 14 Feb 2007 13:58:42 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1EIwgVo019011;
	Wed, 14 Feb 2007 13:58:42 -0500
Date: Wed, 14 Feb 2007 13:58:42 -0500
Message-Id: <200702141858.l1EIwgVo019011@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <45D1405E.30300@cisco.com> (pkyzivat@cisco.com)
Subject: Re: [Sipping] Service Identification in SIP
References: <45C0AEF2.8010101@ericsson.com><532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com><1170835642.29763.54.camel@n95.nomadiclab.com><45CA08B2.4050603@cisco.com><532B18E13CF9E64380EF5FDDE265E071B6192E@blr-m2-msg.wipro.com><45CC7697.1060204@cisco.com><1171031930.3464.92.camel@localhost.localdomain>	<200702110449.l1B4nfk2020329@dragon.ariadne.com>	<24CCCC428EFEA2469BF046DB3C7A8D224538F2@namail5.corp.adobe.com>
	<200702122256.l1CMueqK020621@dragon.ariadne.com>
	<45D1405E.30300@cisco.com>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: cab78e1e39c4b328567edb48482b6a69
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   From: Paul Kyzivat <pkyzivat@cisco.com>

   What you describe is largely consistent with the infrastruction of IMS, 
   although there are a million ways that infrastructure could be applied.

What I'm trying to do is get a real application in front of us.
Unfortunately, the IMS people have not been very forthcoming
discussing this.

   > AFAICT, for many service providers, an incoming call (regardless of
   > the destination E.164 number) will get routed to one of a set of
   > "application servers" (AS), each one of which handles a particular
   > "service".  The application server is responsible for managing the
   > final link to the phone.  From an Internet perspective, what is really
   > happening is that the UA is partitioned into two parts, part on the AS
   > and part on the real phone.

   It is interesting to view the partitioning of the function between two 
   servers as you suggest. I think it would be much more straightforward if 
   such an AS was actually treated as a first class UAS that has registered 
   and is selected in the same way that other registered UAs are selected 
   for a call.

Well, in a sense the AS is -- the call reaches it through the AOR.

   Notably, that callerprefs could be used to make the choice.

Yes, the crux is how the call routing is done.  I'm not as familiar
with callerprefs as I should be.  Could you outline how to use
callerprefs to route, say, PoC calls, to the correct AS to support a
UA that can't handle PoC on its own?

   Unfortunately, IMS has chosen to use a different algorithm to decide how 
   to involve ASs.

I'd like to see how callerprefs avoids the "siloing" risks that the
current proposal has.

   Actually in IMSs case, the call can in fact be chained to multiple ASs. 
   But of course they have to be applied in some order.

I assume that callerprefs can be set up so that routing can be done on
some priority ordering of the features.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 14 17:30:43 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHSd0-0006dy-JN; Wed, 14 Feb 2007 17:29:30 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHScz-0006az-49
	for sipping@ietf.org; Wed, 14 Feb 2007 17:29:29 -0500
Received: from mailgate.siemenscomms.co.uk ([195.171.110.225]
	helo=bemg01.siemenscomms.co.uk)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHScu-0005un-LT
	for sipping@ietf.org; Wed, 14 Feb 2007 17:29:29 -0500
Received: from ntht207e.uksgcs.siemenscomms.co.uk ([137.223.247.82])
	by siemenscomms.co.uk (PMDF V6.0-24 #40642)
	with ESMTP id <0JDG00L3J932XT@siemenscomms.co.uk> for sipping@ietf.org;
	Wed, 14 Feb 2007 10:57:02 +0000 (GMT)
Received: by ntht207e.uksgcs.siemenscomms.co.uk with Internet Mail Service
	(5.5.2657.72)	id <D2K3XZC2>; Wed, 14 Feb 2007 10:57:02 +0000
Content-return: allowed
Date: Wed, 14 Feb 2007 10:57:01 +0000
From: "Elwell, John" <john.elwell@siemens.com>
Subject: RE: [Sipping] Service Identification in SIP
To: Henry Sinnreich <hsinnrei@adobe.com>, sipping@ietf.org
Message-id: <50B1CBA96870A34799A506B2313F26670B0A8D4E@ntht201e.siemenscomms.co.uk>
MIME-version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
Content-type: text/plain
Content-transfer-encoding: 7BIT
X-Spam-Score: 0.0 (/)
X-Scan-Signature: b4a0a5f5992e2a4954405484e7717d8c
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Having followed this thread, I too find the chess example very contrived. If
chess moves are sent as voice commands, then surely it would be acceptable
for the normal telephone application to answer the call, i.e., the remote
player listens to the move, performs this move on a conventional or
electronic chess board and speaks his next move. The calling UA should not
care whether the call is answered by a computer application or a person.
Furthermore, even if the audio needs to consumed by a particular chess
application at the called UA, then other media in the same session such as
video or messaging might not be. So any qualification of the nature of media
to be exchanged probably should be at the media level, not at the session
level or SIP level. As yet another variation, what if the players set up a
conventional telephone call and then mutually agree to switch in their chess
applications?

Now, let's consider another case. I establish an MSRP session. Within this
MSRP session I can send messages with various MIME attachments that are
intended for consumption by various applications. I should not need to know
in advance what applications I will use during the course of the MSRP
session. All I should put in the INVITE request is the fact that I want to
establish an MSRP session.

John 

> -----Original Message-----
> From: Henry Sinnreich [mailto:hsinnrei@adobe.com] 
> Sent: 13 February 2007 18:19
> To: sipping@ietf.org
> Subject: RE: [Sipping] Service Identification in SIP
> 
> It must be my lack of imagination for not being able to come up with a
> credible usage scenario, given the following:
> 
> - Is playing chess by voice a credible usage scenario? :-)
> Especially since voice recognition is now so often found in PCs and in
> mobile phones as well. In other words, chess playing by voice may also
> be an endpoint application, without any network involvement. Actually,
> if placed in the endpoint it works just as well on poor 
> bandwidth links
> by contrast to chess in the network that may require adequate 
> broadband.
> 
> - Is anyone's boss or spouse going to approve extra phone 
> charges above
> the flat rate for playing chess at a time when the gaming industry has
> pioneered free VoIP?
> Will such a service be taxed extra and if yes using by what criteria?
> Gambling comes to my mind :-)
> 
> - VoIP is reported in the press as a give-away and I have seen no
> reports of any VoIP provider recording a profit on VoIP. I assume the
> losses are often hidden by bundling in triple play, etc.
> 
> Given my lack of imagination, can anyone share a more credible usage
> scenario for Service Identification in SIP?
> 
> What's the use of a standard without a credible usage scenario?
> 
> Thanks, Henry
> 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 14 17:38:40 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHSll-0007xW-3L; Wed, 14 Feb 2007 17:38:33 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHSli-0007xE-Rw; Wed, 14 Feb 2007 17:38:30 -0500
Received: from mx12.bbn.com ([128.33.0.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HHSlh-0006vV-Bb; Wed, 14 Feb 2007 17:38:30 -0500
Received: from dommiel.bbn.com ([192.1.122.15] helo=[127.0.0.1])
	by mx12.bbn.com with esmtp (Exim 4.60)
	(envelope-from <rbarnes@bbn.com>)
	id 1HHSlZ-0007QE-3O; Wed, 14 Feb 2007 17:38:21 -0500
Message-ID: <45D38F5C.9050901@bbn.com>
Date: Wed, 14 Feb 2007 17:38:20 -0500
From: Richard Barnes <rbarnes@bbn.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Francois Audet <audet@nortel.com>
Subject: Re: [Sipping] Re: [Sip] New draft on early media
References: <1ECE0EB50388174790F9694F77522CCF0F281DFC@zrc2hxm0.corp.nortel.com>
In-Reply-To: <1ECE0EB50388174790F9694F77522CCF0F281DFC@zrc2hxm0.corp.nortel.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: cdb443e3957ca9b4c5b55e78cfcf4b26
Cc: sip@ietf.org, sipping@ietf.org,
	"Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Thanks for the clarification.  I know that choosing media streams to 
render is a big problem, but it's one I'm viewing as beyond the scope of 
what our draft is explicitly trying to address.

That said, I think that ensuring that media are preceded by an SDP 
answer is a big step in enabling better management of media streams, 
since it gives the offeror (media recipient) SDP metadata on which to 
base its decision.  Once an SDP answer is guaranteed to come before 
media, the issue can be treated more at the SDP layer (e.g. by giving 
SDP sufficient expressiveness to support decisions between media 
streams), rather than at the SIP layer.

With respect to forking in particular: You're correct that even if all 
streams from forked endpoints are preceded by SDP answers, the UAC still 
has to figure out how to deal with them.  But I think that once you 
ensure that the offer/answer finishes before media flows, SRTP keying 
gets a lot simpler, since the UAC can just do a DH exchange through your 
favorite SDP mechanism with each endpoint.

Clipping is certainly something that could be an issue, but I think it 
can be prevented.  The solution I think will work in the end is that an 
SDP answer will be sent immediately by the UAS on receipt of the INVITE, 
before any other action is taken (e.g., before the INVITE is passed to 
the TU).  For example, a PSTN gateway receiving a SIP INVITE would 
return its answer before it initiates the outbound PSTN call.  It seems 
to me like the negotion of an SDP association should precede SIP session 
establishment just like TCP negotiation would precede initiation of a 
higher-layer protocol session.

Does that make sense?
--Richard




Francois Audet wrote:
> I mean play incoming media. 
> 
>> -----Original Message-----
>> From: Richard Barnes [mailto:rbarnes@bbn.com] 
>> Sent: Tuesday, February 13, 2007 11:41
>> To: Audet, Francois (SC100:3055)
>> Cc: Christer Holmberg (JO/LMF); sip@ietf.org; sipping@ietf.org
>> Subject: Re: [Sipping] Re: [Sip] New draft on early media
>>
>> Francois,
>>
>> I think I have some responses, but I'd like to ask you first: 
>> What do you mean by "media cut-through"?
>>
>> Thanks,
>> --Richard
>>
>> Francois Audet wrote:
>>> Hi Richard,
>>>
>>> Forcing media cut-through to occur only after an Answer has been 
>>> received does solve SOME of the early media problems. As 
>> others have 
>>> mentioned, there are environments that operate like this already. I 
>>> also know many products that do not cut-through media before the 
>>> answer is received.
>>>
>>> But there are other problems related with early media. 
>> Brian Stucker's 
>>> draft listed a bunch of them.
>>>
>>> Forking for example comes to mind: even if we did mandate 
>> cut-through 
>>> after answer, we would still have problems related with 
>> forking (i.e., 
>>> how to pick the stream that needs to be played, and how to 
>> correlate 
>>> the stream with the signaling). How to deal with SRTP key 
>> exchange in 
>>> that scenario is an even bigger problem.
>>>
>>> Clipping is a possible side-effect of your proposal. We'd need some 
>>> mechanism to mitigate it, in particular for auto-answer scenarios.
>>>
>>>
>>>> -----Original Message-----
>>>> From: Richard Barnes [mailto:rbarnes@bbn.com]
>>>> Sent: Monday, February 12, 2007 07:08
>>>> To: Christer Holmberg (JO/LMF)
>>>> Cc: sip@ietf.org; sipping@ietf.org
>>>> Subject: [Sipping] Re: [Sip] New draft on early media
>>>>
>>>> Christer,
>>>>
>>>> Thanks for your comments.  Responses inline.
>>>>
>>>> Cheers,
>>>> --Richard
>>>>
>>>>> I haven't had time to read the document yet, but I am not
>>>> sure I agree
>>>>> with the statement saying that the biggest problem is media
>>>> arriving
>>>>> before the SDP answer. In some environments such media will
>>>> not even
>>>>> reach the terminal, due to gating etc.
>>>> Perhaps a better term would be "media sent before an SDP 
>> answer has 
>>>> been received".  As you point out, sending media before an SDP 
>>>> offer/answer exchange has completed poses the risk that that media 
>>>> will not arrive.
>>>> To put it in the terms used in the document, media clipping is a 
>>>> problem caused by early (pre-SDP-answer) media, and thus one which 
>>>> this document seeks to solve.
>>>>
>>>>> Many implementations DO send an SDP answer (or, at least an
>>>> unreliable
>>>>> "preview" of the SDP answer) before (or at the same time 
>> as) early 
>>>>> media is sent.
>>>> This is the behavior we would like to standardize -- that an SDP 
>>>> answer is *always* sent before media of any type.  What 
>> we're trying 
>>>> to do with this document is describe the requirements for how that 
>>>> answer is sent, and consolidate the many ways people have 
>> proposed to 
>>>> send it.
>>>>
>>>>> And, in the cases where implementations do NOT send an SDP answer 
>>>>> before they start sending early media - what behavior 
>> would you be 
>>>>> able to standardize?
>>>> We would prefer that standards not allow this behavior -- 
>> that seems 
>>>> like the right thing to me.  What's a little trickier is 
>> that you'd 
>>>> like to not have the answerer sending media before the answer has 
>>>> *arrived*.
>>>>   When the answer is carried in a reliable message, you could just 
>>>> forbid sending media before receipt has been confirmed.
>>>> When the transport is unreliable, the situation is a little less 
>>>> clear, but I think there's still possibilities for 
>> standardization, 
>>>> e.g., by forbidding answerers to send media until they've received 
>>>> media.  This is discussed in more length in the document.
>>>>
>>>>> Regards,
>>>>>
>>>>> Christer
>>>>>
>>>>>
>>>>>> -----Original Message-----
>>>>>> From: Richard Barnes [mailto:rbarnes@bbn.com]
>>>>>> Sent: 8. helmikuuta 2007 21:47
>>>>>> To: sip@ietf.org; sipping@ietf.org
>>>>>> Subject: [Sip] New draft on early media
>>>>>>
>>>>>> After the big discussions on SIPPING last summer about 
>> how to fix 
>>>>>> early media, and the reaction at the last couple of 
>> IETFs to Brian 
>>>>>> Stucker's draft on how to cope with it, I've put together
>>>> a draft to
>>>>>> try to consolidate all of the discussions and move toward
>>>> a solution.
>>>>>> draft-barnes-sip-em-ps-req-sol includes three important parts:
>>>>>>
>>>>>> 1. Problem statement: Although there's lots of definitions
>>>> of "early
>>>>>> media" in existing documents, I think there's broad 
>> agreement that 
>>>>>> the stuff that causes problems is media that arrives 
>> before an SDP 
>>>>>> offer/answer exchange has completed.
>>>>>> The problem is therefore to enable the SDP exchange(s) in
>>>> the INVITE
>>>>>> transaction to complete more quickly, and make it more
>>>> difficult (if
>>>>>> not impossible) to send media before it has completed.  We are 
>>>>>> explicitly NOT trying to prevent the sending of media
>>>> before a 200/OK
>>>>>> or other closing of an INVITE transaction.
>>>>>>
>>>>>> 2. Requirements: Here, we try to create a consolidated list of 
>>>>>> requirements for an early media fix.
>>>>>>
>>>>>> 3. Analysis of Solutions: Lots of folks have put forward
>>>> proposals to
>>>>>> fix the early media problem.  This section tries to
>>>> enumerate most of
>>>>>> them and evaluate them against the requirements in section 2.
>>>>>>
>>>>>> It hasn't shown up yet on the I-D tracker, so I've 
>> attached a copy.
>>>>>> --Richard
>>>>>>
>>>> _______________________________________________
>>>> Sipping mailing list  
>> https://www1.ietf.org/mailman/listinfo/sipping
>>>> This list is for NEW development of the application of SIP Use 
>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>> sip@ietf.org for new developments of core SIP
>>>>
>>>
>>
> 
> 



_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 14 17:40:51 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHSnq-0000In-O5; Wed, 14 Feb 2007 17:40:42 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHSnp-0000II-Dy
	for sipping@ietf.org; Wed, 14 Feb 2007 17:40:41 -0500
Received: from rwcrmhc12.comcast.net ([204.127.192.82])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHSno-000798-6Z
	for sipping@ietf.org; Wed, 14 Feb 2007 17:40:41 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (rwcrmhc12) with ESMTP
	id <20070214224036m12005ddqbe>; Wed, 14 Feb 2007 22:40:36 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1EMePAM002712
	for <sipping@ietf.org>; Wed, 14 Feb 2007 17:40:25 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1EMePpr002708;
	Wed, 14 Feb 2007 17:40:25 -0500
Date: Wed, 14 Feb 2007 17:40:25 -0500
Message-Id: <200702142240.l1EMePpr002708@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <50B1CBA96870A34799A506B2313F26670B05FEF4@ntht201e.siemenscomms.co.uk>
	(john.elwell@siemens.com)
Subject: Re: [SIPPING] SDP offer/answer	inearly media:draft-sawada-sipping
	-sip-offeranswer-01.txt
References: <50B1CBA96870A34799A506B2313F26670B05FEF4@ntht201e.siemenscomms.co.uk>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 7a6398bf8aaeabc7a7bb696b6b0a2aad
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   From: "Elwell, John" <john.elwell@siemens.com>

   [JRE] I don't think 180 means that. It means play ringback tone until media
   starts to arrive.

It doesn't mean that, either.  It means that the UAS that was reached
is declaring that it is notifying a user.  Whether or how the UAC
wishes to notify its user of this fact is its business.

Currently, SIP has no defined means for indicating "Stop playing the
ringback and start playing this media stream."

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 14 17:41:48 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHSoo-0001aq-74; Wed, 14 Feb 2007 17:41:42 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHSom-0001al-Hc
	for sipping@ietf.org; Wed, 14 Feb 2007 17:41:40 -0500
Received: from sccrmhc15.comcast.net ([204.127.200.85])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHSol-0007C5-C1
	for sipping@ietf.org; Wed, 14 Feb 2007 17:41:40 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (sccrmhc15) with ESMTP
	id <200702142241210150097ua2e>; Wed, 14 Feb 2007 22:41:39 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1EMfKAM002801
	for <sipping@ietf.org>; Wed, 14 Feb 2007 17:41:21 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1EMfKWg002796;
	Wed, 14 Feb 2007 17:41:20 -0500
Date: Wed, 14 Feb 2007 17:41:20 -0500
Message-Id: <200702142241.l1EMfKWg002796@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <7374777208BDC7449D5620EF9423256702047D9A@esealmw113.eemea.ericsson.se>
	(christer.holmberg@ericsson.com)
Subject: Re: [SIPPING] SDP offer/answer
	inearly	media:draft-sawada-sipping-sip-offeranswer-01.txt
References: <7374777208BDC7449D5620EF9423256702047D9A@esealmw113.eemea.ericsson.se>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 1ac7cc0a4cd376402b85bc1961a86ac2
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>

   I once proposed that it should be possible to terminate an early dialog
   with an 19x provisional response.

   Then, if "fake forking" is used, the dialog not used anymore could be
   terminted and the UAC knows he doesn't have to care about it (and
   possible media coming on it).

That seems like it might be useful, given the number of times I've
seen "fake forking" proposed for dealing with one situation or
another.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP

From sipping-bounces@ietf.org Wed Feb 14 17:41:48 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHSok-0001ZM-HH; Wed, 14 Feb 2007 17:41:38 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHSoi-0001UO-W6; Wed, 14 Feb 2007 17:41:36 -0500
Received: from zcars04f.nortel.com ([47.129.242.57])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HHSog-0007Bv-Fc; Wed, 14 Feb 2007 17:41:36 -0500
Received: from zrc2hxm0.corp.nortel.com (zrc2hxm0.corp.nortel.com
	[47.103.123.71])
	by zcars04f.nortel.com (Switch-2.2.6/Switch-2.2.0) with ESMTP id
	l1EMfVY26126; Wed, 14 Feb 2007 17:41:31 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Re: [Sip] New draft on early media
Date: Wed, 14 Feb 2007 16:41:05 -0600
Message-ID: <1ECE0EB50388174790F9694F77522CCF0F31AD3F@zrc2hxm0.corp.nortel.com>
In-Reply-To: <45D38F5C.9050901@bbn.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
TFrom sipping-bounces@ietf.org Wed Feb 14 17:41:48 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHSoo-0001aq-74; Wed, 14 Feb 2007 17:41:42 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHSom-0001al-Hc
	for sipping@ietf.org; Wed, 14 Feb 2007 17:41:40 -0500
Received: from sccrmhc15.comcast.net ([204.127.200.85])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHSol-0007C5-C1
	for sipping@ietf.org; Wed, 14 Feb 2007 17:41:40 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (sccrmhc15) with ESMTP
	id <200702142241210150097ua2e>; Wed, 14 Feb 2007 22:41:39 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1EMfKAM002801
	for <sipping@ietf.org>; Wed, 14 Feb 2007 17:41:21 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1EMfKWg002796;
	Wed, 14 Feb 2007 17:41:20 -0500
Date: Wed, 14 Feb 2007 17:41:20 -0500
Message-Id: <200702142241.l1EMfKWg002796@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <7374777208BDC7449D5620EF9423256702047D9A@esealmw113.eemea.ericsson.se>
	(christer.holmberg@ericsson.com)
Subject: Re: [SIPPING] SDP offer/answer
	inearly	media:draft-sawada-sipping-sip-offeranswer-01.txt
References: <7374777208BDC7449D5620EF9423256702047D9A@esealmw113.eemea.ericsson.se>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 1ac7cc0a4cd376402b85bc1961a86ac2
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>

   I once proposed that it should be possible to terminate an early dialog
   with an 19x provisional response.

   Then, if "fake forking" is used, the dialog not used anymore could be
   terminted and the UAC knows he doesn't have to care about it (and
   possible media coming on it).

That seems like it might be useful, given the number of times I've
seen "fake forking" proposed for dealing with one situation or
another.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP

From sipping-bounces@ietf.org Wed Feb 14 17:41:48 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHSok-0001ZM-HH; Wed, 14 Feb 2007 17:41:38 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHSoi-0001UO-W6; Wed, 14 Feb 2007 17:41:36 -0500
Received: from zcars04f.nortel.com ([47.129.242.57])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HHSog-0007Bv-Fc; Wed, 14 Feb 2007 17:41:36 -0500
Received: from zrc2hxm0.corp.nortel.com (zrc2hxm0.corp.nortel.com
	[47.103.123.71])
	by zcars04f.nortel.com (Switch-2.2.6/Switch-2.2.0) with ESMTP id
	l1EMfVY26126; Wed, 14 Feb 2007 17:41:31 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Re: [Sip] New draft on early media
Date: Wed, 14 Feb 2007 16:41:05 -0600
Message-ID: <1ECE0EB50388174790F9694F77522CCF0F31AD3F@zrc2hxm0.corp.nortel.com>
In-Reply-To: <45D38F5C.9050901@bbn.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Re: [Sip] New draft on early media
Thread-Index: AcdQiNr9Ar+iFqn6S/WNCFhVSRunbAAAE4nw
From: "Francois Audet" <audet@nortel.com>
To: "Richard Barnes" <rbarnes@bbn.com>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: bcd240e64c427d3d3617cfc704e7fd7f
Cc: sip@ietf.org, sipping@ietf.org,
	"Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I think it makes sense as a starting point for a partial solution, yes.=20

> -----Original Message-----
> From: Richard Barnes [mailto:rbarnes@bbn.com]=20
> Sent: Wednesday, February 14, 2007 14:38
> To: Audet, Francois (SC100:3055)
> Cc: Christer Holmberg (JO/LMF); sip@ietf.org; sipping@ietf.org
> Subject: Re: [Sipping] Re: [Sip] New draft on early media
>=20
> Thanks for the clarification.  I know that choosing media=20
> streams to render is a big problem, but it's one I'm viewing=20
> as beyond the scope of what our draft is explicitly trying to address.
>=20
> That said, I think that ensuring that media are preceded by=20
> an SDP answer is a big step in enabling better management of=20
> media streams, since it gives the offeror (media recipient)=20
> SDP metadata on which to base its decision.  Once an SDP=20
> answer is guaranteed to come before media, the issue can be=20
> treated more at the SDP layer (e.g. by giving SDP sufficient=20
> expressiveness to support decisions between media streams),=20
> rather than at the SIP layer.
>=20
> With respect to forking in particular: You're correct that=20
> even if all streams from forked endpoints are preceded by SDP=20
> answers, the UAC still has to figure out how to deal with=20
> them.  But I think that once you ensure that the offer/answer=20
> finishes before media flows, SRTP keying gets a lot simpler,=20
> since the UAC can just do a DH exchange through your favorite=20
> SDP mechanism with each endpoint.
>=20
> Clipping is certainly something that could be an issue, but I=20
> think it can be prevented.  The solution I think will work in=20
> the end is that an SDP answer will be sent immediately by the=20
> UAS on receipt of the INVITE, before any other action is=20
> taken (e.g., before the INVITE is passed to the TU).  For=20
> example, a PSTN gateway receiving a SIP INVITE would return=20
> its answer before it initiates the outbound PSTN call.  It=20
> seems to me like the negotion of an SDP association should=20
> precede SIP session establishment just like TCP negotiation=20
> would precede initiation of a higher-layer protocol session.
>=20
> Does that make sense?
> --Richard
>=20
>=20
>=20
>=20
> Francois Audet wrote:
> > I mean play incoming media.=20
> >=20
> >> -----Original Message-----
> >> From: Richard Barnes [mailto:rbarnes@bbn.com]
> >> Sent: Tuesday, February 13, 2007 11:41
> >> To: Audet, Francois (SC100:3055)
> >> Cc: Christer Holmberg (JO/LMF); sip@ietf.org; sipping@ietf.org
> >> Subject: Re: [Sipping] Re: [Sip] New draft on early media
> >>
> >> Francois,
> >>
> >> I think I have some responses, but I'd like to ask you first:=20
> >> What do you mean by "media cut-through"?
> >>
> >> Thanks,
> >> --Richard
> >>
> >> Francois Audet wrote:
> >>> Hi Richard,
> >>>
> >>> Forcing media cut-through to occur only after an Answer has been=20
> >>> received does solve SOME of the early media problems. As
> >> others have
> >>> mentioned, there are environments that operate like this=20
> already. I=20
> >>> also know many products that do not cut-through media before the=20
> >>> answer is received.
> >>>
> >>> But there are other problems related with early media.=20
> >> Brhread-Topic: [Sipping] Re: [Sip] New draft on early media
Thread-Index: AcdQiNr9Ar+iFqn6S/WNCFhVSRunbAAAE4nw
From: "Francois Audet" <audet@nortel.com>
To: "Richard Barnes" <rbarnes@bbn.com>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: bcd240e64c427d3d3617cfc704e7fd7f
Cc: sip@ietf.org, sipping@ietf.org,
	"Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I think it makes sense as a starting point for a partial solution, yes.=20

> -----Original Message-----
> From: Richard Barnes [mailto:rbarnes@bbn.com]=20
> Sent: Wednesday, February 14, 2007 14:38
> To: Audet, Francois (SC100:3055)
> Cc: Christer Holmberg (JO/LMF); sip@ietf.org; sipping@ietf.org
> Subject: Re: [Sipping] Re: [Sip] New draft on early media
>=20
> Thanks for the clarification.  I know that choosing media=20
> streams to render is a big problem, but it's one I'm viewing=20
> as beyond the scope of what our draft is explicitly trying to address.
>=20
> That said, I think that ensuring that media are preceded by=20
> an SDP answer is a big step in enabling better management of=20
> media streams, since it gives the offeror (media recipient)=20
> SDP metadata on which to base its decision.  Once an SDP=20
> answer is guaranteed to come before media, the issue can be=20
> treated more at the SDP layer (e.g. by giving SDP sufficient=20
> expressiveness to support decisions between media streams),=20
> rather than at the SIP layer.
>=20
> With respect to forking in particular: You're correct that=20
> even if all streams from forked endpoints are preceded by SDP=20
> answers, the UAC still has to figure out how to deal with=20
> them.  But I think that once you ensure that the offer/answer=20
> finishes before media flows, SRTP keying gets a lot simpler,=20
> since the UAC can just do a DH exchange through your favorite=20
> SDP mechanism with each endpoint.
>=20
> Clipping is certainly something that could be an issue, but I=20
> think it can be prevented.  The solution I think will work in=20
> the end is that an SDP answer will be sent immediately by the=20
> UAS on receipt of the INVITE, before any other action is=20
> taken (e.g., before the INVITE is passed to the TU).  For=20
> example, a PSTN gateway receiving a SIP INVITE would return=20
> its answer before it initiates the outbound PSTN call.  It=20
> seems to me like the negotion of an SDP association should=20
> precede SIP session establishment just like TCP negotiation=20
> would precede initiation of a higher-layer protocol session.
>=20
> Does that make sense?
> --Richard
>=20
>=20
>=20
>=20
> Francois Audet wrote:
> > I mean play incoming media.=20
> >=20
> >> -----Original Message-----
> >> From: Richard Barnes [mailto:rbarnes@bbn.com]
> >> Sent: Tuesday, February 13, 2007 11:41
> >> To: Audet, Francois (SC100:3055)
> >> Cc: Christer Holmberg (JO/LMF); sip@ietf.org; sipping@ietf.org
> >> Subject: Re: [Sipping] Re: [Sip] New draft on early media
> >>
> >> Francois,
> >>
> >> I think I have some responses, but I'd like to ask you first:=20
> >> What do you mean by "media cut-through"?
> >>
> >> Thanks,
> >> --Richard
> >>
> >> Francois Audet wrote:
> >>> Hi Richard,
> >>>
> >>> Forcing media cut-through to occur only after an Answer has been=20
> >>> received does solve SOME of the early media problems. As
> >> others have
> >>> mentioned, there are environments that operate like this=20
> already. I=20
> >>> also know many products that do not cut-through media before the=20
> >>> answer is received.
> >>>
> >>> But there are other problems related with early media.=20
> >> Brian Stucker's
> >>> draft listed a bunch of them.
> >>>
> >>> Forking for example comes to mind: even if we did mandate
> >> cut-through
> >>> after answer, we would still have problems related with
> >> forking (i.e.,
> >>> how to pick the stream that needs to be played, and how to
> >> correlate
> >>> the stream with the signaling). How to deal with SRTP key
> >> exchange in
> >>> that scenario is an even bigger problem.
> >>>
> >>> Clipping is a possible side-effect of your proposal. We'd=20
> need some=20
> >>> mechanism to mitigate it, in particular for auto-answer scenarios.
> >>>
> >>>
> >>>> -----Original Message-----
> >>>> From: Richard Barnes [mailto:rbarnes@bbn.com]
> >>>> Sent: Monday, February 12, 2007 07:08
> >>>> To: Christer Holmberg (JO/LMF)
> >>>> Cc: sip@ietf.org; sipping@ietf.org
> >>>> Subject: [Sipping] Re: [Sip] New draft on early media
> >>>>
> >>>> Christer,
> >>>>
> >>>> Thanks for your comments.  Responses inline.
> >>>>
> >>>> Cheers,
> >>>> --Richard
> >>>>
> >>>>> I haven't had time to read the document yet, but I am not
> >>>> sure I agree
> >>>>> with the statement saying that the biggest problem is media
> >>>> arriving
> >>>>> before the SDP answer. In some environments such media will
> >>>> not even
> >>>>> reach the terminal, due to gating etc.
> >>>> Perhaps a better term would be "media sent before an SDP
> >> answer has
> >>>> been received".  As you point out, sending media before an SDP=20
> >>>> offer/answer exchange has completed poses the risk that=20
> that media=20
> >>>> will not arrive.
> >>>> To put it in the terms used in the document, media clipping is a=20
> >>>> problem caused by early (pre-SDP-answer) media, and thus=20
> one which=20
> >>>> this document seeks to solve.
> >>>>
> >>>>> Many implementations DO send an SDP answer (or, at least an
> >>>> unreliable
> >>>>> "preview" of the SDP answer) before (or at the same time
> >> as) early
> >>>>> media is sent.
> >>>> This is the behavior we would like to standardize -- that an SDP=20
> >>>> answer is *always* sent before media of any type.  What
> >> we're trying
> >>>> to do with this document is describe the requirements=20
> for how that=20
> >>>> answer is sent, and consolidate the many ways people have
> >> proposed to
> >>>> send it.
> >>>>
> >>>>> And, in the cases where implementations do NOT send an=20
> SDP answer=20
> >>>>> before they start sending early media - what behavior
> >> would you be
> >>>>> able to standardize?
> >>>> We would prefer that standards not allow this behavior --
> >> that seems
> >>>> like the right thing to me.  What's a little trickier is
> >> that you'd
> >>>> like to not have the answerer sending media before the=20
> answer has=20
> >>>> *arrived*.
> >>>>   When the answer is carried in a reliable message, you=20
> could just=20
> >>>> forbid sending media before receipt has been confirmed.
> >>>> When the transport is unreliable, the situation is a little less=20
> >>>> clear, but I think there's still possibilities for
> >> standardization,
> >>>> e.g., by forbidding answerers to send media until=20
> they've received=20
> >>>> media.  This is discussed in more length in the document.
> >>>>
> >>>>> Regards,
> >>>>>
> >>>>> Christer
> >>>>>
> >>>>>
> >>>>>> -----Original Message-----
> >>>>>> From: Richard Barnes [mailto:rbarnes@bbn.com]
> >>>>>> Sent: 8. helmikuuta 2007 21:47
> >>>>>> To: sip@ietf.org; sipping@ietf.org
> >>>>>> Subject: [Sip] New draft on early media
> >>>>>>
> >>>>>> After the big discussions on SIPPING last summer about
> >> how to fix
> >>>>>> early media, and the reaction at the last couple of
> >> IETFs to Brian
> >>>>>> Stucker's draft on how to cope with it, I've put together
> >>>> a draft to
> >>>>>> try to consolidate all of the discussions and move toward
> >>>> a solution.
> >>>>>> draft-barnes-sip-em-ps-req-sol includes three important parts:
> >>>>>>
> >>>>>> 1. Problem statement: Although there's lots of definitions
> >>>> of "early
> >>>>>> media" in existing documents, I think there's broad
> >> agreement that
> >>>>>> the stuff that caian Stucker's
> >>> draft listed a bunch of them.
> >>>
> >>> Forking for example comes to mind: even if we did mandate
> >> cut-through
> >>> after answer, we would still have problems related with
> >> forking (i.e.,
> >>> how to pick the stream that needs to be played, and how to
> >> correlate
> >>> the stream with the signaling). How to deal with SRTP key
> >> exchange in
> >>> that scenario is an even bigger problem.
> >>>
> >>> Clipping is a possible side-effect of your proposal. We'd=20
> need some=20
> >>> mechanism to mitigate it, in particular for auto-answer scenarios.
> >>>
> >>>
> >>>> -----Original Message-----
> >>>> From: Richard Barnes [mailto:rbarnes@bbn.com]
> >>>> Sent: Monday, February 12, 2007 07:08
> >>>> To: Christer Holmberg (JO/LMF)
> >>>> Cc: sip@ietf.org; sipping@ietf.org
> >>>> Subject: [Sipping] Re: [Sip] New draft on early media
> >>>>
> >>>> Christer,
> >>>>
> >>>> Thanks for your comments.  Responses inline.
> >>>>
> >>>> Cheers,
> >>>> --Richard
> >>>>
> >>>>> I haven't had time to read the document yet, but I am not
> >>>> sure I agree
> >>>>> with the statement saying that the biggest problem is media
> >>>> arriving
> >>>>> before the SDP answer. In some environments such media will
> >>>> not even
> >>>>> reach the terminal, due to gating etc.
> >>>> Perhaps a better term would be "media sent before an SDP
> >> answer has
> >>>> been received".  As you point out, sending media before an SDP=20
> >>>> offer/answer exchange has completed poses the risk that=20
> that media=20
> >>>> will not arrive.
> >>>> To put it in the terms used in the document, media clipping is a=20
> >>>> problem caused by early (pre-SDP-answer) media, and thus=20
> one which=20
> >>>> this document seeks to solve.
> >>>>
> >>>>> Many implementations DO send an SDP answer (or, at least an
> >>>> unreliable
> >>>>> "preview" of the SDP answer) before (or at the same time
> >> as) early
> >>>>> media is sent.
> >>>> This is the behavior we would like to standardize -- that an SDP=20
> >>>> answer is *always* sent before media of any type.  What
> >> we're trying
> >>>> to do with this document is describe the requirements=20
> for how that=20
> >>>> answer is sent, and consolidate the many ways people have
> >> proposed to
> >>>> send it.
> >>>>
> >>>>> And, in the cases where implementations do NOT send an=20
> SDP answer=20
> >>>>> before they start sending early media - what behavior
> >> would you be
> >>>>> able to standardize?
> >>>> We would prefer that standards not allow this behavior --
> >> that seems
> >>>> like the right thing to me.  What's a little trickier is
> >> that you'd
> >>>> like to not have the answerer sending media before the=20
> answer has=20
> >>>> *arrived*.
> >>>>   When the answer is carried in a reliable message, you=20
> could just=20
> >>>> forbid sending media before receipt has been confirmed.
> >>>> When the transport is unreliable, the situation is a little less=20
> >>>> clear, but I think there's still possibilities for
> >> standardization,
> >>>> e.g., by forbidding answerers to send media until=20
> they've received=20
> >>>> media.  This is discussed in more length in the document.
> >>>>
> >>>>> Regards,
> >>>>>
> >>>>> Christer
> >>>>>
> >>>>>
> >>>>>> -----Original Message-----
> >>>>>> From: Richard Barnes [mailto:rbarnes@bbn.com]
> >>>>>> Sent: 8. helmikuuta 2007 21:47
> >>>>>> To: sip@ietf.org; sipping@ietf.org
> >>>>>> Subject: [Sip] New draft on early media
> >>>>>>
> >>>>>> After the big discussions on SIPPING last summer about
> >> how to fix
> >>>>>> early media, and the reaction at the last couple of
> >> IETFs to Brian
> >>>>>> Stucker's draft on how to cope with it, I've put together
> >>>> a draft to
> >>>>>> try to consolidate all of the discussions and move toward
> >>>> a solution.
> >>>>>> draft-barnes-sip-em-ps-req-sol includes three important parts:
> >>>>>>
> >>>>>> 1. Problem statement: Although there's lots of definitions
> >>>> of "early
> >>>>>> media" in existing documents, I think there's broad
> >> agreement that
> >>>>>> the stuff that causes problems is media that arrives
> >> before an SDP
> >>>>>> offer/answer exchange has completed.
> >>>>>> The problem is therefore to enable the SDP exchange(s) in
> >>>> the INVITE
> >>>>>> transaction to complete more quickly, and make it more
> >>>> difficult (if
> >>>>>> not impossible) to send media before it has completed.  We are=20
> >>>>>> explicitly NOT trying to prevent the sending of media
> >>>> before a 200/OK
> >>>>>> or other closing of an INVITE transaction.
> >>>>>>
> >>>>>> 2. Requirements: Here, we try to create a consolidated list of=20
> >>>>>> requirements for an early media fix.
> >>>>>>
> >>>>>> 3. Analysis of Solutions: Lots of folks have put forward
> >>>> proposals to
> >>>>>> fix the early media problem.  This section tries to
> >>>> enumerate most of
> >>>>>> them and evaluate them against the requirements in section 2.
> >>>>>>
> >>>>>> It hasn't shown up yet on the I-D tracker, so I've
> >> attached a copy.
> >>>>>> --Richard
> >>>>>>
> >>>> _______________________________________________
> >>>> Sipping mailing list
> >> https://www1.ietf.org/mailman/listinfo/sipping
> >>>> This list is for NEW development of the application of SIP Use=20
> >>>> sip-implementors@cs.columbia.edu for questions on=20
> current sip Use=20
> >>>> sip@ietf.org for new developments of core SIP
> >>>>
> >>>
> >>
> >=20
> >=20
>=20
>=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP





uses problems is media that arrives
> >> before an SDP
> >>>>>> offer/answer exchange has completed.
> >>>>>> The problem is therefore to enable the SDP exchange(s) in
> >>>> the INVITE
> >>>>>> transaction to complete more quickly, and make it more
> >>>> difficult (if
> >>>>>> not impossible) to send media before it has completed.  We are=20
> >>>>>> explicitly NOT trying to prevent the sending of media
> >>>> before a 200/OK
> >>>>>> or other closing of an INVITE transaction.
> >>>>>>
> >>>>>> 2. Requirements: Here, we try to create a consolidated list of=20
> >>>>>> requirements for an early media fix.
> >>>>>>
> >>>>>> 3. Analysis of Solutions: Lots of folks have put forward
> >>>> proposals to
> >>>>>> fix the early media problem.  This section tries to
> >>>> enumerate most of
> >>>>>> them and evaluate them against the requirements in section 2.
> >>>>>>
> >>>>>> It hasn't shown up yet on the I-D tracker, so I've
> >> attached a copy.
> >>>>>> --Richard
> >>>>>>
> >>>> _______________________________________________
> >>>> Sipping mailing list
> >> https://www1.ietf.org/mailman/listinfo/sipping
> >>>> This list is for NEW development of the application of SIP Use=20
> >>>> sip-implementors@cs.columbia.edu for questions on=20
> current sip Use=20
> >>>> sip@ietf.org for new developments of core SIP
> >>>>
> >>>
> >>
> >=20
> >=20
>=20
>=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP





From sipping-bounces@ietf.org Wed Feb 14 18:22:55 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHTRz-0004EE-FX; Wed, 14 Feb 2007 18:22:11 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHTRy-0004D9-8t
	for sipping@ietf.org; Wed, 14 Feb 2007 18:22:10 -0500
Received: from ihemail1.lucent.com ([135.245.0.33])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHTRu-0004N7-W6
	for sipping@ietf.org; Wed, 14 Feb 2007 18:22:10 -0500
Received: from ilexp01.ndc.lucent.com (h135-3-39-1.lucent.com [135.3.39.1])
	by ihemail1.lucent.com (8.13.8/IER-o) with ESMTP id l1ENM5oN017785;
	Wed, 14 Feb 2007 17:22:05 -0600 (CST)
Received: from ILEXC1U02.ndc.lucent.com ([135.3.39.5]) by
	ilexp01.ndc.lucent.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 14 Feb 2007 17:22:05 -0600
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [SIPPING] SDP offer/answer	inearly
	media:draft-sawada-sipping-sip-offeranswer-01.txt
Date: Wed, 14 Feb 2007 17:22:07 -0600
Message-ID: <077439DFC93C83439E56A4EBDA4FF7944C06AD@ILEXC1U02.ndc.lucent.com>
In-Reply-To: <200702142240.l1EMePpr002708@dragon.ariadne.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [SIPPING] SDP offer/answer	inearly
	media:draft-sawada-sipping-sip-offeranswer-01.txt
Thread-Index: AcdQiTtQiOit9PAeTXiJbL8RnLfSOQAAXxCA
From: "Ejzak, Richard P \(Richard\)" <ejzak@alcatel-lucent.com>
To: <Dale.Worley@comcast.net>, <sipping@ietf.org>
X-OriginalArrivalTime: 14 Feb 2007 23:22:05.0697 (UTC)
	FILETIME=[F0AFA710:01C7508E]
X-Scanned-By: MIMEDefang 2.57 on 135.245.2.33
X-Spam-Score: 0.0 (/)
X-Scan-Signature: e5ba305d0e64821bf3d8bc5d3bb07228
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

=20

> From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net]=20
>=20
>    From: "Elwell, John" <john.elwell@siemens.com>
>=20
>    [JRE] I don't think 180 means that. It means play ringback=20
> tone until media
>    starts to arrive.
>=20
> It doesn't mean that, either.  It means that the UAS that was=20
> reached is declaring that it is notifying a user.  Whether or=20
> how the UAC wishes to notify its user of this fact is its business.
>=20
> Currently, SIP has no defined means for indicating "Stop=20
> playing the ringback and start playing this media stream."

Actually, that is one of the features of the P-Early-Media header in
draft-ejzak-sipping-p-em-auth-03.txt, which Gonzalo tells me has been
requested for publication.  By explicitly authorizing or not authorizing
early media, and changing that authorization if necessary, the UAC can
know when it should play ringback or the media stream.  Strictly
speaking, however, RFC 3960 does recommend the UAC policy that John
describes.  It's just not normative.

Regards,
Richard

>=20
> Dale
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 14 18:35:55 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHTe5-0004EL-Tm; Wed, 14 Feb 2007 18:34:41 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHTe3-0004EF-KC
	for sipping@ietf.org; Wed, 14 Feb 2007 18:34:39 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHTdv-0006DW-RW
	for sipping@ietf.org; Wed, 14 Feb 2007 18:34:39 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	48BC12057A; Thu, 15 Feb 2007 00:34:29 +0100 (CET)
X-AuditID: c1b4fb3c-b0fcabb0000007de-cf-45d39c85c05b 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	2E2F320534; Thu, 15 Feb 2007 00:34:29 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 15 Feb 2007 00:34:28 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Subject: Device architecture? WAS: RE: [Sipping] Service Identification in SIP
Date: Thu, 15 Feb 2007 00:34:28 +0100
Message-ID: <B0AF4CCDCE7F4D439EAC01F3331B1329265BBE@esealmw111>
In-Reply-To: <50B1CBA96870A34799A506B2313F26670B0A8D4E@ntht201e.siemenscomms.co.uk>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Device architecture? WAS: RE: [Sipping] Service Identification
	in SIP
Thread-Index: AcdQh+oX5uAth8ViTQaBZUsL/3x7LwAABNiw
From: =?iso-8859-1?Q?G=F6ran_Eriksson_AP_=28KI/EAB=29?=
	<goran.ap.eriksson@ericsson.com>
To: "Elwell, John" <john.elwell@siemens.com>
X-OriginalArrivalTime: 14 Feb 2007 23:34:28.0532 (UTC)
	FILETIME=[AB734740:01C75090]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: ff03b0075c3fc728d7d60a15b4ee1ad2
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

My humble, perhaps faulty, understanding of the "chess game" was not =
that voice commands were used to control the chess itself, rather
that the chess game, while having some IP-based proprietary gaming =
interaction protocol, also had a feature letting
the users talk to each others while playing. The scenario are two, or =
more, users play a chess game of the (Inter)net, talkin
to eachother from time to time during the game (yeah, don't ask me if I =
think people will love this!!).

The purpose, as I understand it, was a)to show that the 'voice =
component' in the chess game application is similar to that
in a stand alone POC-Only application when it comes to exchange and =
transfer of voice samples and associated floor control
and b) to point to differences, for instance when it comes to =
supplementary services, which may not apply in the chess game case.

Perhaps it is important to be clear on the assumed user device =
realisation, as far as I understand it:

 The "chess game + voice" example is, as I understand it, discussed =
under the assumption that voice and chess game is
 managed by custome designed appliction, or client, which is a =
necessary/likely approach in a mobile device given the
 limited MMI (and also other factors in the mobile context).

 In a PC, I could easily handle this as two different =
applications/client, like when I use Skype/MSN to talk to my friends =
while
 playing some turn based game, but again, the discussion is held under =
the assumption that usability requirements
 leads to an application client integrating voice and chess game events.

The alternative device realisation You describe- different clients =
managing different portions of what is negotiated in the SDP or MSRP =
MIME attachment-=20
is, I think, a good illustration of the need to be precise and explicit =
about the assumed device architecture when discussing the "service =
identification" issue(s).

Regards
G=F6ran




-----Original Message-----
From: Elwell, John [mailto:john.elwell@siemens.com]=20
Sent: den 14 februari 2007 11:57
To: Henry Sinnreich; sipping@ietf.org
Subject: RE: [Sipping] Service Identification in SIP

Having followed this thread, I too find the chess example very =
contrived. If chess moves are sent as voice commands, then surely it =
would be acceptable for the normal telephone application to answer the =
call, i.e., the remote player listens to the move, performs this move on =
a conventional or electronic chess board and speaks his next move. The =
calling UA should not care whether the call is answered by a computer =
application or a person.
Furthermore, even if the audio needs to consumed by a particular chess =
application at the called UA, then other media in the same session such =
as video or messaging might not be. So any qualification of the nature =
of media to be exchanged probably should be at the media level, not at =
the session level or SIP level. As yet another variation, what if the =
players set up a conventional telephone call and then mutually agree to =
switch in their chess applications?

Now, let's consider another case. I establish an MSRP session. Within =
this MSRP session I can send messages with various MIME attachments that =
are intended for consumption by various applications. I should not need =
to know in advance what applications I will use during the course of the =
MSRP session. All I should put in the INVITE request is the fact that I =
want to establish an MSRP session.

John=20

> -----Original Message-----
> From: Henry Sinnreich [mailto:hsinnrei@adobe.com]
> Sent: 13 February 2007 18:19
> To: sipping@ietf.org
> Subject: RE: [Sipping] Service Identification in SIP
>=20
> It must be my lack of imagination for not being able to come up with a =

> credible usage scenario, given the following:
>=20
> - Is playing chess by voice a credible usage scenario? :-) Especially=20
> since voice recognition is now so often found in PCs and in mobile=20
> phones as well. In other words, chess playing by voice may also be an=20
> endpoint application, without any network involvement. Actually, if=20
> placed in the endpoint it works just as well on poor bandwidth links=20
> by contrast to chess in the network that may require adequate=20
> broadband.
>=20
> - Is anyone's boss or spouse going to approve extra phone charges=20
> above the flat rate for playing chess at a time when the gaming=20
> industry has pioneered free VoIP?
> Will such a service be taxed extra and if yes using by what criteria?
> Gambling comes to my mind :-)
>=20
> - VoIP is reported in the press as a give-away and I have seen no=20
> reports of any VoIP provider recording a profit on VoIP. I assume the=20
> losses are often hidden by bundling in triple play, etc.
>=20
> Given my lack of imagination, can anyone share a more credible usage=20
> scenario for Service Identification in SIP?
>=20
> What's the use of a standard without a credible usage scenario?
>=20
> Thanks, Henry
>=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use =
sip-implementors@cs.columbia.edu for questions on current sip Use =
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 14 18:44:30 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHTnT-0008J4-Ag; Wed, 14 Feb 2007 18:44:23 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHTnS-0008I6-6o
	for sipping@ietf.org; Wed, 14 Feb 2007 18:44:22 -0500
Received: from zcars04f.nortel.com ([47.129.242.57])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHTnP-0007bQ-Nk
	for sipping@ietf.org; Wed, 14 Feb 2007 18:44:22 -0500
Received: from zrc2hxm2.corp.nortel.com (zrc2hxm2.corp.nortel.com
	[47.103.123.73])
	by zcars04f.nortel.com (Switch-2.2.6/Switch-2.2.0) with ESMTP id
	l1ENiHY09259; Wed, 14 Feb 2007 18:44:17 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Date: Wed, 14 Feb 2007 17:44:12 -0600
Message-ID: <6FC4416DDE56C44DA0AEE67BC7CA437112BBC630@zrc2hxm2.corp.nortel.com>
In-Reply-To: <50B1CBA96870A34799A506B2313F26670B0A8CB9@ntht201e.siemenscomms.co.uk>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: I-D ACTION:draft-stucker-sipping-early-media-indirection-00.txt
Thread-index: AcdQIW6n1rzFmWlNRDGYSogTgUPoGQAbOKHg
From: "Brian Stucker" <bstucker@nortel.com>
To: "Elwell, John" <john.elwell@siemens.com>, <sipping@ietf.org>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 0e9ebc0cbd700a87c0637ad0e2c91610
Cc: 
Subject: [Sipping] RE: I-D
	ACTION:draft-stucker-sipping-early-media-indirection-00.txt
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Thanks for reading the draft John. Replies below.

> -----Original Message-----
> From: Elwell, John [mailto:john.elwell@siemens.com]=20
> Sent: Wednesday, February 14, 2007 7:18 PM
> To: Stucker, Brian (RICH1:AR00); sipping@ietf.org
> Subject: RE: I-D=20
> ACTION:draft-stucker-sipping-early-media-indirection-00.t xt
>=20
> Brian,
>=20
> Although I am interested in this general problem space and=20
> welcome proposals, I had a hard time understanding the=20
> benefits this gives compared to RFC 3960. For example, in=20
> section 6.2.2 it states:
>=20
> "This dialog, known as an early-media dialog, allows the UAC=20
> to receive early media from endpoints other than the called=20
> party's device(s) such as a network media server, does not=20
> suffer from the problems of forking (since each indirect=20
> early-media session creates a separate early media dialog),=20
> allows for the UAC to specify a different RTP payload type=20
> (to facilitate discrimination of early-media versus=20
> final-media), and gives both the UAC and UAS the ability to=20
> explicitly start and stop the early media stream without=20
> complicating the SDP negotiations for the final media."
>=20
> It is not clear which of these properties of the new=20
> mechanism are not available with RFC 3960. For example, when=20
> an INVITE request is forked, each branch of the fork=20
> establishes a separate early dialog according to RFC 3960.=20
> Forking to a media server could either create a separate=20
> early dialog or use a separate SDP offer-answer exchange=20
> (early session disposition) within the context of an existing=20
> early dialog. The separate offer-answer exchange permits=20
> things like use of a different RTP payload type or different=20
> port. Explicitly stopping an early media stream can be done=20
> by a further offer-answer exchange using early session disposition.

At the last IETF, I asked attendees to raise their hands if any of them
had implemented early session disposition: nobody raised their hands,
and yet the problems of early media persist. RFC-3960 suffers from a
couple of problems that I outlined in
draft-stucker-sipping-early-media-coping. Even with early session
disposition, it's difficult to tell what's early media and what's final
media until you get a 200 OK. Even with RFC-3960 it's very difficult to
control multiple forked call legs that are generating early media (ie.
you have two or more people talking to you at once, and others that may
be sending you a 180 which further complicates things).

The proposed mechanism is to get rid of all of these problems by using a
dynamic RTP payload type (if you want to overload the RTP port at the
originator with early and final media) and simply establish a separate
session for each early media stream. This way everything is nice and
neat. You can start and stop it very clearly (INVITE/CANCEL/BYE) no
special logic is needed in the middle of the network to handle injection
of 183's with SDP for network based early media, etc, etc.

>=20
> Section 6.2.3 fails to say how the proxy that instructed the=20
> UAC to play an early media stream can subsequently tell the=20
> UAC to stop that stream (assuming the dialog between the UAC=20
> and the media server does not pass through that proxy).

Why would it? It's a proxy. Do you have a usecase? I'm sure something
could be proposed if it's necessary.

>=20
> The Early-Media header field definition does not appear to=20
> provide any information about the nature of the early media,=20
> although the examples do seem to show something of this=20
> nature, e.g., "service=3Dringback;user=3Duserb", but presumably=20
> this is information for the media server and not for=20
> consumption by the UAC. The UAC will therefore not be able to=20
> determine relative priorities of Early-Media header fields=20
> received on different early dialogs from the same or=20
> different proxies.

This is definitely true. This is a first revision of the proposed
mechanism to see if there's any support for this type of approach in
general. If people don't see value in setting up a separate session to
handle early media interactions, then the presence or absence of
information in the Early-Media header about prioritization is moot.

>=20
> John
>=20
>=20
> > -----Original Message-----
> > From: Internet-Drafts@ietf.org [mailto:Internet-Drafts@ietf.org]
> > Sent: 13 February 2007 20:50
> > To: i-d-announce@ietf.org
> > Subject: I-D
> > ACTION:draft-stucker-sipping-early-media-indirection-00.txt
> >=20
> > A New Internet-Draft is available from the on-line Internet-Drafts=20
> > directories.
> >=20
> >=20
> > 	Title		: Early Media INDirection (EMIND) in=20
> > the Session Initiation Protocol (SIP)
> > 	Author(s)	: B. Stucker
> > 	Filename	:=20
> > draft-stucker-sipping-early-media-indirection-00.txt
> > 	Pages		: 18
> > 	Date		: 2007-2-13
> > =09
> >    This document describes models that can be used to=20
> manage multiple
> >    early media streams in the Session Initiation Protocol=20
> (SIP).  The
> >    model seeks to allow calling party endpoints to be able to better
> >    control the rendering of early media to the user, and distinguish
> >    early media from final media.
> >=20
> >    Although some of early media challenges are likely to never be
> >    overcome, (e.g. when interworking with an ISUP PSTN gateway that=20
> > does
> >    not take into account CPG or ACM messages), the potential to=20
> > improve
> >    on what is already there does exist.  The recommendations in this
> >    document are intended to be an update to RFC-3960=20
> recommendations=20
> > and
> >    to avoid many of the complications outlined in
> >    draft-stucker-sipping-early-media-coping.
> >=20
> >=20
> >=20
> > A URL for this Internet-Draft is:
> > http://www.ietf.org/internet-drafts/draft-stucker-sipping-earl
> > y-media-indirection-00.txt
> >=20
> > To remove yourself from the I-D Announcement list, send a=20
> message to=20
> > i-d-announce-request@ietf.org with the word unsubscribe in=20
> the body of=20
> > the message.
> > You can also visit
> > https://www1.ietf.org/mailman/listinfo/I-D-announce
> > to change your subscription settings.
> >=20
> > Internet-Drafts are also available by anonymous FTP. Login with the=20
> > username "anonymous" and a password of your e-mail address. After=20
> > logging in, type "cd internet-drafts" and then "get=20
> > draft-stucker-sipping-early-media-indirection-00.txt".
> >=20
> > A list of Internet-Drafts directories can be found in=20
> > http://www.ietf.org/shadow.html or=20
> > ftp://ftp.ietf.org/ietf/1shadow-sites.txt
> >=20
> > Internet-Drafts can also be obtained by e-mail.
> >=20
> > Send a message to:
> > 	mailserv@ietf.org.
> > In the body type:
> > 	"FILE
> > /internet-drafts/draft-stucker-sipping-early-media-indirection
> > -00.txt".
> > =09
> > NOTE:	The mail server at ietf.org can return the document in
> > 	MIME-encoded form by using the "mpack" utility.  To use this
> > 	feature, insert the command "ENCODING mime" before the "FILE"
> > 	command.  To decode the response(s), you will need "munpack" or
> > 	a MIME-compliant mail reader.  Different MIME-compliant=20
> mail readers
> > 	exhibit different behavior, especially when dealing with
> > 	"multipart" MIME messages (i.e. documents which have been split
> > 	up into multiple messages), so check your local documentation on
> > 	how to manipulate these messages.
> >=20
> > Below is the data which will enable a MIME compliant mail reader=20
> > implementation to automatically retrieve the ASCII version of the=20
> > Internet-Draft.
> >=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 14 19:01:54 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHU3y-0003Ml-Qu; Wed, 14 Feb 2007 19:01:26 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHU3w-0003JD-KU
	for sipping@ietf.org; Wed, 14 Feb 2007 19:01:24 -0500
Received: from exprod6og56.obsmtp.com ([64.18.1.208])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1HHU3w-0002xM-1e
	for sipping@ietf.org; Wed, 14 Feb 2007 19:01:24 -0500
Received: from source ([192.150.20.142]) by exprod6ob56.postini.com
	([64.18.5.12]) with SMTP; Wed, 14 Feb 2007 16:01:20 PST
Received: from inner-relay-3.eur.adobe.com (inner-relay-3b [10.128.4.236])
	by outbound-smtp-2.corp.adobe.com (8.12.10/8.12.10) with ESMTP id
	l1F01dUh013122; Wed, 14 Feb 2007 16:01:41 -0800 (PST)
Received: from fe1.corp.adobe.com (fe1.corp.adobe.com [10.8.192.70])
	by inner-relay-3.eur.adobe.com (8.12.10/8.12.9) with ESMTP id
	l1F01BmA020142; Wed, 14 Feb 2007 16:01:12 -0800 (PST)
Received: from namail5.corp.adobe.com ([10.8.192.88]) by fe1.corp.adobe.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 14 Feb 2007 16:01:11 -0800
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Subject: RE: Device architecture? WAS: RE: [Sipping] Service Identification in
	SIP
Date: Wed, 14 Feb 2007 16:01:07 -0800
Message-ID: <24CCCC428EFEA2469BF046DB3C7A8D2249ED47@namail5.corp.adobe.com>
In-Reply-To: <B0AF4CCDCE7F4D439EAC01F3331B1329265BBE@esealmw111>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Device architecture? WAS: RE: [Sipping] Service Identification
	in SIP
Thread-Index: AcdQh+oX5uAth8ViTQaBZUsL/3x7LwAABNiwAAKV3SA=
References: <50B1CBA96870A34799A506B2313F26670B0A8D4E@ntht201e.siemenscomms.co.uk>
	<B0AF4CCDCE7F4D439EAC01F3331B1329265BBE@esealmw111>
From: "Henry Sinnreich" <hsinnrei@adobe.com>
To: =?iso-8859-1?Q?G=F6ran_Eriksson_AP_=28KI/EAB=29?=
	<goran.ap.eriksson@ericsson.com>, "Elwell, John" <john.elwell@siemens.com>
X-OriginalArrivalTime: 15 Feb 2007 00:01:11.0266 (UTC)
	FILETIME=[66C11420:01C75094]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 3d7f2f6612d734db849efa86ea692407
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

G=F6ran Eriksson wrote:

> having some IP-based proprietary gaming interaction protocol

Yes, your understanding is correct and there is a booming gaming =
industry out there working exactly this way.

Where we disagree is that there is a need felt in the gaming industry to =
have application level network infrastructure in the middle and helpful =
operators to charge extra for IP based games.

But there no sense in us trying to convince anyone to avoid planting =
network intermediaries for extra charging for IP based games.=20

Could I just get a ticket to the show to watch the enthusiastic =
reception by users and game manufacturers? :-)

Thanks, Henry

-----Original Message-----
From: G=F6ran Eriksson AP (KI/EAB) =
[mailto:goran.ap.eriksson@ericsson.com]=20
Sent: Wednesday, February 14, 2007 3:34 PM
To: Elwell, John
Cc: sipping@ietf.org
Subject: Device architecture? WAS: RE: [Sipping] Service Identification =
in SIP

My humble, perhaps faulty, understanding of the "chess game" was not =
that voice commands were used to control the chess itself, rather
that the chess game, while having some IP-based proprietary gaming =
interaction protocol, also had a feature letting
the users talk to each others while playing. The scenario are two, or =
more, users play a chess game of the (Inter)net, talkin
to eachother from time to time during the game (yeah, don't ask me if I =
think people will love this!!).

The purpose, as I understand it, was a)to show that the 'voice =
component' in the chess game application is similar to that
in a stand alone POC-Only application when it comes to exchange and =
transfer of voice samples and associated floor control
and b) to point to differences, for instance when it comes to =
supplementary services, which may not apply in the chess game case.

Perhaps it is important to be clear on the assumed user device =
realisation, as far as I understand it:

 The "chess game + voice" example is, as I understand it, discussed =
under the assumption that voice and chess game is
 managed by custome designed appliction, or client, which is a =
necessary/likely approach in a mobile device given the
 limited MMI (and also other factors in the mobile context).

 In a PC, I could easily handle this as two different =
applications/client, like when I use Skype/MSN to talk to my friends =
while
 playing some turn based game, but again, the discussion is held under =
the assumption that usability requirements
 leads to an application client integrating voice and chess game events.

The alternative device realisation You describe- different clients =
managing different portions of what is negotiated in the SDP or MSRP =
MIME attachment-=20
is, I think, a good illustration of the need to be precise and explicit =
about the assumed device architecture when discussing the "service =
identification" issue(s).

Regards
G=F6ran




-----Original Message-----
From: Elwell, John [mailto:john.elwell@siemens.com]=20
Sent: den 14 februari 2007 11:57
To: Henry Sinnreich; sipping@ietf.org
Subject: RE: [Sipping] Service Identification in SIP

Having followed this thread, I too find the chess example very =
contrived. If chess moves are sent as voice commands, then surely it =
would be acceptable for the normal telephone application to answer the =
call, i.e., the remote player listens to the move, performs this move on =
a conventional or electronic chess board and speaks his next move. The =
calling UA should not care whether the call is answered by a computer =
application or a person.
Furthermore, even if the audio needs to consumed by a particular chess =
application at the called UA, then other media in the same session such =
as video or messaging might not be. So any qualification of the nature =
of media to be exchanged probably should be at the media level, not at =
the session level or SIP level. As yet another variation, what if the =
players set up a conventional telephone call and then mutually agree to =
switch in their chess applications?

Now, let's consider another case. I establish an MSRP session. Within =
this MSRP session I can send messages with various MIME attachments that =
are intended for consumption by various applications. I should not need =
to know in advance what applications I will use during the course of the =
MSRP session. All I should put in the INVITE request is the fact that I =
want to establish an MSRP session.

John=20

> -----Original Message-----
> From: Henry Sinnreich [mailto:hsinnrei@adobe.com]
> Sent: 13 February 2007 18:19
> To: sipping@ietf.org
> Subject: RE: [Sipping] Service Identification in SIP
>=20
> It must be my lack of imagination for not being able to come up with a =

> credible usage scenario, given the following:
>=20
> - Is playing chess by voice a credible usage scenario? :-) Especially=20
> since voice recognition is now so often found in PCs and in mobile=20
> phones as well. In other words, chess playing by voice may also be an=20
> endpoint application, without any network involvement. Actually, if=20
> placed in the endpoint it works just as well on poor bandwidth links=20
> by contrast to chess in the network that may require adequate=20
> broadband.
>=20
> - Is anyone's boss or spouse going to approve extra phone charges=20
> above the flat rate for playing chess at a time when the gaming=20
> industry has pioneered free VoIP?
> Will such a service be taxed extra and if yes using by what criteria?
> Gambling comes to my mind :-)
>=20
> - VoIP is reported in the press as a give-away and I have seen no=20
> reports of any VoIP provider recording a profit on VoIP. I assume the=20
> losses are often hidden by bundling in triple play, etc.
>=20
> Given my lack of imagination, can anyone share a more credible usage=20
> scenario for Service Identification in SIP?
>=20
> What's the use of a standard without a credible usage scenario?
>=20
> Thanks, Henry
>=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use =
sip-implementors@cs.columbia.edu for questions on current sip Use =
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 14 19:40:43 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHUf1-0002RP-0Q; Wed, 14 Feb 2007 19:39:43 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHUey-0002RD-R8
	for sipping@ietf.org; Wed, 14 Feb 2007 19:39:40 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHUes-0000bx-W5
	for sipping@ietf.org; Wed, 14 Feb 2007 19:39:40 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	631DF20151; Thu, 15 Feb 2007 01:39:32 +0100 (CET)
X-AuditID: c1b4fb3e-af6d2bb0000007e1-f1-45d3abc4b251 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	37C7420006; Thu, 15 Feb 2007 01:39:32 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 15 Feb 2007 01:39:31 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Subject: RE: Device architecture? WAS: RE: [Sipping] Service Identification in
	SIP
Date: Thu, 15 Feb 2007 01:39:31 +0100
Message-ID: <B0AF4CCDCE7F4D439EAC01F3331B1329265BBF@esealmw111>
In-Reply-To: <24CCCC428EFEA2469BF046DB3C7A8D2249ED47@namail5.corp.adobe.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Device architecture? WAS: RE: [Sipping] Service Identification
	in SIP
Thread-Index: AcdQh+oX5uAth8ViTQaBZUsL/3x7LwAABNiwAAKV3SAAAItz0A==
From: =?iso-8859-1?Q?G=F6ran_Eriksson_AP_=28KI/EAB=29?=
	<goran.ap.eriksson@ericsson.com>
To: "Henry Sinnreich" <hsinnrei@adobe.com>,
	"Elwell, John" <john.elwell@siemens.com>
X-OriginalArrivalTime: 15 Feb 2007 00:39:31.0752 (UTC)
	FILETIME=[C1F36280:01C75099]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8b6657e60309a1317174c9db2ae5f227
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I very much agree with You there Henry! I do recognise the need to give =
*some classes of applications*, read telephony replacement, the =
possibility
to make whatever nethead, monolithic, tower of Babel, architecture that =
that industry need BUT I DO NOT, personally, what
any particular application class enforce/prevent/impose such an =
architecture on other applications, such as game, chat, etc.

The issue is of course that there is a great interest out there for =
telephony as such, and of course all telco operators are concerned with
the future business. This may easily lead to application/industry =
specific requriements getting into the (SIP) protocol(s), something that =
at least should
be avoided as much as possible, and if not possible, then at least =
contained.

And You're right- they will (try to) plant such charging devices there, =
because the name of the (market economy) game is they will if they can.

The funny side of this is that the IMS must probably also support gaming =
traffic, but not in a way that kills gaming traffic for instance by =
stupid charging.=20
Telco operators are more and more under pressure to provide real =
consumer value and charge reasonably. We'll see...


Regards
G=F6ran

PS. The show I would like to see is the one where users get to know that =
they will not have to pay for neither telephony, broadband access, =
mobile access,
expensive PC programs or OS's, music, IPTV, games, PC, mobile phone, etc =
for that matter! And yes, please no commercials also, :-)!!



-----Original Message-----
From: Henry Sinnreich [mailto:hsinnrei@adobe.com]=20
Sent: den 15 februari 2007 01:01
To: G=F6ran Eriksson AP (KI/EAB); Elwell, John
Cc: sipping@ietf.org
Subject: RE: Device architecture? WAS: RE: [Sipping] Service =
Identification in SIP

G=F6ran Eriksson wrote:

> having some IP-based proprietary gaming interaction protocol

Yes, your understanding is correct and there is a booming gaming =
industry out there working exactly this way.

Where we disagree is that there is a need felt in the gaming industry to =
have application level network infrastructure in the middle and helpful =
operators to charge extra for IP based games.

But there no sense in us trying to convince anyone to avoid planting =
network intermediaries for extra charging for IP based games.=20

Could I just get a ticket to the show to watch the enthusiastic =
reception by users and game manufacturers? :-)

Thanks, Henry

-----Original Message-----
From: G=F6ran Eriksson AP (KI/EAB) =
[mailto:goran.ap.eriksson@ericsson.com]
Sent: Wednesday, February 14, 2007 3:34 PM
To: Elwell, John
Cc: sipping@ietf.org
Subject: Device architecture? WAS: RE: [Sipping] Service Identification =
in SIP

My humble, perhaps faulty, understanding of the "chess game" was not =
that voice commands were used to control the chess itself, rather that =
the chess game, while having some IP-based proprietary gaming =
interaction protocol, also had a feature letting the users talk to each =
others while playing. The scenario are two, or more, users play a chess =
game of the (Inter)net, talkin to eachother from time to time during the =
game (yeah, don't ask me if I think people will love this!!).

The purpose, as I understand it, was a)to show that the 'voice =
component' in the chess game application is similar to that in a stand =
alone POC-Only application when it comes to exchange and transfer of =
voice samples and associated floor control and b) to point to =
differences, for instance when it comes to supplementary services, which =
may not apply in the chess game case.

Perhaps it is important to be clear on the assumed user device =
realisation, as far as I understand it:

 The "chess game + voice" example is, as I understand it, discussed =
under the assumption that voice and chess game is  managed by custome =
designed appliction, or client, which is a necessary/likely approach in =
a mobile device given the  limited MMI (and also other factors in the =
mobile context).

 In a PC, I could easily handle this as two different =
applications/client, like when I use Skype/MSN to talk to my friends =
while  playing some turn based game, but again, the discussion is held =
under the assumption that usability requirements  leads to an =
application client integrating voice and chess game events.

The alternative device realisation You describe- different clients =
managing different portions of what is negotiated in the SDP or MSRP =
MIME attachment- is, I think, a good illustration of the need to be =
precise and explicit about the assumed device architecture when =
discussing the "service identification" issue(s).

Regards
G=F6ran




-----Original Message-----
From: Elwell, John [mailto:john.elwell@siemens.com]
Sent: den 14 februari 2007 11:57
To: Henry Sinnreich; sipping@ietf.org
Subject: RE: [Sipping] Service Identification in SIP

Having followed this thread, I too find the chess example very =
contrived. If chess moves are sent as voice commands, then surely it =
would be acceptable for the normal telephone application to answer the =
call, i.e., the remote player listens to the move, performs this move on =
a conventional or electronic chess board and speaks his next move. The =
calling UA should not care whether the call is answered by a computer =
application or a person.
Furthermore, even if the audio needs to consumed by a particular chess =
application at the called UA, then other media in the same session such =
as video or messaging might not be. So any qualification of the nature =
of media to be exchanged probably should be at the media level, not at =
the session level or SIP level. As yet another variation, what if the =
players set up a conventional telephone call and then mutually agree to =
switch in their chess applications?

Now, let's consider another case. I establish an MSRP session. Within =
this MSRP session I can send messages with various MIME attachments that =
are intended for consumption by various applications. I should not need =
to know in advance what applications I will use during the course of the =
MSRP session. All I should put in the INVITE request is the fact that I =
want to establish an MSRP session.

John=20

> -----Original Message-----
> From: Henry Sinnreich [mailto:hsinnrei@adobe.com]
> Sent: 13 February 2007 18:19
> To: sipping@ietf.org
> Subject: RE: [Sipping] Service Identification in SIP
>=20
> It must be my lack of imagination for not being able to come up with a =

> credible usage scenario, given the following:
>=20
> - Is playing chess by voice a credible usage scenario? :-) Especially=20
> since voice recognition is now so often found in PCs and in mobile=20
> phones as well. In other words, chess playing by voice may also be an=20
> endpoint application, without any network involvement. Actually, if=20
> placed in the endpoint it works just as well on poor bandwidth links=20
> by contrast to chess in the network that may require adequate=20
> broadband.
>=20
> - Is anyone's boss or spouse going to approve extra phone charges=20
> above the flat rate for playing chess at a time when the gaming=20
> industry has pioneered free VoIP?
> Will such a service be taxed extra and if yes using by what criteria?
> Gambling comes to my mind :-)
>=20
> - VoIP is reported in the press as a give-away and I have seen no=20
> reports of any VoIP provider recording a profit on VoIP. I assume the=20
> losses are often hidden by bundling in triple play, etc.
>=20
> Given my lack of imagination, can anyone share a more credible usage=20
> scenario for Service Identification in SIP?
>=20
> What's the use of a standard without a credible usage scenario?
>=20
> Thanks, Henry
>=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use =
sip-implementors@cs.columbia.edu for questions on current sip Use =
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use =
sip-implementors@cs.columbia.edu for questions on current sip Use =
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 14 21:25:16 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHWI3-0001U1-BD; Wed, 14 Feb 2007 21:24:07 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHWI1-0001Tj-VM
	for sipping@ietf.org; Wed, 14 Feb 2007 21:24:05 -0500
Received: from sj-iport-2-in.cisco.com ([171.71.176.71]
	helo=sj-iport-2.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HHWI0-0004h2-B8
	for sipping@ietf.org; Wed, 14 Feb 2007 21:24:05 -0500
Received: from sj-dkim-4.cisco.com ([171.71.179.196])
	by sj-iport-2.cisco.com with ESMTP; 14 Feb 2007 18:24:04 -0800
X-IronPort-AV: i="4.14,170,1170662400"; 
	d="scan'208"; a="361027314:sNHT60694492"
Received: from sj-core-3.cisco.com (sj-core-3.cisco.com [171.68.223.137])
	by sj-dkim-4.cisco.com (8.12.11/8.12.11) with ESMTP id l1F2O3Z2022739; 
	Wed, 14 Feb 2007 18:24:03 -0800
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by sj-core-3.cisco.com (8.12.10/8.12.6) with ESMTP id l1F2O3ho015816;
	Wed, 14 Feb 2007 18:24:03 -0800 (PST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 14 Feb 2007 21:24:02 -0500
Received: from [10.86.240.161] ([10.86.240.161]) by xfe-rtp-201.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 14 Feb 2007 21:24:02 -0500
Message-ID: <45D3C441.1040709@cisco.com>
Date: Wed, 14 Feb 2007 21:24:01 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: =?ISO-8859-1?Q?=22G=F6ran_Eriksson_AP_=28KI/EAB=29=22?=
	<goran.ap.eriksson@ericsson.com>
Subject: Re: Device architecture? WAS: RE: [Sipping] Service Identification
	in SIP
References: <B0AF4CCDCE7F4D439EAC01F3331B1329265BBE@esealmw111>
In-Reply-To: <B0AF4CCDCE7F4D439EAC01F3331B1329265BBE@esealmw111>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 8bit
X-OriginalArrivalTime: 15 Feb 2007 02:24:02.0467 (UTC)
	FILETIME=[5B968F30:01C750A8]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=6710; t=1171506243;
	x=1172370243; c=relaxed/simple; s=sjdkim4002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20Device=20architecture?=20WAS=3A=20RE=3A=20[Sipping]=2
	0Service=20Identification=0A=20in=20SIP |Sender:=20;
	bh=Mk3yQSnR7lxD4t55Qa2r6+INZ1apHw5MOBHx6WQLcZM=;
	b=rw7F1WIjplDPUFDUQB+RFmDcPBnZzEPdZjQZywrHdZxyePB2mb+YEIzAxSO9EaCRtfczebRX
	KhnnyTt9JU5rFxT9Xjo4Yyeys2/heqDLIBFKiMWPeMO6SiWBfBQ05aws;
Authentication-Results: sj-dkim-4; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim4002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a7d2e37451f7f22841e3b6f40c67db0f
Cc: sipping@ietf.org, "Elwell, John" <john.elwell@siemens.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Göran Eriksson AP (KI/EAB) wrote:
> My humble, perhaps faulty, understanding of the "chess game" was not that voice commands were used to control the chess itself, rather
> that the chess game, while having some IP-based proprietary gaming interaction protocol, also had a feature letting
> the users talk to each others while playing. The scenario are two, or more, users play a chess game of the (Inter)net, talkin
> to eachother from time to time during the game (yeah, don't ask me if I think people will love this!!).

Well, that is not my understanding of the example Jonathan had used as 
an example. But nevertheless, lets go with it for a moment...

> The purpose, as I understand it, was a)to show that the 'voice component' in the chess game application is similar to that
> in a stand alone POC-Only application when it comes to exchange and transfer of voice samples and associated floor control
> and b) to point to differences, for instance when it comes to supplementary services, which may not apply in the chess game case.
> 
> Perhaps it is important to be clear on the assumed user device realisation, as far as I understand it:
> 
>  The "chess game + voice" example is, as I understand it, discussed under the assumption that voice and chess game is
>  managed by custome designed appliction, or client, which is a necessary/likely approach in a mobile device given the
>  limited MMI (and also other factors in the mobile context).
> 
>  In a PC, I could easily handle this as two different applications/client, like when I use Skype/MSN to talk to my friends while
>  playing some turn based game, but again, the discussion is held under the assumption that usability requirements
>  leads to an application client integrating voice and chess game events.

I don't think the analogy holds. Presumably the chess control protocol 
would be another media stream in the same session. It seems that the 
only key requirement here is that both media streams be accepted by 
something that understands how to deal with them. You may have a very 
special application for this in your mobile device. But it would be 
perfectly reasonable for the callee to have a PC that has a UA that 
delegates the chess protocol to a chess game and the voice stream to a 
general purpose media player.

> The alternative device realisation You describe- different clients managing different portions of what is negotiated in the SDP or MSRP MIME attachment- 
> is, I think, a good illustration of the need to be precise and explicit about the assumed device architecture when discussing the "service identification" issue(s).
> 
> Regards
> Göran
> 
> 
> 
> 
> -----Original Message-----
> From: Elwell, John [mailto:john.elwell@siemens.com] 
> Sent: den 14 februari 2007 11:57
> To: Henry Sinnreich; sipping@ietf.org
> Subject: RE: [Sipping] Service Identification in SIP
> 
> Having followed this thread, I too find the chess example very contrived. If chess moves are sent as voice commands, then surely it would be acceptable for the normal telephone application to answer the call, i.e., the remote player listens to the move, performs this move on a conventional or electronic chess board and speaks his next move. The calling UA should not care whether the call is answered by a computer application or a person.
> Furthermore, even if the audio needs to consumed by a particular chess application at the called UA, then other media in the same session such as video or messaging might not be. So any qualification of the nature of media to be exchanged probably should be at the media level, not at the session level or SIP level. As yet another variation, what if the players set up a conventional telephone call and then mutually agree to switch in their chess applications?
> 
> Now, let's consider another case. I establish an MSRP session. Within this MSRP session I can send messages with various MIME attachments that are intended for consumption by various applications. I should not need to know in advance what applications I will use during the course of the MSRP session. All I should put in the INVITE request is the fact that I want to establish an MSRP session.
> 
> John 
> 
>> -----Original Message-----
>> From: Henry Sinnreich [mailto:hsinnrei@adobe.com]
>> Sent: 13 February 2007 18:19
>> To: sipping@ietf.org
>> Subject: RE: [Sipping] Service Identification in SIP
>>
>> It must be my lack of imagination for not being able to come up with a 
>> credible usage scenario, given the following:
>>
>> - Is playing chess by voice a credible usage scenario? :-) Especially 
>> since voice recognition is now so often found in PCs and in mobile 
>> phones as well. In other words, chess playing by voice may also be an 
>> endpoint application, without any network involvement. Actually, if 
>> placed in the endpoint it works just as well on poor bandwidth links 
>> by contrast to chess in the network that may require adequate 
>> broadband.
>>
>> - Is anyone's boss or spouse going to approve extra phone charges 
>> above the flat rate for playing chess at a time when the gaming 
>> industry has pioneered free VoIP?
>> Will such a service be taxed extra and if yes using by what criteria?
>> Gambling comes to my mind :-)
>>
>> - VoIP is reported in the press as a give-away and I have seen no 
>> reports of any VoIP provider recording a profit on VoIP. I assume the 
>> losses are often hidden by bundling in triple play, etc.
>>
>> Given my lack of imagination, can anyone share a more credible usage 
>> scenario for Service Identification in SIP?
>>
>> What's the use of a standard without a credible usage scenario?
>>
>> Thanks, Henry
>>
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP Use 
>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>> sip@ietf.org for new developments of core SIP
>>
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use sip-implementors@cs.columbia.edu for questions on current sip Use sip@ietf.org for new developments of core SIP
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 15 03:47:36 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHcFF-0006Vm-Vz; Thu, 15 Feb 2007 03:45:37 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHcFE-0006V9-7m
	for sipping@ietf.org; Thu, 15 Feb 2007 03:45:36 -0500
Received: from mailgate.siemenscomms.co.uk ([195.171.110.225]
	helo=bemg01.siemenscomms.co.uk)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHcFA-00021I-L5
	for sipping@ietf.org; Thu, 15 Feb 2007 03:45:36 -0500
Received: from ntht207e.uksgcs.siemenscomms.co.uk ([137.223.247.82])
	by siemenscomms.co.uk (PMDF V6.0-24 #40642)
	with ESMTP id <0JDH00DB0XNKBA@siemenscomms.co.uk> for sipping@ietf.org;
	Thu, 15 Feb 2007 08:45:20 +0000 (GMT)
Received: by ntht207e.uksgcs.siemenscomms.co.uk with Internet Mail Service
	(5.5.2657.72)	id <D2K3YJ93>; Thu, 15 Feb 2007 08:45:20 +0000
Content-return: allowed
Date: Thu, 15 Feb 2007 08:45:19 +0000
From: "Elwell, John" <john.elwell@siemens.com>
Subject: RE: Device architecture? WAS: RE: [Sipping] Service Identificatio	n in
	SIP
To: =?iso-8859-1?Q?G=F6ran_Eriksson_AP_=28KI/EAB=29?=
	<goran.ap.eriksson@ericsson.com>
Message-id: <50B1CBA96870A34799A506B2313F26670B0A9358@ntht201e.siemenscomms.co.uk>
MIME-version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
Content-type: text/plain; charset=iso-8859-1
Content-transfer-encoding: quoted-printable
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a92270ba83d7ead10c5001bb42ec3221
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

G=F6ran,

Your interpretation is different from what is stated in the Rosenberg =
draft.
Anyway, let's not dwell on this particular use case, which may or may =
not be
realistic. The important thing is whether the media being sent is =
suitable
for general purpose devices or requires or could be used by some =
special
class of application.

If a medium is somehow specialised (e.g., audio modulated for =
interpretation
by a modem or fax machine), then the necessary information needs to be
conveyed along with the m-line in the SDP. Perhaps even if it is
human-intelligible voice but with a restricted repertoire of words that =
make
it suitable for a particular class of application, maybe that =
information
too should be conveyed with the m-line. Perhaps the case described in
Jonathan's draft fits this latter category: I believe the medium is =
just
regular voice that can be interpreted by a human, but is also =
constrained in
its repertoire of words such that it can be used by a chess =
application.

The important thing is that there be sufficient information in the SDP =
to
allow appropriate application selection at the end device. I am not
convinced by a need for information at the SIP level during call
establishment, because of course SDP can be re-negotiated during the =
call
(e.g., when the users decide to use a particular application) and what =
is
indicated in SIP at call establishment time might no longer be valid =
later
in the call.

John
=20

> -----Original Message-----
> From: G=F6ran Eriksson AP (KI/EAB)=20
> [mailto:goran.ap.eriksson@ericsson.com]=20
> Sent: 14 February 2007 23:34
> To: Elwell, John
> Cc: sipping@ietf.org
> Subject: Device architecture? WAS: RE: [Sipping] Service=20
> Identification in SIP
>=20
> My humble, perhaps faulty, understanding of the "chess game"=20
> was not that voice commands were used to control the chess=20
> itself, rather
> that the chess game, while having some IP-based proprietary=20
> gaming interaction protocol, also had a feature letting
> the users talk to each others while playing. The scenario are=20
> two, or more, users play a chess game of the (Inter)net, talkin
> to eachother from time to time during the game (yeah, don't=20
> ask me if I think people will love this!!).
>=20
> The purpose, as I understand it, was a)to show that the=20
> 'voice component' in the chess game application is similar to that
> in a stand alone POC-Only application when it comes to=20
> exchange and transfer of voice samples and associated floor control
> and b) to point to differences, for instance when it comes to=20
> supplementary services, which may not apply in the chess game case.
>=20
> Perhaps it is important to be clear on the assumed user=20
> device realisation, as far as I understand it:
>=20
>  The "chess game + voice" example is, as I understand it,=20
> discussed under the assumption that voice and chess game is
>  managed by custome designed appliction, or client, which is=20
> a necessary/likely approach in a mobile device given the
>  limited MMI (and also other factors in the mobile context).
>=20
>  In a PC, I could easily handle this as two different=20
> applications/client, like when I use Skype/MSN to talk to my=20
> friends while
>  playing some turn based game, but again, the discussion is=20
> held under the assumption that usability requirements
>  leads to an application client integrating voice and chess=20
> game events.
>=20
> The alternative device realisation You describe- different=20
> clients managing different portions of what is negotiated in=20
> the SDP or MSRP MIME attachment-=20
> is, I think, a good illustration of the need to be precise=20
> and explicit about the assumed device architecture when=20
> discussing the "service identification" issue(s).
>=20
> Regards
> G=F6ran
>=20
>=20
>=20
>=20
> -----Original Message-----
> From: Elwell, John [mailto:john.elwell@siemens.com]=20
> Sent: den 14 februari 2007 11:57
> To: Henry Sinnreich; sipping@ietf.org
> Subject: RE: [Sipping] Service Identification in SIP
>=20
> Having followed this thread, I too find the chess example=20
> very contrived. If chess moves are sent as voice commands,=20
> then surely it would be acceptable for the normal telephone=20
> application to answer the call, i.e., the remote player=20
> listens to the move, performs this move on a conventional or=20
> electronic chess board and speaks his next move. The calling=20
> UA should not care whether the call is answered by a computer=20
> application or a person.
> Furthermore, even if the audio needs to consumed by a=20
> particular chess application at the called UA, then other=20
> media in the same session such as video or messaging might=20
> not be. So any qualification of the nature of media to be=20
> exchanged probably should be at the media level, not at the=20
> session level or SIP level. As yet another variation, what if=20
> the players set up a conventional telephone call and then=20
> mutually agree to switch in their chess applications?
>=20
> Now, let's consider another case. I establish an MSRP=20
> session. Within this MSRP session I can send messages with=20
> various MIME attachments that are intended for consumption by=20
> various applications. I should not need to know in advance=20
> what applications I will use during the course of the MSRP=20
> session. All I should put in the INVITE request is the fact=20
> that I want to establish an MSRP session.
>=20
> John=20
>=20
> > -----Original Message-----
> > From: Henry Sinnreich [mailto:hsinnrei@adobe.com]
> > Sent: 13 February 2007 18:19
> > To: sipping@ietf.org
> > Subject: RE: [Sipping] Service Identification in SIP
> >=20
> > It must be my lack of imagination for not being able to=20
> come up with a=20
> > credible usage scenario, given the following:
> >=20
> > - Is playing chess by voice a credible usage scenario? :-)=20
> Especially=20
> > since voice recognition is now so often found in PCs and in mobile=20
> > phones as well. In other words, chess playing by voice may=20
> also be an=20
> > endpoint application, without any network involvement. Actually, if =

> > placed in the endpoint it works just as well on poor=20
> bandwidth links=20
> > by contrast to chess in the network that may require adequate=20
> > broadband.
> >=20
> > - Is anyone's boss or spouse going to approve extra phone charges=20
> > above the flat rate for playing chess at a time when the gaming=20
> > industry has pioneered free VoIP?
> > Will such a service be taxed extra and if yes using by what=20
> criteria?
> > Gambling comes to my mind :-)
> >=20
> > - VoIP is reported in the press as a give-away and I have seen no=20
> > reports of any VoIP provider recording a profit on VoIP. I=20
> assume the=20
> > losses are often hidden by bundling in triple play, etc.
> >=20
> > Given my lack of imagination, can anyone share a more=20
> credible usage=20
> > scenario for Service Identification in SIP?
> >=20
> > What's the use of a standard without a credible usage scenario?
> >=20
> > Thanks, Henry
> >=20
> >=20
> > _______________________________________________
> > Sipping mailing list  =
https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP Use=20
> > sip-implementors@cs.columbia.edu for questions on current sip Use=20
> > sip@ietf.org for new developments of core SIP
> >=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 15 05:18:10 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHdg6-0005af-2J; Thu, 15 Feb 2007 05:17:26 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHdg4-0005aa-Rg
	for sipping@ietf.org; Thu, 15 Feb 2007 05:17:24 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHdfx-0001ji-V2
	for sipping@ietf.org; Thu, 15 Feb 2007 05:17:24 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	7728321021; Thu, 15 Feb 2007 11:17:17 +0100 (CET)
X-AuditID: c1b4fb3e-ae6d0bb0000007e1-80-45d4332d3f5c 
Received: from esealmw127.eemea.ericsson.se (unknown [153.88.254.122])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	5751A200C8; Thu, 15 Feb 2007 11:17:17 +0100 (CET)
Received: from eesmdmw020.eemea.ericsson.se ([159.107.3.34]) by
	esealmw127.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 15 Feb 2007 11:17:17 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Thu, 15 Feb 2007 11:17:15 +0100
Message-ID: <BA43A5623E3B9142B2E39AF1E796BB84C84073@eesmdmw020.eemea.ericsson.se>
In-Reply-To: <50B1CBA96870A34799A506B2313F26670B0A8D4E@ntht201e.siemenscomms.co.uk>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdQ6ncyEQ3cE0suQDuCWpDndOEvDw==
From: "Jesus Javier Arauz \(MI/EEM\)" <jesus.javier.arauz@ericsson.com>
To: "Elwell, John" <john.elwell@siemens.com>,
	"Henry Sinnreich" <hsinnrei@adobe.com>, <sipping@ietf.org>
X-OriginalArrivalTime: 15 Feb 2007 10:17:17.0181 (UTC)
	FILETIME=[782812D0:01C750EA]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 944ecb6e61f753561f559a497458fb4f
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi,

I've been following this thread too, and I think perhaps it's useful to
put the problem -whatever it is- in context to focus things a bit.

The pattern we're looking at here is the typical "application uses
underlying protocol". To scope it down a little bit I'll consider two
far ends of this pattern:

* stupidly simple underlying protocol (SSUP): TCP, where TCP offers one
single "service" to the application: reliable bi-directional data stream
exchange. All applications using TCP know the kind of service they're
getting, so the only artifact that each protocol agent needs to earn the
day is the local and remote port numbers. Each application "registers"
with a local socket bound to those ports, and if multiple applications
want to share the ports they need an intermediary that performs the
multiplexing/demultiplexing using artifacts opaque to the TCP agent.

* unbounded complexity underlying protocol (UCUP): WS, where the range
of services available to an application is unlimited. Thus, in order to
use the available services an application needs a detailed formal
description of those services and their providers. Interoperability is
guaranteed by those descriptions, and an application does not need to
support a whole service specification by the letter in order to
use/provide that service successfully.

I think that a fair amount of the debate around SIP services stems from
each one's standpoint. Some stand more on the SSUP side, where SIP just
provides well-defined and scoped "basic multi-media communication", and
each application out there must make its own arrangements if something
more rich or complex wants to be used/provided. Others, however, stand
more on the UCUP side, where SIP is a mechanism on top of which a
potentially unbounded number of inter-operable richer services could be
provided.

I'd say that IETF people tend to be SSUP heads, while telecom people
tend to be UCUP heads. I think this is caused by the different business
models each group makes a living for.

Regards,
/Javier

Elwell, John wrote:
> Having followed this thread, I too find the chess example very
> contrived. If chess moves are sent as voice commands, then surely it
> would be acceptable for the normal telephone application to answer
> the call, i.e., the remote player listens to the move, performs this
> move on a conventional or electronic chess board and speaks his next
> move. The calling UA should not care whether the call is answered by
> a computer application or a person.     =20
> Furthermore, even if the audio needs to consumed by a particular
> chess application at the called UA, then other media in the same
> session such as video or messaging might not be. So any qualification
> of the nature of media to be exchanged probably should be at the
> media level, not at the session level or SIP level. As yet another
> variation, what if the players set up a conventional telephone call
> and then mutually agree to switch in their chess applications?     =20
>=20
> Now, let's consider another case. I establish an MSRP session. Within
> this MSRP session I can send messages with various MIME attachments
> that are intended for consumption by various applications. I should
> not need to know in advance what applications I will use during the
> course of the MSRP session. All I should put in the INVITE request is
> the fact that I want to establish an MSRP session.    =20
>=20
> John
>=20
>> -----Original Message-----
>> From: Henry Sinnreich [mailto:hsinnrei@adobe.com]
>> Sent: 13 February 2007 18:19
>> To: sipping@ietf.org
>> Subject: RE: [Sipping] Service Identification in SIP
>>=20
>> It must be my lack of imagination for not being able to come up with
>> a credible usage scenario, given the following:
>>=20
>> - Is playing chess by voice a credible usage scenario? :-) Especially
>> since voice recognition is now so often found in PCs and in mobile
>> phones as well. In other words, chess playing by voice may also be an
>> endpoint application, without any network involvement. Actually, if
>> placed in the endpoint it works just as well on poor bandwidth links
>> by contrast to chess in the network that may require adequate
>> broadband.=20
>>=20
>> - Is anyone's boss or spouse going to approve extra phone charges
>> above the flat rate for playing chess at a time when the gaming
>> industry has pioneered free VoIP?
>> Will such a service be taxed extra and if yes using by what
>> criteria? Gambling comes to my mind :-)=20
>>=20
>> - VoIP is reported in the press as a give-away and I have seen no
>> reports of any VoIP provider recording a profit on VoIP. I assume the
>> losses are often hidden by bundling in triple play, etc.
>>=20
>> Given my lack of imagination, can anyone share a more credible usage
>> scenario for Service Identification in SIP?
>>=20
>> What's the use of a standard without a credible usage scenario?
>>=20
>> Thanks, Henry
>>=20
>>=20
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP Use
>> sip-implementors@cs.columbia.edu for questions on current sip Use
>> sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 15 05:56:05 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHeFY-0005Z3-PJ; Thu, 15 Feb 2007 05:54:04 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHeFX-0005WM-G5
	for sipping@ietf.org; Thu, 15 Feb 2007 05:54:03 -0500
Received: from nf-out-0910.google.com ([64.233.182.185])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHeFW-0000PJ-Ak
	for sipping@ietf.org; Thu, 15 Feb 2007 05:54:03 -0500
Received: by nf-out-0910.google.com with SMTP id l36so1097121nfa
	for <sipping@ietf.org>; Thu, 15 Feb 2007 02:54:01 -0800 (PST)
DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta;
	h=received:message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:references;
	b=PQkbWqU70IURKgdHMpEI8T2Fq4katYKyOQ8dIjJRiZ1bYTt3GJKewXd6OXtTyAY+ij5zdG2NhE+dfPhTc9Npz18dDwOWQbz+5sr/xw2Yt22f5uD/EdjXV/k/8AGJKsI+Gl6+sThVXEpC5l9K0U30Hpis0b4ObtVAdWvIXRyN6Rg=
Received: by 10.82.153.5 with SMTP id a5mr2484729bue.1171536841395;
	Thu, 15 Feb 2007 02:54:01 -0800 (PST)
Received: by 10.82.185.16 with HTTP; Thu, 15 Feb 2007 02:54:01 -0800 (PST)
Message-ID: <3fe6b8640702150254w2abc2e73q5eedb8a1c9bdee35@mail.gmail.com>
Date: Thu, 15 Feb 2007 16:24:01 +0530
From: "Malleswara Rao Ankem" <malleshavn@gmail.com>
To: "Elwell, John" <john.elwell@siemens.com>
Subject: Re: Device architecture? WAS: RE: [Sipping] Service Identificatio n
	in SIP
In-Reply-To: <50B1CBA96870A34799A506B2313F26670B0A9358@ntht201e.siemenscomms.co.uk>
MIME-Version: 1.0
References: <50B1CBA96870A34799A506B2313F26670B0A9358@ntht201e.siemenscomms.co.uk>
X-Spam-Score: 0.5 (/)
X-Scan-Signature: 2f46977da373544777a750d5247d6ccc
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============0590673881=="
Errors-To: sipping-bounces@ietf.org

--===============0590673881==
Content-Type: multipart/alternative; 
	boundary="----=_Part_165786_23985264.1171536841327"

------=_Part_165786_23985264.1171536841327
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

Inorder to identify a endpoint that has the capabilities to serve a very
specialized functionality, I guess it is simpler to name the functionality.
Naming the functionality not only helps in decision making while routing th=
e
request to that specific endpoint but also helps the endpoint identify whic=
h
specialized application on the endpoint can take up the responsibility.
Instead of all the routing entities looking into the SDP/message bodies it
should be easier/faster if they can look for a specific header.
I thought feature tag mechanisms indicated in the RFC3841/3840 combination
should have sufficed but implementing those RFCs seems to be a overkill for
describing/selecting/routing requests to the specific endpoint and also
choosing a service/feature/application on the endpoint.

-Mallesh



On 2/15/07, Elwell, John <john.elwell@siemens.com> wrote:
>
> G=F6ran,
>
> Your interpretation is different from what is stated in the Rosenberg
> draft.
> Anyway, let's not dwell on this particular use case, which may or may not
> be
> realistic. The important thing is whether the media being sent is suitabl=
e
> for general purpose devices or requires or could be used by some special
> class of application.
>
> If a medium is somehow specialised ( e.g., audio modulated for
> interpretation
> by a modem or fax machine), then the necessary information needs to be
> conveyed along with the m-line in the SDP. Perhaps even if it is
> human-intelligible voice but with a restricted repertoire of words that
> make
> it suitable for a particular class of application, maybe that information
> too should be conveyed with the m-line. Perhaps the case described in
> Jonathan's draft fits this latter category: I believe the medium is just
> regular voice that can be interpreted by a human, but is also constrained
> in
> its repertoire of words such that it can be used by a chess application.
>
> The important thing is that there be sufficient information in the SDP to
> allow appropriate application selection at the end device. I am not
> convinced by a need for information at the SIP level during call
> establishment, because of course SDP can be re-negotiated during the call
> ( e.g., when the users decide to use a particular application) and what i=
s
> indicated in SIP at call establishment time might no longer be valid late=
r
> in the call.
>
> John
>
>
> > -----Original Message-----
> > From: G=F6ran Eriksson AP (KI/EAB)
> > [mailto:goran.ap.eriksson@ericsson.com]
> > Sent: 14 February 2007 23:34
> > To: Elwell, John
> > Cc: sipping@ietf.org
> > Subject: Device architecture? WAS: RE: [Sipping] Service
> > Identification in SIP
> >
> > My humble, perhaps faulty, understanding of the "chess game"
> > was not that voice commands were used to control the chess
> > itself, rather
> > that the chess game, while having some IP-based proprietary
> > gaming interaction protocol, also had a feature letting
> > the users talk to each others while playing. The scenario are
> > two, or more, users play a chess game of the (Inter)net, talkin
> > to eachother from time to time during the game (yeah, don't
> > ask me if I think people will love this!!).
> >
> > The purpose, as I understand it, was a)to show that the
> > 'voice component' in the chess game application is similar to that
> > in a stand alone POC-Only application when it comes to
> > exchange and transfer of voice samples and associated floor control
> > and b) to point to differences, for instance when it comes to
> > supplementary services, which may not apply in the chess game case.
> >
> > Perhaps it is important to be clear on the assumed user
> > device realisation, as far as I understand it:
> >
> >  The "chess game + voice" example is, as I understand it,
> > discussed under the assumption that voice and chess game is
> >  managed by custome designed appliction, or client, which is
> > a necessary/likely approach in a mobile device given the
> >  limited MMI (and also other factors in the mobile context).
> >
> >  In a PC, I could easily handle this as two different
> > applications/client, like when I use Skype/MSN to talk to my
> > friends while
> >  playing some turn based game, but again, the discussion is
> > held under the assumption that usability requirements
> >  leads to an application client integrating voice and chess
> > game events.
> >
> > The alternative device realisation You describe- different
> > clients managing different portions of what is negotiated in
> > the SDP or MSRP MIME attachment-
> > is, I think, a good illustration of the need to be precise
> > and explicit about the assumed device architecture when
> > discussing the "service identification" issue(s).
> >
> > Regards
> > G=F6ran
> >
> >
> >
> >
> > -----Original Message-----
> > From: Elwell, John [mailto:john.elwell@siemens.com]
> > Sent: den 14 februari 2007 11:57
> > To: Henry Sinnreich; sipping@ietf.org
> > Subject: RE: [Sipping] Service Identification in SIP
> >
> > Having followed this thread, I too find the chess example
> > very contrived. If chess moves are sent as voice commands,
> > then surely it would be acceptable for the normal telephone
> > application to answer the call, i.e., the remote player
> > listens to the move, performs this move on a conventional or
> > electronic chess board and speaks his next move. The calling
> > UA should not care whether the call is answered by a computer
> > application or a person.
> > Furthermore, even if the audio needs to consumed by a
> > particular chess application at the called UA, then other
> > media in the same session such as video or messaging might
> > not be. So any qualification of the nature of media to be
> > exchanged probably should be at the media level, not at the
> > session level or SIP level. As yet another variation, what if
> > the players set up a conventional telephone call and then
> > mutually agree to switch in their chess applications?
> >
> > Now, let's consider another case. I establish an MSRP
> > session. Within this MSRP session I can send messages with
> > various MIME attachments that are intended for consumption by
> > various applications. I should not need to know in advance
> > what applications I will use during the course of the MSRP
> > session. All I should put in the INVITE request is the fact
> > that I want to establish an MSRP session.
> >
> > John
> >
> > > -----Original Message-----
> > > From: Henry Sinnreich [mailto:hsinnrei@adobe.com]
> > > Sent: 13 February 2007 18:19
> > > To: sipping@ietf.org
> > > Subject: RE: [Sipping] Service Identification in SIP
> > >
> > > It must be my lack of imagination for not being able to
> > come up with a
> > > credible usage scenario, given the following:
> > >
> > > - Is playing chess by voice a credible usage scenario? :-)
> > Especially
> > > since voice recognition is now so often found in PCs and in mobile
> > > phones as well. In other words, chess playing by voice may
> > also be an
> > > endpoint application, without any network involvement. Actually, if
> > > placed in the endpoint it works just as well on poor
> > bandwidth links
> > > by contrast to chess in the network that may require adequate
> > > broadband.
> > >
> > > - Is anyone's boss or spouse going to approve extra phone charges
> > > above the flat rate for playing chess at a time when the gaming
> > > industry has pioneered free VoIP?
> > > Will such a service be taxed extra and if yes using by what
> > criteria?
> > > Gambling comes to my mind :-)
> > >
> > > - VoIP is reported in the press as a give-away and I have seen no
> > > reports of any VoIP provider recording a profit on VoIP. I
> > assume the
> > > losses are often hidden by bundling in triple play, etc.
> > >
> > > Given my lack of imagination, can anyone share a more
> > credible usage
> > > scenario for Service Identification in SIP?
> > >
> > > What's the use of a standard without a credible usage scenario?
> > >
> > > Thanks, Henry
> > >
> > >
> > > _______________________________________________
> > > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > > This list is for NEW development of the application of SIP Use
> > > sip-implementors@cs.columbia.edu for questions on current sip Use
> > > sip@ietf.org for new developments of core SIP
> > >
> >
> > _______________________________________________
> > Sipping mailing list   https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP
> > Use sip-implementors@cs.columbia.edu for questions on current
> > sip Use sip@ietf.org for new developments of core SIP
> >
>
> _______________________________________________
> Sipping mailing list   https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
>

------=_Part_165786_23985264.1171536841327
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

<div>Inorder to identify a endpoint that has the capabilities to serve a ve=
ry specialized functionality, I guess it is simpler to name the functionali=
ty. Naming the functionality not only helps in decision making while routin=
g the request to that specific endpoint but also helps the endpoint identif=
y which specialized application on the endpoint can take up the responsibil=
ity. Instead of all the routing entities looking into the SDP/message bodie=
s it should be easier/faster if they can look for a specific header.
</div>
<div>I thought feature tag mechanisms indicated in the RFC3841/3840 combina=
tion should have sufficed but implementing those RFCs seems to be a overkil=
l for describing/selecting/routing requests&nbsp;to the specific endpoint a=
nd also choosing a service/feature/application on the endpoint.
</div>
<div>&nbsp;</div>
<div>-Mallesh</div>
<div>&nbsp;</div>
<div>&nbsp;</div>
<div>&nbsp;</div>
<div><span class=3D"gmail_quote">On 2/15/07, <b class=3D"gmail_sendername">=
Elwell, John</b> &lt;<a onclick=3D"return top.js.OpenExtLink(window,event,t=
his)" href=3D"mailto:john.elwell@siemens.com" target=3D"_blank">john.elwell=
@siemens.com
</a>&gt; wrote:</span>=20
<blockquote class=3D"gmail_quote" style=3D"PADDING-LEFT: 1ex; MARGIN: 0px 0=
px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">G=F6ran,<br><br>Your interpretat=
ion is different from what is stated in the Rosenberg draft.<br>Anyway, let=
&#39;s not dwell on this particular use case, which may or may not be=20
<br>realistic. The important thing is whether the media being sent is suita=
ble<br>for general purpose devices or requires or could be used by some spe=
cial<br>class of application.<br><br>If a medium is somehow specialised (=
=20
e.g., audio modulated for interpretation<br>by a modem or fax machine), the=
n the necessary information needs to be<br>conveyed along with the m-line i=
n the SDP. Perhaps even if it is<br>human-intelligible voice but with a res=
tricted repertoire of words that make=20
<br>it suitable for a particular class of application, maybe that informati=
on<br>too should be conveyed with the m-line. Perhaps the case described in=
<br>Jonathan&#39;s draft fits this latter category: I believe the medium is=
 just=20
<br>regular voice that can be interpreted by a human, but is also constrain=
ed in<br>its repertoire of words such that it can be used by a chess applic=
ation.<br><br>The important thing is that there be sufficient information i=
n the SDP to=20
<br>allow appropriate application selection at the end device. I am not<br>=
convinced by a need for information at the SIP level during call<br>establi=
shment, because of course SDP can be re-negotiated during the call<br>(=20
e.g., when the users decide to use a particular application) and what is<br=
>indicated in SIP at call establishment time might no longer be valid later=
<br>in the call.<br><br>John<br><br><br>&gt; -----Original Message-----=20
<br>&gt; From: G=F6ran Eriksson AP (KI/EAB)<br>&gt; [mailto:<a onclick=3D"r=
eturn top.js.OpenExtLink(window,event,this)" href=3D"mailto:goran.ap.erikss=
on@ericsson.com" target=3D"_blank">goran.ap.eriksson@ericsson.com</a>]<br>&=
gt; Sent: 14 February 2007 23:34
<br>&gt; To: Elwell, John<br>&gt; Cc: <a onclick=3D"return top.js.OpenExtLi=
nk(window,event,this)" href=3D"mailto:sipping@ietf.org" target=3D"_blank">s=
ipping@ietf.org</a><br>&gt; Subject: Device architecture? WAS: RE: [Sipping=
] Service
<br>&gt; Identification in SIP<br>&gt;<br>&gt; My humble, perhaps faulty, u=
nderstanding of the &quot;chess game&quot;<br>&gt; was not that voice comma=
nds were used to control the chess <br>&gt; itself, rather<br>&gt; that the=
 chess game, while having some IP-based proprietary
<br>&gt; gaming interaction protocol, also had a feature letting<br>&gt; th=
e users talk to each others while playing. The scenario are <br>&gt; two, o=
r more, users play a chess game of the (Inter)net, talkin<br>&gt; to eachot=
her from time to time during the game (yeah, don&#39;t
<br>&gt; ask me if I think people will love this!!).<br>&gt;<br>&gt; The pu=
rpose, as I understand it, was a)to show that the <br>&gt; &#39;voice compo=
nent&#39; in the chess game application is similar to that<br>&gt; in a sta=
nd alone POC-Only application when it comes to
<br>&gt; exchange and transfer of voice samples and associated floor contro=
l <br>&gt; and b) to point to differences, for instance when it comes to<br=
>&gt; supplementary services, which may not apply in the chess game case.
<br>&gt;<br>&gt; Perhaps it is important to be clear on the assumed user<br=
>&gt; device realisation, as far as I understand it:<br>&gt;<br>&gt;&nbsp;&=
nbsp;The &quot;chess game + voice&quot; example is, as I understand it,<br>=
&gt; discussed under the assumption that voice and chess game is
<br>&gt;&nbsp;&nbsp;managed by custome designed appliction, or client, whic=
h is <br>&gt; a necessary/likely approach in a mobile device given the<br>&=
gt;&nbsp;&nbsp;limited MMI (and also other factors in the mobile context).<=
br>&gt;<br>&gt;&nbsp;&nbsp;In a PC, I could easily handle this as two diffe=
rent
<br>&gt; applications/client, like when I use Skype/MSN to talk to my <br>&=
gt; friends while<br>&gt;&nbsp;&nbsp;playing some turn based game, but agai=
n, the discussion is<br>&gt; held under the assumption that usability requi=
rements
<br>&gt;&nbsp;&nbsp;leads to an application client integrating voice and ch=
ess<br>&gt; game events.<br>&gt;<br>&gt; The alternative device realisation=
 You describe- different<br>&gt; clients managing different portions of wha=
t is negotiated in
<br>&gt; the SDP or MSRP MIME attachment-<br>&gt; is, I think, a good illus=
tration of the need to be precise <br>&gt; and explicit about the assumed d=
evice architecture when<br>&gt; discussing the &quot;service identification=
&quot; issue(s).
<br>&gt;<br>&gt; Regards<br>&gt; G=F6ran<br>&gt;<br>&gt;<br>&gt;<br>&gt;<br=
>&gt; -----Original Message----- <br>&gt; From: Elwell, John [mailto:<a onc=
lick=3D"return top.js.OpenExtLink(window,event,this)" href=3D"mailto:john.e=
lwell@siemens.com" target=3D"_blank">
john.elwell@siemens.com</a>]<br>&gt; Sent: den 14 februari 2007 11:57<br>&g=
t; To: Henry Sinnreich; <a onclick=3D"return top.js.OpenExtLink(window,even=
t,this)" href=3D"mailto:sipping@ietf.org" target=3D"_blank">sipping@ietf.or=
g=20
</a><br>&gt; Subject: RE: [Sipping] Service Identification in SIP<br>&gt;<b=
r>&gt; Having followed this thread, I too find the chess example<br>&gt; ve=
ry contrived. If chess moves are sent as voice commands,<br>&gt; then surel=
y it would be acceptable for the normal telephone=20
<br>&gt; application to answer the call, i.e., the remote player<br>&gt; li=
stens to the move, performs this move on a conventional or<br>&gt; electron=
ic chess board and speaks his next move. The calling<br>&gt; UA should not =
care whether the call is answered by a computer=20
<br>&gt; application or a person.<br>&gt; Furthermore, even if the audio ne=
eds to consumed by a<br>&gt; particular chess application at the called UA,=
 then other<br>&gt; media in the same session such as video or messaging mi=
ght=20
<br>&gt; not be. So any qualification of the nature of media to be<br>&gt; =
exchanged probably should be at the media level, not at the<br>&gt; session=
 level or SIP level. As yet another variation, what if<br>&gt; the players =
set up a conventional telephone call and then=20
<br>&gt; mutually agree to switch in their chess applications?<br>&gt;<br>&=
gt; Now, let&#39;s consider another case. I establish an MSRP<br>&gt; sessi=
on. Within this MSRP session I can send messages with<br>&gt; various MIME =
attachments that are intended for consumption by=20
<br>&gt; various applications. I should not need to know in advance<br>&gt;=
 what applications I will use during the course of the MSRP<br>&gt; session=
. All I should put in the INVITE request is the fact<br>&gt; that I want to=
 establish an MSRP session.=20
<br>&gt;<br>&gt; John<br>&gt;<br>&gt; &gt; -----Original Message-----<br>&g=
t; &gt; From: Henry Sinnreich [mailto:<a onclick=3D"return top.js.OpenExtLi=
nk(window,event,this)" href=3D"mailto:hsinnrei@adobe.com" target=3D"_blank"=
>hsinnrei@adobe.com
</a>]<br>&gt; &gt; Sent: 13 February 2007 18:19<br>&gt; &gt; To: <a onclick=
=3D"return top.js.OpenExtLink(window,event,this)" href=3D"mailto:sipping@ie=
tf.org" target=3D"_blank">sipping@ietf.org</a><br>&gt; &gt; Subject: RE: [S=
ipping] Service Identification in SIP
<br>&gt; &gt;<br>&gt; &gt; It must be my lack of imagination for not being =
able to<br>&gt; come up with a <br>&gt; &gt; credible usage scenario, given=
 the following:<br>&gt; &gt;<br>&gt; &gt; - Is playing chess by voice a cre=
dible usage scenario? :-)
<br>&gt; Especially<br>&gt; &gt; since voice recognition is now so often fo=
und in PCs and in mobile <br>&gt; &gt; phones as well. In other words, ches=
s playing by voice may<br>&gt; also be an<br>&gt; &gt; endpoint application=
, without any network involvement. Actually, if
<br>&gt; &gt; placed in the endpoint it works just as well on poor <br>&gt;=
 bandwidth links<br>&gt; &gt; by contrast to chess in the network that may =
require adequate<br>&gt; &gt; broadband.<br>&gt; &gt;<br>&gt; &gt; - Is any=
one&#39;s boss or spouse going to approve extra phone charges
<br>&gt; &gt; above the flat rate for playing chess at a time when the gami=
ng<br>&gt; &gt; industry has pioneered free VoIP?<br>&gt; &gt; Will such a =
service be taxed extra and if yes using by what<br>&gt; criteria?<br>&gt; &=
gt; Gambling comes to my mind :-)=20
<br>&gt; &gt;<br>&gt; &gt; - VoIP is reported in the press as a give-away a=
nd I have seen no<br>&gt; &gt; reports of any VoIP provider recording a pro=
fit on VoIP. I<br>&gt; assume the<br>&gt; &gt; losses are often hidden by b=
undling in triple play, etc.=20
<br>&gt; &gt;<br>&gt; &gt; Given my lack of imagination, can anyone share a=
 more<br>&gt; credible usage<br>&gt; &gt; scenario for Service Identificati=
on in SIP?<br>&gt; &gt;<br>&gt; &gt; What&#39;s the use of a standard witho=
ut a credible usage scenario?=20
<br>&gt; &gt;<br>&gt; &gt; Thanks, Henry<br>&gt; &gt;<br>&gt; &gt;<br>&gt; =
&gt; _______________________________________________<br>&gt; &gt; Sipping m=
ailing list&nbsp;&nbsp;<a onclick=3D"return top.js.OpenExtLink(window,event=
,this)" href=3D"https://www1.ietf.org/mailman/listinfo/sipping" target=3D"_=
blank">
https://www1.ietf.org/mailman/listinfo/sipping </a><br>&gt; &gt; This list =
is for NEW development of the application of SIP Use<br>&gt; &gt; <a onclic=
k=3D"return top.js.OpenExtLink(window,event,this)" href=3D"mailto:sip-imple=
mentors@cs.columbia.edu" target=3D"_blank">
sip-implementors@cs.columbia.edu</a> for questions on current sip Use<br>&g=
t; &gt; <a onclick=3D"return top.js.OpenExtLink(window,event,this)" href=3D=
"mailto:sip@ietf.org" target=3D"_blank">sip@ietf.org</a> for new developmen=
ts of core SIP
<br>&gt; &gt;<br>&gt;<br>&gt; _____________________________________________=
__<br>&gt; Sipping mailing list&nbsp;&nbsp;<a onclick=3D"return top.js.Open=
ExtLink(window,event,this)" href=3D"https://www1.ietf.org/mailman/listinfo/=
sipping" target=3D"_blank">
 https://www1.ietf.org/mailman/listinfo/sipping</a><br>&gt; This list is fo=
r NEW development of the application of SIP<br>&gt; Use <a onclick=3D"retur=
n top.js.OpenExtLink(window,event,this)" href=3D"mailto:sip-implementors@cs=
.columbia.edu" target=3D"_blank">
sip-implementors@cs.columbia.edu</a> for questions on current <br>&gt; sip =
Use <a onclick=3D"return top.js.OpenExtLink(window,event,this)" href=3D"mai=
lto:sip@ietf.org" target=3D"_blank">sip@ietf.org</a> for new developments o=
f core SIP
<br>&gt;<br><br>_______________________________________________<br>Sipping =
mailing list&nbsp;&nbsp;<a onclick=3D"return top.js.OpenExtLink(window,even=
t,this)" href=3D"https://www1.ietf.org/mailman/listinfo/sipping" target=3D"=
_blank"> https://www1.ietf.org/mailman/listinfo/sipping
</a><br>This list is for NEW development of the application of SIP<br>Use <=
a onclick=3D"return top.js.OpenExtLink(window,event,this)" href=3D"mailto:s=
ip-implementors@cs.columbia.edu" target=3D"_blank">sip-implementors@cs.colu=
mbia.edu
</a> for questions on current sip <br>Use <a onclick=3D"return top.js.OpenE=
xtLink(window,event,this)" href=3D"mailto:sip@ietf.org" target=3D"_blank">s=
ip@ietf.org</a> for new developments of core SIP<br></blockquote></div><br>

------=_Part_165786_23985264.1171536841327--


--===============0590673881==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============0590673881==--




From sipping-bounces@ietf.org Thu Feb 15 06:17:14 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHeaY-0001Ds-Ak; Thu, 15 Feb 2007 06:15:46 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHeaW-0001Dh-C3
	for sipping@ietf.org; Thu, 15 Feb 2007 06:15:44 -0500
Received: from mailgate.siemenscomms.co.uk ([195.171.110.225]
	helo=bemg01.siemenscomms.co.uk)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHeaV-0005DY-50
	for sipping@ietf.org; Thu, 15 Feb 2007 06:15:44 -0500
Received: from ntht207e.uksgcs.siemenscomms.co.uk ([137.223.247.82])
	by siemenscomms.co.uk (PMDF V6.0-24 #40642)
	with ESMTP id <0JDI0097L4M02C@siemenscomms.co.uk> for sipping@ietf.org;
	Thu, 15 Feb 2007 11:15:36 +0000 (GMT)
Received: by ntht207e.uksgcs.siemenscomms.co.uk with Internet Mail Service
	(5.5.2657.72)	id <D2K3YPAD>; Thu, 15 Feb 2007 11:15:31 +0000
Content-return: allowed
Date: Thu, 15 Feb 2007 11:15:30 +0000
From: "Elwell, John" <john.elwell@siemens.com>
Subject: RE: Device architecture? WAS: RE: [Sipping] Service Identificatio	n in
	SIP
To: Malleswara Rao Ankem <malleshavn@gmail.com>
Message-id: <50B1CBA96870A34799A506B2313F26670B0A9615@ntht201e.siemenscomms.co.uk>
MIME-version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
X-Spam-Score: 0.1 (/)
X-Scan-Signature: f45599941caef2d9ac4ed98df73589d5
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1235688201=="
Errors-To: sipping-bounces@ietf.org

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

--===============1235688201==
Content-return: allowed
Content-type: multipart/alternative;
	boundary="Boundary_(ID_5zQmlkOS6FkRuxAeQ41q8g)"

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

--Boundary_(ID_5zQmlkOS6FkRuxAeQ41q8g)
Content-type: text/plain; charset=iso-8859-1
Content-transfer-encoding: quoted-printable

Mallesh,
=20
Yes, I agree that some means of selecting a preferred device might be
needed, but that is indeed what RFC3841/3840 does. Why do you consider =
this
overkill?
=20
John


  _____ =20

From: Malleswara Rao Ankem [mailto:malleshavn@gmail.com]=20
Sent: 15 February 2007 10:54
To: Elwell, John
Cc: G=F6ran Eriksson AP (KI/EAB); sipping@ietf.org
Subject: Re: Device architecture? WAS: RE: [Sipping] Service =
Identificatio n
in SIP


Inorder to identify a endpoint that has the capabilities to serve a =
very
specialized functionality, I guess it is simpler to name the =
functionality.
Naming the functionality not only helps in decision making while =
routing the
request to that specific endpoint but also helps the endpoint identify =
which
specialized application on the endpoint can take up the responsibility.
Instead of all the routing entities looking into the SDP/message bodies =
it
should be easier/faster if they can look for a specific header.=20
I thought feature tag mechanisms indicated in the RFC3841/3840 =
combination
should have sufficed but implementing those RFCs seems to be a overkill =
for
describing/selecting/routing requests to the specific endpoint and also
choosing a service/feature/application on the endpoint.=20
=20
-Mallesh
=20
=20
=20
On 2/15/07, Elwell, John <john.elwell@siemens.com
<mailto:john.elwell@siemens.com> > wrote:=20

G=F6ran,

Your interpretation is different from what is stated in the Rosenberg =
draft.
Anyway, let's not dwell on this particular use case, which may or may =
not be

realistic. The important thing is whether the media being sent is =
suitable
for general purpose devices or requires or could be used by some =
special
class of application.

If a medium is somehow specialised ( e.g., audio modulated for
interpretation
by a modem or fax machine), then the necessary information needs to be
conveyed along with the m-line in the SDP. Perhaps even if it is
human-intelligible voice but with a restricted repertoire of words that =
make

it suitable for a particular class of application, maybe that =
information
too should be conveyed with the m-line. Perhaps the case described in
Jonathan's draft fits this latter category: I believe the medium is =
just=20
regular voice that can be interpreted by a human, but is also =
constrained in
its repertoire of words such that it can be used by a chess =
application.

The important thing is that there be sufficient information in the SDP =
to=20
allow appropriate application selection at the end device. I am not
convinced by a need for information at the SIP level during call
establishment, because of course SDP can be re-negotiated during the =
call
( e.g., when the users decide to use a particular application) and what =
is
indicated in SIP at call establishment time might no longer be valid =
later
in the call.

John


> -----Original Message-----=20
> From: G=F6ran Eriksson AP (KI/EAB)
> [mailto:goran.ap.eriksson@ericsson.com
<mailto:goran.ap.eriksson@ericsson.com> ]
> Sent: 14 February 2007 23:34=20
> To: Elwell, John
> Cc: sipping@ietf.org <mailto:sipping@ietf.org>=20
> Subject: Device architecture? WAS: RE: [Sipping] Service=20
> Identification in SIP
>
> My humble, perhaps faulty, understanding of the "chess game"
> was not that voice commands were used to control the chess=20
> itself, rather
> that the chess game, while having some IP-based proprietary=20
> gaming interaction protocol, also had a feature letting
> the users talk to each others while playing. The scenario are=20
> two, or more, users play a chess game of the (Inter)net, talkin
> to eachother from time to time during the game (yeah, don't=20
> ask me if I think people will love this!!).
>
> The purpose, as I understand it, was a)to show that the=20
> 'voice component' in the chess game application is similar to that
> in a stand alone POC-Only application when it comes to=20
> exchange and transfer of voice samples and associated floor control=20
> and b) to point to differences, for instance when it comes to
> supplementary services, which may not apply in the chess game case.=20
>
> Perhaps it is important to be clear on the assumed user
> device realisation, as far as I understand it:
>
>  The "chess game + voice" example is, as I understand it,
> discussed under the assumption that voice and chess game is=20
>  managed by custome designed appliction, or client, which is=20
> a necessary/likely approach in a mobile device given the
>  limited MMI (and also other factors in the mobile context).
>
>  In a PC, I could easily handle this as two different=20
> applications/client, like when I use Skype/MSN to talk to my=20
> friends while
>  playing some turn based game, but again, the discussion is
> held under the assumption that usability requirements=20
>  leads to an application client integrating voice and chess
> game events.
>
> The alternative device realisation You describe- different
> clients managing different portions of what is negotiated in=20
> the SDP or MSRP MIME attachment-
> is, I think, a good illustration of the need to be precise=20
> and explicit about the assumed device architecture when
> discussing the "service identification" issue(s).=20
>
> Regards
> G=F6ran
>
>
>
>
> -----Original Message-----=20
> From: Elwell, John [mailto:  <mailto:john.elwell@siemens.com>
john.elwell@siemens.com]
> Sent: den 14 februari 2007 11:57
> To: Henry Sinnreich; sipping@ietf.org  <mailto:sipping@ietf.org>=20
> Subject: RE: [Sipping] Service Identification in SIP
>
> Having followed this thread, I too find the chess example
> very contrived. If chess moves are sent as voice commands,
> then surely it would be acceptable for the normal telephone=20
> application to answer the call, i.e., the remote player
> listens to the move, performs this move on a conventional or
> electronic chess board and speaks his next move. The calling
> UA should not care whether the call is answered by a computer=20
> application or a person.
> Furthermore, even if the audio needs to consumed by a
> particular chess application at the called UA, then other
> media in the same session such as video or messaging might=20
> not be. So any qualification of the nature of media to be
> exchanged probably should be at the media level, not at the
> session level or SIP level. As yet another variation, what if
> the players set up a conventional telephone call and then=20
> mutually agree to switch in their chess applications?
>
> Now, let's consider another case. I establish an MSRP
> session. Within this MSRP session I can send messages with
> various MIME attachments that are intended for consumption by=20
> various applications. I should not need to know in advance
> what applications I will use during the course of the MSRP
> session. All I should put in the INVITE request is the fact
> that I want to establish an MSRP session.=20
>
> John
>
> > -----Original Message-----
> > From: Henry Sinnreich [mailto:hsinnrei@adobe.com
<mailto:hsinnrei@adobe.com> ]
> > Sent: 13 February 2007 18:19
> > To: sipping@ietf.org <mailto:sipping@ietf.org>=20
> > Subject: RE: [Sipping] Service Identification in SIP=20
> >
> > It must be my lack of imagination for not being able to
> come up with a=20
> > credible usage scenario, given the following:
> >
> > - Is playing chess by voice a credible usage scenario? :-)=20
> Especially
> > since voice recognition is now so often found in PCs and in mobile=20
> > phones as well. In other words, chess playing by voice may
> also be an
> > endpoint application, without any network involvement. Actually, if =

> > placed in the endpoint it works just as well on poor=20
> bandwidth links
> > by contrast to chess in the network that may require adequate
> > broadband.
> >
> > - Is anyone's boss or spouse going to approve extra phone charges=20
> > above the flat rate for playing chess at a time when the gaming
> > industry has pioneered free VoIP?
> > Will such a service be taxed extra and if yes using by what
> criteria?
> > Gambling comes to my mind :-)=20
> >
> > - VoIP is reported in the press as a give-away and I have seen no
> > reports of any VoIP provider recording a profit on VoIP. I
> assume the
> > losses are often hidden by bundling in triple play, etc.=20
> >
> > Given my lack of imagination, can anyone share a more
> credible usage
> > scenario for Service Identification in SIP?
> >
> > What's the use of a standard without a credible usage scenario?=20
> >
> > Thanks, Henry
> >
> >
> > _______________________________________________
> > Sipping mailing list    =
<https://www1.ietf.org/mailman/listinfo/sipping>
https://www1.ietf.org/mailman/listinfo/sipping=20
> > This list is for NEW development of the application of SIP Use
> > sip-implementors@cs.columbia.edu
<mailto:sip-implementors@cs.columbia.edu>  for questions on current sip =
Use
> > sip@ietf.org <mailto:sip@ietf.org>  for new developments of core =
SIP=20
> >
>
> _______________________________________________
> Sipping mailing list    =
<https://www1.ietf.org/mailman/listinfo/sipping>
https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu
<mailto:sip-implementors@cs.columbia.edu>  for questions on current=20
> sip Use sip@ietf.org <mailto:sip@ietf.org>  for new developments of =
core
SIP=20
>

_______________________________________________
Sipping mailing list    =
<https://www1.ietf.org/mailman/listinfo/sipping>
https://www1.ietf.org/mailman/listinfo/sipping=20
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu
<mailto:sip-implementors@cs.columbia.edu> for questions on current sip=20
Use sip@ietf.org <mailto:sip@ietf.org>  for new developments of core =
SIP




--Boundary_(ID_5zQmlkOS6FkRuxAeQ41q8g)
Content-type: text/html; charset=iso-8859-1
Content-transfer-encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3Diso-8859-1">


<META content=3D"MSHTML 6.00.2900.3020" name=3DGENERATOR></HEAD>
<BODY>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>Mallesh,</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>Yes, I agree that some means of selecting a =
preferred=20
device might be needed, but that is indeed what RFC3841/3840 does. Why =
do you=20
consider this overkill?</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>John</FONT></SPAN></DIV><BR>
<BLOCKQUOTE=20
style=3D"PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #0000ff 2px =
solid; MARGIN-RIGHT: 0px">
  <DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft>
  <HR tabIndex=3D-1>
  <FONT face=3DTahoma size=3D2><B>From:</B> Malleswara Rao Ankem=20
  [mailto:malleshavn@gmail.com] <BR><B>Sent:</B> 15 February 2007=20
  10:54<BR><B>To:</B> Elwell, John<BR><B>Cc:</B> G=F6ran Eriksson AP =
(KI/EAB);=20
  sipping@ietf.org<BR><B>Subject:</B> Re: Device architecture? WAS: RE: =

  [Sipping] Service Identificatio n in SIP<BR></FONT><BR></DIV>
  <DIV></DIV>
  <DIV>Inorder to identify a endpoint that has the capabilities to =
serve a very=20
  specialized functionality, I guess it is simpler to name the =
functionality.=20
  Naming the functionality not only helps in decision making while =
routing the=20
  request to that specific endpoint but also helps the endpoint =
identify which=20
  specialized application on the endpoint can take up the =
responsibility.=20
  Instead of all the routing entities looking into the SDP/message =
bodies it=20
  should be easier/faster if they can look for a specific header. =
</DIV>
  <DIV>I thought feature tag mechanisms indicated in the RFC3841/3840=20
  combination should have sufficed but implementing those RFCs seems to =
be a=20
  overkill for describing/selecting/routing requests&nbsp;to the =
specific=20
  endpoint and also choosing a service/feature/application on the =
endpoint.=20
  </DIV>
  <DIV>&nbsp;</DIV>
  <DIV>-Mallesh</DIV>
  <DIV>&nbsp;</DIV>
  <DIV>&nbsp;</DIV>
  <DIV>&nbsp;</DIV>
  <DIV><SPAN class=3Dgmail_quote>On 2/15/07, <B =
class=3Dgmail_sendername>Elwell,=20
  John</B> &lt;<A onclick=3D"return =
top.js.OpenExtLink(window,event,this)"=20
  href=3D"mailto:john.elwell@siemens.com" =
target=3D_blank>john.elwell@siemens.com=20
  </A>&gt; wrote:</SPAN>=20
  <BLOCKQUOTE class=3Dgmail_quote=20
  style=3D"PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: =
#ccc 1px solid">G=F6ran,<BR><BR>Your=20
    interpretation is different from what is stated in the Rosenberg=20
    draft.<BR>Anyway, let's not dwell on this particular use case, =
which may or=20
    may not be <BR>realistic. The important thing is whether the media =
being=20
    sent is suitable<BR>for general purpose devices or requires or =
could be used=20
    by some special<BR>class of application.<BR><BR>If a medium is =
somehow=20
    specialised ( e.g., audio modulated for interpretation<BR>by a =
modem or fax=20
    machine), then the necessary information needs to be<BR>conveyed =
along with=20
    the m-line in the SDP. Perhaps even if it is<BR>human-intelligible =
voice but=20
    with a restricted repertoire of words that make <BR>it suitable for =
a=20
    particular class of application, maybe that information<BR>too =
should be=20
    conveyed with the m-line. Perhaps the case described =
in<BR>Jonathan's draft=20
    fits this latter category: I believe the medium is just <BR>regular =
voice=20
    that can be interpreted by a human, but is also constrained =
in<BR>its=20
    repertoire of words such that it can be used by a chess=20
    application.<BR><BR>The important thing is that there be sufficient =

    information in the SDP to <BR>allow appropriate application =
selection at the=20
    end device. I am not<BR>convinced by a need for information at the =
SIP level=20
    during call<BR>establishment, because of course SDP can be =
re-negotiated=20
    during the call<BR>( e.g., when the users decide to use a =
particular=20
    application) and what is<BR>indicated in SIP at call establishment =
time=20
    might no longer be valid later<BR>in the =
call.<BR><BR>John<BR><BR><BR>&gt;=20
    -----Original Message----- <BR>&gt; From: G=F6ran Eriksson AP =
(KI/EAB)<BR>&gt;=20
    [mailto:<A onclick=3D"return top.js.OpenExtLink(window,event,this)" =

    href=3D"mailto:goran.ap.eriksson@ericsson.com"=20
    target=3D_blank>goran.ap.eriksson@ericsson.com</A>]<BR>&gt; Sent: =
14 February=20
    2007 23:34 <BR>&gt; To: Elwell, John<BR>&gt; Cc: <A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sipping@ietf.org" =
target=3D_blank>sipping@ietf.org</A><BR>&gt;=20
    Subject: Device architecture? WAS: RE: [Sipping] Service <BR>&gt;=20
    Identification in SIP<BR>&gt;<BR>&gt; My humble, perhaps faulty,=20
    understanding of the "chess game"<BR>&gt; was not that voice =
commands were=20
    used to control the chess <BR>&gt; itself, rather<BR>&gt; that the =
chess=20
    game, while having some IP-based proprietary <BR>&gt; gaming =
interaction=20
    protocol, also had a feature letting<BR>&gt; the users talk to each =
others=20
    while playing. The scenario are <BR>&gt; two, or more, users play a =
chess=20
    game of the (Inter)net, talkin<BR>&gt; to eachother from time to =
time during=20
    the game (yeah, don't <BR>&gt; ask me if I think people will love=20
    this!!).<BR>&gt;<BR>&gt; The purpose, as I understand it, was a)to =
show that=20
    the <BR>&gt; 'voice component' in the chess game application is =
similar to=20
    that<BR>&gt; in a stand alone POC-Only application when it comes to =
<BR>&gt;=20
    exchange and transfer of voice samples and associated floor control =
<BR>&gt;=20
    and b) to point to differences, for instance when it comes =
to<BR>&gt;=20
    supplementary services, which may not apply in the chess game case. =

    <BR>&gt;<BR>&gt; Perhaps it is important to be clear on the assumed =

    user<BR>&gt; device realisation, as far as I understand=20
    it:<BR>&gt;<BR>&gt;&nbsp;&nbsp;The "chess game + voice" example is, =
as I=20
    understand it,<BR>&gt; discussed under the assumption that voice =
and chess=20
    game is <BR>&gt;&nbsp;&nbsp;managed by custome designed appliction, =
or=20
    client, which is <BR>&gt; a necessary/likely approach in a mobile =
device=20
    given the<BR>&gt;&nbsp;&nbsp;limited MMI (and also other factors in =
the=20
    mobile context).<BR>&gt;<BR>&gt;&nbsp;&nbsp;In a PC, I could easily =
handle=20
    this as two different <BR>&gt; applications/client, like when I use =

    Skype/MSN to talk to my <BR>&gt; friends =
while<BR>&gt;&nbsp;&nbsp;playing=20
    some turn based game, but again, the discussion is<BR>&gt; held =
under the=20
    assumption that usability requirements <BR>&gt;&nbsp;&nbsp;leads to =
an=20
    application client integrating voice and chess<BR>&gt; game=20
    events.<BR>&gt;<BR>&gt; The alternative device realisation You =
describe-=20
    different<BR>&gt; clients managing different portions of what is =
negotiated=20
    in <BR>&gt; the SDP or MSRP MIME attachment-<BR>&gt; is, I think, a =
good=20
    illustration of the need to be precise <BR>&gt; and explicit about =
the=20
    assumed device architecture when<BR>&gt; discussing the "service=20
    identification" issue(s). <BR>&gt;<BR>&gt; Regards<BR>&gt;=20
    G=F6ran<BR>&gt;<BR>&gt;<BR>&gt;<BR>&gt;<BR>&gt; -----Original =
Message-----=20
    <BR>&gt; From: Elwell, John [mailto:<A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:john.elwell@siemens.com" target=3D_blank>=20
    john.elwell@siemens.com</A>]<BR>&gt; Sent: den 14 februari 2007=20
    11:57<BR>&gt; To: Henry Sinnreich; <A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sipping@ietf.org" target=3D_blank>sipping@ietf.org =
</A><BR>&gt;=20
    Subject: RE: [Sipping] Service Identification in =
SIP<BR>&gt;<BR>&gt; Having=20
    followed this thread, I too find the chess example<BR>&gt; very =
contrived.=20
    If chess moves are sent as voice commands,<BR>&gt; then surely it =
would be=20
    acceptable for the normal telephone <BR>&gt; application to answer =
the call,=20
    i.e., the remote player<BR>&gt; listens to the move, performs this =
move on a=20
    conventional or<BR>&gt; electronic chess board and speaks his next =
move. The=20
    calling<BR>&gt; UA should not care whether the call is answered by =
a=20
    computer <BR>&gt; application or a person.<BR>&gt; Furthermore, =
even if the=20
    audio needs to consumed by a<BR>&gt; particular chess application =
at the=20
    called UA, then other<BR>&gt; media in the same session such as =
video or=20
    messaging might <BR>&gt; not be. So any qualification of the nature =
of media=20
    to be<BR>&gt; exchanged probably should be at the media level, not =
at=20
    the<BR>&gt; session level or SIP level. As yet another variation, =
what=20
    if<BR>&gt; the players set up a conventional telephone call and =
then=20
    <BR>&gt; mutually agree to switch in their chess=20
    applications?<BR>&gt;<BR>&gt; Now, let's consider another case. I =
establish=20
    an MSRP<BR>&gt; session. Within this MSRP session I can send =
messages=20
    with<BR>&gt; various MIME attachments that are intended for =
consumption by=20
    <BR>&gt; various applications. I should not need to know in =
advance<BR>&gt;=20
    what applications I will use during the course of the MSRP<BR>&gt; =
session.=20
    All I should put in the INVITE request is the fact<BR>&gt; that I =
want to=20
    establish an MSRP session. <BR>&gt;<BR>&gt; John<BR>&gt;<BR>&gt; =
&gt;=20
    -----Original Message-----<BR>&gt; &gt; From: Henry Sinnreich =
[mailto:<A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:hsinnrei@adobe.com" =
target=3D_blank>hsinnrei@adobe.com=20
    </A>]<BR>&gt; &gt; Sent: 13 February 2007 18:19<BR>&gt; &gt; To: <A =

    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sipping@ietf.org" =
target=3D_blank>sipping@ietf.org</A><BR>&gt;=20
    &gt; Subject: RE: [Sipping] Service Identification in SIP <BR>&gt;=20
    &gt;<BR>&gt; &gt; It must be my lack of imagination for not being =
able=20
    to<BR>&gt; come up with a <BR>&gt; &gt; credible usage scenario, =
given the=20
    following:<BR>&gt; &gt;<BR>&gt; &gt; - Is playing chess by voice a =
credible=20
    usage scenario? :-) <BR>&gt; Especially<BR>&gt; &gt; since voice =
recognition=20
    is now so often found in PCs and in mobile <BR>&gt; &gt; phones as =
well. In=20
    other words, chess playing by voice may<BR>&gt; also be an<BR>&gt; =
&gt;=20
    endpoint application, without any network involvement. Actually, if =
<BR>&gt;=20
    &gt; placed in the endpoint it works just as well on poor <BR>&gt; =
bandwidth=20
    links<BR>&gt; &gt; by contrast to chess in the network that may =
require=20
    adequate<BR>&gt; &gt; broadband.<BR>&gt; &gt;<BR>&gt; &gt; - Is =
anyone's=20
    boss or spouse going to approve extra phone charges <BR>&gt; &gt; =
above the=20
    flat rate for playing chess at a time when the gaming<BR>&gt; &gt; =
industry=20
    has pioneered free VoIP?<BR>&gt; &gt; Will such a service be taxed =
extra and=20
    if yes using by what<BR>&gt; criteria?<BR>&gt; &gt; Gambling comes =
to my=20
    mind :-) <BR>&gt; &gt;<BR>&gt; &gt; - VoIP is reported in the press =
as a=20
    give-away and I have seen no<BR>&gt; &gt; reports of any VoIP =
provider=20
    recording a profit on VoIP. I<BR>&gt; assume the<BR>&gt; &gt; =
losses are=20
    often hidden by bundling in triple play, etc. <BR>&gt; &gt;<BR>&gt; =
&gt;=20
    Given my lack of imagination, can anyone share a more<BR>&gt; =
credible=20
    usage<BR>&gt; &gt; scenario for Service Identification in =
SIP?<BR>&gt;=20
    &gt;<BR>&gt; &gt; What's the use of a standard without a credible =
usage=20
    scenario? <BR>&gt; &gt;<BR>&gt; &gt; Thanks, Henry<BR>&gt; =
&gt;<BR>&gt;=20
    &gt;<BR>&gt; &gt; =
_______________________________________________<BR>&gt;=20
    &gt; Sipping mailing list&nbsp;&nbsp;<A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"https://www1.ietf.org/mailman/listinfo/sipping" =
target=3D_blank>=20
    https://www1.ietf.org/mailman/listinfo/sipping </A><BR>&gt; &gt; =
This list=20
    is for NEW development of the application of SIP Use<BR>&gt; &gt; =
<A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sip-implementors@cs.columbia.edu"=20
    target=3D_blank>sip-implementors@cs.columbia.edu</A> for questions =
on current=20
    sip Use<BR>&gt; &gt; <A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sip@ietf.org" target=3D_blank>sip@ietf.org</A> for =
new=20
    developments of core SIP <BR>&gt; &gt;<BR>&gt;<BR>&gt;=20
    _______________________________________________<BR>&gt; Sipping =
mailing=20
    list&nbsp;&nbsp;<A onclick=3D"return =
top.js.OpenExtLink(window,event,this)"=20
    href=3D"https://www1.ietf.org/mailman/listinfo/sipping" =
target=3D_blank>=20
    https://www1.ietf.org/mailman/listinfo/sipping</A><BR>&gt; This =
list is for=20
    NEW development of the application of SIP<BR>&gt; Use <A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sip-implementors@cs.columbia.edu"=20
    target=3D_blank>sip-implementors@cs.columbia.edu</A> for questions =
on current=20
    <BR>&gt; sip Use <A onclick=3D"return =
top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sip@ietf.org" target=3D_blank>sip@ietf.org</A> for =
new=20
    developments of core SIP=20
    =
<BR>&gt;<BR><BR>_______________________________________________<BR>Sippi=
ng=20
    mailing list&nbsp;&nbsp;<A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"https://www1.ietf.org/mailman/listinfo/sipping" =
target=3D_blank>=20
    https://www1.ietf.org/mailman/listinfo/sipping </A><BR>This list is =
for NEW=20
    development of the application of SIP<BR>Use <A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sip-implementors@cs.columbia.edu"=20
    target=3D_blank>sip-implementors@cs.columbia.edu </A>for questions =
on current=20
    sip <BR>Use <A onclick=3D"return =
top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sip@ietf.org" target=3D_blank>sip@ietf.org</A> for =
new=20
    developments of core =
SIP<BR></BLOCKQUOTE></DIV><BR></BLOCKQUOTE></BODY></HTML>

--Boundary_(ID_5zQmlkOS6FkRuxAeQ41q8g)--


--===============1235688201==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1235688201==--




From sipping-bounces@ietf.org Thu Feb 15 06:31:30 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHeo3-0002rR-Ph; Thu, 15 Feb 2007 06:29:43 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHeo0-0002qc-TJ; Thu, 15 Feb 2007 06:29:40 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HHenv-0006lB-0b; Thu, 15 Feb 2007 06:29:40 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	22C1020502; Thu, 15 Feb 2007 12:29:20 +0100 (CET)
X-AuditID: c1b4fb3c-b0fcabb0000007de-fb-45d44410e02d 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	F0F0720495; Thu, 15 Feb 2007 12:29:19 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 15 Feb 2007 12:29:19 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Re: [Sip] New draft on early media
Date: Thu, 15 Feb 2007 12:29:16 +0100
Message-ID: <7374777208BDC7449D5620EF94232567020F1D72@esealmw113.eemea.ericsson.se>
In-Reply-To: <45D38F5C.9050901@bbn.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Re: [Sip] New draft on early media
Thread-Index: AcdQiNjklU3kkfFTTdy2L5b4YWhIEgAariYQ
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Richard Barnes" <rbarnes@bbn.com>,
	"Francois Audet" <audet@nortel.com>
X-OriginalArrivalTime: 15 Feb 2007 11:29:19.0782 (UTC)
	FILETIME=[88A0A460:01C750F4]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 162d87dc0b780d17da9b1934777fd451
Cc: sip@ietf.org, sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,=20

>Thanks for the clarification.  I know that choosing media=20
>streams to render is a big problem, but it's one I'm viewing=20
>as beyond the scope of what our draft is explicitly trying to address.

Fair enough. However, I do agree with Francois that forking is one of
the biggest issue.

>That said, I think that ensuring that media are preceded by=20
>an SDP answer is a big step in enabling better management of=20
>media streams, since it gives the offeror (media recipient)=20
>SDP metadata on which to base its decision.

I agree.

>Once an SDP answer is guaranteed to come before media, the issue can be

>treated more at the SDP layer (e.g. by giving SDP sufficient=20
>expressiveness to support decisions between media streams),=20
>rather than at the SIP layer.

Due to different paths for signalling and media, it may be difficult to
guarantee that the SDP answer reaches the UAC. But, of course, the
sooner the SDP can be sent before the media, the bigger the chance is.

>With respect to forking in particular: You're correct that=20
>even if all streams from forked endpoints are preceded by SDP=20
>answers, the UAC still has to figure out how to deal with=20
>them.  But I think that once you ensure that the offer/answer=20
>finishes before media flows, SRTP keying gets a lot simpler,=20
>since the UAC can just do a DH exchange through your favorite=20
>SDP mechanism with each endpoint.
>=20
>Clipping is certainly something that could be an issue, but I=20
>think it can be prevented.  The solution I think will work in=20
>the end is that an SDP answer will be sent immediately by the=20
>UAS on receipt of the INVITE, before any other action is=20
>taken (e.g., before the INVITE is passed to the TU).  For=20
>example, a PSTN gateway receiving a SIP INVITE would return=20
>its answer before it initiates the outbound PSTN call.

That may not always be feasible/possible, but of course it can be used
as an example.=20

I am not sure we need to go so much into details. I think the most
important thing is to try to provide the SDP answer as soon as possible
before the media is sent - or at latest at the same thing when the media
is sent.

For audio I guess a short clipping may not always be a big issue. But,
for video it may be, because if the initial packages are lost it may
take a relatively long time before the UAC is able to create a correct
picture.

>It seems to me like the negotion of an SDP association should=20
>precede SIP session establishment just like TCP negotiation=20
>would precede initiation of a higher-layer protocol session.

Or the other way around, like H.323 and H.245 ;)

Regards,

Christer







> --Richard
>=20
>=20
>=20
>=20
> Francois Audet wrote:
> > I mean play incoming media.=20
> >=20
> >> -----Original Message-----
> >> From: Richard Barnes [mailto:rbarnes@bbn.com]
> >> Sent: Tuesday, February 13, 2007 11:41
> >> To: Audet, Francois (SC100:3055)
> >> Cc: Christer Holmberg (JO/LMF); sip@ietf.org; sipping@ietf.org
> >> Subject: Re: [Sipping] Re: [Sip] New draft on early media
> >>
> >> Francois,
> >>
> >> I think I have some responses, but I'd like to ask you first:=20
> >> What do you mean by "media cut-through"?
> >>
> >> Thanks,
> >> --Richard
> >>
> >> Francois Audet wrote:
> >>> Hi Richard,
> >>>
> >>> Forcing media cut-through to occur only after an Answer has been=20
> >>> received does solve SOME of the early media problems. As
> >> others have
> >>> mentioned, there are environments that operate like this=20
> already. I=20
> >>> also know many products that do not cut-through media before the=20
> >>> answer is received.
> >>>
> >>> But there are other problems related with early media.=20
> >> Brian Stucker's
> >>> draft listed a bunch of them.
> >>>
> >>> Forking for example comes to mind: even if we did mandate
> >> cut-through
> >>> after answer, we would still have problems related with
> >> forking (i.e.,
> >>> how to pick the stream that needs to be played, and how to
> >> correlate
> >>> the stream with the signaling). How to deal with SRTP key
> >> exchange in
> >>> that scenario is an even bigger problem.
> >>>
> >>> Clipping is a possible side-effect of your proposal. We'd=20
> need some=20
> >>> mechanism to mitigate it, in particular for auto-answer scenarios.
> >>>
> >>>
> >>>> -----Original Message-----
> >>>> From: Richard Barnes [mailto:rbarnes@bbn.com]
> >>>> Sent: Monday, February 12, 2007 07:08
> >>>> To: Christer Holmberg (JO/LMF)
> >>>> Cc: sip@ietf.org; sipping@ietf.org
> >>>> Subject: [Sipping] Re: [Sip] New draft on early media
> >>>>
> >>>> Christer,
> >>>>
> >>>> Thanks for your comments.  Responses inline.
> >>>>
> >>>> Cheers,
> >>>> --Richard
> >>>>
> >>>>> I haven't had time to read the document yet, but I am not
> >>>> sure I agree
> >>>>> with the statement saying that the biggest problem is media
> >>>> arriving
> >>>>> before the SDP answer. In some environments such media will
> >>>> not even
> >>>>> reach the terminal, due to gating etc.
> >>>> Perhaps a better term would be "media sent before an SDP
> >> answer has
> >>>> been received".  As you point out, sending media before an SDP=20
> >>>> offer/answer exchange has completed poses the risk that=20
> that media=20
> >>>> will not arrive.
> >>>> To put it in the terms used in the document, media clipping is a=20
> >>>> problem caused by early (pre-SDP-answer) media, and thus=20
> one which=20
> >>>> this document seeks to solve.
> >>>>
> >>>>> Many implementations DO send an SDP answer (or, at least an
> >>>> unreliable
> >>>>> "preview" of the SDP answer) before (or at the same time
> >> as) early
> >>>>> media is sent.
> >>>> This is the behavior we would like to standardize -- that an SDP=20
> >>>> answer is *always* sent before media of any type.  What
> >> we're trying
> >>>> to do with this document is describe the requirements=20
> for how that=20
> >>>> answer is sent, and consolidate the many ways people have
> >> proposed to
> >>>> send it.
> >>>>
> >>>>> And, in the cases where implementations do NOT send an=20
> SDP answer=20
> >>>>> before they start sending early media - what behavior
> >> would you be
> >>>>> able to standardize?
> >>>> We would prefer that standards not allow this behavior --
> >> that seems
> >>>> like the right thing to me.  What's a little trickier is
> >> that you'd
> >>>> like to not have the answerer sending media before the=20
> answer has=20
> >>>> *arrived*.
> >>>>   When the answer is carried in a reliable message, you=20
> could just=20
> >>>> forbid sending media before receipt has been confirmed.
> >>>> When the transport is unreliable, the situation is a little less=20
> >>>> clear, but I think there's still possibilities for
> >> standardization,
> >>>> e.g., by forbidding answerers to send media until=20
> they've received=20
> >>>> media.  This is discussed in more length in the document.
> >>>>
> >>>>> Regards,
> >>>>>
> >>>>> Christer
> >>>>>
> >>>>>
> >>>>>> -----Original Message-----
> >>>>>> From: Richard Barnes [mailto:rbarnes@bbn.com]
> >>>>>> Sent: 8. helmikuuta 2007 21:47
> >>>>>> To: sip@ietf.org; sipping@ietf.org
> >>>>>> Subject: [Sip] New draft on early media
> >>>>>>
> >>>>>> After the big discussions on SIPPING last summer about
> >> how to fix
> >>>>>> early media, and the reaction at the last couple of
> >> IETFs to Brian
> >>>>>> Stucker's draft on how to cope with it, I've put together
> >>>> a draft to
> >>>>>> try to consolidate all of the discussions and move toward
> >>>> a solution.
> >>>>>> draft-barnes-sip-em-ps-req-sol includes three important parts:
> >>>>>>
> >>>>>> 1. Problem statement: Although there's lots of definitions
> >>>> of "early
> >>>>>> media" in existing documents, I think there's broad
> >> agreement that
> >>>>>> the stuff that causes problems is media that arrives
> >> before an SDP
> >>>>>> offer/answer exchange has completed.
> >>>>>> The problem is therefore to enable the SDP exchange(s) in
> >>>> the INVITE
> >>>>>> transaction to complete more quickly, and make it more
> >>>> difficult (if
> >>>>>> not impossible) to send media before it has completed.  We are=20
> >>>>>> explicitly NOT trying to prevent the sending of media
> >>>> before a 200/OK
> >>>>>> or other closing of an INVITE transaction.
> >>>>>>
> >>>>>> 2. Requirements: Here, we try to create a consolidated list of=20
> >>>>>> requirements for an early media fix.
> >>>>>>
> >>>>>> 3. Analysis of Solutions: Lots of folks have put forward
> >>>> proposals to
> >>>>>> fix the early media problem.  This section tries to
> >>>> enumerate most of
> >>>>>> them and evaluate them against the requirements in section 2.
> >>>>>>
> >>>>>> It hasn't shown up yet on the I-D tracker, so I've
> >> attached a copy.
> >>>>>> --Richard
> >>>>>>
> >>>> _______________________________________________
> >>>> Sipping mailing list
> >> https://www1.ietf.org/mailman/listinfo/sipping
> >>>> This list is for NEW development of the application of SIP Use=20
> >>>> sip-implementors@cs.columbia.edu for questions on=20
> current sip Use=20
> >>>> sip@ietf.org for new developments of core SIP
> >>>>
> >>>
> >>
> >=20
> >=20
>=20
>=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 15 06:32:08 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHepd-0004Jq-O2; Thu, 15 Feb 2007 06:31:21 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHepc-0004Jd-4s
	for sipping@ietf.org; Thu, 15 Feb 2007 06:31:20 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHepZ-0006qs-RO
	for sipping@ietf.org; Thu, 15 Feb 2007 06:31:20 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	4525F203F9; Thu, 15 Feb 2007 12:31:13 +0100 (CET)
X-AuditID: c1b4fb3e-ae6d0bb0000007e1-a3-45d444812658 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	396CA2016C; Thu, 15 Feb 2007 12:31:13 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 15 Feb 2007 12:31:12 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [SIPPING] SDP offer/answer	inearly
	media:draft-sawada-sipping-sip-offeranswer-01.txt
Date: Thu, 15 Feb 2007 12:31:12 +0100
Message-ID: <7374777208BDC7449D5620EF94232567020F1D85@esealmw113.eemea.ericsson.se>
In-Reply-To: <200702142240.l1EMePpr002708@dragon.ariadne.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [SIPPING] SDP offer/answer	inearly
	media:draft-sawada-sipping-sip-offeranswer-01.txt
Thread-Index: AcdQiTcxnI4a5jjpS4ygifv8DmpI4wAa2+eA
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: <Dale.Worley@comcast.net>,
	<sipping@ietf.org>
X-OriginalArrivalTime: 15 Feb 2007 11:31:13.0045 (UTC)
	FILETIME=[CC233450:01C750F4]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: d6b246023072368de71562c0ab503126
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,=20

>[JRE] I don't think 180 means that. It means play ringback=20
>tone until media starts to arrive.
>=20
>It doesn't mean that, either.  It means that the UAS that was=20
>reached is declaring that it is notifying a user.  Whether or=20
>how the UAC wishes to notify its user of this fact is its business.
>=20
>Currently, SIP has no defined means for indicating "Stop=20
>playing the ringback and start playing this media stream."

In TISPAN the usage of P-Early-Media header solves that, at least to
some extent.

Regards,

Christer

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 15 06:34:15 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHerb-0005mb-7w; Thu, 15 Feb 2007 06:33:23 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHera-0005mM-Cx
	for sipping@ietf.org; Thu, 15 Feb 2007 06:33:22 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHerU-000779-Tw
	for sipping@ietf.org; Thu, 15 Feb 2007 06:33:22 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	751BD20FE6; Thu, 15 Feb 2007 12:33:16 +0100 (CET)
X-AuditID: c1b4fb3e-aeed1bb0000007e1-c9-45d444fcb42a 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	65F02201E5; Thu, 15 Feb 2007 12:33:16 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 15 Feb 2007 12:33:16 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [SIPPING] SDP
	offer/answerinearly	media:draft-sawada-sipping-sip-offeranswer-01.txt
Date: Thu, 15 Feb 2007 12:33:15 +0100
Message-ID: <7374777208BDC7449D5620EF94232567020F1D9B@esealmw113.eemea.ericsson.se>
In-Reply-To: <200702142241.l1EMfKWg002796@dragon.ariadne.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [SIPPING] SDP
	offer/answerinearly	media:draft-sawada-sipping-sip-offeranswer-01.txt
Thread-Index: AcdQiWNBlJ7hUL6ZSwaLA+XzsWv1QQAa4i0g
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: <Dale.Worley@comcast.net>,
	<sipping@ietf.org>
X-OriginalArrivalTime: 15 Feb 2007 11:33:16.0175 (UTC)
	FILETIME=[158759F0:01C750F5]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a7d6aff76b15f3f56fcb94490e1052e4
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,

It seems like a number of people now seem to be interested in at least
investigating the possible usage of 19x, so I will try to put together a
draft which we can use as base for the discussion.

Regards,

Christer

> -----Original Message-----
> From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net]=20
> Sent: 15. helmikuuta 2007 0:41
> To: sipping@ietf.org
> Subject: Re: [SIPPING] SDP offer/answerinearly=20
> media:draft-sawada-sipping-sip-offeranswer-01.txt
>=20
>    From: "Christer Holmberg \(JO/LMF\)"=20
> <christer.holmberg@ericsson.com>
>=20
>    I once proposed that it should be possible to terminate an=20
> early dialog
>    with an 19x provisional response.
>=20
>    Then, if "fake forking" is used, the dialog not used=20
> anymore could be
>    terminted and the UAC knows he doesn't have to care about it (and
>    possible media coming on it).
>=20
> That seems like it might be useful, given the number of times=20
> I've seen "fake forking" proposed for dealing with one=20
> situation or another.
>=20
> Dale
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 15 07:09:29 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHfQ4-0004KZ-On; Thu, 15 Feb 2007 07:09:00 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHfQ3-0004KU-KS
	for sipping@ietf.org; Thu, 15 Feb 2007 07:08:59 -0500
Received: from mailgate.siemenscomms.co.uk ([195.171.110.225]
	helo=bemg01.siemenscomms.co.uk)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHfQ1-0003er-0h
	for sipping@ietf.org; Thu, 15 Feb 2007 07:08:59 -0500
Received: from ntht207e.uksgcs.siemenscomms.co.uk ([137.223.247.82])
	by siemenscomms.co.uk (PMDF V6.0-24 #40642)
	with ESMTP id <0JDI00M7472UOR@siemenscomms.co.uk> for sipping@ietf.org;
	Thu, 15 Feb 2007 12:08:54 +0000 (GMT)
Received: by ntht207e.uksgcs.siemenscomms.co.uk with Internet Mail Service
	(5.5.2657.72)	id <D2K3YQ43>; Thu, 15 Feb 2007 12:08:54 +0000
Content-return: allowed
Date: Thu, 15 Feb 2007 12:08:53 +0000
From: "Elwell, John" <john.elwell@siemens.com>
To: Brian Stucker <bstucker@nortel.com>, sipping@ietf.org
Message-id: <50B1CBA96870A34799A506B2313F26670B0A96F0@ntht201e.siemenscomms.co.uk>
MIME-version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
Content-type: text/plain
Content-transfer-encoding: 7BIT
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 4b66a1e94d7d92973ece9e5da449ff80
Cc: 
Subject: [Sipping] RE: I-D
	ACTION:draft-stucker-sipping-early-media-indirection-00.t	xt
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Brian,

See below.

John 

> -----Original Message-----
> From: Brian Stucker [mailto:bstucker@nortel.com] 
> Sent: 14 February 2007 23:44
> To: Elwell, John; sipping@ietf.org
> Subject: RE: I-D 
> ACTION:draft-stucker-sipping-early-media-indirection-00.txt
> 
> Thanks for reading the draft John. Replies below.
> 
> > -----Original Message-----
> > From: Elwell, John [mailto:john.elwell@siemens.com] 
> > Sent: Wednesday, February 14, 2007 7:18 PM
> > To: Stucker, Brian (RICH1:AR00); sipping@ietf.org
> > Subject: RE: I-D 
> > ACTION:draft-stucker-sipping-early-media-indirection-00.t xt
> > 
> > Brian,
> > 
> > Although I am interested in this general problem space and 
> > welcome proposals, I had a hard time understanding the 
> > benefits this gives compared to RFC 3960. For example, in 
> > section 6.2.2 it states:
> > 
> > "This dialog, known as an early-media dialog, allows the UAC 
> > to receive early media from endpoints other than the called 
> > party's device(s) such as a network media server, does not 
> > suffer from the problems of forking (since each indirect 
> > early-media session creates a separate early media dialog), 
> > allows for the UAC to specify a different RTP payload type 
> > (to facilitate discrimination of early-media versus 
> > final-media), and gives both the UAC and UAS the ability to 
> > explicitly start and stop the early media stream without 
> > complicating the SDP negotiations for the final media."
> > 
> > It is not clear which of these properties of the new 
> > mechanism are not available with RFC 3960. For example, when 
> > an INVITE request is forked, each branch of the fork 
> > establishes a separate early dialog according to RFC 3960. 
> > Forking to a media server could either create a separate 
> > early dialog or use a separate SDP offer-answer exchange 
> > (early session disposition) within the context of an existing 
> > early dialog. The separate offer-answer exchange permits 
> > things like use of a different RTP payload type or different 
> > port. Explicitly stopping an early media stream can be done 
> > by a further offer-answer exchange using early session disposition.
> 
> At the last IETF, I asked attendees to raise their hands if 
> any of them
> had implemented early session disposition: nobody raised their hands,
> and yet the problems of early media persist.
[JRE] The reason it hasn't been implemented is perhaps its complexity or
perceived complexity, but I don't see that the new proposal is significantly
less complex.

> RFC-3960 suffers from a
> couple of problems that I outlined in
> draft-stucker-sipping-early-media-coping. Even with early session
> disposition, it's difficult to tell what's early media and 
> what's final
> media until you get a 200 OK. Even with RFC-3960 it's very 
> difficult to
> control multiple forked call legs that are generating early media (ie.
> you have two or more people talking to you at once, and 
> others that may
> be sending you a 180 which further complicates things).
[JRE] I don't see how the present proposal helps with forked calls where you
have multiple early dialogs created, any one of which can potentially send
an Early-Media header field asking the UAC to contact a media server. Is it
really any easier for the UAC to sort this out than to sort out RFC 3959/60?

> 
> The proposed mechanism is to get rid of all of these problems 
> by using a
> dynamic RTP payload type (if you want to overload the RTP port at the
> originator with early and final media) and simply establish a separate
> session for each early media stream. This way everything is nice and
> neat. You can start and stop it very clearly (INVITE/CANCEL/BYE) no
> special logic is needed in the middle of the network to 
> handle injection
> of 183's with SDP for network based early media, etc, etc.
[JRE] Also with RFC 3959/60 you can assign a different payload type to the
early session, if you wish. You can turn early media on and off with further
offer-answer cycles in UPDATE.

> 
> > 
> > Section 6.2.3 fails to say how the proxy that instructed the 
> > UAC to play an early media stream can subsequently tell the 
> > UAC to stop that stream (assuming the dialog between the UAC 
> > and the media server does not pass through that proxy).
> 
> Why would it? It's a proxy. Do you have a usecase? I'm sure something
> could be proposed if it's necessary.
[JRE] In your example it is proxy B that invokes early media by inserting
the Early-Media header field in the 180 response. If proxy B causes early
media to be played, then surely proxy B should have the capability of
stopping early media. A use case might be that proxy B is scripted to route
incoming INVITE requests to a particular target, which requires early media
to be played. Then, if there is no answer within a certain time, it might be
scripted to retarget to somewhere where early media is no longer required or
different early media is to be played.

> 
> > 
> > The Early-Media header field definition does not appear to 
> > provide any information about the nature of the early media, 
> > although the examples do seem to show something of this 
> > nature, e.g., "service=ringback;user=userb", but presumably 
> > this is information for the media server and not for 
> > consumption by the UAC. The UAC will therefore not be able to 
> > determine relative priorities of Early-Media header fields 
> > received on different early dialogs from the same or 
> > different proxies.
> 
> This is definitely true. This is a first revision of the proposed
> mechanism to see if there's any support for this type of approach in
> general. If people don't see value in setting up a separate session to
> handle early media interactions, then the presence or absence of
> information in the Early-Media header about prioritization is moot.
> 
> > 
> > John
> > 
> > 
> > > -----Original Message-----
> > > From: Internet-Drafts@ietf.org [mailto:Internet-Drafts@ietf.org]
> > > Sent: 13 February 2007 20:50
> > > To: i-d-announce@ietf.org
> > > Subject: I-D
> > > ACTION:draft-stucker-sipping-early-media-indirection-00.txt
> > > 
> > > A New Internet-Draft is available from the on-line 
> Internet-Drafts 
> > > directories.
> > > 
> > > 
> > > 	Title		: Early Media INDirection (EMIND) in 
> > > the Session Initiation Protocol (SIP)
> > > 	Author(s)	: B. Stucker
> > > 	Filename	: 
> > > draft-stucker-sipping-early-media-indirection-00.txt
> > > 	Pages		: 18
> > > 	Date		: 2007-2-13
> > > 	
> > >    This document describes models that can be used to 
> > manage multiple
> > >    early media streams in the Session Initiation Protocol 
> > (SIP).  The
> > >    model seeks to allow calling party endpoints to be 
> able to better
> > >    control the rendering of early media to the user, and 
> distinguish
> > >    early media from final media.
> > > 
> > >    Although some of early media challenges are likely to never be
> > >    overcome, (e.g. when interworking with an ISUP PSTN 
> gateway that 
> > > does
> > >    not take into account CPG or ACM messages), the potential to 
> > > improve
> > >    on what is already there does exist.  The 
> recommendations in this
> > >    document are intended to be an update to RFC-3960 
> > recommendations 
> > > and
> > >    to avoid many of the complications outlined in
> > >    draft-stucker-sipping-early-media-coping.
> > > 
> > > 
> > > 
> > > A URL for this Internet-Draft is:
> > > http://www.ietf.org/internet-drafts/draft-stucker-sipping-earl
> > > y-media-indirection-00.txt
> > > 
> > > To remove yourself from the I-D Announcement list, send a 
> > message to 
> > > i-d-announce-request@ietf.org with the word unsubscribe in 
> > the body of 
> > > the message.
> > > You can also visit
> > > https://www1.ietf.org/mailman/listinfo/I-D-announce
> > > to change your subscription settings.
> > > 
> > > Internet-Drafts are also available by anonymous FTP. 
> Login with the 
> > > username "anonymous" and a password of your e-mail address. After 
> > > logging in, type "cd internet-drafts" and then "get 
> > > draft-stucker-sipping-early-media-indirection-00.txt".
> > > 
> > > A list of Internet-Drafts directories can be found in 
> > > http://www.ietf.org/shadow.html or 
> > > ftp://ftp.ietf.org/ietf/1shadow-sites.txt
> > > 
> > > Internet-Drafts can also be obtained by e-mail.
> > > 
> > > Send a message to:
> > > 	mailserv@ietf.org.
> > > In the body type:
> > > 	"FILE
> > > /internet-drafts/draft-stucker-sipping-early-media-indirection
> > > -00.txt".
> > > 	
> > > NOTE:	The mail server at ietf.org can return the document in
> > > 	MIME-encoded form by using the "mpack" utility.  To use this
> > > 	feature, insert the command "ENCODING mime" before the "FILE"
> > > 	command.  To decode the response(s), you will need "munpack" or
> > > 	a MIME-compliant mail reader.  Different MIME-compliant 
> > mail readers
> > > 	exhibit different behavior, especially when dealing with
> > > 	"multipart" MIME messages (i.e. documents which have been split
> > > 	up into multiple messages), so check your local documentation on
> > > 	how to manipulate these messages.
> > > 
> > > Below is the data which will enable a MIME compliant mail reader 
> > > implementation to automatically retrieve the ASCII version of the 
> > > Internet-Draft.
> > > 
> > 
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 15 08:07:01 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHgJF-0003CO-E9; Thu, 15 Feb 2007 08:06:01 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHgJD-0003CG-SW
	for sipping@ietf.org; Thu, 15 Feb 2007 08:05:59 -0500
Received: from nf-out-0910.google.com ([64.233.182.190])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHgJB-0002UY-2D
	for sipping@ietf.org; Thu, 15 Feb 2007 08:05:59 -0500
Received: by nf-out-0910.google.com with SMTP id l36so1137821nfa
	for <sipping@ietf.org>; Thu, 15 Feb 2007 05:05:56 -0800 (PST)
DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta;
	h=received:message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:references;
	b=e15Xu5daIGGsLachbRJQkh7f0wVrfDnGJuYxHnnH+l21Ag6H0uBi298Hhs4WcEXuz1MLsktKoEUMU64UPdhlYboz5WOh/fu3xms2qNVeosVGvSjosDs8zB/08VtGyqWSKX2TSnA09ifW0VyN2yfulc6fSkjjemM4b19FVc38NyE=
Received: by 10.82.163.13 with SMTP id l13mr2720881bue.1171544756056;
	Thu, 15 Feb 2007 05:05:56 -0800 (PST)
Received: by 10.82.185.16 with HTTP; Thu, 15 Feb 2007 05:05:55 -0800 (PST)
Message-ID: <3fe6b8640702150505pa2f6a90o3ee907080b0e510a@mail.gmail.com>
Date: Thu, 15 Feb 2007 18:35:55 +0530
From: "Malleswara Rao Ankem" <malleshavn@gmail.com>
To: "Elwell, John" <john.elwell@siemens.com>
Subject: Re: Device architecture? WAS: RE: [Sipping] Service Identificatio n
	in SIP
In-Reply-To: <50B1CBA96870A34799A506B2313F26670B0A9615@ntht201e.siemenscomms.co.uk>
MIME-Version: 1.0
References: <50B1CBA96870A34799A506B2313F26670B0A9615@ntht201e.siemenscomms.co.uk>
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 210770d71723b650f9c8e3db4e95b596
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============0247663828=="
Errors-To: sipping-bounces@ietf.org

--===============0247663828==
Content-Type: multipart/alternative; 
	boundary="----=_Part_167195_33456593.1171544755960"

------=_Part_167195_33456593.1171544755960
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

>From a "dumb" terminal perspective it is a overkill. The reason being those
RFCs (3841/3840) not only provide macro-capability identification (through
feature tags) but also very granular, micro-capability (method based or
other parameters) based routing and service selection. A endpoint claiming
support for the same need to implement all the recommendations or has to be
partially be non-compliant for most other parts of those RFCs.
I understand that a service can be described by a feature tag or can be
described by the session parameters that would be involved in the session. =
I
guess you would agree that while former is less complex when I dont intend
to be interoperable with all the devices on the earth while latter though
powerful but that device would still need to understand the feature tags cu=
z
the RFCs talk about it.
If feature tags can be accepted as service identifiers then why not service
urns ?

-Mallesh



On 2/15/07, Elwell, John <john.elwell@siemens.com> wrote:
>
>  Mallesh,
>
> Yes, I agree that some means of selecting a preferred device might be
> needed, but that is indeed what RFC3841/3840 does. Why do you consider th=
is
> overkill?
>
> John
>
>  ------------------------------
> *From:* Malleswara Rao Ankem [mailto:malleshavn@gmail.com]
> *Sent:* 15 February 2007 10:54
> *To:* Elwell, John
> *Cc:* G=F6ran Eriksson AP (KI/EAB); sipping@ietf.org
> *Subject:* Re: Device architecture? WAS: RE: [Sipping] Service
> Identificatio n in SIP
>
>
>  Inorder to identify a endpoint that has the capabilities to serve a very
> specialized functionality, I guess it is simpler to name the functionalit=
y.
> Naming the functionality not only helps in decision making while routing =
the
> request to that specific endpoint but also helps the endpoint identify wh=
ich
> specialized application on the endpoint can take up the responsibility.
> Instead of all the routing entities looking into the SDP/message bodies i=
t
> should be easier/faster if they can look for a specific header.
> I thought feature tag mechanisms indicated in the RFC3841/3840 combinatio=
n
> should have sufficed but implementing those RFCs seems to be a overkill f=
or
> describing/selecting/routing requests to the specific endpoint and also
> choosing a service/feature/application on the endpoint.
>
> -Mallesh
>
>
>
> On 2/15/07, Elwell, John <john.elwell@siemens.com > wrote:
> >
> > G=F6ran,
> >
> > Your interpretation is different from what is stated in the Rosenberg
> > draft.
> > Anyway, let's not dwell on this particular use case, which may or may
> > not be
> > realistic. The important thing is whether the media being sent is
> > suitable
> > for general purpose devices or requires or could be used by some specia=
l
> > class of application.
> >
> > If a medium is somehow specialised ( e.g., audio modulated for
> > interpretation
> > by a modem or fax machine), then the necessary information needs to be
> > conveyed along with the m-line in the SDP. Perhaps even if it is
> > human-intelligible voice but with a restricted repertoire of words that
> > make
> > it suitable for a particular class of application, maybe that
> > information
> > too should be conveyed with the m-line. Perhaps the case described in
> > Jonathan's draft fits this latter category: I believe the medium is jus=
t
> >
> > regular voice that can be interpreted by a human, but is also
> > constrained in
> > its repertoire of words such that it can be used by a chess application=
.
> >
> > The important thing is that there be sufficient information in the SDP
> > to
> > allow appropriate application selection at the end device. I am not
> > convinced by a need for information at the SIP level during call
> > establishment, because of course SDP can be re-negotiated during the
> > call
> > ( e.g., when the users decide to use a particular application) and what
> > is
> > indicated in SIP at call establishment time might no longer be valid
> > later
> > in the call.
> >
> > John
> >
> >
> > > -----Original Message-----
> > > From: G=F6ran Eriksson AP (KI/EAB)
> > > [mailto:goran.ap.eriksson@ericsson.com]
> > > Sent: 14 February 2007 23:34
> > > To: Elwell, John
> > > Cc: sipping@ietf.org
> > > Subject: Device architecture? WAS: RE: [Sipping] Service
> > > Identification in SIP
> > >
> > > My humble, perhaps faulty, understanding of the "chess game"
> > > was not that voice commands were used to control the chess
> > > itself, rather
> > > that the chess game, while having some IP-based proprietary
> > > gaming interaction protocol, also had a feature letting
> > > the users talk to each others while playing. The scenario are
> > > two, or more, users play a chess game of the (Inter)net, talkin
> > > to eachother from time to time during the game (yeah, don't
> > > ask me if I think people will love this!!).
> > >
> > > The purpose, as I understand it, was a)to show that the
> > > 'voice component' in the chess game application is similar to that
> > > in a stand alone POC-Only application when it comes to
> > > exchange and transfer of voice samples and associated floor control
> > > and b) to point to differences, for instance when it comes to
> > > supplementary services, which may not apply in the chess game case.
> > >
> > > Perhaps it is important to be clear on the assumed user
> > > device realisation, as far as I understand it:
> > >
> > >  The "chess game + voice" example is, as I understand it,
> > > discussed under the assumption that voice and chess game is
> > >  managed by custome designed appliction, or client, which is
> > > a necessary/likely approach in a mobile device given the
> > >  limited MMI (and also other factors in the mobile context).
> > >
> > >  In a PC, I could easily handle this as two different
> > > applications/client, like when I use Skype/MSN to talk to my
> > > friends while
> > >  playing some turn based game, but again, the discussion is
> > > held under the assumption that usability requirements
> > >  leads to an application client integrating voice and chess
> > > game events.
> > >
> > > The alternative device realisation You describe- different
> > > clients managing different portions of what is negotiated in
> > > the SDP or MSRP MIME attachment-
> > > is, I think, a good illustration of the need to be precise
> > > and explicit about the assumed device architecture when
> > > discussing the "service identification" issue(s).
> > >
> > > Regards
> > > G=F6ran
> > >
> > >
> > >
> > >
> > > -----Original Message-----
> > > From: Elwell, John [mailto: john.elwell@siemens.com]
> > > Sent: den 14 februari 2007 11:57
> > > To: Henry Sinnreich; sipping@ietf.org
> > > Subject: RE: [Sipping] Service Identification in SIP
> > >
> > > Having followed this thread, I too find the chess example
> > > very contrived. If chess moves are sent as voice commands,
> > > then surely it would be acceptable for the normal telephone
> > > application to answer the call, i.e., the remote player
> > > listens to the move, performs this move on a conventional or
> > > electronic chess board and speaks his next move. The calling
> > > UA should not care whether the call is answered by a computer
> > > application or a person.
> > > Furthermore, even if the audio needs to consumed by a
> > > particular chess application at the called UA, then other
> > > media in the same session such as video or messaging might
> > > not be. So any qualification of the nature of media to be
> > > exchanged probably should be at the media level, not at the
> > > session level or SIP level. As yet another variation, what if
> > > the players set up a conventional telephone call and then
> > > mutually agree to switch in their chess applications?
> > >
> > > Now, let's consider another case. I establish an MSRP
> > > session. Within this MSRP session I can send messages with
> > > various MIME attachments that are intended for consumption by
> > > various applications. I should not need to know in advance
> > > what applications I will use during the course of the MSRP
> > > session. All I should put in the INVITE request is the fact
> > > that I want to establish an MSRP session.
> > >
> > > John
> > >
> > > > -----Original Message-----
> > > > From: Henry Sinnreich [mailto:hsinnrei@adobe.com ]
> > > > Sent: 13 February 2007 18:19
> > > > To: sipping@ietf.org
> > > > Subject: RE: [Sipping] Service Identification in SIP
> > > >
> > > > It must be my lack of imagination for not being able to
> > > come up with a
> > > > credible usage scenario, given the following:
> > > >
> > > > - Is playing chess by voice a credible usage scenario? :-)
> > > Especially
> > > > since voice recognition is now so often found in PCs and in mobile
> > > > phones as well. In other words, chess playing by voice may
> > > also be an
> > > > endpoint application, without any network involvement. Actually, if
> > > > placed in the endpoint it works just as well on poor
> > > bandwidth links
> > > > by contrast to chess in the network that may require adequate
> > > > broadband.
> > > >
> > > > - Is anyone's boss or spouse going to approve extra phone charges
> > > > above the flat rate for playing chess at a time when the gaming
> > > > industry has pioneered free VoIP?
> > > > Will such a service be taxed extra and if yes using by what
> > > criteria?
> > > > Gambling comes to my mind :-)
> > > >
> > > > - VoIP is reported in the press as a give-away and I have seen no
> > > > reports of any VoIP provider recording a profit on VoIP. I
> > > assume the
> > > > losses are often hidden by bundling in triple play, etc.
> > > >
> > > > Given my lack of imagination, can anyone share a more
> > > credible usage
> > > > scenario for Service Identification in SIP?
> > > >
> > > > What's the use of a standard without a credible usage scenario?
> > > >
> > > > Thanks, Henry
> > > >
> > > >
> > > > _______________________________________________
> > > > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sippin=
g
> > > > This list is for NEW development of the application of SIP Use
> > > > sip-implementors@cs.columbia.edu for questions on current sip Use
> > > > sip@ietf.org for new developments of core SIP
> > > >
> > >
> > > _______________________________________________
> > > Sipping mailing list   https://www1.ietf.org/mailman/listinfo/sipping
> > > This list is for NEW development of the application of SIP
> > > Use sip-implementors@cs.columbia.edu for questions on current
> > > sip Use sip@ietf.org for new developments of core SIP
> > >
> >
> > _______________________________________________
> > Sipping mailing list   https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP
> > Use sip-implementors@cs.columbia.edu for questions on current sip
> > Use sip@ietf.org for new developments of core SIP
> >
>
>

------=_Part_167195_33456593.1171544755960
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

<div>From a &quot;dumb&quot; terminal perspective it is a overkill. The rea=
son being those RFCs (3841/3840) not only provide macro-capability identifi=
cation (through feature tags)&nbsp;but also very granular, micro-capability=
 (method based or other parameters) based routing and service selection. A =
endpoint claiming support for the same need to implement all the recommenda=
tions or has to be partially be non-compliant for most other parts of those=
 RFCs.
</div>
<div>I understand that&nbsp;a service can be described by a feature tag or =
can be described by the session parameters that would&nbsp;be involved in t=
he session. I guess you would agree that while former is less complex when =
I dont intend to be interoperable with all the devices on the earth while l=
atter though powerful but that device would still need to understand the fe=
ature tags cuz the RFCs talk about it.
</div>
<div>If feature tags can be accepted as service identifiers then why not se=
rvice urns ?</div>
<div>&nbsp;</div>
<div>-Mallesh</div>
<div>&nbsp;</div>
<div><br>&nbsp;</div>
<div><span class=3D"gmail_quote">On 2/15/07, <b class=3D"gmail_sendername">=
Elwell, John</b> &lt;<a href=3D"mailto:john.elwell@siemens.com">john.elwell=
@siemens.com</a>&gt; wrote:</span>
<blockquote class=3D"gmail_quote" style=3D"PADDING-LEFT: 1ex; MARGIN: 0px 0=
px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">
<div>
<div dir=3D"ltr" align=3D"left"><span><font face=3D"Arial" color=3D"#0000ff=
" size=3D"2">Mallesh,</font></span></div>
<div dir=3D"ltr" align=3D"left"><span><font face=3D"Arial" color=3D"#0000ff=
" size=3D"2"></font></span>&nbsp;</div>
<div dir=3D"ltr" align=3D"left"><span><font face=3D"Arial" color=3D"#0000ff=
" size=3D"2">Yes, I agree that some means of selecting a preferred device m=
ight be needed, but that is indeed what RFC3841/3840 does. Why do you consi=
der this overkill?
</font></span></div>
<div dir=3D"ltr" align=3D"left"><span><font face=3D"Arial" color=3D"#0000ff=
" size=3D"2"></font></span>&nbsp;</div>
<div dir=3D"ltr" align=3D"left"><span><font face=3D"Arial" color=3D"#0000ff=
" size=3D"2">John</font></span></div><br>
<blockquote style=3D"PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #000=
0ff 2px solid; MARGIN-RIGHT: 0px">
<div lang=3D"en-us" dir=3D"ltr" align=3D"left">
<hr>
<font face=3D"Tahoma" size=3D"2"><b>From:</b> Malleswara Rao Ankem [mailto:=
<a onclick=3D"return top.js.OpenExtLink(window,event,this)" href=3D"mailto:=
malleshavn@gmail.com" target=3D"_blank">malleshavn@gmail.com</a>] <br><b>Se=
nt:</b>
 15 February 2007 10:54<br><b>To:</b> Elwell, John<br><b>Cc:</b> G=F6ran Er=
iksson AP (KI/EAB); <a onclick=3D"return top.js.OpenExtLink(window,event,th=
is)" href=3D"mailto:sipping@ietf.org" target=3D"_blank">sipping@ietf.org</a=
><br>
<b>Subject:</b> Re: Device architecture? WAS: RE: [Sipping] Service Identif=
icatio n in SIP<br></font><br>&nbsp;</div>
<div><span class=3D"e" id=3D"q_110c51d56fecfe1e_1">
<div></div>
<div>Inorder to identify a endpoint that has the capabilities to serve a ve=
ry specialized functionality, I guess it is simpler to name the functionali=
ty. Naming the functionality not only helps in decision making while routin=
g the request to that specific endpoint but also helps the endpoint identif=
y which specialized application on the endpoint can take up the responsibil=
ity. Instead of all the routing entities looking into the SDP/message bodie=
s it should be easier/faster if they can look for a specific header.=20
</div>
<div>I thought feature tag mechanisms indicated in the RFC3841/3840 combina=
tion should have sufficed but implementing those RFCs seems to be a overkil=
l for describing/selecting/routing requests&nbsp;to the specific endpoint a=
nd also choosing a service/feature/application on the endpoint.=20
</div>
<div>&nbsp;</div>
<div>-Mallesh</div>
<div>&nbsp;</div>
<div>&nbsp;</div>
<div>&nbsp;</div>
<div><span class=3D"gmail_quote">On 2/15/07, <b class=3D"gmail_sendername">=
Elwell, John</b> &lt;<a onclick=3D"return top.js.OpenExtLink(window,event,t=
his)" href=3D"mailto:john.elwell@siemens.com" target=3D"_blank">john.elwell=
@siemens.com=20
</a>&gt; wrote:</span>=20
<blockquote class=3D"gmail_quote" style=3D"PADDING-LEFT: 1ex; MARGIN: 0px 0=
px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">G=F6ran,<br><br>Your interpretat=
ion is different from what is stated in the Rosenberg draft.<br>Anyway, let=
&#39;s not dwell on this particular use case, which may or may not be=20
<br>realistic. The important thing is whether the media being sent is suita=
ble<br>for general purpose devices or requires or could be used by some spe=
cial<br>class of application.<br><br>If a medium is somehow specialised (=
=20
e.g., audio modulated for interpretation<br>by a modem or fax machine), the=
n the necessary information needs to be<br>conveyed along with the m-line i=
n the SDP. Perhaps even if it is<br>human-intelligible voice but with a res=
tricted repertoire of words that make=20
<br>it suitable for a particular class of application, maybe that informati=
on<br>too should be conveyed with the m-line. Perhaps the case described in=
<br>Jonathan&#39;s draft fits this latter category: I believe the medium is=
 just=20
<br>regular voice that can be interpreted by a human, but is also constrain=
ed in<br>its repertoire of words such that it can be used by a chess applic=
ation.<br><br>The important thing is that there be sufficient information i=
n the SDP to=20
<br>allow appropriate application selection at the end device. I am not<br>=
convinced by a need for information at the SIP level during call<br>establi=
shment, because of course SDP can be re-negotiated during the call<br>(=20
e.g., when the users decide to use a particular application) and what is<br=
>indicated in SIP at call establishment time might no longer be valid later=
<br>in the call.<br><br>John<br><br><br>&gt; -----Original Message-----=20
<br>&gt; From: G=F6ran Eriksson AP (KI/EAB)<br>&gt; [mailto:<a onclick=3D"r=
eturn top.js.OpenExtLink(window,event,this)" href=3D"mailto:goran.ap.erikss=
on@ericsson.com" target=3D"_blank">goran.ap.eriksson@ericsson.com</a>]<br>&=
gt; Sent: 14 February 2007 23:34=20
<br>&gt; To: Elwell, John<br>&gt; Cc: <a onclick=3D"return top.js.OpenExtLi=
nk(window,event,this)" href=3D"mailto:sipping@ietf.org" target=3D"_blank">s=
ipping@ietf.org</a><br>&gt; Subject: Device architecture? WAS: RE: [Sipping=
] Service=20
<br>&gt; Identification in SIP<br>&gt;<br>&gt; My humble, perhaps faulty, u=
nderstanding of the &quot;chess game&quot;<br>&gt; was not that voice comma=
nds were used to control the chess <br>&gt; itself, rather<br>&gt; that the=
 chess game, while having some IP-based proprietary=20
<br>&gt; gaming interaction protocol, also had a feature letting<br>&gt; th=
e users talk to each others while playing. The scenario are <br>&gt; two, o=
r more, users play a chess game of the (Inter)net, talkin<br>&gt; to eachot=
her from time to time during the game (yeah, don&#39;t=20
<br>&gt; ask me if I think people will love this!!).<br>&gt;<br>&gt; The pu=
rpose, as I understand it, was a)to show that the <br>&gt; &#39;voice compo=
nent&#39; in the chess game application is similar to that<br>&gt; in a sta=
nd alone POC-Only application when it comes to=20
<br>&gt; exchange and transfer of voice samples and associated floor contro=
l <br>&gt; and b) to point to differences, for instance when it comes to<br=
>&gt; supplementary services, which may not apply in the chess game case.=
=20
<br>&gt;<br>&gt; Perhaps it is important to be clear on the assumed user<br=
>&gt; device realisation, as far as I understand it:<br>&gt;<br>&gt;&nbsp;&=
nbsp;The &quot;chess game + voice&quot; example is, as I understand it,<br>=
&gt; discussed under the assumption that voice and chess game is=20
<br>&gt;&nbsp;&nbsp;managed by custome designed appliction, or client, whic=
h is <br>&gt; a necessary/likely approach in a mobile device given the<br>&=
gt;&nbsp;&nbsp;limited MMI (and also other factors in the mobile context).<=
br>&gt;<br>&gt;&nbsp;&nbsp;In a PC, I could easily handle this as two diffe=
rent=20
<br>&gt; applications/client, like when I use Skype/MSN to talk to my <br>&=
gt; friends while<br>&gt;&nbsp;&nbsp;playing some turn based game, but agai=
n, the discussion is<br>&gt; held under the assumption that usability requi=
rements=20
<br>&gt;&nbsp;&nbsp;leads to an application client integrating voice and ch=
ess<br>&gt; game events.<br>&gt;<br>&gt; The alternative device realisation=
 You describe- different<br>&gt; clients managing different portions of wha=
t is negotiated in=20
<br>&gt; the SDP or MSRP MIME attachment-<br>&gt; is, I think, a good illus=
tration of the need to be precise <br>&gt; and explicit about the assumed d=
evice architecture when<br>&gt; discussing the &quot;service identification=
&quot; issue(s).=20
<br>&gt;<br>&gt; Regards<br>&gt; G=F6ran<br>&gt;<br>&gt;<br>&gt;<br>&gt;<br=
>&gt; -----Original Message----- <br>&gt; From: Elwell, John [mailto:<a onc=
lick=3D"return top.js.OpenExtLink(window,event,this)" href=3D"mailto:john.e=
lwell@siemens.com" target=3D"_blank">
 john.elwell@siemens.com</a>]<br>&gt; Sent: den 14 februari 2007 11:57<br>&=
gt; To: Henry Sinnreich; <a onclick=3D"return top.js.OpenExtLink(window,eve=
nt,this)" href=3D"mailto:sipping@ietf.org" target=3D"_blank">sipping@ietf.o=
rg=20
</a><br>&gt; Subject: RE: [Sipping] Service Identification in SIP<br>&gt;<b=
r>&gt; Having followed this thread, I too find the chess example<br>&gt; ve=
ry contrived. If chess moves are sent as voice commands,<br>&gt; then surel=
y it would be acceptable for the normal telephone=20
<br>&gt; application to answer the call, i.e., the remote player<br>&gt; li=
stens to the move, performs this move on a conventional or<br>&gt; electron=
ic chess board and speaks his next move. The calling<br>&gt; UA should not =
care whether the call is answered by a computer=20
<br>&gt; application or a person.<br>&gt; Furthermore, even if the audio ne=
eds to consumed by a<br>&gt; particular chess application at the called UA,=
 then other<br>&gt; media in the same session such as video or messaging mi=
ght=20
<br>&gt; not be. So any qualification of the nature of media to be<br>&gt; =
exchanged probably should be at the media level, not at the<br>&gt; session=
 level or SIP level. As yet another variation, what if<br>&gt; the players =
set up a conventional telephone call and then=20
<br>&gt; mutually agree to switch in their chess applications?<br>&gt;<br>&=
gt; Now, let&#39;s consider another case. I establish an MSRP<br>&gt; sessi=
on. Within this MSRP session I can send messages with<br>&gt; various MIME =
attachments that are intended for consumption by=20
<br>&gt; various applications. I should not need to know in advance<br>&gt;=
 what applications I will use during the course of the MSRP<br>&gt; session=
. All I should put in the INVITE request is the fact<br>&gt; that I want to=
 establish an MSRP session.=20
<br>&gt;<br>&gt; John<br>&gt;<br>&gt; &gt; -----Original Message-----<br>&g=
t; &gt; From: Henry Sinnreich [mailto:<a onclick=3D"return top.js.OpenExtLi=
nk(window,event,this)" href=3D"mailto:hsinnrei@adobe.com" target=3D"_blank"=
>hsinnrei@adobe.com=20
</a>]<br>&gt; &gt; Sent: 13 February 2007 18:19<br>&gt; &gt; To: <a onclick=
=3D"return top.js.OpenExtLink(window,event,this)" href=3D"mailto:sipping@ie=
tf.org" target=3D"_blank">sipping@ietf.org</a><br>&gt; &gt; Subject: RE: [S=
ipping] Service Identification in SIP=20
<br>&gt; &gt;<br>&gt; &gt; It must be my lack of imagination for not being =
able to<br>&gt; come up with a <br>&gt; &gt; credible usage scenario, given=
 the following:<br>&gt; &gt;<br>&gt; &gt; - Is playing chess by voice a cre=
dible usage scenario? :-)=20
<br>&gt; Especially<br>&gt; &gt; since voice recognition is now so often fo=
und in PCs and in mobile <br>&gt; &gt; phones as well. In other words, ches=
s playing by voice may<br>&gt; also be an<br>&gt; &gt; endpoint application=
, without any network involvement. Actually, if=20
<br>&gt; &gt; placed in the endpoint it works just as well on poor <br>&gt;=
 bandwidth links<br>&gt; &gt; by contrast to chess in the network that may =
require adequate<br>&gt; &gt; broadband.<br>&gt; &gt;<br>&gt; &gt; - Is any=
one&#39;s boss or spouse going to approve extra phone charges=20
<br>&gt; &gt; above the flat rate for playing chess at a time when the gami=
ng<br>&gt; &gt; industry has pioneered free VoIP?<br>&gt; &gt; Will such a =
service be taxed extra and if yes using by what<br>&gt; criteria?<br>&gt; &=
gt; Gambling comes to my mind :-)=20
<br>&gt; &gt;<br>&gt; &gt; - VoIP is reported in the press as a give-away a=
nd I have seen no<br>&gt; &gt; reports of any VoIP provider recording a pro=
fit on VoIP. I<br>&gt; assume the<br>&gt; &gt; losses are often hidden by b=
undling in triple play, etc.=20
<br>&gt; &gt;<br>&gt; &gt; Given my lack of imagination, can anyone share a=
 more<br>&gt; credible usage<br>&gt; &gt; scenario for Service Identificati=
on in SIP?<br>&gt; &gt;<br>&gt; &gt; What&#39;s the use of a standard witho=
ut a credible usage scenario?=20
<br>&gt; &gt;<br>&gt; &gt; Thanks, Henry<br>&gt; &gt;<br>&gt; &gt;<br>&gt; =
&gt; _______________________________________________<br>&gt; &gt; Sipping m=
ailing list&nbsp;&nbsp;<a onclick=3D"return top.js.OpenExtLink(window,event=
,this)" href=3D"https://www1.ietf.org/mailman/listinfo/sipping" target=3D"_=
blank">
 https://www1.ietf.org/mailman/listinfo/sipping </a><br>&gt; &gt; This list=
 is for NEW development of the application of SIP Use<br>&gt; &gt; <a oncli=
ck=3D"return top.js.OpenExtLink(window,event,this)" href=3D"mailto:sip-impl=
ementors@cs.columbia.edu" target=3D"_blank">
sip-implementors@cs.columbia.edu</a> for questions on current sip Use<br>&g=
t; &gt; <a onclick=3D"return top.js.OpenExtLink(window,event,this)" href=3D=
"mailto:sip@ietf.org" target=3D"_blank">sip@ietf.org</a> for new developmen=
ts of core SIP=20
<br>&gt; &gt;<br>&gt;<br>&gt; _____________________________________________=
__<br>&gt; Sipping mailing list&nbsp;&nbsp;<a onclick=3D"return top.js.Open=
ExtLink(window,event,this)" href=3D"https://www1.ietf.org/mailman/listinfo/=
sipping" target=3D"_blank">
 https://www1.ietf.org/mailman/listinfo/sipping</a><br>&gt; This list is fo=
r NEW development of the application of SIP<br>&gt; Use <a onclick=3D"retur=
n top.js.OpenExtLink(window,event,this)" href=3D"mailto:sip-implementors@cs=
.columbia.edu" target=3D"_blank">
sip-implementors@cs.columbia.edu</a> for questions on current <br>&gt; sip =
Use <a onclick=3D"return top.js.OpenExtLink(window,event,this)" href=3D"mai=
lto:sip@ietf.org" target=3D"_blank">sip@ietf.org</a> for new developments o=
f core SIP=20
<br>&gt;<br><br>_______________________________________________<br>Sipping =
mailing list&nbsp;&nbsp;<a onclick=3D"return top.js.OpenExtLink(window,even=
t,this)" href=3D"https://www1.ietf.org/mailman/listinfo/sipping" target=3D"=
_blank"> https://www1.ietf.org/mailman/listinfo/sipping=20
</a><br>This list is for NEW development of the application of SIP<br>Use <=
a onclick=3D"return top.js.OpenExtLink(window,event,this)" href=3D"mailto:s=
ip-implementors@cs.columbia.edu" target=3D"_blank">sip-implementors@cs.colu=
mbia.edu=20
</a>for questions on current sip <br>Use <a onclick=3D"return top.js.OpenEx=
tLink(window,event,this)" href=3D"mailto:sip@ietf.org" target=3D"_blank">si=
p@ietf.org</a> for new developments of core SIP<br></blockquote></div><br><=
/span>
</div></blockquote></div></blockquote></div><br>

------=_Part_167195_33456593.1171544755960--


--===============0247663828==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============0247663828==--




From sipping-bounces@ietf.org Thu Feb 15 10:04:22 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHi9F-00041R-7F; Thu, 15 Feb 2007 10:03:49 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHi9C-00041M-FQ
	for sipping@ietf.org; Thu, 15 Feb 2007 10:03:46 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHi94-0003Zo-Bz
	for sipping@ietf.org; Thu, 15 Feb 2007 10:03:46 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	B8AD121192
	for <sipping@ietf.org>; Thu, 15 Feb 2007 16:03:33 +0100 (CET)
X-AuditID: c1b4fb3e-aeed1bb0000007e1-25-45d47645aafc 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	9F06C20FF3
	for <sipping@ietf.org>; Thu, 15 Feb 2007 16:03:33 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 15 Feb 2007 16:03:33 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Thu, 15 Feb 2007 16:03:32 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD028861AF@esealmw111>
In-Reply-To: <200702110449.l1B4nfk2020329@dragon.ariadne.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdNmH89O4T46iGoRjicZ3hYknVpawB8E6oQ
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 15 Feb 2007 15:03:33.0131 (UTC)
	FILETIME=[75D20DB0:01C75112]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a069a8e8835d39ce36e425c148267a7b
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi,

Firstly to some examples that come to my mind.  I know that the
separation of PoC and from Telephony has been raised.  It has been
stated a few times that "PoC is just a specialised floor control".  I
don't believe that to be correct.  It is true that both Telephony and
PoC use voice as a media; and that in addition PoC uses a floor control.
In addition to that, an operator may wish to prioritise these
differently in a resource constrained access network (e.g. a radio
network).  For example, voice for Multimedia telephony may be
prioritised higher than Push to talk voice - this is also for more
efficient use of resources in a resource constrained access network.  A
further differentiation between them, is that some of the "sub services"
or "features" or <choose your favourite word> for them are different
(e.g. communicate forwarding on subscriber not reachable) which
ultimately results in the service logic being different for them.

An example of another service is (Instant) messaging.  Instant messaging
may use MSRP to transfer text (e.g. "Hello") between users.  However,
MSRP may also be used in the context of other services, such as
telephony, or to transfer information from one device to another device
(e.g. address book information).

Another example could be a SIP push type of service.  A SIP push service
would be the ability to push e.g. data to a device.  The SIP push may be
just a simple MESSAGE message, or an INVITE; however it should be
distinguished from other uses of that as the information may not be
destined for the end-user.  On top of that, then the exact application
to which the information is being pushed would be useful.

These examples of PoC - Telephony; Messaging - Telephony are examples of
different types of communication services where same media can be used
in different contexts.  I should note that the examples of PoC,
telephony and messaging are examples of bundled services, but what is
wrong with that?  It provides a good way to increase the possibility of
mxn interoperability by ensuring that clients will test against at least
those capabilities.

In my understanding, the lack of the ability to identify these
communication services, implies that "basic SIP communication" in itself
is monolithic service.  That can be quite OK in a small, or controlled
network, but I don't see that in a broader context, in particular a
multiservice network.  I would like to point out here, that this does
not imply anything relating to the implementation.  A vendor can build
the support for the different services out of a bunch of common
components; and there is of course a great deal of commonality that
supports such re-use.

This leads me to my next topic, the example of a chess application using
the voice.  While being discredited by some people, I think it does
point out that that context where the media being used needs to be
understood by some applications.  You could as well take the same
example where the Chess application also established a separate control
channel between the two devices, but the rest of the experience required
was as per telephony.  In the 3GPP documentation, there is a
differentiation between a "Communication service" and an "Application".
The idea is that the "Applications" use the basic rules for
communication defined by the "communication service".  The parallel
example here, is Chess is the "application", while the communication
service itself could be telephony.  We can further note here, that
having an application re-use another existing communication is not new -
it has been used with e.g. SMS for a while, where the SMS infrastructure
has been used for delivering messages to the user, as well as to the
device.  We use it in IP all the time as well ;-)

Moving onto the proposed solution draft itself, in addition for
providing for the identification of the "application" and the
"communication service", I have a few additional thoughts.

The draft seems to have the implication that all services (communication
services in other terminology) that are to be identified have to be
standardised within the IETF.  This is clearly an restriction that will
not be acceptable to the industry as a whole.  My guess is that if such
a statement is mandated by the IETF, it will just be words on paper
which would likely be ignored in practice by other standardisation
organisations.

The draft also only recognises the case where all explicitly knowledge
of the service is required.  I think it would also be good to allow the
case where knowledge of the service would be helpful, but not explicitly
required.  This could be a case where the service to the end-user would
not be degraded if this was not supported.  This functionality seems to
be missing.

The draft seems to indicate that that application servers in the network
are involved only at the request of the originating terminal.  This
seems a strange restriction.  For some network support that supports the
originating user, such as the one provided in the example "I would like
this call recorded for later retrieval", such mechanisms may make sense,
however as a general approach it seems a bit limited.  For example, in
3GPP networks there is the need to invoke application logic in the
originating side and terminating side, not only at the request of the
originating terminal, but as a consequence of the requested service.  If
there is an SIP message contains indication of the requested service, a
proxy can do this (or loose route to an AS that contains the necessary
service logic).  I note here that the proposed service URN does not
contain the address of a specific server containing the application
logic.  An AS can look at that, and loose route to an application server
that does contain such logic (If it is separated from the proxy),
however I fail to see the difference for that proxy if it looks at a
feature tag, or it looks at the service URN for this.  In both cases, it
has to parse the message for the data, and modify the routing
information.  Further, the route header could be removed in the
originating network; and it is still required to identify the service in
the terminating network.  As proxies may remove the route, it can't be
used to transport the requested service.  From this perspective it is
not clear to me the advantage of using a service URN in a route header
compared to the use of a feature tag.

The draft seems to mandate GRUU - is this really so, is that really
necessary.  It seems strange to place such a dependency, it also starts
to move the GRUU as being the primary means of which to contact a user.
This implies that the originating user decides which terminal I should
receive my requested service on, even if I have two devices for that.
As such, I don't believe that there should be a dependency on GRUU.

Parts of the draft seems to rely on the use of presence.  While presence
could be a supportive mechanism, I doubt that we can mandate that the
number you want to dial has be available in presence as I assume not all
entities are available in presence (diallable network services, users in
the PSTN).  This would be strange to mandated.

Moving on from the draft I have seen the questioned raised about "why
can't we expand the power of SDP".  Such an approach would mandate that
SIP methods use SDP, I don't think that is appropriate.=20

I hear the appeals that the mechanisms should not be abused, and that we
should not have 1000s of separate "communication services", and I fully
agree with this. I see a smaller number of communication services, not a
larger number; where those communication services are used to access
other end-user-services.

*flame away*

Regards,

Steve

---


-----Original Message-----
From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net]=20
Sent: Sunday, February 11, 2007 5:50 AM
To: sipping@ietf.org
Subject: Re: [Sipping] Service Identification in SIP

I've attempted to re-read the threads that Gonzalo referred to, and the
major limitation that I see is the lack of any concrete example of why
these identifiers are needed.  Probably the IMS people have discussed
this so many times that they've forgotten the detailed rationale behind
the requirement.  But I'm a net-head, (and most of the people here are
net-heads) and the SIP systems that I've dealt with show no evidence of
needing "service identification", they're pure proxies, and requests are
passed transparently end-to-end.

Could somebody please lay out an example showing the utility of service
identification?

   From: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>

   I think that everyone here agree that there is the possibility to use
a
   media in different ways, and that we need a distinction about what
the
   capabilities are and how they are used, whatever will be the
mechanism
   that at the end we will identify to do that.

Respectfully, I disagree.  We do need ways of identifying the
capabilities of SIP agents and transports, but I see no necessity of
creating specified bundles of capabailities, and then constraining
dialogs to require an initial agreement as to which bundle of
capabilities will be used.

One of the central features of Internet architecture is that agents
negotiate compatibility on a feature-by-feature basis; there are no
"profiles" which must be known on a network-wide basis.  (And I have
observed that this feature is of significant commercial value.)

In regard to identifying an "application" to process the dialog (and I
use "application" in a standard computing way), I see no reason that a
user cannot select URI-parameters to append to his AOR, and arrange for
his UA to route invites with those URI-parameters to the appropriate
application.  (And given the rich mechanisms for redirection in SIP, the
application need not even reside at a UA which is registered for the
AOR.)  These URI-parameters can be formally or informally standardized,
or established ad-hoc by each user.

In regard to PoC (as someone said, "a specialized floor-control
protocol") and similar features, SIP currently has mechanisms for
indicating support of features and requiring that a recipient support a
feature before accepting a call (the Supported and Requires headers, and
simply using the feature).  The only limitation I can see that might
need to be addressed is that there is no way to create an option tag for
a new feature that one has invented that is not yet thoroughly
standardized.

But I now discover RFC 3261 section 8.1.1.9:

   The option tags listed MUST only refer to extensions defined in
   standards-track RFCs.  This is to prevent servers from insisting that
   clients implement non-standard, vendor-defined features in order to
   receive service.  Extensions defined by experimental and
   informational RFCs are explicitly excluded from usage with the
   Supported header field in a request, since they too are often used to
   document vendor-defined extensions.

So I see that as a fundamental architectural decision, SIP does not
permit sending INVITEs that require specialized recipients, excepting if
the specialization is standards-track.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 15 11:40:52 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHjeg-0004Oe-1P; Thu, 15 Feb 2007 11:40:22 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHjee-0004ON-IY
	for sipping@ietf.org; Thu, 15 Feb 2007 11:40:20 -0500
Received: from rtp-iport-2.cisco.com ([64.102.122.149])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHjec-00081V-RS
	for sipping@ietf.org; Thu, 15 Feb 2007 11:40:20 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by rtp-iport-2.cisco.com with ESMTP; 15 Feb 2007 11:40:19 -0500
X-IronPort-AV: i="4.14,176,1170651600"; 
	d="scan'208"; a="114009944:sNHT66764988"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1FGeIoo020333; 
	Thu, 15 Feb 2007 11:40:18 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1FGeIVT026611; 
	Thu, 15 Feb 2007 11:40:18 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 15 Feb 2007 11:40:18 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 15 Feb 2007 11:40:18 -0500
Message-ID: <45D48CEF.8070504@cisco.com>
Date: Thu, 15 Feb 2007 11:40:15 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Stephen Terrill (MI/EEM)" <stephen.terrill@ericsson.com>
Subject: Re: [Sipping] Service Identification in SIP
References: <989763567E51AB48BE015DF7BB960CAD028861AF@esealmw111>
In-Reply-To: <989763567E51AB48BE015DF7BB960CAD028861AF@esealmw111>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 15 Feb 2007 16:40:18.0120 (UTC)
	FILETIME=[F9DD1480:01C7511F]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=13003; t=1171557618;
	x=1172421618; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20
	|To:=20=22Stephen=20Terrill=20(MI/EEM)=22=20<stephen.terrill@ericsson.com
	>; bh=H4EBpCo0+sk748NW0G32+WXK2eVxJY2qRiPLB0Uc+Rk=;
	b=kIymc1zJdr4qZDId2VX1CAWyMSJBBpRSjdFSVDItRsEiZJNQEirmQ/ZhZxHgKWqN3W5gaAh5
	1Ybva+sqF9r4m1DbKAkVvINHbhmV8Z3XGuTJm5yhOi/kw0TTTHLV1Jmh;
Authentication-Results: rtp-dkim-2; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: ce732c7d36989a1bd55104ba259c40a1
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Steve,

Thank you for providing some insight here. I know you are deeply into 
IMS, so I'm hoping we can take these as representative of the intended 
usage by those asking for service identification in SIP.

I think now it will be helpful to discuss each of these in more depth, 
to understand them better and that plausible ways to achieve these ends. 
(Or perhaps to consider if these ends are consistent with internet goals.)

I can't start that right this minute, but will post something before long.

	Thanks,
	Paul

Stephen Terrill (MI/EEM) wrote:
> Hi,
> 
> Firstly to some examples that come to my mind.  I know that the
> separation of PoC and from Telephony has been raised.  It has been
> stated a few times that "PoC is just a specialised floor control".  I
> don't believe that to be correct.  It is true that both Telephony and
> PoC use voice as a media; and that in addition PoC uses a floor control.
> In addition to that, an operator may wish to prioritise these
> differently in a resource constrained access network (e.g. a radio
> network).  For example, voice for Multimedia telephony may be
> prioritised higher than Push to talk voice - this is also for more
> efficient use of resources in a resource constrained access network.  A
> further differentiation between them, is that some of the "sub services"
> or "features" or <choose your favourite word> for them are different
> (e.g. communicate forwarding on subscriber not reachable) which
> ultimately results in the service logic being different for them.
> 
> An example of another service is (Instant) messaging.  Instant messaging
> may use MSRP to transfer text (e.g. "Hello") between users.  However,
> MSRP may also be used in the context of other services, such as
> telephony, or to transfer information from one device to another device
> (e.g. address book information).
> 
> Another example could be a SIP push type of service.  A SIP push service
> would be the ability to push e.g. data to a device.  The SIP push may be
> just a simple MESSAGE message, or an INVITE; however it should be
> distinguished from other uses of that as the information may not be
> destined for the end-user.  On top of that, then the exact application
> to which the information is being pushed would be useful.
> 
> These examples of PoC - Telephony; Messaging - Telephony are examples of
> different types of communication services where same media can be used
> in different contexts.  I should note that the examples of PoC,
> telephony and messaging are examples of bundled services, but what is
> wrong with that?  It provides a good way to increase the possibility of
> mxn interoperability by ensuring that clients will test against at least
> those capabilities.
> 
> In my understanding, the lack of the ability to identify these
> communication services, implies that "basic SIP communication" in itself
> is monolithic service.  That can be quite OK in a small, or controlled
> network, but I don't see that in a broader context, in particular a
> multiservice network.  I would like to point out here, that this does
> not imply anything relating to the implementation.  A vendor can build
> the support for the different services out of a bunch of common
> components; and there is of course a great deal of commonality that
> supports such re-use.
> 
> This leads me to my next topic, the example of a chess application using
> the voice.  While being discredited by some people, I think it does
> point out that that context where the media being used needs to be
> understood by some applications.  You could as well take the same
> example where the Chess application also established a separate control
> channel between the two devices, but the rest of the experience required
> was as per telephony.  In the 3GPP documentation, there is a
> differentiation between a "Communication service" and an "Application".
> The idea is that the "Applications" use the basic rules for
> communication defined by the "communication service".  The parallel
> example here, is Chess is the "application", while the communication
> service itself could be telephony.  We can further note here, that
> having an application re-use another existing communication is not new -
> it has been used with e.g. SMS for a while, where the SMS infrastructure
> has been used for delivering messages to the user, as well as to the
> device.  We use it in IP all the time as well ;-)
> 
> Moving onto the proposed solution draft itself, in addition for
> providing for the identification of the "application" and the
> "communication service", I have a few additional thoughts.
> 
> The draft seems to have the implication that all services (communication
> services in other terminology) that are to be identified have to be
> standardised within the IETF.  This is clearly an restriction that will
> not be acceptable to the industry as a whole.  My guess is that if such
> a statement is mandated by the IETF, it will just be words on paper
> which would likely be ignored in practice by other standardisation
> organisations.
> 
> The draft also only recognises the case where all explicitly knowledge
> of the service is required.  I think it would also be good to allow the
> case where knowledge of the service would be helpful, but not explicitly
> required.  This could be a case where the service to the end-user would
> not be degraded if this was not supported.  This functionality seems to
> be missing.
> 
> The draft seems to indicate that that application servers in the network
> are involved only at the request of the originating terminal.  This
> seems a strange restriction.  For some network support that supports the
> originating user, such as the one provided in the example "I would like
> this call recorded for later retrieval", such mechanisms may make sense,
> however as a general approach it seems a bit limited.  For example, in
> 3GPP networks there is the need to invoke application logic in the
> originating side and terminating side, not only at the request of the
> originating terminal, but as a consequence of the requested service.  If
> there is an SIP message contains indication of the requested service, a
> proxy can do this (or loose route to an AS that contains the necessary
> service logic).  I note here that the proposed service URN does not
> contain the address of a specific server containing the application
> logic.  An AS can look at that, and loose route to an application server
> that does contain such logic (If it is separated from the proxy),
> however I fail to see the difference for that proxy if it looks at a
> feature tag, or it looks at the service URN for this.  In both cases, it
> has to parse the message for the data, and modify the routing
> information.  Further, the route header could be removed in the
> originating network; and it is still required to identify the service in
> the terminating network.  As proxies may remove the route, it can't be
> used to transport the requested service.  From this perspective it is
> not clear to me the advantage of using a service URN in a route header
> compared to the use of a feature tag.
> 
> The draft seems to mandate GRUU - is this really so, is that really
> necessary.  It seems strange to place such a dependency, it also starts
> to move the GRUU as being the primary means of which to contact a user.
> This implies that the originating user decides which terminal I should
> receive my requested service on, even if I have two devices for that.
> As such, I don't believe that there should be a dependency on GRUU.
> 
> Parts of the draft seems to rely on the use of presence.  While presence
> could be a supportive mechanism, I doubt that we can mandate that the
> number you want to dial has be available in presence as I assume not all
> entities are available in presence (diallable network services, users in
> the PSTN).  This would be strange to mandated.
> 
> Moving on from the draft I have seen the questioned raised about "why
> can't we expand the power of SDP".  Such an approach would mandate that
> SIP methods use SDP, I don't think that is appropriate. 
> 
> I hear the appeals that the mechanisms should not be abused, and that we
> should not have 1000s of separate "communication services", and I fully
> agree with this. I see a smaller number of communication services, not a
> larger number; where those communication services are used to access
> other end-user-services.
> 
> *flame away*
> 
> Regards,
> 
> Steve
> 
> ---
> 
> 
> -----Original Message-----
> From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net] 
> Sent: Sunday, February 11, 2007 5:50 AM
> To: sipping@ietf.org
> Subject: Re: [Sipping] Service Identification in SIP
> 
> I've attempted to re-read the threads that Gonzalo referred to, and the
> major limitation that I see is the lack of any concrete example of why
> these identifiers are needed.  Probably the IMS people have discussed
> this so many times that they've forgotten the detailed rationale behind
> the requirement.  But I'm a net-head, (and most of the people here are
> net-heads) and the SIP systems that I've dealt with show no evidence of
> needing "service identification", they're pure proxies, and requests are
> passed transparently end-to-end.
> 
> Could somebody please lay out an example showing the utility of service
> identification?
> 
>    From: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>
> 
>    I think that everyone here agree that there is the possibility to use
> a
>    media in different ways, and that we need a distinction about what
> the
>    capabilities are and how they are used, whatever will be the
> mechanism
>    that at the end we will identify to do that.
> 
> Respectfully, I disagree.  We do need ways of identifying the
> capabilities of SIP agents and transports, but I see no necessity of
> creating specified bundles of capabailities, and then constraining
> dialogs to require an initial agreement as to which bundle of
> capabilities will be used.
> 
> One of the central features of Internet architecture is that agents
> negotiate compatibility on a feature-by-feature basis; there are no
> "profiles" which must be known on a network-wide basis.  (And I have
> observed that this feature is of significant commercial value.)
> 
> In regard to identifying an "application" to process the dialog (and I
> use "application" in a standard computing way), I see no reason that a
> user cannot select URI-parameters to append to his AOR, and arrange for
> his UA to route invites with those URI-parameters to the appropriate
> application.  (And given the rich mechanisms for redirection in SIP, the
> application need not even reside at a UA which is registered for the
> AOR.)  These URI-parameters can be formally or informally standardized,
> or established ad-hoc by each user.
> 
> In regard to PoC (as someone said, "a specialized floor-control
> protocol") and similar features, SIP currently has mechanisms for
> indicating support of features and requiring that a recipient support a
> feature before accepting a call (the Supported and Requires headers, and
> simply using the feature).  The only limitation I can see that might
> need to be addressed is that there is no way to create an option tag for
> a new feature that one has invented that is not yet thoroughly
> standardized.
> 
> But I now discover RFC 3261 section 8.1.1.9:
> 
>    The option tags listed MUST only refer to extensions defined in
>    standards-track RFCs.  This is to prevent servers from insisting that
>    clients implement non-standard, vendor-defined features in order to
>    receive service.  Extensions defined by experimental and
>    informational RFCs are explicitly excluded from usage with the
>    Supported header field in a request, since they too are often used to
>    document vendor-defined extensions.
> 
> So I see that as a fundamental architectural decision, SIP does not
> permit sending INVITEs that require specialized recipients, excepting if
> the specialization is standards-track.
> 
> Dale
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use
> sip-implementors@cs.columbia.edu for questions on current sip Use
> sip@ietf.org for new developments of core SIP
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 15 12:58:15 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHkqc-0004dv-53; Thu, 15 Feb 2007 12:56:46 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHkqa-0004We-DI
	for sipping@ietf.org; Thu, 15 Feb 2007 12:56:44 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHkqU-0007rI-DW
	for sipping@ietf.org; Thu, 15 Feb 2007 12:56:44 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	B20842087D; Thu, 15 Feb 2007 18:56:29 +0100 (CET)
X-AuditID: c1b4fb3c-b17cbbb0000007de-40-45d49ecd466e 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	98EF32048F; Thu, 15 Feb 2007 18:56:29 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 15 Feb 2007 18:56:29 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Thu, 15 Feb 2007 18:56:28 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD028862D1@esealmw111>
In-Reply-To: <45D48CEF.8070504@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdRH//zGEXkhrGeTpe2K2o6WSEqVAACayig
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>
X-OriginalArrivalTime: 15 Feb 2007 17:56:29.0133 (UTC)
	FILETIME=[9E664FD0:01C7512A]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: cbb41f2dbf0f142369614756642005e3
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi Paul,

Thanks.  On ... "as representative of the intended usage by those asking
for service identification in SIP" .. Each community is made up of
participants, so of course I can't speak for everyone, but it gives at
least my understanding.  Maybe I have overlooked something -> for that,
time will tell.

BR,

..//Steve=20

-----Original Message-----
From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
Sent: Thursday, February 15, 2007 5:40 PM
To: Stephen Terrill (MI/EEM)
Cc: sipping@ietf.org
Subject: Re: [Sipping] Service Identification in SIP

Steve,

Thank you for providing some insight here. I know you are deeply into
IMS, so I'm hoping we can take these as representative of the intended
usage by those asking for service identification in SIP.

I think now it will be helpful to discuss each of these in more depth,
to understand them better and that plausible ways to achieve these ends.

(Or perhaps to consider if these ends are consistent with internet
goals.)

I can't start that right this minute, but will post something before
long.

	Thanks,
	Paul

Stephen Terrill (MI/EEM) wrote:
> Hi,
>=20
> Firstly to some examples that come to my mind.  I know that the=20
> separation of PoC and from Telephony has been raised.  It has been=20
> stated a few times that "PoC is just a specialised floor control".  I=20
> don't believe that to be correct.  It is true that both Telephony and=20
> PoC use voice as a media; and that in addition PoC uses a floor
control.
> In addition to that, an operator may wish to prioritise these=20
> differently in a resource constrained access network (e.g. a radio=20
> network).  For example, voice for Multimedia telephony may be=20
> prioritised higher than Push to talk voice - this is also for more=20
> efficient use of resources in a resource constrained access network. =20
> A further differentiation between them, is that some of the "sub
services"
> or "features" or <choose your favourite word> for them are different=20
> (e.g. communicate forwarding on subscriber not reachable) which=20
> ultimately results in the service logic being different for them.
>=20
> An example of another service is (Instant) messaging.  Instant=20
> messaging may use MSRP to transfer text (e.g. "Hello") between users.

> However, MSRP may also be used in the context of other services, such=20
> as telephony, or to transfer information from one device to another=20
> device (e.g. address book information).
>=20
> Another example could be a SIP push type of service.  A SIP push=20
> service would be the ability to push e.g. data to a device.  The SIP=20
> push may be just a simple MESSAGE message, or an INVITE; however it=20
> should be distinguished from other uses of that as the information may

> not be destined for the end-user.  On top of that, then the exact=20
> application to which the information is being pushed would be useful.
>=20
> These examples of PoC - Telephony; Messaging - Telephony are examples=20
> of different types of communication services where same media can be=20
> used in different contexts.  I should note that the examples of PoC,=20
> telephony and messaging are examples of bundled services, but what is=20
> wrong with that?  It provides a good way to increase the possibility=20
> of mxn interoperability by ensuring that clients will test against at=20
> least those capabilities.
>=20
> In my understanding, the lack of the ability to identify these=20
> communication services, implies that "basic SIP communication" in=20
> itself is monolithic service.  That can be quite OK in a small, or=20
> controlled network, but I don't see that in a broader context, in=20
> particular a multiservice network.  I would like to point out here,=20
> that this does not imply anything relating to the implementation.  A=20
> vendor can build the support for the different services out of a bunch

> of common components; and there is of course a great deal of=20
> commonality that supports such re-use.
>=20
> This leads me to my next topic, the example of a chess application=20
> using the voice.  While being discredited by some people, I think it=20
> does point out that that context where the media being used needs to=20
> be understood by some applications.  You could as well take the same=20
> example where the Chess application also established a separate=20
> control channel between the two devices, but the rest of the=20
> experience required was as per telephony.  In the 3GPP documentation,=20
> there is a differentiation between a "Communication service" and an
"Application".
> The idea is that the "Applications" use the basic rules for=20
> communication defined by the "communication service".  The parallel=20
> example here, is Chess is the "application", while the communication=20
> service itself could be telephony.  We can further note here, that=20
> having an application re-use another existing communication is not new

> - it has been used with e.g. SMS for a while, where the SMS=20
> infrastructure has been used for delivering messages to the user, as=20
> well as to the device.  We use it in IP all the time as well ;-)
>=20
> Moving onto the proposed solution draft itself, in addition for=20
> providing for the identification of the "application" and the=20
> "communication service", I have a few additional thoughts.
>=20
> The draft seems to have the implication that all services=20
> (communication services in other terminology) that are to be=20
> identified have to be standardised within the IETF.  This is clearly=20
> an restriction that will not be acceptable to the industry as a whole.

> My guess is that if such a statement is mandated by the IETF, it will=20
> just be words on paper which would likely be ignored in practice by=20
> other standardisation organisations.
>=20
> The draft also only recognises the case where all explicitly knowledge

> of the service is required.  I think it would also be good to allow=20
> the case where knowledge of the service would be helpful, but not=20
> explicitly required.  This could be a case where the service to the=20
> end-user would not be degraded if this was not supported.  This=20
> functionality seems to be missing.
>=20
> The draft seems to indicate that that application servers in the=20
> network are involved only at the request of the originating terminal.

> This seems a strange restriction.  For some network support that=20
> supports the originating user, such as the one provided in the example

> "I would like this call recorded for later retrieval", such mechanisms

> may make sense, however as a general approach it seems a bit limited.

> For example, in 3GPP networks there is the need to invoke application=20
> logic in the originating side and terminating side, not only at the=20
> request of the originating terminal, but as a consequence of the=20
> requested service.  If there is an SIP message contains indication of=20
> the requested service, a proxy can do this (or loose route to an AS=20
> that contains the necessary service logic).  I note here that the=20
> proposed service URN does not contain the address of a specific server

> containing the application logic.  An AS can look at that, and loose=20
> route to an application server that does contain such logic (If it is=20
> separated from the proxy), however I fail to see the difference for=20
> that proxy if it looks at a feature tag, or it looks at the service=20
> URN for this.  In both cases, it has to parse the message for the=20
> data, and modify the routing information.  Further, the route header=20
> could be removed in the originating network; and it is still required=20
> to identify the service in the terminating network.  As proxies may=20
> remove the route, it can't be used to transport the requested service.

> From this perspective it is not clear to me the advantage of using a=20
> service URN in a route header compared to the use of a feature tag.
>=20
> The draft seems to mandate GRUU - is this really so, is that really=20
> necessary.  It seems strange to place such a dependency, it also=20
> starts to move the GRUU as being the primary means of which to contact
a user.
> This implies that the originating user decides which terminal I should

> receive my requested service on, even if I have two devices for that.
> As such, I don't believe that there should be a dependency on GRUU.
>=20
> Parts of the draft seems to rely on the use of presence.  While=20
> presence could be a supportive mechanism, I doubt that we can mandate=20
> that the number you want to dial has be available in presence as I=20
> assume not all entities are available in presence (diallable network=20
> services, users in the PSTN).  This would be strange to mandated.
>=20
> Moving on from the draft I have seen the questioned raised about "why=20
> can't we expand the power of SDP".  Such an approach would mandate=20
> that SIP methods use SDP, I don't think that is appropriate.
>=20
> I hear the appeals that the mechanisms should not be abused, and that=20
> we should not have 1000s of separate "communication services", and I=20
> fully agree with this. I see a smaller number of communication=20
> services, not a larger number; where those communication services are=20
> used to access other end-user-services.
>=20
> *flame away*
>=20
> Regards,
>=20
> Steve
>=20
> ---
>=20
>=20
> -----Original Message-----
> From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net]
> Sent: Sunday, February 11, 2007 5:50 AM
> To: sipping@ietf.org
> Subject: Re: [Sipping] Service Identification in SIP
>=20
> I've attempted to re-read the threads that Gonzalo referred to, and=20
> the major limitation that I see is the lack of any concrete example of

> why these identifiers are needed.  Probably the IMS people have=20
> discussed this so many times that they've forgotten the detailed=20
> rationale behind the requirement.  But I'm a net-head, (and most of=20
> the people here are
> net-heads) and the SIP systems that I've dealt with show no evidence=20
> of needing "service identification", they're pure proxies, and=20
> requests are passed transparently end-to-end.
>=20
> Could somebody please lay out an example showing the utility of=20
> service identification?
>=20
>    From: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>
>=20
>    I think that everyone here agree that there is the possibility to=20
> use a
>    media in different ways, and that we need a distinction about what=20
> the
>    capabilities are and how they are used, whatever will be the=20
> mechanism
>    that at the end we will identify to do that.
>=20
> Respectfully, I disagree.  We do need ways of identifying the=20
> capabilities of SIP agents and transports, but I see no necessity of=20
> creating specified bundles of capabailities, and then constraining=20
> dialogs to require an initial agreement as to which bundle of=20
> capabilities will be used.
>=20
> One of the central features of Internet architecture is that agents=20
> negotiate compatibility on a feature-by-feature basis; there are no=20
> "profiles" which must be known on a network-wide basis.  (And I have=20
> observed that this feature is of significant commercial value.)
>=20
> In regard to identifying an "application" to process the dialog (and I

> use "application" in a standard computing way), I see no reason that a

> user cannot select URI-parameters to append to his AOR, and arrange=20
> for his UA to route invites with those URI-parameters to the=20
> appropriate application.  (And given the rich mechanisms for=20
> redirection in SIP, the application need not even reside at a UA which

> is registered for the
> AOR.)  These URI-parameters can be formally or informally=20
> standardized, or established ad-hoc by each user.
>=20
> In regard to PoC (as someone said, "a specialized floor-control
> protocol") and similar features, SIP currently has mechanisms for=20
> indicating support of features and requiring that a recipient support=20
> a feature before accepting a call (the Supported and Requires headers,

> and simply using the feature).  The only limitation I can see that=20
> might need to be addressed is that there is no way to create an option

> tag for a new feature that one has invented that is not yet thoroughly

> standardized.
>=20
> But I now discover RFC 3261 section 8.1.1.9:
>=20
>    The option tags listed MUST only refer to extensions defined in
>    standards-track RFCs.  This is to prevent servers from insisting
that
>    clients implement non-standard, vendor-defined features in order to
>    receive service.  Extensions defined by experimental and
>    informational RFCs are explicitly excluded from usage with the
>    Supported header field in a request, since they too are often used
to
>    document vendor-defined extensions.
>=20
> So I see that as a fundamental architectural decision, SIP does not=20
> permit sending INVITEs that require specialized recipients, excepting=20
> if the specialization is standards-track.
>=20
> Dale
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 15 14:09:48 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHlx0-0004fb-RD; Thu, 15 Feb 2007 14:07:26 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHlwy-0004dQ-OY
	for sipping@ietf.org; Thu, 15 Feb 2007 14:07:24 -0500
Received: from exprod6og50.obsmtp.com ([64.18.1.181])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1HHlwx-00005A-36
	for sipping@ietf.org; Thu, 15 Feb 2007 14:07:24 -0500
Received: from source ([192.150.11.134]) by exprod6ob50.postini.com
	([64.18.5.12]) with SMTP; Thu, 15 Feb 2007 11:06:44 PST
Received: from inner-relay-3.eur.adobe.com (inner-relay-3.adobe.com
	[192.150.20.198] (may be forged))
	by outbound-smtp-1.corp.adobe.com (8.12.10/8.12.10) with ESMTP id
	l1FJ6UvV014074; Thu, 15 Feb 2007 11:06:30 -0800 (PST)
Received: from fe2.corp.adobe.com (fe2.corp.adobe.com [10.8.192.72])
	by inner-relay-3.eur.adobe.com (8.12.10/8.12.9) with ESMTP id
	l1FJ64mg004998; Thu, 15 Feb 2007 11:06:34 -0800 (PST)
Received: from namail5.corp.adobe.com ([10.8.192.88]) by fe2.corp.adobe.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 15 Feb 2007 11:06:27 -0800
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Subject: RE: Device architecture? WAS: RE: [Sipping] Service Identification in
	SIP
Date: Thu, 15 Feb 2007 11:05:40 -0800
Message-ID: <24CCCC428EFEA2469BF046DB3C7A8D2249EED3@namail5.corp.adobe.com>
In-Reply-To: <B0AF4CCDCE7F4D439EAC01F3331B1329265BBF@esealmw111>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Device architecture? WAS: RE: [Sipping] Service Identification
	in SIP
Thread-Index: AcdQh+oX5uAth8ViTQaBZUsL/3x7LwAABNiwAAKV3SAAAItz0AAnlufQ
References: <24CCCC428EFEA2469BF046DB3C7A8D2249ED47@namail5.corp.adobe.com>
	<B0AF4CCDCE7F4D439EAC01F3331B1329265BBF@esealmw111>
From: "Henry Sinnreich" <hsinnrei@adobe.com>
To: =?iso-8859-1?Q?G=F6ran_Eriksson_AP_=28KI/EAB=29?=
	<goran.ap.eriksson@ericsson.com>, "Elwell, John" <john.elwell@siemens.com>
X-OriginalArrivalTime: 15 Feb 2007 19:06:27.0310 (UTC)
	FILETIME=[64B548E0:01C75134]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8a85b14f27c9dcbe0719e27d46abc1f8
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

G=F6ran,

I really appreciate very much even the limited part of agreement we have =
here and the right of operators to try to monetize IMS the best they =
can.

Besides the extra charges, I had forgotten to mention the control over =
features and innovation that such an approach would give the operators.
=20
I suspect that endpoint application developers will dislike this even =
more. Endpoint application developers will have even more incentives to =
resist such control by operators over their product development.

This is already felt by mobile phone manufacturers who are not that =
happy to have their features controlled by operators. This bubble is =
going to burst one way or another.

Most likely that this issue will be decide in the Internet marketplace =
and by regulators, so please file this exchange so we can draw some =
insight on predictions in the near future.

Thanks, Henry

-----Original Message-----
From: G=F6ran Eriksson AP (KI/EAB) =
[mailto:goran.ap.eriksson@ericsson.com]=20
Sent: Wednesday, February 14, 2007 4:40 PM
To: Henry Sinnreich; Elwell, John
Cc: sipping@ietf.org
Subject: RE: Device architecture? WAS: RE: [Sipping] Service =
Identification in SIP

I very much agree with You there Henry! I do recognise the need to give =
*some classes of applications*, read telephony replacement, the =
possibility
to make whatever nethead, monolithic, tower of Babel, architecture that =
that industry need BUT I DO NOT, personally, what
any particular application class enforce/prevent/impose such an =
architecture on other applications, such as game, chat, etc.

The issue is of course that there is a great interest out there for =
telephony as such, and of course all telco operators are concerned with
the future business. This may easily lead to application/industry =
specific requriements getting into the (SIP) protocol(s), something that =
at least should
be avoided as much as possible, and if not possible, then at least =
contained.

And You're right- they will (try to) plant such charging devices there, =
because the name of the (market economy) game is they will if they can.

The funny side of this is that the IMS must probably also support gaming =
traffic, but not in a way that kills gaming traffic for instance by =
stupid charging.=20
Telco operators are more and more under pressure to provide real =
consumer value and charge reasonably. We'll see...


Regards
G=F6ran

PS. The show I would like to see is the one where users get to know that =
they will not have to pay for neither telephony, broadband access, =
mobile access,
expensive PC programs or OS's, music, IPTV, games, PC, mobile phone, etc =
for that matter! And yes, please no commercials also, :-)!!



-----Original Message-----
From: Henry Sinnreich [mailto:hsinnrei@adobe.com]=20
Sent: den 15 februari 2007 01:01
To: G=F6ran Eriksson AP (KI/EAB); Elwell, John
Cc: sipping@ietf.org
Subject: RE: Device architecture? WAS: RE: [Sipping] Service =
Identification in SIP

G=F6ran Eriksson wrote:

> having some IP-based proprietary gaming interaction protocol

Yes, your understanding is correct and there is a booming gaming =
industry out there working exactly this way.

Where we disagree is that there is a need felt in the gaming industry to =
have application level network infrastructure in the middle and helpful =
operators to charge extra for IP based games.

But there no sense in us trying to convince anyone to avoid planting =
network intermediaries for extra charging for IP based games.=20

Could I just get a ticket to the show to watch the enthusiastic =
reception by users and game manufacturers? :-)

Thanks, Henry

-----Original Message-----
From: G=F6ran Eriksson AP (KI/EAB) =
[mailto:goran.ap.eriksson@ericsson.com]
Sent: Wednesday, February 14, 2007 3:34 PM
To: Elwell, John
Cc: sipping@ietf.org
Subject: Device architecture? WAS: RE: [Sipping] Service Identification =
in SIP

My humble, perhaps faulty, understanding of the "chess game" was not =
that voice commands were used to control the chess itself, rather that =
the chess game, while having some IP-based proprietary gaming =
interaction protocol, also had a feature letting the users talk to each =
others while playing. The scenario are two, or more, users play a chess =
game of the (Inter)net, talkin to eachother from time to time during the =
game (yeah, don't ask me if I think people will love this!!).

The purpose, as I understand it, was a)to show that the 'voice =
component' in the chess game application is similar to that in a stand =
alone POC-Only application when it comes to exchange and transfer of =
voice samples and associated floor control and b) to point to =
differences, for instance when it comes to supplementary services, which =
may not apply in the chess game case.

Perhaps it is important to be clear on the assumed user device =
realisation, as far as I understand it:

 The "chess game + voice" example is, as I understand it, discussed =
under the assumption that voice and chess game is  managed by custome =
designed appliction, or client, which is a necessary/likely approach in =
a mobile device given the  limited MMI (and also other factors in the =
mobile context).

 In a PC, I could easily handle this as two different =
applications/client, like when I use Skype/MSN to talk to my friends =
while  playing some turn based game, but again, the discussion is held =
under the assumption that usability requirements  leads to an =
application client integrating voice and chess game events.

The alternative device realisation You describe- different clients =
managing different portions of what is negotiated in the SDP or MSRP =
MIME attachment- is, I think, a good illustration of the need to be =
precise and explicit about the assumed device architecture when =
discussing the "service identification" issue(s).

Regards
G=F6ran




-----Original Message-----
From: Elwell, John [mailto:john.elwell@siemens.com]
Sent: den 14 februari 2007 11:57
To: Henry Sinnreich; sipping@ietf.org
Subject: RE: [Sipping] Service Identification in SIP

Having followed this thread, I too find the chess example very =
contrived. If chess moves are sent as voice commands, then surely it =
would be acceptable for the normal telephone application to answer the =
call, i.e., the remote player listens to the move, performs this move on =
a conventional or electronic chess board and speaks his next move. The =
calling UA should not care whether the call is answered by a computer =
application or a person.
Furthermore, even if the audio needs to consumed by a particular chess =
application at the called UA, then other media in the same session such =
as video or messaging might not be. So any qualification of the nature =
of media to be exchanged probably should be at the media level, not at =
the session level or SIP level. As yet another variation, what if the =
players set up a conventional telephone call and then mutually agree to =
switch in their chess applications?

Now, let's consider another case. I establish an MSRP session. Within =
this MSRP session I can send messages with various MIME attachments that =
are intended for consumption by various applications. I should not need =
to know in advance what applications I will use during the course of the =
MSRP session. All I should put in the INVITE request is the fact that I =
want to establish an MSRP session.

John=20

> -----Original Message-----
> From: Henry Sinnreich [mailto:hsinnrei@adobe.com]
> Sent: 13 February 2007 18:19
> To: sipping@ietf.org
> Subject: RE: [Sipping] Service Identification in SIP
>=20
> It must be my lack of imagination for not being able to come up with a =

> credible usage scenario, given the following:
>=20
> - Is playing chess by voice a credible usage scenario? :-) Especially=20
> since voice recognition is now so often found in PCs and in mobile=20
> phones as well. In other words, chess playing by voice may also be an=20
> endpoint application, without any network involvement. Actually, if=20
> placed in the endpoint it works just as well on poor bandwidth links=20
> by contrast to chess in the network that may require adequate=20
> broadband.
>=20
> - Is anyone's boss or spouse going to approve extra phone charges=20
> above the flat rate for playing chess at a time when the gaming=20
> industry has pioneered free VoIP?
> Will such a service be taxed extra and if yes using by what criteria?
> Gambling comes to my mind :-)
>=20
> - VoIP is reported in the press as a give-away and I have seen no=20
> reports of any VoIP provider recording a profit on VoIP. I assume the=20
> losses are often hidden by bundling in triple play, etc.
>=20
> Given my lack of imagination, can anyone share a more credible usage=20
> scenario for Service Identification in SIP?
>=20
> What's the use of a standard without a credible usage scenario?
>=20
> Thanks, Henry
>=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use =
sip-implementors@cs.columbia.edu for questions on current sip Use =
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use =
sip-implementors@cs.columbia.edu for questions on current sip Use =
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 15 17:18:40 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHouq-0004GT-5Q; Thu, 15 Feb 2007 17:17:24 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHouo-0004GN-N1
	for sipping@ietf.org; Thu, 15 Feb 2007 17:17:22 -0500
Received: from c243-111.rim.net ([216.9.243.111] helo=MHS01YKF.rim.net)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHoul-0002uu-3Q
	for sipping@ietf.org; Thu, 15 Feb 2007 17:17:22 -0500
Importance: normal
Priority: normal
Received: from ngw02ykf.rim.net ([10.102.108.31]) by MHS01YKF.rim.net with
	Microsoft SMTPSVC(6.0.3790.211); Thu, 15 Feb 2007 17:17:08 -0500
Received: from XCH21YKF.rim.net ([10.102.100.36]) by ngw02ykf.rim.net (SAVSMTP
	3.1.6.45) with SMTP id M2007021517170815805 ;
	Thu, 15 Feb 2007 17:17:08 -0500
Received: from XCH47YKF.rim.net ([10.64.31.217]) by XCH21YKF.rim.net with
	Microsoft SMTPSVC(5.0.2195.6713); Thu, 15 Feb 2007 17:17:08 -0500
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.181
Content-Class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Thu, 15 Feb 2007 17:17:06 -0500
Message-ID: <61968779B8AC4C4BAB421D4C12F008C00915140A@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
thread-index: AcdK2+Kqf8XxF4RlRBKez3RRQPczsAAAHOqwAZlFC4A=
From: "Andrew Allen" <aallen@rim.com>
To: "Henry Sinnreich" <hsinnrei@adobe.com>,
	"Paul Kyzivat" <pkyzivat@cisco.com>,
	"Salvatore Loreto \(JO/LMF\)" <salvatore.loreto@ericsson.com>
X-OriginalArrivalTime: 15 Feb 2007 22:17:08.0457 (UTC)
	FILETIME=[0829C590:01C7514F]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 9f79b8e383fd3af2b1b5b1d0910f6094
Cc: Andrew Allen <aallen@rim.com>, ravishankar.shiroor@wipro.com,
	sipping@ietf.org, Gonzalo.Camarillo@ericsson.com, mary.barnes@nortel.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


=20
IN response to Henry:

RFC 4485 also states:

   Session initiation also depends on the ability of the called party to
   have enough information about the session itself to make a decision
   on whether to join.  That information includes data about the caller,
   the purpose for the invitation, and parameters of the session itself.
   For this reason, SIP includes this kind of information.

IMHO "the purpose for the invitation" also includes information about
what type of service or application this session initiation request is
related to and SDP doesn't provide sufficent information to definitively
determine this.

I think we can all debate the advantages and disadvantages of building
intelligence in the network as opposed to distributing it to the
endpoints and I feel that I am also generally on the distributed side of
that particular argument however SIP allows for both models and many
services/applications utilise both intelligence in the endpoint and
intelligence in Application Servers in the network. A prime example is
Presence where Presence can be provided with SIP using only SIP
endpoints however many practical implementations make use of network
based Presence Servers and Resource List Servers in order to create a
scalable Presence Service and significant IETF work has been done on
creating SIP based and other related mechanisms (e.g XCAP) to support
Presence using Presence Servers and Resource List Servers. Being purist
philosophical about moving all intelligence to the very edge doesn't
always result in the most optimal solution as the scalability advantages
of having all the intelligence in the endpoints has to be considered
along with the scalability disadvantages of having all the signalling
transported to the endpoints. In limited and high cost bandwidth access
networks like cellular the scalability disadvantages of having all
signaling transported to the endpoint can sometimes outweigh the
advantages of distributed intelligence.=20

My understanding is that SIP doesn't take any position on the business
models used for communication service provision and IETF has been
cooperating closely with those who are using SIP for carrier based
services. Some people may not think that the carrier centric service
model is the way to go in the future but with an increasing number of
competing service providers in the market this argument will be settled
by market competition and is not really the role of engineering to
settle.=20

Also don't think that failing to develop or delaying an IETF based
solution will stop 3GPP from developing their own solution. There are
already proposals being agreed within 3GPP based on Accept-Contact
header and feature tags to implement this in what I believe is a way
that potentially breaks interoperability between IMS and other SIP
networks. This is because ithis proposalt requires pre knowledge of
these service identifiers and their use in all networks for
communication without providing a mechanism to obtain them and also
through its use of Accept-Contact potentially signiifcantly limits the
flexibility use of that header for its intended purpose. However RFC
3427 doesn't require that new feature tags be discussed in SIPPING even
when the proposed usage is required for interoperability. So basically
we have a choice between working on an open internet friendly solution
to service identification or allowing by default a 3GPP specific
solution that potentially will create interoperability issues between
internet users and about 2 Billion Mobile users and several hundred
million fixed carrier device users. Jonthan's proposal addresses these
issues.

Something else from RFC 4485:

   Proxies can ignore bodies:  In order for proxies to scale well, they
      must be able to operate with minimal message processing.  SIP has
      been engineered so that proxies can always ignore bodies.
      Extensions SHOULD NOT require proxies to examine bodies.

As already stated I don't believe that SDP always provides sufficent
information to definitively determine the application or service that
the session initiation request is related to and even when it does this
requires that a Proxy examine the body in order to determine whether the
request needs to be routed to some intermediate server that performs
service logic or make other service based routing decisions. I think it
is far preferable to have some kind of service identifier that
specifcally identifies the service logic processing needed included in
an appropriate place in the request as proposed by Jonathan and make
route decisions based on that than to have proxies attempt to do this
based on examing SDP or based on the content type. As well as being more
efficient in terms of proxy processing this avoids routing failures due
to encrypted bodies or unexpected multipart MIME content types or new
alternative content types not yet understood by the proxy causing the
service routing to fail.

So you can count me in the supporting the work on service identification
based on Jonathan's draft column.

Andrew



-----Original Message-----
From: Henry Sinnreich [mailto:hsinnrei@adobe.com]=20
Sent: Wednesday, February 07, 2007 12:34 PM
To: Paul Kyzivat; Salvatore Loreto (JO/LMF)
Cc: Gonzalo.Camarillo@ericsson.com; sipping@ietf.org;
mary.barnes@nortel.com; ravishankar.shiroor@wipro.com
Subject: RE: [Sipping] Service Identification in SIP

> I still don't get the intent of identifying the application. It seems
to > me this is evil - akin to sending an email message and identifying
that > the recipient must use Outlook to receive it, not Thunderbird,
Eudora,=20
> gmail, or whatever.

This goes back to the old discussion about the IN versus the "dumb
network" and we know how this has played out.

Placing intelligence (AKA services) in the network is just "dumb design"
IMHO for all the good reasons, but makes no sense to discuss on this
list.

SIP was designed as a rendezvous protocol for multimedia communications
on the Internet and not to replicate ISDN, etc., as has been so well
documented in RFC 4485:

   "SIP is not meant to be used as a strict Public Switched Telephone
   Network (PSTN) signaling replacement.  It is not a superset of the
   Integrated Services Digital Network (ISDN) User Part (ISUP).
   Although it can support gatewaying of PSTN signaling and can provide
   many features present in the PSTN, the mere existence of a feature or
   capability in the PSTN is not a justification for its inclusion in
   SIP."

Which makes me wonder how many folks have read RFC 4485 and taken it to
heart or may have chosen to forget it.

Thanks, Henry

-----Original Message-----
From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
Sent: Wednesday, February 07, 2007 11:13 AM
To: Salvatore Loreto (JO/LMF)
Cc: mary.barnes@nortel.com; sipping@ietf.org;
ravishankar.shiroor@wipro.com; Gonzalo.Camarillo@ericsson.com
Subject: Re: [Sipping] Service Identification in SIP

I still don't get the intent of identifying the application. It seems to

me this is evil - akin to sending an email message and identifying that
the recipient must use Outlook to receive it, not Thunderbird, Eudora,
gmail, or whatever.

I also wish somebody could give better examples of a communication
service. Chess is too artificial to be believable, and IMO PoC is only a

service because of the way the implementation has been done.=20
Conceptually PoC could be nothing but a special floor control protocol
accompanying a voice session. In that case none of the Service mechanism

would be needed.

	Paul

Salvatore Loreto (JO/LMF) wrote:
> hi Ravi,
>=20
> in IMS a Service is a "type of communication defined by a service=20
> definition that specifies the rules and procedures and allowed media,
if
> any, for a specific type of communication and that utilises the IMS=20
> enabler. An IMS communication service defines what is possible within
a
> single SIP session or standalone transaction."
> So a Service implies some combination of capabilities and ways to use=20
> those capabilities.
> If you want the Service specifies the context for a communication.
>=20
> Instead the Application define the specific communication (or if you=20
> want application) that you want run in the specific context. So you=20
> should have several different applications designed to run on a
specific
> environment or that have different performance or behavior in
different
> context (running on top of different service).
>=20
> Lets go to Jonathan's example of the Chess.
> In his example the Chess is a service, for IMS instead is an
application
> that you can run on top of different service.
> For example you can run Chess application on top of PoC service or on=20
> top of the "standard" service. So let say you run Chess on top of PoC=20
> service, then you can use only voice in the way that PoC specifies.
>=20
> /sal
>=20
>=20
>=20
>=20
> On Thu, 2007-02-01 at 16:39 +0530, ravishankar.shiroor@wipro.com
wrote:
>> Hello,
>>
>> IMO, Service Identification is definitely needed and should be taken
up
>> as a WG as proposed below.
>>
>> Further, I think Service Identification scope should be as per the=20
>> requirement in Jonathan's draft, as I still do not understand the
need
>> for having a Communication Service Identifier and another Application

>> Identifier as given in the 3gpp-ics-requirements document.
>>
>> Regards,
>> Ravi.
>>
>> --
>> =20
>> Ravishankar. G. Shiroor
>> Wipro Technologies, Bangalore.
>> =20
>> ravishankar.shiroor@wipro.com
>> --
>> =20
>>
>>> -----Original Message-----
>>> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
>>> Sent: Wednesday, January 31, 2007 8:30 PM
>>> To: sipping
>>> Cc: Mary Barnes
>>> Subject: [Sipping] Service Identification in SIP
>>>
>>> Folks,
>>>
>>> during the last year, there have been a number of threads on the=20
>>> mailing list discussing service identification in SIP. In=20
>>> particular, the following thread discussed 3GPP requirements on this

>>> area:
>>>
>>> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
>>>
>>> At a later point, those 3GPP requirements were documented in the=20
>>> following (expired by now) Internet Draft, which was presented in a=20
>>> face-to-face IETF meeting.
>>>
>>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
>>> cs-requirements-00.txt
>>>
>>> Lately, the following draft has been submitted to the archives:
>>>
>>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
>>> rvice-identification-00.txt
>>>
>>> This draft proposes a solution that intends to meet known=20
>>> requirements in this area.
>>>
>>> Given the amount of discussions on issues related to service=20
>>> identification we have seen during the last year, we believe this is

>>> a fairly important topic for the SIPPING WG to tackle. At some=20
>>> point, we would like to have a milestone with its associated WG item

>>> dealing with this issue.
>>>
>>> Therefore, we would like to encourage people to read this draft and=20
>>> send comments to the list. Additionally, if you believe that this=20
>>> work is (or is not) important, we would like to hear that as well.
>>>
>>> Cheers,
>>>
>>> Gonzalo
>>> SIPPING co-chair
>>>
>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP Use=20
>>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
>>> sip@ietf.org for new developments of core SIP
>>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP Use=20
>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
>> sip@ietf.org for new developments of core SIP
>=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential =
information, privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 15 18:37:45 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHq92-00020Y-Sp; Thu, 15 Feb 2007 18:36:08 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHq91-00020T-Py
	for sipping@ietf.org; Thu, 15 Feb 2007 18:36:07 -0500
Received: from c243-111.rim.net ([216.9.243.111] helo=MHS02YKF.rim.net)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHq8y-00081a-5n
	for sipping@ietf.org; Thu, 15 Feb 2007 18:36:07 -0500
Importance: normal
Priority: normal
Received: from ngw01ykf.rim.net ([10.102.108.30]) by MHS02YKF.rim.net with
	Microsoft SMTPSVC(6.0.3790.211); Thu, 15 Feb 2007 18:35:53 -0500
Received: from XCH21YKF.rim.net ([10.102.100.36]) by ngw01ykf.rim.net (SAVSMTP
	3.1.6.45) with SMTP id M2007021518355323485 ;
	Thu, 15 Feb 2007 18:35:53 -0500
Received: from XCH47YKF.rim.net ([10.64.31.217]) by XCH21YKF.rim.net with
	Microsoft SMTPSVC(5.0.2195.6713); Thu, 15 Feb 2007 18:35:53 -0500
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.181
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Thu, 15 Feb 2007 18:35:51 -0500
Message-ID: <61968779B8AC4C4BAB421D4C12F008C009151456@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdMWCOCuK5QoXPJRKyMHDPmXeQ6uQE+1Q3Q
From: "Andrew Allen" <aallen@rim.com>
To: "Salvatore Loreto \(JO/LMF\)" <salvatore.loreto@ericsson.com>,
	"Paul Kyzivat" <pkyzivat@cisco.com>
X-OriginalArrivalTime: 15 Feb 2007 23:35:53.0569 (UTC)
	FILETIME=[088CA510:01C7515A]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: bc102ac530ba955ef81f1f75b8bebe44
Cc: Gonzalo.Camarillo@ericsson.com, sipping@ietf.org, mary.barnes@nortel.com,
	ravishankar.shiroor@wipro.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


I agree with Salvatore that IM is a better example of where service
identification may have a role in expressing a preference for a
particular application. Certainly there are different emoticons and
other value added features proprietary to different messaging
applications so having the ability to express preferences to use a
preferred application seems to me to have value. I think that email is a
bad counter example as with email the email client interfaces to the
email server not directly to the other email client.

Where I disagree with Salvatore is that the fact that the use of
callerprefs doesn't support well, if at all, the inclusion of
intermediate servers into the call path is not a problem for 3GPP as
well. From my viewpoint 3GPP needs to make use of Callerprefs in the
same way as was intended in RFC 3840/3841 and not extend this with an
entire bunch of semantics that were not defined for Accept-Contact. To
do otherwise is only likely to create problems with services that work
in the internet but do not work with mobile and other devices in IMS
networks. Also service identification needs to work in both IMS networks
and non-IMS networks so that the the same services can be deployed in
both environments. The fact that IMS filter criteria enables routing on
just about anything (Using iFC I could route based on my favourite
number appearing in a Call-ID or Cseq etc) doesn't mean that it's Ok or
a good idea to simply add a new token anywhere you can and route based
on that. Services should not depend on the use of filter criteria and
the IMS architecture and routing decisions should be based on the SIP
routing mechanisms that use the contents of the Route header and
Request-URI for routing decisions. IMHO Accept-Contact is not a routing
header per say but simply allows a filter to be specified that filters
the list of registered contacts in order to determine which contacts of
the AOR to fork the request to. Building additional semantics on top of
this filter potentially significantly restricts the use of this filter
mechanism as originally intended.

I think it is in the interest of both 3GPP and the internet that a
general mechanism that works also in a pure SIP environment is
developed.

Andrew


-----Original Message-----
From: Salvatore Loreto (JO/LMF) [mailto:salvatore.loreto@ericsson.com]=20
Sent: Friday, February 09, 2007 9:39 AM
To: Paul Kyzivat
Cc: mary.barnes@nortel.com; sipping@ietf.org;
ravishankar.shiroor@wipro.com; Gonzalo.Camarillo@ericsson.com
Subject: Re: [Sipping] Service Identification in SIP

Hi,

In my opinion we should also consider the user expectation.
I fully agree with you about the mail service, but if I have look at the
Instant Messaging service where at present there are different and
incompatible programs, then maybe the necessity of a identification for
the specific IM service become a little bit more clear.


I think that everyone here agree that there is the possibility to use a
media in different ways, and that we need a distinction about what the
capabilities are and how they are used, whatever will be the mechanism
that at the end we will identify to do that.
What we do need is a way to identify a preference.

then what I don't completely understand, surely it is my fault, that a
combination of capabilities and ways to use those capabilities it is not
itself a capability, perhaps something we can call a *macro-capability*
Even if I agree on the opportunity of don't create a real Tower of Babel
and avoid . So definitely this macro-capability should be well defined
and standardized.


Of course we need also discuss the opportunity to use the collerpref
mechanism to address this capability.
I agree with Paul that "callerprefs doesn't support well, if at all, is
the inclusion of intermediate servers into the call path."
This of course is not a problem from 3GPP point of view, because in a
IMS network it is present the IFC mechanism that resolve this
"callerpref" problem.=20
But I do think we need a general mechanism that works also in a pure SIP
environment.



/sal






On Fri, 2007-02-09 at 08:26 -0500, Paul Kyzivat wrote:
>=20
> ravishankar.shiroor@wipro.com wrote:
> > Hi Paul,
> >=20
> > IMO, service identification is different from identifying the tool
that
> > should be used for executing that service on the end device. In the
> > analogy that you have given, e-mail is the service. Eudora/gmail/...
Are
> > the tools that are used to provide the service.
>=20
> The comment below was about identifying the application in the end=20
> system, not the protocol.
>=20
> In the case of email, the "service" is determined by the protocol and=20
> the address/port of the destination. Nobody has suggested adding
headers=20
> to mail messages to identify the service/application in the end system

> that should process the mail message. But here we see proposals to add
a=20
> header to a sip message to identify the service/application in the
UAS.
>=20
> 	Paul
>=20
> > Regards,
> > Ravi.
> >=20
> >=20
> > --
> > =20
> > Ravishankar. G. Shiroor
> > Wipro Technologies, Bangalore.
> > =20
> > ravishankar.shiroor@wipro.com
> > --
> > =20
> >=20
> >> -----Original Message-----
> >> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
> >> Sent: Wednesday, February 07, 2007 10:43 PM
> >> To: Salvatore Loreto (JO/LMF)
> >> Cc: RaviShankar Shiroor (WT01 - Product Engineering=20
> >> Solutions); sipping@ietf.org; Gonzalo.Camarillo@ericsson.com;=20
> >> mary.barnes@nortel.com
> >> Subject: Re: [Sipping] Service Identification in SIP
> >>
> >> I still don't get the intent of identifying the application.=20
> >> It seems to me this is evil - akin to sending an email=20
> >> message and identifying that the recipient must use Outlook=20
> >> to receive it, not Thunderbird, Eudora, gmail, or whatever.
> >>
> >> I also wish somebody could give better examples of a=20
> >> communication service. Chess is too artificial to be=20
> >> believable, and IMO PoC is only a service because of the way=20
> >> the implementation has been done.=20
> >> Conceptually PoC could be nothing but a special floor control=20
> >> protocol accompanying a voice session. In that case none of=20
> >> the Service mechanism would be needed.
> >>
> >> 	Paul
> >>
> >> Salvatore Loreto (JO/LMF) wrote:
> >>> hi Ravi,
> >>>
> >>> in IMS a Service is a "type of communication defined by a service=20
> >>> definition that specifies the rules and procedures and=20
> >> allowed media,=20
> >>> if any, for a specific type of communication and that=20
> >> utilises the IMS=20
> >>> enabler. An IMS communication service defines what is=20
> >> possible within=20
> >>> a single SIP session or standalone transaction."
> >>> So a Service implies some combination of capabilities and=20
> >> ways to use=20
> >>> those capabilities.
> >>> If you want the Service specifies the context for a communication.
> >>>
> >>> Instead the Application define the specific communication=20
> >> (or if you=20
> >>> want application) that you want run in the specific context. So
you=20
> >>> should have several different applications designed to run on a=20
> >>> specific environment or that have different performance or=20
> >> behavior in=20
> >>> different context (running on top of different service).
> >>>
> >>> Lets go to Jonathan's example of the Chess.
> >>> In his example the Chess is a service, for IMS instead is an=20
> >>> application that you can run on top of different service.
> >>> For example you can run Chess application on top of PoC=20
> >> service or on=20
> >>> top of the "standard" service. So let say you run Chess on=20
> >> top of PoC=20
> >>> service, then you can use only voice in the way that PoC
specifies.
> >>>
> >>> /sal
> >>>
> >>>
> >>>
> >>>
> >>> On Thu, 2007-02-01 at 16:39 +0530,=20
> >> ravishankar.shiroor@wipro.com wrote:
> >>>> Hello,
> >>>>
> >>>> IMO, Service Identification is definitely needed and=20
> >> should be taken=20
> >>>> up as a WG as proposed below.
> >>>>
> >>>> Further, I think Service Identification scope should be as per
the=20
> >>>> requirement in Jonathan's draft, as I still do not understand the

> >>>> need for having a Communication Service Identifier and another=20
> >>>> Application Identifier as given in the=20
> >> 3gpp-ics-requirements document.
> >>>> Regards,
> >>>> Ravi.
> >>>>
> >>>> --
> >>>> =20
> >>>> Ravishankar. G. Shiroor
> >>>> Wipro Technologies, Bangalore.
> >>>> =20
> >>>> ravishankar.shiroor@wipro.com
> >>>> --
> >>>> =20
> >>>>
> >>>>> -----Original Message-----
> >>>>> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
> >>>>> Sent: Wednesday, January 31, 2007 8:30 PM
> >>>>> To: sipping
> >>>>> Cc: Mary Barnes
> >>>>> Subject: [Sipping] Service Identification in SIP
> >>>>>
> >>>>> Folks,
> >>>>>
> >>>>> during the last year, there have been a number of threads on the

> >>>>> mailing list discussing service identification in SIP. In=20
> >>>>> particular, the following thread discussed 3GPP=20
> >> requirements on this=20
> >>>>> area:
> >>>>>
> >>>>>
> >> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
> >>>>> At a later point, those 3GPP requirements were documented in the

> >>>>> following (expired by now) Internet Draft, which was=20
> >> presented in a=20
> >>>>> face-to-face IETF meeting.
> >>>>>
> >>>>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
> >>>>> cs-requirements-00.txt
> >>>>>
> >>>>> Lately, the following draft has been submitted to the archives:
> >>>>>
> >>>>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
> >>>>> rvice-identification-00.txt
> >>>>>
> >>>>> This draft proposes a solution that intends to meet known=20
> >>>>> requirements in this area.
> >>>>>
> >>>>> Given the amount of discussions on issues related to service=20
> >>>>> identification we have seen during the last year, we=20
> >> believe this is=20
> >>>>> a fairly important topic for the SIPPING WG to tackle. At some=20
> >>>>> point, we would like to have a milestone with its=20
> >> associated WG item=20
> >>>>> dealing with this issue.
> >>>>>
> >>>>> Therefore, we would like to encourage people to read this=20
> >> draft and=20
> >>>>> send comments to the list. Additionally, if you believe that
this=20
> >>>>> work is (or is not) important, we would like to hear that as
well.
> >>>>>
> >>>>> Cheers,
> >>>>>
> >>>>> Gonzalo
> >>>>> SIPPING co-chair
> >>>>>
> >>>>>
> >>>>> _______________________________________________
> >>>>> Sipping mailing list =20
> >> https://www1.ietf.org/mailman/listinfo/sipping
> >>>>> This list is for NEW development of the application of SIP Use=20
> >>>>> sip-implementors@cs.columbia.edu for questions on current sip
Use=20
> >>>>> sip@ietf.org for new developments of core SIP
> >>>>>
> >>>> _______________________________________________
> >>>> Sipping mailing list =20
> >> https://www1.ietf.org/mailman/listinfo/sipping
> >>>> This list is for NEW development of the application of SIP Use=20
> >>>> sip-implementors@cs.columbia.edu for questions on current sip Use

> >>>> sip@ietf.org for new developments of core SIP
> >>>
> >>> _______________________________________________
> >>> Sipping mailing list
https://www1.ietf.org/mailman/listinfo/sipping
> >>> This list is for NEW development of the application of SIP Use=20
> >>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> >>> sip@ietf.org for new developments of core SIP
> >>>
> >=20
> >=20
> > The information contained in this electronic message and any
attachments to this message are intended for the exclusive use of the
addressee(s) and may contain proprietary, confidential or privileged
information. If you are not the intended recipient, you should not
disseminate, distribute or copy this e-mail. Please notify the sender
immediately and destroy all copies of this message and any attachments.=20
> >=20
> > WARNING: Computer viruses can be transmitted via email. The
recipient should check this email and any attachments for the presence
of viruses. The company accepts no liability for any damage caused by
any virus transmitted by this email.
> > =20
> > www.wipro.com
> >=20


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential =
information, privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 15 20:42:10 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHs6N-0006bq-Sx; Thu, 15 Feb 2007 20:41:31 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHs6L-0006bl-IX
	for sipping@ietf.org; Thu, 15 Feb 2007 20:41:29 -0500
Received: from rtp-iport-2.cisco.com ([64.102.122.149])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHs6G-00050M-Qk
	for sipping@ietf.org; Thu, 15 Feb 2007 20:41:28 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by rtp-iport-2.cisco.com with ESMTP; 15 Feb 2007 20:41:25 -0500
X-IronPort-AV: i="4.14,178,1170651600"; 
	d="scan'208"; a="114055001:sNHT69566860"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1G1fO9n004497; 
	Thu, 15 Feb 2007 20:41:24 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1G1fOOA009177; 
	Thu, 15 Feb 2007 20:41:24 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 15 Feb 2007 20:41:23 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 15 Feb 2007 20:41:23 -0500
Message-ID: <45D50BC1.5060604@cisco.com>
Date: Thu, 15 Feb 2007 20:41:21 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Andrew Allen <aallen@rim.com>
Subject: Re: [Sipping] Service Identification in SIP
References: <61968779B8AC4C4BAB421D4C12F008C009151456@XCH47YKF.rim.net>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C009151456@XCH47YKF.rim.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 16 Feb 2007 01:41:23.0440 (UTC)
	FILETIME=[90B3C700:01C7516B]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=13813; t=1171590084;
	x=1172454084; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20 |To:=20Andrew=20Allen=20<aallen@rim.com>;
	bh=27mLTd1SgaxLf8+Y29Up7XBfCyRsBjlj3rMvzsVVUfY=;
	b=BKPHqTlgPkGzpWI8KiFyWTnzhcb6D7ARi8ktZdQMgVFkLExnrtvAAFlnDlzbFq22QmWfbPX5
	B3iAPNdjf2HcDLUn4xczXt03NNMduKuDUP76G3yZvgSW4rCBGxDG9TU5;
Authentication-Results: rtp-dkim-2; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: fac892abe0c719c7bb99f6e7c710cdae
Cc: Gonzalo.Camarillo@ericsson.com, sipping@ietf.org, mary.barnes@nortel.com,
	ravishankar.shiroor@wipro.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Andrew Allen wrote:
> I agree with Salvatore that IM is a better example of where service
> identification may have a role in expressing a preference for a
> particular application. Certainly there are different emoticons and
> other value added features proprietary to different messaging
> applications so having the ability to express preferences to use a
> preferred application seems to me to have value. I think that email is a
> bad counter example as with email the email client interfaces to the
> email server not directly to the other email client.

If the kind of usage above were to take hold in any significant way, it 
would be a huge step towards balkanizing the internet. Part of agreeing 
on a common protocol is actually supporting interoperation.

But I agree with what Andrew says below.

	Paul

> Where I disagree with Salvatore is that the fact that the use of
> callerprefs doesn't support well, if at all, the inclusion of
> intermediate servers into the call path is not a problem for 3GPP as
> well. From my viewpoint 3GPP needs to make use of Callerprefs in the
> same way as was intended in RFC 3840/3841 and not extend this with an
> entire bunch of semantics that were not defined for Accept-Contact. To
> do otherwise is only likely to create problems with services that work
> in the internet but do not work with mobile and other devices in IMS
> networks. Also service identification needs to work in both IMS networks
> and non-IMS networks so that the the same services can be deployed in
> both environments. The fact that IMS filter criteria enables routing on
> just about anything (Using iFC I could route based on my favourite
> number appearing in a Call-ID or Cseq etc) doesn't mean that it's Ok or
> a good idea to simply add a new token anywhere you can and route based
> on that. Services should not depend on the use of filter criteria and
> the IMS architecture and routing decisions should be based on the SIP
> routing mechanisms that use the contents of the Route header and
> Request-URI for routing decisions. IMHO Accept-Contact is not a routing
> header per say but simply allows a filter to be specified that filters
> the list of registered contacts in order to determine which contacts of
> the AOR to fork the request to. Building additional semantics on top of
> this filter potentially significantly restricts the use of this filter
> mechanism as originally intended.
> 
> I think it is in the interest of both 3GPP and the internet that a
> general mechanism that works also in a pure SIP environment is
> developed.
> 
> Andrew
> 
> 
> -----Original Message-----
> From: Salvatore Loreto (JO/LMF) [mailto:salvatore.loreto@ericsson.com] 
> Sent: Friday, February 09, 2007 9:39 AM
> To: Paul Kyzivat
> Cc: mary.barnes@nortel.com; sipping@ietf.org;
> ravishankar.shiroor@wipro.com; Gonzalo.Camarillo@ericsson.com
> Subject: Re: [Sipping] Service Identification in SIP
> 
> Hi,
> 
> In my opinion we should also consider the user expectation.
> I fully agree with you about the mail service, but if I have look at the
> Instant Messaging service where at present there are different and
> incompatible programs, then maybe the necessity of a identification for
> the specific IM service become a little bit more clear.
> 
> 
> I think that everyone here agree that there is the possibility to use a
> media in different ways, and that we need a distinction about what the
> capabilities are and how they are used, whatever will be the mechanism
> that at the end we will identify to do that.
> What we do need is a way to identify a preference.
> 
> then what I don't completely understand, surely it is my fault, that a
> combination of capabilities and ways to use those capabilities it is not
> itself a capability, perhaps something we can call a *macro-capability*
> Even if I agree on the opportunity of don't create a real Tower of Babel
> and avoid . So definitely this macro-capability should be well defined
> and standardized.
> 
> 
> Of course we need also discuss the opportunity to use the collerpref
> mechanism to address this capability.
> I agree with Paul that "callerprefs doesn't support well, if at all, is
> the inclusion of intermediate servers into the call path."
> This of course is not a problem from 3GPP point of view, because in a
> IMS network it is present the IFC mechanism that resolve this
> "callerpref" problem. 
> But I do think we need a general mechanism that works also in a pure SIP
> environment.
> 
> 
> 
> /sal
> 
> 
> 
> 
> 
> 
> On Fri, 2007-02-09 at 08:26 -0500, Paul Kyzivat wrote:
>> ravishankar.shiroor@wipro.com wrote:
>>> Hi Paul,
>>>
>>> IMO, service identification is different from identifying the tool
> that
>>> should be used for executing that service on the end device. In the
>>> analogy that you have given, e-mail is the service. Eudora/gmail/...
> Are
>>> the tools that are used to provide the service.
>> The comment below was about identifying the application in the end 
>> system, not the protocol.
>>
>> In the case of email, the "service" is determined by the protocol and 
>> the address/port of the destination. Nobody has suggested adding
> headers 
>> to mail messages to identify the service/application in the end system
> 
>> that should process the mail message. But here we see proposals to add
> a 
>> header to a sip message to identify the service/application in the
> UAS.
>> 	Paul
>>
>>> Regards,
>>> Ravi.
>>>
>>>
>>> --
>>>  
>>> Ravishankar. G. Shiroor
>>> Wipro Technologies, Bangalore.
>>>  
>>> ravishankar.shiroor@wipro.com
>>> --
>>>  
>>>
>>>> -----Original Message-----
>>>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
>>>> Sent: Wednesday, February 07, 2007 10:43 PM
>>>> To: Salvatore Loreto (JO/LMF)
>>>> Cc: RaviShankar Shiroor (WT01 - Product Engineering 
>>>> Solutions); sipping@ietf.org; Gonzalo.Camarillo@ericsson.com; 
>>>> mary.barnes@nortel.com
>>>> Subject: Re: [Sipping] Service Identification in SIP
>>>>
>>>> I still don't get the intent of identifying the application. 
>>>> It seems to me this is evil - akin to sending an email 
>>>> message and identifying that the recipient must use Outlook 
>>>> to receive it, not Thunderbird, Eudora, gmail, or whatever.
>>>>
>>>> I also wish somebody could give better examples of a 
>>>> communication service. Chess is too artificial to be 
>>>> believable, and IMO PoC is only a service because of the way 
>>>> the implementation has been done. 
>>>> Conceptually PoC could be nothing but a special floor control 
>>>> protocol accompanying a voice session. In that case none of 
>>>> the Service mechanism would be needed.
>>>>
>>>> 	Paul
>>>>
>>>> Salvatore Loreto (JO/LMF) wrote:
>>>>> hi Ravi,
>>>>>
>>>>> in IMS a Service is a "type of communication defined by a service 
>>>>> definition that specifies the rules and procedures and 
>>>> allowed media, 
>>>>> if any, for a specific type of communication and that 
>>>> utilises the IMS 
>>>>> enabler. An IMS communication service defines what is 
>>>> possible within 
>>>>> a single SIP session or standalone transaction."
>>>>> So a Service implies some combination of capabilities and 
>>>> ways to use 
>>>>> those capabilities.
>>>>> If you want the Service specifies the context for a communication.
>>>>>
>>>>> Instead the Application define the specific communication 
>>>> (or if you 
>>>>> want application) that you want run in the specific context. So
> you 
>>>>> should have several different applications designed to run on a 
>>>>> specific environment or that have different performance or 
>>>> behavior in 
>>>>> different context (running on top of different service).
>>>>>
>>>>> Lets go to Jonathan's example of the Chess.
>>>>> In his example the Chess is a service, for IMS instead is an 
>>>>> application that you can run on top of different service.
>>>>> For example you can run Chess application on top of PoC 
>>>> service or on 
>>>>> top of the "standard" service. So let say you run Chess on 
>>>> top of PoC 
>>>>> service, then you can use only voice in the way that PoC
> specifies.
>>>>> /sal
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> On Thu, 2007-02-01 at 16:39 +0530, 
>>>> ravishankar.shiroor@wipro.com wrote:
>>>>>> Hello,
>>>>>>
>>>>>> IMO, Service Identification is definitely needed and 
>>>> should be taken 
>>>>>> up as a WG as proposed below.
>>>>>>
>>>>>> Further, I think Service Identification scope should be as per
> the 
>>>>>> requirement in Jonathan's draft, as I still do not understand the
> 
>>>>>> need for having a Communication Service Identifier and another 
>>>>>> Application Identifier as given in the 
>>>> 3gpp-ics-requirements document.
>>>>>> Regards,
>>>>>> Ravi.
>>>>>>
>>>>>> --
>>>>>>  
>>>>>> Ravishankar. G. Shiroor
>>>>>> Wipro Technologies, Bangalore.
>>>>>>  
>>>>>> ravishankar.shiroor@wipro.com
>>>>>> --
>>>>>>  
>>>>>>
>>>>>>> -----Original Message-----
>>>>>>> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
>>>>>>> Sent: Wednesday, January 31, 2007 8:30 PM
>>>>>>> To: sipping
>>>>>>> Cc: Mary Barnes
>>>>>>> Subject: [Sipping] Service Identification in SIP
>>>>>>>
>>>>>>> Folks,
>>>>>>>
>>>>>>> during the last year, there have been a number of threads on the
> 
>>>>>>> mailing list discussing service identification in SIP. In 
>>>>>>> particular, the following thread discussed 3GPP 
>>>> requirements on this 
>>>>>>> area:
>>>>>>>
>>>>>>>
>>>> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
>>>>>>> At a later point, those 3GPP requirements were documented in the
> 
>>>>>>> following (expired by now) Internet Draft, which was 
>>>> presented in a 
>>>>>>> face-to-face IETF meeting.
>>>>>>>
>>>>>>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
>>>>>>> cs-requirements-00.txt
>>>>>>>
>>>>>>> Lately, the following draft has been submitted to the archives:
>>>>>>>
>>>>>>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
>>>>>>> rvice-identification-00.txt
>>>>>>>
>>>>>>> This draft proposes a solution that intends to meet known 
>>>>>>> requirements in this area.
>>>>>>>
>>>>>>> Given the amount of discussions on issues related to service 
>>>>>>> identification we have seen during the last year, we 
>>>> believe this is 
>>>>>>> a fairly important topic for the SIPPING WG to tackle. At some 
>>>>>>> point, we would like to have a milestone with its 
>>>> associated WG item 
>>>>>>> dealing with this issue.
>>>>>>>
>>>>>>> Therefore, we would like to encourage people to read this 
>>>> draft and 
>>>>>>> send comments to the list. Additionally, if you believe that
> this 
>>>>>>> work is (or is not) important, we would like to hear that as
> well.
>>>>>>> Cheers,
>>>>>>>
>>>>>>> Gonzalo
>>>>>>> SIPPING co-chair
>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> Sipping mailing list  
>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>>>> This list is for NEW development of the application of SIP Use 
>>>>>>> sip-implementors@cs.columbia.edu for questions on current sip
> Use 
>>>>>>> sip@ietf.org for new developments of core SIP
>>>>>>>
>>>>>> _______________________________________________
>>>>>> Sipping mailing list  
>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>>> This list is for NEW development of the application of SIP Use 
>>>>>> sip-implementors@cs.columbia.edu for questions on current sip Use
> 
>>>>>> sip@ietf.org for new developments of core SIP
>>>>> _______________________________________________
>>>>> Sipping mailing list
> https://www1.ietf.org/mailman/listinfo/sipping
>>>>> This list is for NEW development of the application of SIP Use 
>>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>>> sip@ietf.org for new developments of core SIP
>>>>>
>>>
>>> The information contained in this electronic message and any
> attachments to this message are intended for the exclusive use of the
> addressee(s) and may contain proprietary, confidential or privileged
> information. If you are not the intended recipient, you should not
> disseminate, distribute or copy this e-mail. Please notify the sender
> immediately and destroy all copies of this message and any attachments. 
>>> WARNING: Computer viruses can be transmitted via email. The
> recipient should check this email and any attachments for the presence
> of viruses. The company accepts no liability for any damage caused by
> any virus transmitted by this email.
>>>  
>>> www.wipro.com
>>>
> 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 
> 
> 
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential information, privileged material (including material protected by the solicitor-client or other applicable privileges), or constitute non-public information. Any use of this information by anyone other than the intended recipient is prohibited. If you have received this transmission in error, please immediately reply to the sender and delete this information from your system. Use, dissemination, distribution, or reproduction of this transmission by unintended recipients is not authorized and may be unlawful.
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 15 21:04:17 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHsRC-0004SA-Tg; Thu, 15 Feb 2007 21:03:02 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHsRB-0004LY-36
	for sipping@ietf.org; Thu, 15 Feb 2007 21:03:01 -0500
Received: from dnsmx2pya.telcordia.com ([128.96.20.32])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHsR9-0007Df-Ok
	for sipping@ietf.org; Thu, 15 Feb 2007 21:03:01 -0500
Received: from pya-dte-ieg01.cc.telcordia.com (pya-dte-ieg01.cc.telcordia.com
	[128.96.20.21])
	by dnsmx2pya.telcordia.com (8.11.7+Sun/8.9.3) with SMTP id l1G22xq20684
	for <sipping@ietf.org>; Thu, 15 Feb 2007 21:02:59 -0500 (EST)
Received: from rrc-dte-exbh01.dte.telcordia.com ([128.96.150.31])
	by pya-dte-ieg01.cc.telcordia.com (SMSSMTP 4.1.9.35) with SMTP id
	M2007021521032012129
	for <sipping@ietf.org>; Thu, 15 Feb 2007 21:03:20 -0500
Received: from rrc-dte-exs01.dte.telcordia.com ([128.96.150.34]) by
	rrc-dte-exbh01.dte.telcordia.com with Microsoft
	SMTPSVC(6.0.3790.1830); Thu, 15 Feb 2007 21:03:20 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5.7226.0
Content-class: urn:content-classes:message
MIME-Version: 1.0
Date: Thu, 15 Feb 2007 21:03:19 -0500
Message-ID: <A09345776B6C7A4985573569C0F300430EAE5290@rrc-dte-exs01.dte.telcordia.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: draft-haluska-sipping-directory-assistance-02.txt
Thread-Index: AcdRbqFXZX/0yE6tRnKwu4nbta1czA==
From: "Haluska, John J" <jhaluska@telcordia.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 16 Feb 2007 02:03:20.0424 (UTC)
	FILETIME=[A1AF9680:01C7516E]
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 00e94c813bef7832af255170dca19e36
Subject: [Sipping] draft-haluska-sipping-directory-assistance-02.txt
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============0326898135=="
Errors-To: sipping-bounces@ietf.org

This is a multi-part message in MIME format.

--===============0326898135==
Content-class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C7516E.A1689E7B"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C7516E.A1689E7B
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

Hello All,=20


We are seeking comments on the latest revision of =
draft-haluska-sipping-directory-assistance-02.txt

This revision goes into much more discussion about applying specific SIP =
mechanisms for these kinds of services.=20

We would appreciate any technical input on the choices of mechanisms, =
and the usage of these mechanisms.=20


Until it is available in the ID repository, it is available at

ftp://ftp.research.telcordia.com/pub/world/haluska/draft-haluska-sipping-=
directory-assistance-02.txt


Thanks much

John Haluska

------_=_NextPart_001_01C7516E.A1689E7B
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3Diso-8859-1">
<META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
6.5.7226.0">
<TITLE>draft-haluska-sipping-directory-assistance-02.txt</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->

<P><FONT SIZE=3D2>Hello All,<BR>
<BR>
<BR>
We are seeking comments on the latest revision of =
draft-haluska-sipping-directory-assistance-02.txt<BR>
<BR>
This revision goes into much more discussion about applying specific SIP =
mechanisms for these kinds of services.<BR>
<BR>
We would appreciate any technical input on the choices of mechanisms, =
and the usage of these mechanisms.<BR>
<BR>
<BR>
Until it is available in the ID repository, it is available at<BR>
<BR>
<A =
HREF=3D"ftp://ftp.research.telcordia.com/pub/world/haluska/draft-haluska-=
sipping-directory-assistance-02.txt">ftp://ftp.research.telcordia.com/pub=
/world/haluska/draft-haluska-sipping-directory-assistance-02.txt</A><BR>
<BR>
<BR>
Thanks much<BR>
<BR>
John Haluska</FONT>
</P>

</BODY>
</HTML>
------_=_NextPart_001_01C7516E.A1689E7B--


--===============0326898135==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============0326898135==--




From sipping-bounces@ietf.org Thu Feb 15 21:05:11 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHsTG-0007RO-S0; Thu, 15 Feb 2007 21:05:10 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHsTE-0007Mq-MM
	for sipping@ietf.org; Thu, 15 Feb 2007 21:05:08 -0500
Received: from exprod6og50.obsmtp.com ([64.18.1.181])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1HHsTC-0007Jt-Q5
	for sipping@ietf.org; Thu, 15 Feb 2007 21:05:08 -0500
Received: from source ([192.150.11.134]) by exprod6ob50.postini.com
	([64.18.5.12]) with SMTP; Thu, 15 Feb 2007 18:04:38 PST
Received: from inner-relay-3.eur.adobe.com (inner-relay-3.adobe.com
	[192.150.20.198] (may be forged))
	by outbound-smtp-1.corp.adobe.com (8.12.10/8.12.10) with ESMTP id
	l1G24W2s001370; Thu, 15 Feb 2007 18:04:32 -0800 (PST)
Received: from fe2.corp.adobe.com (fe2.corp.adobe.com [10.8.192.72])
	by inner-relay-3.eur.adobe.com (8.12.10/8.12.9) with ESMTP id
	l1G24XmC005570; Thu, 15 Feb 2007 18:04:35 -0800 (PST)
Received: from namail5.corp.adobe.com ([10.8.192.88]) by fe2.corp.adobe.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 15 Feb 2007 18:04:34 -0800
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Thu, 15 Feb 2007 18:03:34 -0800
Message-ID: <24CCCC428EFEA2469BF046DB3C7A8D2249F0AA@namail5.corp.adobe.com>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C00915140A@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdK2+Kqf8XxF4RlRBKez3RRQPczsAAAHOqwAZlFC4AACrTVoA==
References: <61968779B8AC4C4BAB421D4C12F008C00915140A@XCH47YKF.rim.net>
From: "Henry Sinnreich" <hsinnrei@adobe.com>
To: "Andrew Allen" <aallen@rim.com>, "Paul Kyzivat" <pkyzivat@cisco.com>,
	"Salvatore Loreto \(JO/LMF\)" <salvatore.loreto@ericsson.com>
X-OriginalArrivalTime: 16 Feb 2007 02:04:34.0879 (UTC)
	FILETIME=[CE1084F0:01C7516E]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 4c358d334afcd91b425d436ca5722f22
Cc: ravishankar.shiroor@wipro.com, sipping@ietf.org,
	Gonzalo.Camarillo@ericsson.com, mary.barnes@nortel.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Andrew,

It is interesting to note that there is similar thinking going on in
other WGs, see for example:

http://www.ietf.org/internet-drafts/draft-fairhurst-dccp-serv-codes-02.t
xt=20

and

http://www.ietf.org/internet-drafts/draft-andreasen-sipping-rfc3603bis-0
2.txt=20

Along with your correct observation:

> with an increasing number of competing service providers in the market

 One has to wonder about the scalability and complexity all levels of
defining so many services for so many providers, not only for SIP but
also in the transport area.=20

Coming back to scalability, even plain services could have promotions,
coupons "free video for frequent flyers", etc. that interestingly enough
may not be a good idea to trumpet to the competition or get the
agreement first from the competition. Perish the thought...

It is also correct that=20

> this argument will be settled by market competition and is not really=20
> the role of engineering to settle. =20

Internet engineers may however share the discomfort that service
identification facilitates the following on the Internet:

- Possible discrimination of unwanted services by the competition,
- Restricting the freedom of development for edge devices and apps,
- Charging for the two above, to top it off.

Thanks again to all for the thoughtful contributions to this discussion.

Henry


-----Original Message-----
From: Andrew Allen [mailto:aallen@rim.com]=20
Sent: Thursday, February 15, 2007 2:17 PM
To: Henry Sinnreich; Paul Kyzivat; Salvatore Loreto (JO/LMF)
Cc: Gonzalo.Camarillo@ericsson.com; sipping@ietf.org;
mary.barnes@nortel.com; ravishankar.shiroor@wipro.com; Andrew Allen
Subject: RE: [Sipping] Service Identification in SIP


=20
IN response to Henry:

RFC 4485 also states:

   Session initiation also depends on the ability of the called party to
   have enough information about the session itself to make a decision
   on whether to join.  That information includes data about the caller,
   the purpose for the invitation, and parameters of the session itself.
   For this reason, SIP includes this kind of information.

IMHO "the purpose for the invitation" also includes information about
what type of service or application this session initiation request is
related to and SDP doesn't provide sufficent information to definitively
determine this.

I think we can all debate the advantages and disadvantages of building
intelligence in the network as opposed to distributing it to the
endpoints and I feel that I am also generally on the distributed side of
that particular argument however SIP allows for both models and many
services/applications utilise both intelligence in the endpoint and
intelligence in Application Servers in the network. A prime example is
Presence where Presence can be provided with SIP using only SIP
endpoints however many practical implementations make use of network
based Presence Servers and Resource List Servers in order to create a
scalable Presence Service and significant IETF work has been done on
creating SIP based and other related mechanisms (e.g XCAP) to support
Presence using Presence Servers and Resource List Servers. Being purist
philosophical about moving all intelligence to the very edge doesn't
always result in the most optimal solution as the scalability advantages
of having all the intelligence in the endpoints has to be considered
along with the scalability disadvantages of having all the signalling
transported to the endpoints. In limited and high cost bandwidth access
networks like cellular the scalability disadvantages of having all
signaling transported to the endpoint can sometimes outweigh the
advantages of distributed intelligence.=20

My understanding is that SIP doesn't take any position on the business
models used for communication service provision and IETF has been
cooperating closely with those who are using SIP for carrier based
services. Some people may not think that the carrier centric service
model is the way to go in the future but with an increasing number of
competing service providers in the market this argument will be settled
by market competition and is not really the role of engineering to
settle.=20

Also don't think that failing to develop or delaying an IETF based
solution will stop 3GPP from developing their own solution. There are
already proposals being agreed within 3GPP based on Accept-Contact
header and feature tags to implement this in what I believe is a way
that potentially breaks interoperability between IMS and other SIP
networks. This is because ithis proposalt requires pre knowledge of
these service identifiers and their use in all networks for
communication without providing a mechanism to obtain them and also
through its use of Accept-Contact potentially signiifcantly limits the
flexibility use of that header for its intended purpose. However RFC
3427 doesn't require that new feature tags be discussed in SIPPING even
when the proposed usage is required for interoperability. So basically
we have a choice between working on an open internet friendly solution
to service identification or allowing by default a 3GPP specific
solution that potentially will create interoperability issues between
internet users and about 2 Billion Mobile users and several hundred
million fixed carrier device users. Jonthan's proposal addresses these
issues.

Something else from RFC 4485:

   Proxies can ignore bodies:  In order for proxies to scale well, they
      must be able to operate with minimal message processing.  SIP has
      been engineered so that proxies can always ignore bodies.
      Extensions SHOULD NOT require proxies to examine bodies.

As already stated I don't believe that SDP always provides sufficent
information to definitively determine the application or service that
the session initiation request is related to and even when it does this
requires that a Proxy examine the body in order to determine whether the
request needs to be routed to some intermediate server that performs
service logic or make other service based routing decisions. I think it
is far preferable to have some kind of service identifier that
specifcally identifies the service logic processing needed included in
an appropriate place in the request as proposed by Jonathan and make
route decisions based on that than to have proxies attempt to do this
based on examing SDP or based on the content type. As well as being more
efficient in terms of proxy processing this avoids routing failures due
to encrypted bodies or unexpected multipart MIME content types or new
alternative content types not yet understood by the proxy causing the
service routing to fail.

So you can count me in the supporting the work on service identification
based on Jonathan's draft column.

Andrew



-----Original Message-----
From: Henry Sinnreich [mailto:hsinnrei@adobe.com]=20
Sent: Wednesday, February 07, 2007 12:34 PM
To: Paul Kyzivat; Salvatore Loreto (JO/LMF)
Cc: Gonzalo.Camarillo@ericsson.com; sipping@ietf.org;
mary.barnes@nortel.com; ravishankar.shiroor@wipro.com
Subject: RE: [Sipping] Service Identification in SIP

> I still don't get the intent of identifying the application. It seems
to > me this is evil - akin to sending an email message and identifying
that > the recipient must use Outlook to receive it, not Thunderbird,
Eudora,=20
> gmail, or whatever.

This goes back to the old discussion about the IN versus the "dumb
network" and we know how this has played out.

Placing intelligence (AKA services) in the network is just "dumb design"
IMHO for all the good reasons, but makes no sense to discuss on this
list.

SIP was designed as a rendezvous protocol for multimedia communications
on the Internet and not to replicate ISDN, etc., as has been so well
documented in RFC 4485:

   "SIP is not meant to be used as a strict Public Switched Telephone
   Network (PSTN) signaling replacement.  It is not a superset of the
   Integrated Services Digital Network (ISDN) User Part (ISUP).
   Although it can support gatewaying of PSTN signaling and can provide
   many features present in the PSTN, the mere existence of a feature or
   capability in the PSTN is not a justification for its inclusion in
   SIP."

Which makes me wonder how many folks have read RFC 4485 and taken it to
heart or may have chosen to forget it.

Thanks, Henry

-----Original Message-----
From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
Sent: Wednesday, February 07, 2007 11:13 AM
To: Salvatore Loreto (JO/LMF)
Cc: mary.barnes@nortel.com; sipping@ietf.org;
ravishankar.shiroor@wipro.com; Gonzalo.Camarillo@ericsson.com
Subject: Re: [Sipping] Service Identification in SIP

I still don't get the intent of identifying the application. It seems to

me this is evil - akin to sending an email message and identifying that
the recipient must use Outlook to receive it, not Thunderbird, Eudora,
gmail, or whatever.

I also wish somebody could give better examples of a communication
service. Chess is too artificial to be believable, and IMO PoC is only a

service because of the way the implementation has been done.=20
Conceptually PoC could be nothing but a special floor control protocol
accompanying a voice session. In that case none of the Service mechanism

would be needed.

	Paul

Salvatore Loreto (JO/LMF) wrote:
> hi Ravi,
>=20
> in IMS a Service is a "type of communication defined by a service=20
> definition that specifies the rules and procedures and allowed media,
if
> any, for a specific type of communication and that utilises the IMS=20
> enabler. An IMS communication service defines what is possible within
a
> single SIP session or standalone transaction."
> So a Service implies some combination of capabilities and ways to use=20
> those capabilities.
> If you want the Service specifies the context for a communication.
>=20
> Instead the Application define the specific communication (or if you=20
> want application) that you want run in the specific context. So you=20
> should have several different applications designed to run on a
specific
> environment or that have different performance or behavior in
different
> context (running on top of different service).
>=20
> Lets go to Jonathan's example of the Chess.
> In his example the Chess is a service, for IMS instead is an
application
> that you can run on top of different service.
> For example you can run Chess application on top of PoC service or on=20
> top of the "standard" service. So let say you run Chess on top of PoC=20
> service, then you can use only voice in the way that PoC specifies.
>=20
> /sal
>=20
>=20
>=20
>=20
> On Thu, 2007-02-01 at 16:39 +0530, ravishankar.shiroor@wipro.com
wrote:
>> Hello,
>>
>> IMO, Service Identification is definitely needed and should be taken
up
>> as a WG as proposed below.
>>
>> Further, I think Service Identification scope should be as per the=20
>> requirement in Jonathan's draft, as I still do not understand the
need
>> for having a Communication Service Identifier and another Application

>> Identifier as given in the 3gpp-ics-requirements document.
>>
>> Regards,
>> Ravi.
>>
>> --
>> =20
>> Ravishankar. G. Shiroor
>> Wipro Technologies, Bangalore.
>> =20
>> ravishankar.shiroor@wipro.com
>> --
>> =20
>>
>>> -----Original Message-----
>>> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
>>> Sent: Wednesday, January 31, 2007 8:30 PM
>>> To: sipping
>>> Cc: Mary Barnes
>>> Subject: [Sipping] Service Identification in SIP
>>>
>>> Folks,
>>>
>>> during the last year, there have been a number of threads on the=20
>>> mailing list discussing service identification in SIP. In=20
>>> particular, the following thread discussed 3GPP requirements on this

>>> area:
>>>
>>> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
>>>
>>> At a later point, those 3GPP requirements were documented in the=20
>>> following (expired by now) Internet Draft, which was presented in a=20
>>> face-to-face IETF meeting.
>>>
>>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
>>> cs-requirements-00.txt
>>>
>>> Lately, the following draft has been submitted to the archives:
>>>
>>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
>>> rvice-identification-00.txt
>>>
>>> This draft proposes a solution that intends to meet known=20
>>> requirements in this area.
>>>
>>> Given the amount of discussions on issues related to service=20
>>> identification we have seen during the last year, we believe this is

>>> a fairly important topic for the SIPPING WG to tackle. At some=20
>>> point, we would like to have a milestone with its associated WG item

>>> dealing with this issue.
>>>
>>> Therefore, we would like to encourage people to read this draft and=20
>>> send comments to the list. Additionally, if you believe that this=20
>>> work is (or is not) important, we would like to hear that as well.
>>>
>>> Cheers,
>>>
>>> Gonzalo
>>> SIPPING co-chair
>>>
>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP Use=20
>>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
>>> sip@ietf.org for new developments of core SIP
>>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP Use=20
>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
>> sip@ietf.org for new developments of core SIP
>=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential
information, privileged material (including material protected by the
solicitor-client or other applicable privileges), or constitute
non-public information. Any use of this information by anyone other than
the intended recipient is prohibited. If you have received this
transmission in error, please immediately reply to the sender and delete
this information from your system. Use, dissemination, distribution, or
reproduction of this transmission by unintended recipients is not
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 15 22:28:06 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHtkF-0004Rh-Um; Thu, 15 Feb 2007 22:26:47 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHtkE-0004Ns-M2
	for sipping@ietf.org; Thu, 15 Feb 2007 22:26:46 -0500
Received: from c243-111.rim.net ([216.9.243.111] helo=MHS02YKF.rim.net)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHtkE-0002KQ-75
	for sipping@ietf.org; Thu, 15 Feb 2007 22:26:46 -0500
Importance: normal
Priority: normal
Received: from ngw01ykf.rim.net ([10.102.108.30]) by MHS02YKF.rim.net with
	Microsoft SMTPSVC(6.0.3790.211); Thu, 15 Feb 2007 22:26:37 -0500
Received: from XCH20YKF.rim.net ([10.102.100.35]) by ngw01ykf.rim.net (SAVSMTP
	3.1.6.45) with SMTP id M2007021522263431303 ;
	Thu, 15 Feb 2007 22:26:34 -0500
Received: from XCH47YKF.rim.net ([10.64.31.217]) by XCH20YKF.rim.net with
	Microsoft SMTPSVC(5.0.2195.6713); Thu, 15 Feb 2007 22:26:34 -0500
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.181
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Thu, 15 Feb 2007 22:26:33 -0500
Message-ID: <61968779B8AC4C4BAB421D4C12F008C009151469@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdNmFyZ6rDpmUNAS5iWJYKZtamTgQDxmlLQ
From: "Andrew Allen" <aallen@rim.com>
To: <Dale.Worley@comcast.net>,
	<sipping@ietf.org>
X-OriginalArrivalTime: 16 Feb 2007 03:26:34.0694 (UTC)
	FILETIME=[4280B660:01C7517A]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 156eddb66af16eef49a76ae923b15b92
Cc: Andrew Allen <aallen@rim.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


I'll have a crack at laying out examples showing the utility of service
identification.

First a gaming example somewhat similar to Jonathan's. Take a network
based multiplayer game (Chess, D&D, Quake, Pool, Roulette, virtual golf,
etc.....). This game allows multiple users to participate in the game
either as players or as observers. An audio stream is established so
that players and observers can exchange comments on the game along with
the media stream (either using MSRP or an application specific data
transport that is used to exchange game commands and game position
updates between the participants and the game server which also acts as
a focus. The gaming session is established using SIP and SDP over a SIP
network such as IMS that has Application Servers that perform telephony
call features for basic voice only telephony and multimedia "rich"
telephony calls. The multimedia telephony sessions also support combined
audio with MSRP and other application specific data streams (such as
whiteboards) in a call. Now gaming sessions don't require telephony
service call feature handling in fact this isnt even desired however
based on the SIP and SDP it isnt possible to determine whether an INVITE
request is related to a telephony call or a gaming session and hence
whether the request needs to be routed via an Application Server that
performs telephony call features or not. In addition users may have
multiple devices only one of which has the application that understands
the game. Other devices that the user has registered against his AoR may
handle audio for telephony and other voice applications but if Alice
invites Bob to play her favourite game it is not very interesting if Bob
accepts the gaming session on a device that can only accept the audio
stream. Also even when the terminal that has the game application
receives the INVITE the correct application in that terminal needs to be
identified to accept the session. In addition service providers may
charge different rates for telephony and for games. I also don't want to
be charged telephony per minute call rates for playing my game. =20

Here is another example that I think shows that service identification
doesn't just apply to INVITE and sessions but also can apply to other
requests such a MESSAGE as well. SIP MESSAGE method is being used for
Instant messaging applications but it is also being used for many non
Instant Messaging applications as well. These include as a replacement
SIP transport for transporting encapsulated SMS (Short Message Service)
contents (see 3GPP SMS over IP), as a transport for Group Advertisement
documents for Push To Talk (see OMA PoC), as a notification transport
for various Push Services (see OMA SIP Push). Each of these different
service applications of SIP MESSAGE method requires routing to different
Application Servers in the network, delivery to different applications
in the terminal and the ability to be charged for separately. The only
SIP header that differentiates each of these different useages of SIP
MESSAGE is the Content-Type header but routing based on Content-Type is
a very bad idea. Firstly the Content-Type doesn't identify the service
specifically but some content related to a subset of the service. Also
as the service evolves the Content-Types used may change or new ones may
be added and this shouldn't require updating the routing rules of the
proxies or the internal application invocation logic in endpoints. The
possibility of Multipart MIME Content-Types makes parsing of
Content-Types by proxies inefficient and complex and potentially
impossible if the message body is encrypted. What is needed is a means
compliant with standard SIP Routing mechanisms that allow SIP proxies to
efficently route these different MESSAGE requests to the appropriate
servers, endpoints and invoke applications within endpoints. I think
Jonathan's draft is a good basis for doing this.

Another example using INVITE is IPTV. IPTV may be established using
INVITE containing SDP with both Audio and Video streams. How does the
network determine the difference between an IPTV related INVITE and one
related to multimedia telephony that supports video? Certainly it is not
desired that the IPTV INVITE invoke telephony call features in an
Application Server and you probably don't want the IPTV INVITE being
routed to your mobile phone (unless it's a mobile TV service). It is
likely that an SDP offer for IPTV is for a unidirectional media stream
but this is real convoluted to use as the basis for routing and
charging.=20

In adddition it is possible that an INVITE doesn't contain SDP at all
and then there is no basis in SDP for determining the related service.

Andrew



-----Original Message-----
From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net]=20
Sent: Saturday, February 10, 2007 11:50 PM
To: sipping@ietf.org
Subject: Re: [Sipping] Service Identification in SIP

I've attempted to re-read the threads that Gonzalo referred to, and the
major limitation that I see is the lack of any concrete example of why
these identifiers are needed.  Probably the IMS people have discussed
this so many times that they've forgotten the detailed rationale behind
the requirement.  But I'm a net-head, (and most of the people here are
net-heads) and the SIP systems that I've dealt with show no evidence of
needing "service identification", they're pure proxies, and requests are
passed transparently end-to-end.

Could somebody please lay out an example showing the utility of service
identification?

   From: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>

   I think that everyone here agree that there is the possibility to use
a
   media in different ways, and that we need a distinction about what
the
   capabilities are and how they are used, whatever will be the
mechanism
   that at the end we will identify to do that.

Respectfully, I disagree.  We do need ways of identifying the
capabilities of SIP agents and transports, but I see no necessity of
creating specified bundles of capabailities, and then constraining
dialogs to require an initial agreement as to which bundle of
capabilities will be used.

One of the central features of Internet architecture is that agents
negotiate compatibility on a feature-by-feature basis; there are no
"profiles" which must be known on a network-wide basis.  (And I have
observed that this feature is of significant commercial value.)

In regard to identifying an "application" to process the dialog (and I
use "application" in a standard computing way), I see no reason that a
user cannot select URI-parameters to append to his AOR, and arrange for
his UA to route invites with those URI-parameters to the appropriate
application.  (And given the rich mechanisms for redirection in SIP, the
application need not even reside at a UA which is registered for the
AOR.)  These URI-parameters can be formally or informally standardized,
or established ad-hoc by each user.

In regard to PoC (as someone said, "a specialized floor-control
protocol") and similar features, SIP currently has mechanisms for
indicating support of features and requiring that a recipient support a
feature before accepting a call (the Supported and Requires headers, and
simply using the feature).  The only limitation I can see that might
need to be addressed is that there is no way to create an option tag for
a new feature that one has invented that is not yet thoroughly
standardized.

But I now discover RFC 3261 section 8.1.1.9:

   The option tags listed MUST only refer to extensions defined in
   standards-track RFCs.  This is to prevent servers from insisting that
   clients implement non-standard, vendor-defined features in order to
   receive service.  Extensions defined by experimental and
   informational RFCs are explicitly excluded from usage with the
   Supported header field in a request, since they too are often used to
   document vendor-defined extensions.

So I see that as a fundamental architectural decision, SIP does not
permit sending INVITEs that require specialized recipients, excepting if
the specialization is standards-track.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential =
information, privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 15 22:51:21 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHu7t-0005xb-VX; Thu, 15 Feb 2007 22:51:13 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHu7s-0005xT-Oy
	for sipping@ietf.org; Thu, 15 Feb 2007 22:51:12 -0500
Received: from c243-111.rim.net ([216.9.243.111] helo=MHS01YKF.rim.net)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHu7p-000628-N1
	for sipping@ietf.org; Thu, 15 Feb 2007 22:51:12 -0500
Importance: normal
Priority: normal
Received: from ngw01ykf.rim.net ([10.102.108.30]) by MHS01YKF.rim.net with
	Microsoft SMTPSVC(6.0.3790.211); Thu, 15 Feb 2007 22:51:01 -0500
Received: from XCH21YKF.rim.net ([10.102.100.36]) by ngw01ykf.rim.net (SAVSMTP
	3.1.6.45) with SMTP id M2007021522510031853 ;
	Thu, 15 Feb 2007 22:51:00 -0500
Received: from XCH47YKF.rim.net ([10.64.31.217]) by XCH21YKF.rim.net with
	Microsoft SMTPSVC(5.0.2195.6713); Thu, 15 Feb 2007 22:51:01 -0500
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.181
Content-Class: urn:content-classes:message
MIME-Version: 1.0
Subject: RE: Device architecture? WAS: RE: [Sipping] Service Identificatio	n
	inSIP
Date: Thu, 15 Feb 2007 22:50:59 -0500
Message-ID: <61968779B8AC4C4BAB421D4C12F008C00915146B@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Device architecture? WAS: RE: [Sipping] Service Identificatio	n
	inSIP
thread-index: AcdQ8txxUBhdMWq5TIulh7vriiYTRQAib3RQ
From: "Andrew Allen" <aallen@rim.com>
To: "Elwell, John" <john.elwell@siemens.com>,
	"Malleswara Rao Ankem" <malleshavn@gmail.com>
X-OriginalArrivalTime: 16 Feb 2007 03:51:01.0053 (UTC)
	FILETIME=[AC8556D0:01C7517D]
X-Spam-Score: 0.1 (/)
X-Scan-Signature: d62adab73c3674ddf4d2b0d221683819
Cc: Andrew Allen <aallen@rim.com>, sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============2066490588=="
Errors-To: sipping-bounces@ietf.org

This is a multi-part message in MIME format.

--===============2066490588==
Content-Transfer-Encoding: 7bit
Content-Class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C7517D.AC491183"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C7517D.AC491183
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

=20
John,
=20
If Service Identification was only about selecting the appropriate =
contacts from those registered agaisnt the AOR then RFC 3840/3841 would =
be fine however Service Identification is much more than that and may be =
used to route requests to application servers for service logic =
invocation, and used as the basis for charging. Accept-Contact is =
basically just a filter expression to filter a set of registered =
contacts. Including a parameter in a filter expression and using that as =
the basis for charging, authorisation and service invocation is going to =
cause problems.
=20
Andrew
=20
________________________________

From: Elwell, John [mailto:john.elwell@siemens.com]=20
Sent: Thursday, February 15, 2007 6:16 AM
To: Malleswara Rao Ankem
Cc: sipping@ietf.org
Subject: RE: Device architecture? WAS: RE: [Sipping] Service =
Identificatio n inSIP


Mallesh,
=20
Yes, I agree that some means of selecting a preferred device might be =
needed, but that is indeed what RFC3841/3840 does. Why do you consider =
this overkill?
=20
John


________________________________

	From: Malleswara Rao Ankem [mailto:malleshavn@gmail.com]=20
	Sent: 15 February 2007 10:54
	To: Elwell, John
	Cc: G=F6ran Eriksson AP (KI/EAB); sipping@ietf.org
	Subject: Re: Device architecture? WAS: RE: [Sipping] Service =
Identificatio n in SIP
=09
=09
	Inorder to identify a endpoint that has the capabilities to serve a =
very specialized functionality, I guess it is simpler to name the =
functionality. Naming the functionality not only helps in decision =
making while routing the request to that specific endpoint but also =
helps the endpoint identify which specialized application on the =
endpoint can take up the responsibility. Instead of all the routing =
entities looking into the SDP/message bodies it should be easier/faster =
if they can look for a specific header.=20
	I thought feature tag mechanisms indicated in the RFC3841/3840 =
combination should have sufficed but implementing those RFCs seems to be =
a overkill for describing/selecting/routing requests to the specific =
endpoint and also choosing a service/feature/application on the =
endpoint.=20
	=20
	-Mallesh
	=20
	=20
	=20
	On 2/15/07, Elwell, John <john.elwell@siemens.com > wrote:=20

		G=F6ran,
	=09
		Your interpretation is different from what is stated in the Rosenberg =
draft.
		Anyway, let's not dwell on this particular use case, which may or may =
not be=20
		realistic. The important thing is whether the media being sent is =
suitable
		for general purpose devices or requires or could be used by some =
special
		class of application.
	=09
		If a medium is somehow specialised ( e.g., audio modulated for =
interpretation
		by a modem or fax machine), then the necessary information needs to be
		conveyed along with the m-line in the SDP. Perhaps even if it is
		human-intelligible voice but with a restricted repertoire of words =
that make=20
		it suitable for a particular class of application, maybe that =
information
		too should be conveyed with the m-line. Perhaps the case described in
		Jonathan's draft fits this latter category: I believe the medium is =
just=20
		regular voice that can be interpreted by a human, but is also =
constrained in
		its repertoire of words such that it can be used by a chess =
application.
	=09
		The important thing is that there be sufficient information in the SDP =
to=20
		allow appropriate application selection at the end device. I am not
		convinced by a need for information at the SIP level during call
		establishment, because of course SDP can be re-negotiated during the =
call
		( e.g., when the users decide to use a particular application) and =
what is
		indicated in SIP at call establishment time might no longer be valid =
later
		in the call.
	=09
		John
	=09
	=09
		> -----Original Message-----=20
		> From: G=F6ran Eriksson AP (KI/EAB)
		> [mailto:goran.ap.eriksson@ericsson.com]
		> Sent: 14 February 2007 23:34=20
		> To: Elwell, John
		> Cc: sipping@ietf.org
		> Subject: Device architecture? WAS: RE: [Sipping] Service=20
		> Identification in SIP
		>
		> My humble, perhaps faulty, understanding of the "chess game"
		> was not that voice commands were used to control the chess=20
		> itself, rather
		> that the chess game, while having some IP-based proprietary=20
		> gaming interaction protocol, also had a feature letting
		> the users talk to each others while playing. The scenario are=20
		> two, or more, users play a chess game of the (Inter)net, talkin
		> to eachother from time to time during the game (yeah, don't=20
		> ask me if I think people will love this!!).
		>
		> The purpose, as I understand it, was a)to show that the=20
		> 'voice component' in the chess game application is similar to that
		> in a stand alone POC-Only application when it comes to=20
		> exchange and transfer of voice samples and associated floor control=20
		> and b) to point to differences, for instance when it comes to
		> supplementary services, which may not apply in the chess game case.=20
		>
		> Perhaps it is important to be clear on the assumed user
		> device realisation, as far as I understand it:
		>
		>  The "chess game + voice" example is, as I understand it,
		> discussed under the assumption that voice and chess game is=20
		>  managed by custome designed appliction, or client, which is=20
		> a necessary/likely approach in a mobile device given the
		>  limited MMI (and also other factors in the mobile context).
		>
		>  In a PC, I could easily handle this as two different=20
		> applications/client, like when I use Skype/MSN to talk to my=20
		> friends while
		>  playing some turn based game, but again, the discussion is
		> held under the assumption that usability requirements=20
		>  leads to an application client integrating voice and chess
		> game events.
		>
		> The alternative device realisation You describe- different
		> clients managing different portions of what is negotiated in=20
		> the SDP or MSRP MIME attachment-
		> is, I think, a good illustration of the need to be precise=20
		> and explicit about the assumed device architecture when
		> discussing the "service identification" issue(s).=20
		>
		> Regards
		> G=F6ran
		>
		>
		>
		>
		> -----Original Message-----=20
		> From: Elwell, John [mailto: john.elwell@siemens.com =
<mailto:john.elwell@siemens.com> ]
		> Sent: den 14 februari 2007 11:57
		> To: Henry Sinnreich; sipping@ietf.org=20
		> Subject: RE: [Sipping] Service Identification in SIP
		>
		> Having followed this thread, I too find the chess example
		> very contrived. If chess moves are sent as voice commands,
		> then surely it would be acceptable for the normal telephone=20
		> application to answer the call, i.e., the remote player
		> listens to the move, performs this move on a conventional or
		> electronic chess board and speaks his next move. The calling
		> UA should not care whether the call is answered by a computer=20
		> application or a person.
		> Furthermore, even if the audio needs to consumed by a
		> particular chess application at the called UA, then other
		> media in the same session such as video or messaging might=20
		> not be. So any qualification of the nature of media to be
		> exchanged probably should be at the media level, not at the
		> session level or SIP level. As yet another variation, what if
		> the players set up a conventional telephone call and then=20
		> mutually agree to switch in their chess applications?
		>
		> Now, let's consider another case. I establish an MSRP
		> session. Within this MSRP session I can send messages with
		> various MIME attachments that are intended for consumption by=20
		> various applications. I should not need to know in advance
		> what applications I will use during the course of the MSRP
		> session. All I should put in the INVITE request is the fact
		> that I want to establish an MSRP session.=20
		>
		> John
		>
		> > -----Original Message-----
		> > From: Henry Sinnreich [mailto:hsinnrei@adobe.com ]
		> > Sent: 13 February 2007 18:19
		> > To: sipping@ietf.org
		> > Subject: RE: [Sipping] Service Identification in SIP=20
		> >
		> > It must be my lack of imagination for not being able to
		> come up with a=20
		> > credible usage scenario, given the following:
		> >
		> > - Is playing chess by voice a credible usage scenario? :-)=20
		> Especially
		> > since voice recognition is now so often found in PCs and in mobile =

		> > phones as well. In other words, chess playing by voice may
		> also be an
		> > endpoint application, without any network involvement. Actually, =
if=20
		> > placed in the endpoint it works just as well on poor=20
		> bandwidth links
		> > by contrast to chess in the network that may require adequate
		> > broadband.
		> >
		> > - Is anyone's boss or spouse going to approve extra phone charges=20
		> > above the flat rate for playing chess at a time when the gaming
		> > industry has pioneered free VoIP?
		> > Will such a service be taxed extra and if yes using by what
		> criteria?
		> > Gambling comes to my mind :-)=20
		> >
		> > - VoIP is reported in the press as a give-away and I have seen no
		> > reports of any VoIP provider recording a profit on VoIP. I
		> assume the
		> > losses are often hidden by bundling in triple play, etc.=20
		> >
		> > Given my lack of imagination, can anyone share a more
		> credible usage
		> > scenario for Service Identification in SIP?
		> >
		> > What's the use of a standard without a credible usage scenario?=20
		> >
		> > Thanks, Henry
		> >
		> >
		> > _______________________________________________
		> > Sipping mailing list   =
https://www1.ietf.org/mailman/listinfo/sipping =
<https://www1.ietf.org/mailman/listinfo/sipping>=20
		> > This list is for NEW development of the application of SIP Use
		> > sip-implementors@cs.columbia.edu for questions on current sip Use
		> > sip@ietf.org for new developments of core SIP=20
		> >
		>
		> _______________________________________________
		> Sipping mailing list   =
https://www1.ietf.org/mailman/listinfo/sipping =
<https://www1.ietf.org/mailman/listinfo/sipping>=20
		> This list is for NEW development of the application of SIP
		> Use sip-implementors@cs.columbia.edu for questions on current=20
		> sip Use sip@ietf.org for new developments of core SIP=20
		>
	=09
		_______________________________________________
		Sipping mailing list   https://www1.ietf.org/mailman/listinfo/sipping =
<https://www1.ietf.org/mailman/listinfo/sipping>=20
		This list is for NEW development of the application of SIP
		Use sip-implementors@cs.columbia.edu for questions on current sip=20
		Use sip@ietf.org for new developments of core SIP
	=09





---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential =
information, privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.

------_=_NextPart_001_01C7517D.AC491183
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; =
charset=3Diso-8859-1">
<META content=3D"MSHTML 6.00.2900.2963" name=3DGENERATOR></HEAD>
<BODY>
<DIV dir=3Dltr align=3Dleft><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>John,</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>If Service Identification was only about =
selecting the=20
appropriate contacts from those registered agaisnt the AOR then RFC =
3840/3841=20
would be fine however Service Identification is much more than that and =
may be=20
used to route requests to application servers for service logic =
invocation, and=20
used as the basis for charging. Accept-Contact is basically just a =
filter=20
expression to filter a set of registered contacts. Including a parameter =
in a=20
filter expression and using that as the basis for charging,=20
authorisation&nbsp;and service invocation is going to cause=20
problems.</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>Andrew</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN =
class=3D927204303-16022007></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft>
<HR tabIndex=3D-1>
</DIV>
<DIV dir=3Dltr align=3Dleft><FONT face=3DTahoma size=3D2><B>From:</B> =
Elwell, John=20
[mailto:john.elwell@siemens.com] <BR><B>Sent:</B> Thursday, February 15, =
2007=20
6:16 AM<BR><B>To:</B> Malleswara Rao Ankem<BR><B>Cc:</B>=20
sipping@ietf.org<BR><B>Subject:</B> RE: Device architecture? WAS: RE: =
[Sipping]=20
Service Identificatio n inSIP<BR></FONT><BR></DIV>
<DIV></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>Mallesh,</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>Yes, I agree that some means of selecting a =
preferred=20
device might be needed, but that is indeed what RFC3841/3840 does. Why =
do you=20
consider this overkill?</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>John</FONT></SPAN></DIV><BR>
<BLOCKQUOTE=20
style=3D"PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #0000ff 2px =
solid; MARGIN-RIGHT: 0px">
  <DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr align=3Dleft>
  <HR tabIndex=3D-1>
  <FONT face=3DTahoma size=3D2><B>From:</B> Malleswara Rao Ankem=20
  [mailto:malleshavn@gmail.com] <BR><B>Sent:</B> 15 February 2007=20
  10:54<BR><B>To:</B> Elwell, John<BR><B>Cc:</B> G=F6ran Eriksson AP =
(KI/EAB);=20
  sipping@ietf.org<BR><B>Subject:</B> Re: Device architecture? WAS: RE:=20
  [Sipping] Service Identificatio n in SIP<BR></FONT><BR></DIV>
  <DIV></DIV>
  <DIV>Inorder to identify a endpoint that has the capabilities to serve =
a very=20
  specialized functionality, I guess it is simpler to name the =
functionality.=20
  Naming the functionality not only helps in decision making while =
routing the=20
  request to that specific endpoint but also helps the endpoint identify =
which=20
  specialized application on the endpoint can take up the =
responsibility.=20
  Instead of all the routing entities looking into the SDP/message =
bodies it=20
  should be easier/faster if they can look for a specific header. </DIV>
  <DIV>I thought feature tag mechanisms indicated in the RFC3841/3840=20
  combination should have sufficed but implementing those RFCs seems to =
be a=20
  overkill for describing/selecting/routing requests&nbsp;to the =
specific=20
  endpoint and also choosing a service/feature/application on the =
endpoint.=20
  </DIV>
  <DIV>&nbsp;</DIV>
  <DIV>-Mallesh</DIV>
  <DIV>&nbsp;</DIV>
  <DIV>&nbsp;</DIV>
  <DIV>&nbsp;</DIV>
  <DIV><SPAN class=3Dgmail_quote>On 2/15/07, <B =
class=3Dgmail_sendername>Elwell,=20
  John</B> &lt;<A onclick=3D"return =
top.js.OpenExtLink(window,event,this)"=20
  href=3D"mailto:john.elwell@siemens.com" =
target=3D_blank>john.elwell@siemens.com=20
  </A>&gt; wrote:</SPAN>=20
  <BLOCKQUOTE class=3Dgmail_quote=20
  style=3D"PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: =
#ccc 1px solid">G=F6ran,<BR><BR>Your=20
    interpretation is different from what is stated in the Rosenberg=20
    draft.<BR>Anyway, let's not dwell on this particular use case, which =
may or=20
    may not be <BR>realistic. The important thing is whether the media =
being=20
    sent is suitable<BR>for general purpose devices or requires or could =
be used=20
    by some special<BR>class of application.<BR><BR>If a medium is =
somehow=20
    specialised ( e.g., audio modulated for interpretation<BR>by a modem =
or fax=20
    machine), then the necessary information needs to be<BR>conveyed =
along with=20
    the m-line in the SDP. Perhaps even if it is<BR>human-intelligible =
voice but=20
    with a restricted repertoire of words that make <BR>it suitable for =
a=20
    particular class of application, maybe that information<BR>too =
should be=20
    conveyed with the m-line. Perhaps the case described =
in<BR>Jonathan's draft=20
    fits this latter category: I believe the medium is just <BR>regular =
voice=20
    that can be interpreted by a human, but is also constrained =
in<BR>its=20
    repertoire of words such that it can be used by a chess=20
    application.<BR><BR>The important thing is that there be sufficient=20
    information in the SDP to <BR>allow appropriate application =
selection at the=20
    end device. I am not<BR>convinced by a need for information at the =
SIP level=20
    during call<BR>establishment, because of course SDP can be =
re-negotiated=20
    during the call<BR>( e.g., when the users decide to use a particular =

    application) and what is<BR>indicated in SIP at call establishment =
time=20
    might no longer be valid later<BR>in the =
call.<BR><BR>John<BR><BR><BR>&gt;=20
    -----Original Message----- <BR>&gt; From: G=F6ran Eriksson AP =
(KI/EAB)<BR>&gt;=20
    [mailto:<A onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:goran.ap.eriksson@ericsson.com"=20
    target=3D_blank>goran.ap.eriksson@ericsson.com</A>]<BR>&gt; Sent: 14 =
February=20
    2007 23:34 <BR>&gt; To: Elwell, John<BR>&gt; Cc: <A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sipping@ietf.org" =
target=3D_blank>sipping@ietf.org</A><BR>&gt;=20
    Subject: Device architecture? WAS: RE: [Sipping] Service <BR>&gt;=20
    Identification in SIP<BR>&gt;<BR>&gt; My humble, perhaps faulty,=20
    understanding of the "chess game"<BR>&gt; was not that voice =
commands were=20
    used to control the chess <BR>&gt; itself, rather<BR>&gt; that the =
chess=20
    game, while having some IP-based proprietary <BR>&gt; gaming =
interaction=20
    protocol, also had a feature letting<BR>&gt; the users talk to each =
others=20
    while playing. The scenario are <BR>&gt; two, or more, users play a =
chess=20
    game of the (Inter)net, talkin<BR>&gt; to eachother from time to =
time during=20
    the game (yeah, don't <BR>&gt; ask me if I think people will love=20
    this!!).<BR>&gt;<BR>&gt; The purpose, as I understand it, was a)to =
show that=20
    the <BR>&gt; 'voice component' in the chess game application is =
similar to=20
    that<BR>&gt; in a stand alone POC-Only application when it comes to =
<BR>&gt;=20
    exchange and transfer of voice samples and associated floor control =
<BR>&gt;=20
    and b) to point to differences, for instance when it comes =
to<BR>&gt;=20
    supplementary services, which may not apply in the chess game case.=20
    <BR>&gt;<BR>&gt; Perhaps it is important to be clear on the assumed=20
    user<BR>&gt; device realisation, as far as I understand=20
    it:<BR>&gt;<BR>&gt;&nbsp;&nbsp;The "chess game + voice" example is, =
as I=20
    understand it,<BR>&gt; discussed under the assumption that voice and =
chess=20
    game is <BR>&gt;&nbsp;&nbsp;managed by custome designed appliction, =
or=20
    client, which is <BR>&gt; a necessary/likely approach in a mobile =
device=20
    given the<BR>&gt;&nbsp;&nbsp;limited MMI (and also other factors in =
the=20
    mobile context).<BR>&gt;<BR>&gt;&nbsp;&nbsp;In a PC, I could easily =
handle=20
    this as two different <BR>&gt; applications/client, like when I use=20
    Skype/MSN to talk to my <BR>&gt; friends =
while<BR>&gt;&nbsp;&nbsp;playing=20
    some turn based game, but again, the discussion is<BR>&gt; held =
under the=20
    assumption that usability requirements <BR>&gt;&nbsp;&nbsp;leads to =
an=20
    application client integrating voice and chess<BR>&gt; game=20
    events.<BR>&gt;<BR>&gt; The alternative device realisation You =
describe-=20
    different<BR>&gt; clients managing different portions of what is =
negotiated=20
    in <BR>&gt; the SDP or MSRP MIME attachment-<BR>&gt; is, I think, a =
good=20
    illustration of the need to be precise <BR>&gt; and explicit about =
the=20
    assumed device architecture when<BR>&gt; discussing the "service=20
    identification" issue(s). <BR>&gt;<BR>&gt; Regards<BR>&gt;=20
    G=F6ran<BR>&gt;<BR>&gt;<BR>&gt;<BR>&gt;<BR>&gt; -----Original =
Message-----=20
    <BR>&gt; From: Elwell, John [mailto:<A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:john.elwell@siemens.com" target=3D_blank>=20
    john.elwell@siemens.com</A>]<BR>&gt; Sent: den 14 februari 2007=20
    11:57<BR>&gt; To: Henry Sinnreich; <A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sipping@ietf.org" target=3D_blank>sipping@ietf.org =
</A><BR>&gt;=20
    Subject: RE: [Sipping] Service Identification in SIP<BR>&gt;<BR>&gt; =
Having=20
    followed this thread, I too find the chess example<BR>&gt; very =
contrived.=20
    If chess moves are sent as voice commands,<BR>&gt; then surely it =
would be=20
    acceptable for the normal telephone <BR>&gt; application to answer =
the call,=20
    i.e., the remote player<BR>&gt; listens to the move, performs this =
move on a=20
    conventional or<BR>&gt; electronic chess board and speaks his next =
move. The=20
    calling<BR>&gt; UA should not care whether the call is answered by a =

    computer <BR>&gt; application or a person.<BR>&gt; Furthermore, even =
if the=20
    audio needs to consumed by a<BR>&gt; particular chess application at =
the=20
    called UA, then other<BR>&gt; media in the same session such as =
video or=20
    messaging might <BR>&gt; not be. So any qualification of the nature =
of media=20
    to be<BR>&gt; exchanged probably should be at the media level, not =
at=20
    the<BR>&gt; session level or SIP level. As yet another variation, =
what=20
    if<BR>&gt; the players set up a conventional telephone call and then =

    <BR>&gt; mutually agree to switch in their chess=20
    applications?<BR>&gt;<BR>&gt; Now, let's consider another case. I =
establish=20
    an MSRP<BR>&gt; session. Within this MSRP session I can send =
messages=20
    with<BR>&gt; various MIME attachments that are intended for =
consumption by=20
    <BR>&gt; various applications. I should not need to know in =
advance<BR>&gt;=20
    what applications I will use during the course of the MSRP<BR>&gt; =
session.=20
    All I should put in the INVITE request is the fact<BR>&gt; that I =
want to=20
    establish an MSRP session. <BR>&gt;<BR>&gt; John<BR>&gt;<BR>&gt; =
&gt;=20
    -----Original Message-----<BR>&gt; &gt; From: Henry Sinnreich =
[mailto:<A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:hsinnrei@adobe.com" =
target=3D_blank>hsinnrei@adobe.com=20
    </A>]<BR>&gt; &gt; Sent: 13 February 2007 18:19<BR>&gt; &gt; To: <A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sipping@ietf.org" =
target=3D_blank>sipping@ietf.org</A><BR>&gt;=20
    &gt; Subject: RE: [Sipping] Service Identification in SIP <BR>&gt;=20
    &gt;<BR>&gt; &gt; It must be my lack of imagination for not being =
able=20
    to<BR>&gt; come up with a <BR>&gt; &gt; credible usage scenario, =
given the=20
    following:<BR>&gt; &gt;<BR>&gt; &gt; - Is playing chess by voice a =
credible=20
    usage scenario? :-) <BR>&gt; Especially<BR>&gt; &gt; since voice =
recognition=20
    is now so often found in PCs and in mobile <BR>&gt; &gt; phones as =
well. In=20
    other words, chess playing by voice may<BR>&gt; also be an<BR>&gt; =
&gt;=20
    endpoint application, without any network involvement. Actually, if =
<BR>&gt;=20
    &gt; placed in the endpoint it works just as well on poor <BR>&gt; =
bandwidth=20
    links<BR>&gt; &gt; by contrast to chess in the network that may =
require=20
    adequate<BR>&gt; &gt; broadband.<BR>&gt; &gt;<BR>&gt; &gt; - Is =
anyone's=20
    boss or spouse going to approve extra phone charges <BR>&gt; &gt; =
above the=20
    flat rate for playing chess at a time when the gaming<BR>&gt; &gt; =
industry=20
    has pioneered free VoIP?<BR>&gt; &gt; Will such a service be taxed =
extra and=20
    if yes using by what<BR>&gt; criteria?<BR>&gt; &gt; Gambling comes =
to my=20
    mind :-) <BR>&gt; &gt;<BR>&gt; &gt; - VoIP is reported in the press =
as a=20
    give-away and I have seen no<BR>&gt; &gt; reports of any VoIP =
provider=20
    recording a profit on VoIP. I<BR>&gt; assume the<BR>&gt; &gt; losses =
are=20
    often hidden by bundling in triple play, etc. <BR>&gt; &gt;<BR>&gt; =
&gt;=20
    Given my lack of imagination, can anyone share a more<BR>&gt; =
credible=20
    usage<BR>&gt; &gt; scenario for Service Identification in =
SIP?<BR>&gt;=20
    &gt;<BR>&gt; &gt; What's the use of a standard without a credible =
usage=20
    scenario? <BR>&gt; &gt;<BR>&gt; &gt; Thanks, Henry<BR>&gt; =
&gt;<BR>&gt;=20
    &gt;<BR>&gt; &gt; =
_______________________________________________<BR>&gt;=20
    &gt; Sipping mailing list&nbsp;&nbsp;<A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"https://www1.ietf.org/mailman/listinfo/sipping" =
target=3D_blank>=20
    https://www1.ietf.org/mailman/listinfo/sipping </A><BR>&gt; &gt; =
This list=20
    is for NEW development of the application of SIP Use<BR>&gt; &gt; <A =

    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sip-implementors@cs.columbia.edu"=20
    target=3D_blank>sip-implementors@cs.columbia.edu</A> for questions =
on current=20
    sip Use<BR>&gt; &gt; <A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sip@ietf.org" target=3D_blank>sip@ietf.org</A> for =
new=20
    developments of core SIP <BR>&gt; &gt;<BR>&gt;<BR>&gt;=20
    _______________________________________________<BR>&gt; Sipping =
mailing=20
    list&nbsp;&nbsp;<A onclick=3D"return =
top.js.OpenExtLink(window,event,this)"=20
    href=3D"https://www1.ietf.org/mailman/listinfo/sipping" =
target=3D_blank>=20
    https://www1.ietf.org/mailman/listinfo/sipping</A><BR>&gt; This list =
is for=20
    NEW development of the application of SIP<BR>&gt; Use <A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sip-implementors@cs.columbia.edu"=20
    target=3D_blank>sip-implementors@cs.columbia.edu</A> for questions =
on current=20
    <BR>&gt; sip Use <A onclick=3D"return =
top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sip@ietf.org" target=3D_blank>sip@ietf.org</A> for =
new=20
    developments of core SIP=20
    =
<BR>&gt;<BR><BR>_______________________________________________<BR>Sippin=
g=20
    mailing list&nbsp;&nbsp;<A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"https://www1.ietf.org/mailman/listinfo/sipping" =
target=3D_blank>=20
    https://www1.ietf.org/mailman/listinfo/sipping </A><BR>This list is =
for NEW=20
    development of the application of SIP<BR>Use <A=20
    onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sip-implementors@cs.columbia.edu"=20
    target=3D_blank>sip-implementors@cs.columbia.edu </A>for questions =
on current=20
    sip <BR>Use <A onclick=3D"return =
top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:sip@ietf.org" target=3D_blank>sip@ietf.org</A> for =
new=20
    developments of core =
SIP<BR></BLOCKQUOTE></DIV><BR></BLOCKQUOTE><p><hr>This transmission =
(including any attachments) may contain confidential information, =
privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.</BODY></HTML>

------_=_NextPart_001_01C7517D.AC491183--


--===============2066490588==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============2066490588==--




From sipping-bounces@ietf.org Thu Feb 15 23:45:08 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHuwc-0003fR-2s; Thu, 15 Feb 2007 23:43:38 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHuwa-0003fM-Nv
	for sipping@ietf.org; Thu, 15 Feb 2007 23:43:36 -0500
Received: from c243-111.rim.net ([216.9.243.111] helo=MHS01YKF.rim.net)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHuwX-00087M-2S
	for sipping@ietf.org; Thu, 15 Feb 2007 23:43:36 -0500
Importance: normal
Priority: normal
Received: from ngw01ykf.rim.net ([10.102.108.30]) by MHS01YKF.rim.net with
	Microsoft SMTPSVC(6.0.3790.211); Thu, 15 Feb 2007 23:43:22 -0500
Received: from XCH20YKF.rim.net ([10.102.100.35]) by ngw01ykf.rim.net (SAVSMTP
	3.1.6.45) with SMTP id M2007021523432200278 ;
	Thu, 15 Feb 2007 23:43:22 -0500
Received: from XCH47YKF.rim.net ([10.64.31.217]) by XCH20YKF.rim.net with
	Microsoft SMTPSVC(5.0.2195.6713); Thu, 15 Feb 2007 23:43:22 -0500
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.181
Content-Class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Thu, 15 Feb 2007 23:43:20 -0500
Message-ID: <61968779B8AC4C4BAB421D4C12F008C00915146C@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
thread-index: AcdRa5UuVWahqYE5RSSUh5kw9n5XogAFaSXA
From: "Andrew Allen" <aallen@rim.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>
X-OriginalArrivalTime: 16 Feb 2007 04:43:22.0333 (UTC)
	FILETIME=[FCDEA0D0:01C75184]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: efb5d987e2484f3d9a304cc31a003441
Cc: Gonzalo.Camarillo@ericsson.com, sipping@ietf.org, mary.barnes@nortel.com,
	ravishankar.shiroor@wipro.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Paul

Comments inline

Andrew=20

-----Original Message-----
From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
Sent: Thursday, February 15, 2007 8:41 PM
To: Andrew Allen
Cc: Salvatore Loreto (JO/LMF); mary.barnes@nortel.com; sipping@ietf.org;
ravishankar.shiroor@wipro.com; Gonzalo.Camarillo@ericsson.com
Subject: Re: [Sipping] Service Identification in SIP



Andrew Allen wrote:
> I agree with Salvatore that IM is a better example of where service=20
> identification may have a role in expressing a preference for a=20
> particular application. Certainly there are different emoticons and=20
> other value added features proprietary to different messaging=20
> applications so having the ability to express preferences to use a=20
> preferred application seems to me to have value. I think that email is

> a bad counter example as with email the email client interfaces to the

> email server not directly to the other email client.

If the kind of usage above were to take hold in any significant way, it
would be a huge step towards balkanizing the internet. Part of agreeing
on a common protocol is actually supporting interoperation.

[AA] My thinking is the opposite. Currently today with several non
interoperable proprietary IM systems we have balkanizing of the
internet. SIMPLE hopefully will lead to inteoperable IM systems however
I think it likely that many of these different systems will still retain
some proprietary value added features even when they have
inteoperability for basic messaging exchange. Having the ability to
express a preference for an application that supports the same value
added features I think aids inteoperability and removes an excuse for
maintaining a totally proprietary non inteoperable system.

But I agree with what Andrew says below.

	Paul

> Where I disagree with Salvatore is that the fact that the use of=20
> callerprefs doesn't support well, if at all, the inclusion of=20
> intermediate servers into the call path is not a problem for 3GPP as=20
> well. From my viewpoint 3GPP needs to make use of Callerprefs in the=20
> same way as was intended in RFC 3840/3841 and not extend this with an=20
> entire bunch of semantics that were not defined for Accept-Contact. To

> do otherwise is only likely to create problems with services that work

> in the internet but do not work with mobile and other devices in IMS=20
> networks. Also service identification needs to work in both IMS=20
> networks and non-IMS networks so that the the same services can be=20
> deployed in both environments. The fact that IMS filter criteria=20
> enables routing on just about anything (Using iFC I could route based=20
> on my favourite number appearing in a Call-ID or Cseq etc) doesn't=20
> mean that it's Ok or a good idea to simply add a new token anywhere=20
> you can and route based on that. Services should not depend on the use

> of filter criteria and the IMS architecture and routing decisions=20
> should be based on the SIP routing mechanisms that use the contents of

> the Route header and Request-URI for routing decisions. IMHO=20
> Accept-Contact is not a routing header per say but simply allows a=20
> filter to be specified that filters the list of registered contacts in

> order to determine which contacts of the AOR to fork the request to.=20
> Building additional semantics on top of this filter potentially=20
> significantly restricts the use of this filter mechanism as originally
intended.
>=20
> I think it is in the interest of both 3GPP and the internet that a=20
> general mechanism that works also in a pure SIP environment is=20
> developed.
>=20
> Andrew
>=20
>=20
> -----Original Message-----
> From: Salvatore Loreto (JO/LMF) [mailto:salvatore.loreto@ericsson.com]
> Sent: Friday, February 09, 2007 9:39 AM
> To: Paul Kyzivat
> Cc: mary.barnes@nortel.com; sipping@ietf.org;=20
> ravishankar.shiroor@wipro.com; Gonzalo.Camarillo@ericsson.com
> Subject: Re: [Sipping] Service Identification in SIP
>=20
> Hi,
>=20
> In my opinion we should also consider the user expectation.
> I fully agree with you about the mail service, but if I have look at=20
> the Instant Messaging service where at present there are different and

> incompatible programs, then maybe the necessity of a identification=20
> for the specific IM service become a little bit more clear.
>=20
>=20
> I think that everyone here agree that there is the possibility to use=20
> a media in different ways, and that we need a distinction about what=20
> the capabilities are and how they are used, whatever will be the=20
> mechanism that at the end we will identify to do that.
> What we do need is a way to identify a preference.
>=20
> then what I don't completely understand, surely it is my fault, that a

> combination of capabilities and ways to use those capabilities it is=20
> not itself a capability, perhaps something we can call a=20
> *macro-capability* Even if I agree on the opportunity of don't create=20
> a real Tower of Babel and avoid . So definitely this macro-capability=20
> should be well defined and standardized.
>=20
>=20
> Of course we need also discuss the opportunity to use the collerpref=20
> mechanism to address this capability.
> I agree with Paul that "callerprefs doesn't support well, if at all,=20
> is the inclusion of intermediate servers into the call path."
> This of course is not a problem from 3GPP point of view, because in a=20
> IMS network it is present the IFC mechanism that resolve this=20
> "callerpref" problem.
> But I do think we need a general mechanism that works also in a pure=20
> SIP environment.
>=20
>=20
>=20
> /sal
>=20
>=20
>=20
>=20
>=20
>=20
> On Fri, 2007-02-09 at 08:26 -0500, Paul Kyzivat wrote:
>> ravishankar.shiroor@wipro.com wrote:
>>> Hi Paul,
>>>
>>> IMO, service identification is different from identifying the tool
> that
>>> should be used for executing that service on the end device. In the=20
>>> analogy that you have given, e-mail is the service. Eudora/gmail/...
> Are
>>> the tools that are used to provide the service.
>> The comment below was about identifying the application in the end=20
>> system, not the protocol.
>>
>> In the case of email, the "service" is determined by the protocol and

>> the address/port of the destination. Nobody has suggested adding
> headers
>> to mail messages to identify the service/application in the end=20
>> system
>=20
>> that should process the mail message. But here we see proposals to=20
>> add
> a
>> header to a sip message to identify the service/application in the
> UAS.
>> 	Paul
>>
>>> Regards,
>>> Ravi.
>>>
>>>
>>> --
>>> =20
>>> Ravishankar. G. Shiroor
>>> Wipro Technologies, Bangalore.
>>> =20
>>> ravishankar.shiroor@wipro.com
>>> --
>>> =20
>>>
>>>> -----Original Message-----
>>>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
>>>> Sent: Wednesday, February 07, 2007 10:43 PM
>>>> To: Salvatore Loreto (JO/LMF)
>>>> Cc: RaviShankar Shiroor (WT01 - Product Engineering Solutions);=20
>>>> sipping@ietf.org; Gonzalo.Camarillo@ericsson.com;=20
>>>> mary.barnes@nortel.com
>>>> Subject: Re: [Sipping] Service Identification in SIP
>>>>
>>>> I still don't get the intent of identifying the application.=20
>>>> It seems to me this is evil - akin to sending an email message and=20
>>>> identifying that the recipient must use Outlook to receive it, not=20
>>>> Thunderbird, Eudora, gmail, or whatever.
>>>>
>>>> I also wish somebody could give better examples of a communication=20
>>>> service. Chess is too artificial to be believable, and IMO PoC is=20
>>>> only a service because of the way the implementation has been done.
>>>> Conceptually PoC could be nothing but a special floor control=20
>>>> protocol accompanying a voice session. In that case none of the=20
>>>> Service mechanism would be needed.
>>>>
>>>> 	Paul
>>>>
>>>> Salvatore Loreto (JO/LMF) wrote:
>>>>> hi Ravi,
>>>>>
>>>>> in IMS a Service is a "type of communication defined by a service=20
>>>>> definition that specifies the rules and procedures and
>>>> allowed media,
>>>>> if any, for a specific type of communication and that
>>>> utilises the IMS
>>>>> enabler. An IMS communication service defines what is
>>>> possible within
>>>>> a single SIP session or standalone transaction."
>>>>> So a Service implies some combination of capabilities and
>>>> ways to use
>>>>> those capabilities.
>>>>> If you want the Service specifies the context for a communication.
>>>>>
>>>>> Instead the Application define the specific communication
>>>> (or if you
>>>>> want application) that you want run in the specific context. So
> you
>>>>> should have several different applications designed to run on a=20
>>>>> specific environment or that have different performance or
>>>> behavior in
>>>>> different context (running on top of different service).
>>>>>
>>>>> Lets go to Jonathan's example of the Chess.
>>>>> In his example the Chess is a service, for IMS instead is an=20
>>>>> application that you can run on top of different service.
>>>>> For example you can run Chess application on top of PoC
>>>> service or on
>>>>> top of the "standard" service. So let say you run Chess on
>>>> top of PoC
>>>>> service, then you can use only voice in the way that PoC
> specifies.
>>>>> /sal
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> On Thu, 2007-02-01 at 16:39 +0530,
>>>> ravishankar.shiroor@wipro.com wrote:
>>>>>> Hello,
>>>>>>
>>>>>> IMO, Service Identification is definitely needed and
>>>> should be taken
>>>>>> up as a WG as proposed below.
>>>>>>
>>>>>> Further, I think Service Identification scope should be as per
> the
>>>>>> requirement in Jonathan's draft, as I still do not understand the
>=20
>>>>>> need for having a Communication Service Identifier and another=20
>>>>>> Application Identifier as given in the
>>>> 3gpp-ics-requirements document.
>>>>>> Regards,
>>>>>> Ravi.
>>>>>>
>>>>>> --
>>>>>> =20
>>>>>> Ravishankar. G. Shiroor
>>>>>> Wipro Technologies, Bangalore.
>>>>>> =20
>>>>>> ravishankar.shiroor@wipro.com
>>>>>> --
>>>>>> =20
>>>>>>
>>>>>>> -----Original Message-----
>>>>>>> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
>>>>>>> Sent: Wednesday, January 31, 2007 8:30 PM
>>>>>>> To: sipping
>>>>>>> Cc: Mary Barnes
>>>>>>> Subject: [Sipping] Service Identification in SIP
>>>>>>>
>>>>>>> Folks,
>>>>>>>
>>>>>>> during the last year, there have been a number of threads on the
>=20
>>>>>>> mailing list discussing service identification in SIP. In=20
>>>>>>> particular, the following thread discussed 3GPP
>>>> requirements on this
>>>>>>> area:
>>>>>>>
>>>>>>>
>>>> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
>>>>>>> At a later point, those 3GPP requirements were documented in the
>=20
>>>>>>> following (expired by now) Internet Draft, which was
>>>> presented in a
>>>>>>> face-to-face IETF meeting.
>>>>>>>
>>>>>>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
>>>>>>> cs-requirements-00.txt
>>>>>>>
>>>>>>> Lately, the following draft has been submitted to the archives:
>>>>>>>
>>>>>>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
>>>>>>> rvice-identification-00.txt
>>>>>>>
>>>>>>> This draft proposes a solution that intends to meet known=20
>>>>>>> requirements in this area.
>>>>>>>
>>>>>>> Given the amount of discussions on issues related to service=20
>>>>>>> identification we have seen during the last year, we
>>>> believe this is
>>>>>>> a fairly important topic for the SIPPING WG to tackle. At some=20
>>>>>>> point, we would like to have a milestone with its
>>>> associated WG item
>>>>>>> dealing with this issue.
>>>>>>>
>>>>>>> Therefore, we would like to encourage people to read this
>>>> draft and
>>>>>>> send comments to the list. Additionally, if you believe that
> this
>>>>>>> work is (or is not) important, we would like to hear that as
> well.
>>>>>>> Cheers,
>>>>>>>
>>>>>>> Gonzalo
>>>>>>> SIPPING co-chair
>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> Sipping mailing list
>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>>>> This list is for NEW development of the application of SIP Use=20
>>>>>>> sip-implementors@cs.columbia.edu for questions on current sip
> Use
>>>>>>> sip@ietf.org for new developments of core SIP
>>>>>>>
>>>>>> _______________________________________________
>>>>>> Sipping mailing list
>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>>> This list is for NEW development of the application of SIP Use=20
>>>>>> sip-implementors@cs.columbia.edu for questions on current sip Use
>=20
>>>>>> sip@ietf.org for new developments of core SIP
>>>>> _______________________________________________
>>>>> Sipping mailing list
> https://www1.ietf.org/mailman/listinfo/sipping
>>>>> This list is for NEW development of the application of SIP Use=20
>>>>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
>>>>> sip@ietf.org for new developments of core SIP
>>>>>
>>>
>>> The information contained in this electronic message and any
> attachments to this message are intended for the exclusive use of the
> addressee(s) and may contain proprietary, confidential or privileged=20
> information. If you are not the intended recipient, you should not=20
> disseminate, distribute or copy this e-mail. Please notify the sender=20
> immediately and destroy all copies of this message and any
attachments.
>>> WARNING: Computer viruses can be transmitted via email. The
> recipient should check this email and any attachments for the presence

> of viruses. The company accepts no liability for any damage caused by=20
> any virus transmitted by this email.
>>> =20
>>> www.wipro.com
>>>
>=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20
>=20
>=20
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential
information, privileged material (including material protected by the
solicitor-client or other applicable privileges), or constitute
non-public information. Any use of this information by anyone other than
the intended recipient is prohibited. If you have received this
transmission in error, please immediately reply to the sender and delete
this information from your system. Use, dissemination, distribution, or
reproduction of this transmission by unintended recipients is not
authorized and may be unlawful.
>=20



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential =
information, privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 16 01:16:39 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHwN0-00073N-SW; Fri, 16 Feb 2007 01:14:58 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHwMx-00071y-LA
	for sipping@ietf.org; Fri, 16 Feb 2007 01:14:55 -0500
Received: from c243-111.rim.net ([216.9.243.111] helo=MHS01YKF.rim.net)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHwMx-0000cP-1Y
	for sipping@ietf.org; Fri, 16 Feb 2007 01:14:55 -0500
Importance: normal
Priority: normal
Received: from ngw01ykf.rim.net ([10.102.108.30]) by MHS01YKF.rim.net with
	Microsoft SMTPSVC(6.0.3790.211); Fri, 16 Feb 2007 01:14:48 -0500
Received: from XCH20YKF.rim.net ([10.102.100.35]) by ngw01ykf.rim.net (SAVSMTP
	3.1.6.45) with SMTP id M2007021601144802503 ;
	Fri, 16 Feb 2007 01:14:48 -0500
Received: from XCH47YKF.rim.net ([10.64.31.217]) by XCH20YKF.rim.net with
	Microsoft SMTPSVC(5.0.2195.6713); Fri, 16 Feb 2007 01:14:48 -0500
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.181
Content-Class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Fri, 16 Feb 2007 01:14:46 -0500
Message-ID: <61968779B8AC4C4BAB421D4C12F008C00915146D@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
thread-index: AcdNmH89O4T46iGoRjicZ3hYknVpawB8E6oQAH++leA=
From: "Andrew Allen" <aallen@rim.com>
To: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>,
	<sipping@ietf.org>
X-OriginalArrivalTime: 16 Feb 2007 06:14:48.0337 (UTC)
	FILETIME=[C2C86810:01C75191]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8cb9b411340046bf4080a729180a0672
Cc: Andrew Allen <aallen@rim.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Some comments on Stephen's comments on the draft:
=20

-Moving onto the proposed solution draft itself, in addition for
providing for the identification of the "application" and the
"communication service", I have a few additional thoughts.

The draft seems to have the implication that all services (communication
services in other terminology) that are to be identified have to be
standardised within the IETF.  This is clearly an restriction that will
not be acceptable to the industry as a whole.  My guess is that if such
a statement is mandated by the IETF, it will just be words on paper
which would likely be ignored in practice by other standardisation
organisations.

[AA] This I think is an incorrect interpretation of the draft. My
understanding is that the only services according to the draft that
require some standardisation in IETF are those that by their very nature
cannot interoperate unless there is some common additional understanding
between the endpoints outside of what is currently specified in SIP. The
use of SIP option tags with Require header is proposed by the draft for
this purpose as such services create interoperability issues with SIP.

The draft also only recognises the case where all explicitly knowledge
of the service is required.  I think it would also be good to allow the
case where knowledge of the service would be helpful, but not explicitly
required.  This could be a case where the service to the end-user would
not be degraded if this was not supported.  This functionality seems to
be missing.

[AA] Service URIs are defined in the draft for these services where the
support by the other endpoint is not required for the session to be
sucessfully established.

The draft seems to indicate that that application servers in the network
are involved only at the request of the originating terminal.  This
seems a strange restriction.  For some network support that supports the
originating user, such as the one provided in the example "I would like
this call recorded for later retrieval", such mechanisms may make sense,
however as a general approach it seems a bit limited.  For example, in
3GPP networks there is the need to invoke application logic in the
originating side and terminating side, not only at the request of the
originating terminal, but as a consequence of the requested service.  If
there is an SIP message contains indication of the requested service, a
proxy can do this (or loose route to an AS that contains the necessary
service logic).=20

[AA] The draft supports both possibilities using the Route header for
originating network servers that need to be involved in the route and
based on the terminating networks routing policy for the Service URI in
the Request URI.

 I note here that the proposed service URN does not contain the address
of a specific server containing the application logic.  An AS can look
at that, and loose route to an application server that does contain such
logic (If it is separated from the proxy), however I fail to see the
difference for that proxy if it looks at a feature tag, or it looks at
the service URN for this.  In both cases, it has to parse the message
for the data, and modify the routing information. =20

[AA]Route headers are the SIP mechanism for requesting intermediaries to
be includede on the route of the request. There is nothing in RFC
3840/RFC 3841 that addresses using feature tags in Accept-Contact header
for any routing decisions prior to determining which of the registered
contacts the request should be forked to. Accept-Contact is not a Route
header it is merely a filter that is applied to the set of valid
contacts. Using a parameter in the filter as a service identifier has a
potential significant impact on the use of Accept-Contact for its
original purpose.=20


Further, the route header could be removed in the originating network;
and it is still required to identify the service in the terminating
network.  As proxies may remove the route, it can't be used to transport
the requested service. =20

[AA] Proxies arbitarily removing Route headers is a BAD thing akin to a
bull in a china shop - almost certainly going to break something. RFC
3261 summarises what proxies should be doing with Route headers:
16.12 Summary of Proxy Route Processing

   In the absence of local policy to the contrary, the processing a
   proxy performs on a request containing a Route header field can be
   summarized in the following steps.

      1.  The proxy will inspect the Request-URI.  If it indicates a
          resource owned by this proxy, the proxy will replace it with
          the results of running a location service.  Otherwise, the
          proxy will not change the Request-URI.

      2.  The proxy will inspect the URI in the topmost Route header
          field value.  If it indicates this proxy, the proxy removes it
          from the Route header field (this route node has been
          reached).

      3.  The proxy will forward the request to the resource indicated
          by the URI in the topmost Route header field value or in the
          Request-URI if no Route header field is present.  The proxy
          determines the address, port and transport to use when
          forwarding the request by applying the procedures in [4] to
          that URI.

   If no strict-routing elements are encountered on the path of the
   request, the Request-URI will always indicate the target of the
   request.

[AA] Generally a UA wouldn't know the appropriate service identifiers
used in the terminating network or whether a Route header is appropriate
to be used in the terminating network. The draft defines the Service URI
for the purpose of identifying services in the terminating network and
service URIs can be obtained using Presence.

>From this perspective it is not clear to me the advantage of using a
service URN in a route header compared to the use of a feature tag.

[AA] I hope the above responses clarify that somewhat. Also the Service
URN and Service URIs allow different service identifiers to be used in
the originating and terminating network for session compatiable services
instead of forcing everyone to deploy and understand the same service
and service identifier.

The draft seems to mandate GRUU - is this really so, is that really
necessary.  It seems strange to place such a dependency, it also starts
to move the GRUU as being the primary means of which to contact a user.

[AA] My understanding is that GRUU in the draft is used to address how a
request addressed to Service URI is forked only to those contacts that
support the service identified by the that Service URI. A contact would
seem to be fine so long as it had GRUU properties - no need to obtain a
GRUU if you already have one.

This implies that the originating user decides which terminal I should
receive my requested service on, even if I have two devices for that.
As such, I don't believe that there should be a dependency on GRUU.

[AA] If you have obtained a Service URI (probably through presence) for
that user then you can address the request to the Service URI which
represents all the terminals that support that service. However GRUU and
the Service Instance URIs created from GRUUs provide another mechanism
through which a user can obtain a URI to reach that user for the
required service.

Parts of the draft seems to rely on the use of presence.  While presence
could be a supportive mechanism, I doubt that we can mandate that the
number you want to dial has be available in presence as I assume not all
entities are available in presence (diallable network services, users in
the PSTN).  This would be strange to mandated.

[AA] PSTN gateways are presumably in the same domain as the terminating
proxy and therefore these PSTN gateways can be configured with the
appropriate service identifiers for telephony services and can include
this in either a Route header to route to the right application server
or can construct a valid Service URI. The real problem is how do users
in other networks know what the service identifier that's is used in the
terminating network is. The use of Service URIs and Presence provides a
solution to that problem.

Moving on from the draft I have seen the questioned raised about "why
can't we expand the power of SDP".  Such an approach would mandate that
SIP methods use SDP, I don't think that is appropriate.=20

[AA] I agree with you here

I hear the appeals that the mechanisms should not be abused, and that we
should not have 1000s of separate "communication services", and I fully
agree with this. I see a smaller number of communication services, not a
larger number; where those communication services are used to access
other end-user-services.

*flame away*

Regards,

Steve

---


-----Original Message-----
From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net]
Sent: Sunday, February 11, 2007 5:50 AM
To: sipping@ietf.org
Subject: Re: [Sipping] Service Identification in SIP

I've attempted to re-read the threads that Gonzalo referred to, and the
major limitation that I see is the lack of any concrete example of why
these identifiers are needed.  Probably the IMS people have discussed
this so many times that they've forgotten the detailed rationale behind
the requirement.  But I'm a net-head, (and most of the people here are
net-heads) and the SIP systems that I've dealt with show no evidence of
needing "service identification", they're pure proxies, and requests are
passed transparently end-to-end.

Could somebody please lay out an example showing the utility of service
identification?

   From: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>

   I think that everyone here agree that there is the possibility to use
a
   media in different ways, and that we need a distinction about what
the
   capabilities are and how they are used, whatever will be the
mechanism
   that at the end we will identify to do that.

Respectfully, I disagree.  We do need ways of identifying the
capabilities of SIP agents and transports, but I see no necessity of
creating specified bundles of capabailities, and then constraining
dialogs to require an initial agreement as to which bundle of
capabilities will be used.

One of the central features of Internet architecture is that agents
negotiate compatibility on a feature-by-feature basis; there are no
"profiles" which must be known on a network-wide basis.  (And I have
observed that this feature is of significant commercial value.)

In regard to identifying an "application" to process the dialog (and I
use "application" in a standard computing way), I see no reason that a
user cannot select URI-parameters to append to his AOR, and arrange for
his UA to route invites with those URI-parameters to the appropriate
application.  (And given the rich mechanisms for redirection in SIP, the
application need not even reside at a UA which is registered for the
AOR.)  These URI-parameters can be formally or informally standardized,
or established ad-hoc by each user.

In regard to PoC (as someone said, "a specialized floor-control
protocol") and similar features, SIP currently has mechanisms for
indicating support of features and requiring that a recipient support a
feature before accepting a call (the Supported and Requires headers, and
simply using the feature).  The only limitation I can see that might
need to be addressed is that there is no way to create an option tag for
a new feature that one has invented that is not yet thoroughly
standardized.

But I now discover RFC 3261 section 8.1.1.9:

   The option tags listed MUST only refer to extensions defined in
   standards-track RFCs.  This is to prevent servers from insisting that
   clients implement non-standard, vendor-defined features in order to
   receive service.  Extensions defined by experimental and
   informational RFCs are explicitly excluded from usage with the
   Supported header field in a request, since they too are often used to
   document vendor-defined extensions.

So I see that as a fundamental architectural decision, SIP does not
permit sending INVITEs that require specialized recipients, excepting if
the specialization is standards-track.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential =
information, privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 16 03:13:32 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHyC8-0005d7-7G; Fri, 16 Feb 2007 03:11:52 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHyC6-0005cr-8l
	for sipping@ietf.org; Fri, 16 Feb 2007 03:11:50 -0500
Received: from mailgate.siemenscomms.co.uk ([195.171.110.225]
	helo=bemg01.siemenscomms.co.uk)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHyC2-0002SX-Po
	for sipping@ietf.org; Fri, 16 Feb 2007 03:11:50 -0500
Received: from ntht207e.uksgcs.siemenscomms.co.uk ([137.223.247.82])
	by siemenscomms.co.uk (PMDF V6.0-24 #40642)
	with ESMTP id <0JDJ00H94QR9UH@siemenscomms.co.uk> for sipping@ietf.org;
	Fri, 16 Feb 2007 08:11:33 +0000 (GMT)
Received: by ntht207e.uksgcs.siemenscomms.co.uk with Internet Mail Service
	(5.5.2657.72)	id <D2K3Y9QF>; Fri, 16 Feb 2007 08:11:33 +0000
Content-return: allowed
Date: Fri, 16 Feb 2007 08:11:32 +0000
From: "Elwell, John" <john.elwell@siemens.com>
Subject: RE: Device architecture? WAS: RE: [Sipping] Service Identificatio	n
	inSIP
To: Andrew Allen <aallen@rim.com>, Malleswara Rao Ankem <malleshavn@gmail.com>
Message-id: <50B1CBA96870A34799A506B2313F26670B0F806B@ntht201e.siemenscomms.co.uk>
MIME-version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 4eaba8a3f66804313fcc9b62b5e6d8a8
Cc: Andrew Allen <aallen@rim.com>, sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============0400969377=="
Errors-To: sipping-bounces@ietf.org

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

--===============0400969377==
Content-return: allowed
Content-type: multipart/alternative;
	boundary="Boundary_(ID_6cW26/U9o/yFvNnJQnggRQ)"

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

--Boundary_(ID_6cW26/U9o/yFvNnJQnggRQ)
Content-type: text/plain; charset=iso-8859-1
Content-transfer-encoding: quoted-printable

Andrew,
=20
Well, an AS is just another contact (perhaps a permanently registered =
one,
rather than one that uses SIP REGISTER), isn't it? So caller =
preferences can
be used to choose an AS or any other contact that supports the desired
capability.
=20
I agree caller preferences is not a suitable basis for charging. Just
because I say I have a preference for a contact that supports video, =
doesn't
mean that I will necessarily use video during that call. That might be
determined during some mid-call session-re-negotiation. Surely charging =
has
to be based on what is actually used, not what might be used, and you =
can't
always determine this at INVITE time.
=20
So I am saying use caller preferences to select a preferred contact, =
use SDP
to negotiate and re-negotiate the session with the contact that you end =
up
with. I haven't heard any compelling arguments for anything else so =
far.
=20
John


  _____ =20

From: Andrew Allen [mailto:aallen@rim.com]=20
Sent: 16 February 2007 03:51
To: Elwell, John; Malleswara Rao Ankem
Cc: sipping@ietf.org; Andrew Allen
Subject: RE: Device architecture? WAS: RE: [Sipping] Service =
Identificatio n
inSIP


=20
John,
=20
If Service Identification was only about selecting the appropriate =
contacts
from those registered agaisnt the AOR then RFC 3840/3841 would be fine
however Service Identification is much more than that and may be used =
to
route requests to application servers for service logic invocation, and =
used
as the basis for charging. Accept-Contact is basically just a filter
expression to filter a set of registered contacts. Including a =
parameter in
a filter expression and using that as the basis for charging, =
authorisation
and service invocation is going to cause problems.
=20
Andrew
=20
  _____ =20

From: Elwell, John [mailto:john.elwell@siemens.com]=20
Sent: Thursday, February 15, 2007 6:16 AM
To: Malleswara Rao Ankem
Cc: sipping@ietf.org
Subject: RE: Device architecture? WAS: RE: [Sipping] Service =
Identificatio n
inSIP


Mallesh,
=20
Yes, I agree that some means of selecting a preferred device might be
needed, but that is indeed what RFC3841/3840 does. Why do you consider =
this
overkill?
=20
John


  _____ =20

From: Malleswara Rao Ankem [mailto:malleshavn@gmail.com]=20
Sent: 15 February 2007 10:54
To: Elwell, John
Cc: G=F6ran Eriksson AP (KI/EAB); sipping@ietf.org
Subject: Re: Device architecture? WAS: RE: [Sipping] Service =
Identificatio n
in SIP


Inorder to identify a endpoint that has the capabilities to serve a =
very
specialized functionality, I guess it is simpler to name the =
functionality.
Naming the functionality not only helps in decision making while =
routing the
request to that specific endpoint but also helps the endpoint identify =
which
specialized application on the endpoint can take up the responsibility.
Instead of all the routing entities looking into the SDP/message bodies =
it
should be easier/faster if they can look for a specific header.=20
I thought feature tag mechanisms indicated in the RFC3841/3840 =
combination
should have sufficed but implementing those RFCs seems to be a overkill =
for
describing/selecting/routing requests to the specific endpoint and also
choosing a service/feature/application on the endpoint.=20
=20
-Mallesh
=20
=20
=20
On 2/15/07, Elwell, John <john.elwell@siemens.com
<mailto:john.elwell@siemens.com> > wrote:=20

G=F6ran,

Your interpretation is different from what is stated in the Rosenberg =
draft.
Anyway, let's not dwell on this particular use case, which may or may =
not be

realistic. The important thing is whether the media being sent is =
suitable
for general purpose devices or requires or could be used by some =
special
class of application.

If a medium is somehow specialised ( e.g., audio modulated for
interpretation
by a modem or fax machine), then the necessary information needs to be
conveyed along with the m-line in the SDP. Perhaps even if it is
human-intelligible voice but with a restricted repertoire of words that =
make

it suitable for a particular class of application, maybe that =
information
too should be conveyed with the m-line. Perhaps the case described in
Jonathan's draft fits this latter category: I believe the medium is =
just=20
regular voice that can be interpreted by a human, but is also =
constrained in
its repertoire of words such that it can be used by a chess =
application.

The important thing is that there be sufficient information in the SDP =
to=20
allow appropriate application selection at the end device. I am not
convinced by a need for information at the SIP level during call
establishment, because of course SDP can be re-negotiated during the =
call
( e.g., when the users decide to use a particular application) and what =
is
indicated in SIP at call establishment time might no longer be valid =
later
in the call.

John


> -----Original Message-----=20
> From: G=F6ran Eriksson AP (KI/EAB)
> [mailto:goran.ap.eriksson@ericsson.com
<mailto:goran.ap.eriksson@ericsson.com> ]
> Sent: 14 February 2007 23:34=20
> To: Elwell, John
> Cc: sipping@ietf.org <mailto:sipping@ietf.org>=20
> Subject: Device architecture? WAS: RE: [Sipping] Service=20
> Identification in SIP
>
> My humble, perhaps faulty, understanding of the "chess game"
> was not that voice commands were used to control the chess=20
> itself, rather
> that the chess game, while having some IP-based proprietary=20
> gaming interaction protocol, also had a feature letting
> the users talk to each others while playing. The scenario are=20
> two, or more, users play a chess game of the (Inter)net, talkin
> to eachother from time to time during the game (yeah, don't=20
> ask me if I think people will love this!!).
>
> The purpose, as I understand it, was a)to show that the=20
> 'voice component' in the chess game application is similar to that
> in a stand alone POC-Only application when it comes to=20
> exchange and transfer of voice samples and associated floor control=20
> and b) to point to differences, for instance when it comes to
> supplementary services, which may not apply in the chess game case.=20
>
> Perhaps it is important to be clear on the assumed user
> device realisation, as far as I understand it:
>
>  The "chess game + voice" example is, as I understand it,
> discussed under the assumption that voice and chess game is=20
>  managed by custome designed appliction, or client, which is=20
> a necessary/likely approach in a mobile device given the
>  limited MMI (and also other factors in the mobile context).
>
>  In a PC, I could easily handle this as two different=20
> applications/client, like when I use Skype/MSN to talk to my=20
> friends while
>  playing some turn based game, but again, the discussion is
> held under the assumption that usability requirements=20
>  leads to an application client integrating voice and chess
> game events.
>
> The alternative device realisation You describe- different
> clients managing different portions of what is negotiated in=20
> the SDP or MSRP MIME attachment-
> is, I think, a good illustration of the need to be precise=20
> and explicit about the assumed device architecture when
> discussing the "service identification" issue(s).=20
>
> Regards
> G=F6ran
>
>
>
>
> -----Original Message-----=20
> From: Elwell, John [mailto:  <mailto:john.elwell@siemens.com>
john.elwell@siemens.com]
> Sent: den 14 februari 2007 11:57
> To: Henry Sinnreich; sipping@ietf.org  <mailto:sipping@ietf.org>=20
> Subject: RE: [Sipping] Service Identification in SIP
>
> Having followed this thread, I too find the chess example
> very contrived. If chess moves are sent as voice commands,
> then surely it would be acceptable for the normal telephone=20
> application to answer the call, i.e., the remote player
> listens to the move, performs this move on a conventional or
> electronic chess board and speaks his next move. The calling
> UA should not care whether the call is answered by a computer=20
> application or a person.
> Furthermore, even if the audio needs to consumed by a
> particular chess application at the called UA, then other
> media in the same session such as video or messaging might=20
> not be. So any qualification of the nature of media to be
> exchanged probably should be at the media level, not at the
> session level or SIP level. As yet another variation, what if
> the players set up a conventional telephone call and then=20
> mutually agree to switch in their chess applications?
>
> Now, let's consider another case. I establish an MSRP
> session. Within this MSRP session I can send messages with
> various MIME attachments that are intended for consumption by=20
> various applications. I should not need to know in advance
> what applications I will use during the course of the MSRP
> session. All I should put in the INVITE request is the fact
> that I want to establish an MSRP session.=20
>
> John
>
> > -----Original Message-----
> > From: Henry Sinnreich [mailto:hsinnrei@adobe.com
<mailto:hsinnrei@adobe.com> ]
> > Sent: 13 February 2007 18:19
> > To: sipping@ietf.org <mailto:sipping@ietf.org>=20
> > Subject: RE: [Sipping] Service Identification in SIP=20
> >
> > It must be my lack of imagination for not being able to
> come up with a=20
> > credible usage scenario, given the following:
> >
> > - Is playing chess by voice a credible usage scenario? :-)=20
> Especially
> > since voice recognition is now so often found in PCs and in mobile=20
> > phones as well. In other words, chess playing by voice may
> also be an
> > endpoint application, without any network involvement. Actually, if =

> > placed in the endpoint it works just as well on poor=20
> bandwidth links
> > by contrast to chess in the network that may require adequate
> > broadband.
> >
> > - Is anyone's boss or spouse going to approve extra phone charges=20
> > above the flat rate for playing chess at a time when the gaming
> > industry has pioneered free VoIP?
> > Will such a service be taxed extra and if yes using by what
> criteria?
> > Gambling comes to my mind :-)=20
> >
> > - VoIP is reported in the press as a give-away and I have seen no
> > reports of any VoIP provider recording a profit on VoIP. I
> assume the
> > losses are often hidden by bundling in triple play, etc.=20
> >
> > Given my lack of imagination, can anyone share a more
> credible usage
> > scenario for Service Identification in SIP?
> >
> > What's the use of a standard without a credible usage scenario?=20
> >
> > Thanks, Henry
> >
> >
> > _______________________________________________
> > Sipping mailing list    =
<https://www1.ietf.org/mailman/listinfo/sipping>
https://www1.ietf.org/mailman/listinfo/sipping=20
> > This list is for NEW development of the application of SIP Use
> > sip-implementors@cs.columbia.edu
<mailto:sip-implementors@cs.columbia.edu>  for questions on current sip =
Use
> > sip@ietf.org <mailto:sip@ietf.org>  for new developments of core =
SIP=20
> >
>
> _______________________________________________
> Sipping mailing list    =
<https://www1.ietf.org/mailman/listinfo/sipping>
https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu
<mailto:sip-implementors@cs.columbia.edu>  for questions on current=20
> sip Use sip@ietf.org <mailto:sip@ietf.org>  for new developments of =
core
SIP=20
>

_______________________________________________
Sipping mailing list    =
<https://www1.ietf.org/mailman/listinfo/sipping>
https://www1.ietf.org/mailman/listinfo/sipping=20
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu
<mailto:sip-implementors@cs.columbia.edu> for questions on current sip=20
Use sip@ietf.org <mailto:sip@ietf.org>  for new developments of core =
SIP





  _____ =20

This transmission (including any attachments) may contain confidential
information, privileged material (including material protected by the
solicitor-client or other applicable privileges), or constitute =
non-public
information. Any use of this information by anyone other than the =
intended
recipient is prohibited. If you have received this transmission in =
error,
please immediately reply to the sender and delete this information from =
your
system. Use, dissemination, distribution, or reproduction of this
transmission by unintended recipients is not authorized and may be =
unlawful.



--Boundary_(ID_6cW26/U9o/yFvNnJQnggRQ)
Content-type: text/html; charset=iso-8859-1
Content-transfer-encoding: QUOTED-PRINTABLE

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; charset=3Diso=
-8859-1">


<META content=3D"MSHTML 6.00.2900.3020" name=3DGENERATOR></HEAD>
<BODY>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT fa=
ce=3DArial=20
color=3D#0000ff size=3D2>Andrew,</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT fa=
ce=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT fa=
ce=3DArial=20
color=3D#0000ff size=3D2>Well, an AS is just another contact (perhaps=
 a permanently=20
registered one, rather than one that uses SIP REGISTER), isn't it? So=
 caller=20
preferences can be used to choose an AS or any other contact that sup=
ports the=20
desired capability.</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT fa=
ce=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT fa=
ce=3DArial=20
color=3D#0000ff size=3D2>I agree caller preferences&nbsp;is not a sui=
table basis for=20
charging. Just because I say I have a preference for a contact that s=
upports=20
video, doesn't mean that I will necessarily use video during that cal=
l. That=20
might be determined during some mid-call session-re-negotiation. Sure=
ly charging=20
has to be based on what is actually used, not what might be used, and=
 you can't=20
always determine this at INVITE time.</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT fa=
ce=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT fa=
ce=3DArial=20
color=3D#0000ff size=3D2>So I am saying use caller preferences to sel=
ect a preferred=20
contact, use SDP to negotiate and re-negotiate the session with the c=
ontact that=20
you end up with. I haven't heard any compelling arguments for anythin=
g else so=20
far.</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT fa=
ce=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT fa=
ce=3DArial=20
color=3D#0000ff size=3D2>John</FONT></SPAN></DIV><BR>
<BLOCKQUOTE=20
style=3D"PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #0000ff 2p=
x solid; MARGIN-RIGHT: 0px">
  <DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr align=3Dle=
ft>
  <HR tabIndex=3D-1>
  <FONT face=3DTahoma size=3D2><B>From:</B> Andrew Allen [mailto:aall=
en@rim.com]=20
  <BR><B>Sent:</B> 16 February 2007 03:51<BR><B>To:</B> Elwell, John;=
 Malleswara=20
  Rao Ankem<BR><B>Cc:</B> sipping@ietf.org; Andrew Allen<BR><B>Subjec=
t:</B> RE:=20
  Device architecture? WAS: RE: [Sipping] Service Identificatio n=
=20
  inSIP<BR></FONT><BR></DIV>
  <DIV></DIV>
  <DIV dir=3Dltr align=3Dleft><FONT face=3DArial color=3D#0000ff=20
  size=3D2></FONT>&nbsp;</DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2>John,</FONT></SPAN></DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2>If Service Identification was only about s=
electing the=20
  appropriate contacts from those registered agaisnt the AOR then RFC=
 3840/3841=20
  would be fine however Service Identification is much more than that=
 and may be=20
  used to route requests to application servers for service logic inv=
ocation,=20
  and used as the basis for charging. Accept-Contact is basically jus=
t a filter=20
  expression to filter a set of registered contacts. Including a para=
meter in a=20
  filter expression and using that as the basis for charging,=20
  authorisation&nbsp;and service invocation is going to cause=20
  problems.</FONT></SPAN></DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2>Andrew</FONT></SPAN></DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007></SPAN=
>&nbsp;</DIV>
  <DIV dir=3Dltr align=3Dleft>
  <HR tabIndex=3D-1>
  </DIV>
  <DIV dir=3Dltr align=3Dleft><FONT face=3DTahoma size=3D2><B>From:</=
B> Elwell, John=20
  [mailto:john.elwell@siemens.com] <BR><B>Sent:</B> Thursday, Februar=
y 15, 2007=20
  6:16 AM<BR><B>To:</B> Malleswara Rao Ankem<BR><B>Cc:</B>=20
  sipping@ietf.org<BR><B>Subject:</B> RE: Device architecture? WAS: R=
E:=20
  [Sipping] Service Identificatio n inSIP<BR></FONT><BR></DIV>
  <DIV></DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2>Mallesh,</FONT></SPAN></DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2>Yes, I agree that some means of selecting =
a preferred=20
  device might be needed, but that is indeed what RFC3841/3840 does. =
Why do you=20
  consider this overkill?</FONT></SPAN></DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2>John</FONT></SPAN></DIV><BR>
  <BLOCKQUOTE=20
  style=3D"PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #0000ff =
2px solid; MARGIN-RIGHT: 0px">
    <DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr align=
=3Dleft>
    <HR tabIndex=3D-1>
    <FONT face=3DTahoma size=3D2><B>From:</B> Malleswara Rao Ankem=
=20
    [mailto:malleshavn@gmail.com] <BR><B>Sent:</B> 15 February 2007=
=20
    10:54<BR><B>To:</B> Elwell, John<BR><B>Cc:</B> G=F6ran Eriksson A=
P (KI/EAB);=20
    sipping@ietf.org<BR><B>Subject:</B> Re: Device architecture? WAS:=
 RE:=20
    [Sipping] Service Identificatio n in SIP<BR></FONT><BR></DIV>
    <DIV></DIV>
    <DIV>Inorder to identify a endpoint that has the capabilities to =
serve a=20
    very specialized functionality, I guess it is simpler to name the=
=20
    functionality. Naming the functionality not only helps in decisio=
n making=20
    while routing the request to that specific endpoint but also help=
s the=20
    endpoint identify which specialized application on the endpoint c=
an take up=20
    the responsibility. Instead of all the routing entities looking i=
nto the=20
    SDP/message bodies it should be easier/faster if they can look fo=
r a=20
    specific header. </DIV>
    <DIV>I thought feature tag mechanisms indicated in the RFC3841/38=
40=20
    combination should have sufficed but implementing those RFCs seem=
s to be a=20
    overkill for describing/selecting/routing requests&nbsp;to the sp=
ecific=20
    endpoint and also choosing a service/feature/application on the e=
ndpoint.=20
    </DIV>
    <DIV>&nbsp;</DIV>
    <DIV>-Mallesh</DIV>
    <DIV>&nbsp;</DIV>
    <DIV>&nbsp;</DIV>
    <DIV>&nbsp;</DIV>
    <DIV><SPAN class=3Dgmail_quote>On 2/15/07, <B class=3Dgmail_sende=
rname>Elwell,=20
    John</B> &lt;<A onclick=3D"return top.js.OpenExtLink(window,event=
,this)"=20
    href=3D"mailto:john.elwell@siemens.com" target=3D_blank>john.elwe=
ll@siemens.com=20
    </A>&gt; wrote:</SPAN>=20
    <BLOCKQUOTE class=3Dgmail_quote=20
    style=3D"PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEF=
T: #ccc 1px solid">G=F6ran,<BR><BR>Your=20
      interpretation is different from what is stated in the Rosenber=
g=20
      draft.<BR>Anyway, let's not dwell on this particular use case, =
which may=20
      or may not be <BR>realistic. The important thing is whether the=
 media=20
      being sent is suitable<BR>for general purpose devices or requir=
es or could=20
      be used by some special<BR>class of application.<BR><BR>If a me=
dium is=20
      somehow specialised ( e.g., audio modulated for interpretation<=
BR>by a=20
      modem or fax machine), then the necessary information needs to=
=20
      be<BR>conveyed along with the m-line in the SDP. Perhaps even i=
f it=20
      is<BR>human-intelligible voice but with a restricted repertoire=
 of words=20
      that make <BR>it suitable for a particular class of application=
, maybe=20
      that information<BR>too should be conveyed with the m-line. Per=
haps the=20
      case described in<BR>Jonathan's draft fits this latter category=
: I believe=20
      the medium is just <BR>regular voice that can be interpreted by=
 a human,=20
      but is also constrained in<BR>its repertoire of words such that=
 it can be=20
      used by a chess application.<BR><BR>The important thing is that=
 there be=20
      sufficient information in the SDP to <BR>allow appropriate appl=
ication=20
      selection at the end device. I am not<BR>convinced by a need fo=
r=20
      information at the SIP level during call<BR>establishment, beca=
use of=20
      course SDP can be re-negotiated during the call<BR>( e.g., when=
 the users=20
      decide to use a particular application) and what is<BR>indicate=
d in SIP at=20
      call establishment time might no longer be valid later<BR>in th=
e=20
      call.<BR><BR>John<BR><BR><BR>&gt; -----Original Message----- <B=
R>&gt;=20
      From: G=F6ran Eriksson AP (KI/EAB)<BR>&gt; [mailto:<A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:goran.ap.eriksson@ericsson.com"=20
      target=3D_blank>goran.ap.eriksson@ericsson.com</A>]<BR>&gt; Sen=
t: 14=20
      February 2007 23:34 <BR>&gt; To: Elwell, John<BR>&gt; Cc: <A=
=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sipping@ietf.org" target=3D_blank>sipping@ietf.o=
rg</A><BR>&gt;=20
      Subject: Device architecture? WAS: RE: [Sipping] Service <BR>&g=
t;=20
      Identification in SIP<BR>&gt;<BR>&gt; My humble, perhaps faulty=
,=20
      understanding of the "chess game"<BR>&gt; was not that voice co=
mmands were=20
      used to control the chess <BR>&gt; itself, rather<BR>&gt; that =
the chess=20
      game, while having some IP-based proprietary <BR>&gt; gaming in=
teraction=20
      protocol, also had a feature letting<BR>&gt; the users talk to =
each others=20
      while playing. The scenario are <BR>&gt; two, or more, users pl=
ay a chess=20
      game of the (Inter)net, talkin<BR>&gt; to eachother from time t=
o time=20
      during the game (yeah, don't <BR>&gt; ask me if I think people =
will love=20
      this!!).<BR>&gt;<BR>&gt; The purpose, as I understand it, was a=
)to show=20
      that the <BR>&gt; 'voice component' in the chess game applicati=
on is=20
      similar to that<BR>&gt; in a stand alone POC-Only application w=
hen it=20
      comes to <BR>&gt; exchange and transfer of voice samples and as=
sociated=20
      floor control <BR>&gt; and b) to point to differences, for inst=
ance when=20
      it comes to<BR>&gt; supplementary services, which may not apply=
 in the=20
      chess game case. <BR>&gt;<BR>&gt; Perhaps it is important to be=
 clear on=20
      the assumed user<BR>&gt; device realisation, as far as I unders=
tand=20
      it:<BR>&gt;<BR>&gt;&nbsp;&nbsp;The "chess game + voice" example=
 is, as I=20
      understand it,<BR>&gt; discussed under the assumption that voic=
e and chess=20
      game is <BR>&gt;&nbsp;&nbsp;managed by custome designed applict=
ion, or=20
      client, which is <BR>&gt; a necessary/likely approach in a mobi=
le device=20
      given the<BR>&gt;&nbsp;&nbsp;limited MMI (and also other factor=
s in the=20
      mobile context).<BR>&gt;<BR>&gt;&nbsp;&nbsp;In a PC, I could ea=
sily handle=20
      this as two different <BR>&gt; applications/client, like when I=
 use=20
      Skype/MSN to talk to my <BR>&gt; friends while<BR>&gt;&nbsp;&nb=
sp;playing=20
      some turn based game, but again, the discussion is<BR>&gt; held=
 under the=20
      assumption that usability requirements <BR>&gt;&nbsp;&nbsp;lead=
s to an=20
      application client integrating voice and chess<BR>&gt; game=
=20
      events.<BR>&gt;<BR>&gt; The alternative device realisation You =
describe-=20
      different<BR>&gt; clients managing different portions of what i=
s=20
      negotiated in <BR>&gt; the SDP or MSRP MIME attachment-<BR>&gt;=
 is, I=20
      think, a good illustration of the need to be precise <BR>&gt; a=
nd explicit=20
      about the assumed device architecture when<BR>&gt; discussing t=
he "service=20
      identification" issue(s). <BR>&gt;<BR>&gt; Regards<BR>&gt;=20
      G=F6ran<BR>&gt;<BR>&gt;<BR>&gt;<BR>&gt;<BR>&gt; -----Original M=
essage-----=20
      <BR>&gt; From: Elwell, John [mailto:<A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:john.elwell@siemens.com" target=3D_blank>=20
      john.elwell@siemens.com</A>]<BR>&gt; Sent: den 14 februari 2007=
=20
      11:57<BR>&gt; To: Henry Sinnreich; <A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sipping@ietf.org" target=3D_blank>sipping@ietf.o=
rg </A><BR>&gt;=20
      Subject: RE: [Sipping] Service Identification in SIP<BR>&gt;<BR=
>&gt;=20
      Having followed this thread, I too find the chess example<BR>&g=
t; very=20
      contrived. If chess moves are sent as voice commands,<BR>&gt; t=
hen surely=20
      it would be acceptable for the normal telephone <BR>&gt; applic=
ation to=20
      answer the call, i.e., the remote player<BR>&gt; listens to the=
 move,=20
      performs this move on a conventional or<BR>&gt; electronic ches=
s board and=20
      speaks his next move. The calling<BR>&gt; UA should not care wh=
ether the=20
      call is answered by a computer <BR>&gt; application or a person=
.<BR>&gt;=20
      Furthermore, even if the audio needs to consumed by a<BR>&gt; p=
articular=20
      chess application at the called UA, then other<BR>&gt; media in=
 the same=20
      session such as video or messaging might <BR>&gt; not be. So an=
y=20
      qualification of the nature of media to be<BR>&gt; exchanged pr=
obably=20
      should be at the media level, not at the<BR>&gt; session level =
or SIP=20
      level. As yet another variation, what if<BR>&gt; the players se=
t up a=20
      conventional telephone call and then <BR>&gt; mutually agree to=
 switch in=20
      their chess applications?<BR>&gt;<BR>&gt; Now, let's consider a=
nother=20
      case. I establish an MSRP<BR>&gt; session. Within this MSRP ses=
sion I can=20
      send messages with<BR>&gt; various MIME attachments that are in=
tended for=20
      consumption by <BR>&gt; various applications. I should not need=
 to know in=20
      advance<BR>&gt; what applications I will use during the course =
of the=20
      MSRP<BR>&gt; session. All I should put in the INVITE request is=
 the=20
      fact<BR>&gt; that I want to establish an MSRP session. <BR>&gt;=
<BR>&gt;=20
      John<BR>&gt;<BR>&gt; &gt; -----Original Message-----<BR>&gt; &g=
t; From:=20
      Henry Sinnreich [mailto:<A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:hsinnrei@adobe.com" target=3D_blank>hsinnrei@ado=
be.com=20
      </A>]<BR>&gt; &gt; Sent: 13 February 2007 18:19<BR>&gt; &gt; To=
: <A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sipping@ietf.org" target=3D_blank>sipping@ietf.o=
rg</A><BR>&gt;=20
      &gt; Subject: RE: [Sipping] Service Identification in SIP <BR>&=
gt;=20
      &gt;<BR>&gt; &gt; It must be my lack of imagination for not bei=
ng able=20
      to<BR>&gt; come up with a <BR>&gt; &gt; credible usage scenario=
, given the=20
      following:<BR>&gt; &gt;<BR>&gt; &gt; - Is playing chess by voic=
e a=20
      credible usage scenario? :-) <BR>&gt; Especially<BR>&gt; &gt; s=
ince voice=20
      recognition is now so often found in PCs and in mobile <BR>&gt;=
 &gt;=20
      phones as well. In other words, chess playing by voice may<BR>&=
gt; also be=20
      an<BR>&gt; &gt; endpoint application, without any network invol=
vement.=20
      Actually, if <BR>&gt; &gt; placed in the endpoint it works just=
 as well on=20
      poor <BR>&gt; bandwidth links<BR>&gt; &gt; by contrast to chess=
 in the=20
      network that may require adequate<BR>&gt; &gt; broadband.<BR>&g=
t;=20
      &gt;<BR>&gt; &gt; - Is anyone's boss or spouse going to approve=
 extra=20
      phone charges <BR>&gt; &gt; above the flat rate for playing che=
ss at a=20
      time when the gaming<BR>&gt; &gt; industry has pioneered free=
=20
      VoIP?<BR>&gt; &gt; Will such a service be taxed extra and if ye=
s using by=20
      what<BR>&gt; criteria?<BR>&gt; &gt; Gambling comes to my mind :=
-) <BR>&gt;=20
      &gt;<BR>&gt; &gt; - VoIP is reported in the press as a give-awa=
y and I=20
      have seen no<BR>&gt; &gt; reports of any VoIP provider recordin=
g a profit=20
      on VoIP. I<BR>&gt; assume the<BR>&gt; &gt; losses are often hid=
den by=20
      bundling in triple play, etc. <BR>&gt; &gt;<BR>&gt; &gt; Given =
my lack of=20
      imagination, can anyone share a more<BR>&gt; credible usage<BR>=
&gt; &gt;=20
      scenario for Service Identification in SIP?<BR>&gt; &gt;<BR>&gt=
; &gt;=20
      What's the use of a standard without a credible usage scenario?=
 <BR>&gt;=20
      &gt;<BR>&gt; &gt; Thanks, Henry<BR>&gt; &gt;<BR>&gt; &gt;<BR>&g=
t; &gt;=20
      _______________________________________________<BR>&gt; &gt; Si=
pping=20
      mailing list&nbsp;&nbsp;<A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"https://www1.ietf.org/mailman/listinfo/sipping" target=
=3D_blank>=20
      https://www1.ietf.org/mailman/listinfo/sipping </A><BR>&gt; &gt=
; This list=20
      is for NEW development of the application of SIP Use<BR>&gt; &g=
t; <A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sip-implementors@cs.columbia.edu"=20
      target=3D_blank>sip-implementors@cs.columbia.edu</A> for questi=
ons on=20
      current sip Use<BR>&gt; &gt; <A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sip@ietf.org" target=3D_blank>sip@ietf.org</A> f=
or new=20
      developments of core SIP <BR>&gt; &gt;<BR>&gt;<BR>&gt;=20
      _______________________________________________<BR>&gt; Sipping=
 mailing=20
      list&nbsp;&nbsp;<A onclick=3D"return top.js.OpenExtLink(window,=
event,this)"=20
      href=3D"https://www1.ietf.org/mailman/listinfo/sipping" target=
=3D_blank>=20
      https://www1.ietf.org/mailman/listinfo/sipping</A><BR>&gt; This=
 list is=20
      for NEW development of the application of SIP<BR>&gt; Use <A=
=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sip-implementors@cs.columbia.edu"=20
      target=3D_blank>sip-implementors@cs.columbia.edu</A> for questi=
ons on=20
      current <BR>&gt; sip Use <A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sip@ietf.org" target=3D_blank>sip@ietf.org</A> f=
or new=20
      developments of core SIP=20
      <BR>&gt;<BR><BR>_______________________________________________=
<BR>Sipping=20
      mailing list&nbsp;&nbsp;<A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"https://www1.ietf.org/mailman/listinfo/sipping" target=
=3D_blank>=20
      https://www1.ietf.org/mailman/listinfo/sipping </A><BR>This lis=
t is for=20
      NEW development of the application of SIP<BR>Use <A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sip-implementors@cs.columbia.edu"=20
      target=3D_blank>sip-implementors@cs.columbia.edu </A>for questi=
ons on=20
      current sip <BR>Use <A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sip@ietf.org" target=3D_blank>sip@ietf.org</A> f=
or new=20
      developments of core SIP<BR></BLOCKQUOTE></DIV><BR></BLOCKQUOTE=
>
  <P>
  <HR>
  This transmission (including any attachments) may contain confident=
ial=20
  information, privileged material (including material protected by t=
he=20
  solicitor-client or other applicable privileges), or constitute non=
-public=20
  information. Any use of this information by anyone other than the i=
ntended=20
  recipient is prohibited. If you have received this transmission in =
error,=20
  please immediately reply to the sender and delete this information =
=66rom your=20
  system. Use, dissemination, distribution, or reproduction of this t=
ransmission=20
  by unintended recipients is not authorized and may be unlawful.=
=20
</BLOCKQUOTE></BODY></HTML>


--Boundary_(ID_6cW26/U9o/yFvNnJQnggRQ)--


--===============0400969377==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============0400969377==--




From sipping-bounces@ietf.org Fri Feb 16 03:52:42 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HHyoK-0002eZ-4E; Fri, 16 Feb 2007 03:51:20 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HHyoI-0002Zn-Ld
	for sipping@ietf.org; Fri, 16 Feb 2007 03:51:18 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HHyoC-0007jP-IG
	for sipping@ietf.org; Fri, 16 Feb 2007 03:51:18 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	B5354204AC
	for <sipping@ietf.org>; Fri, 16 Feb 2007 09:51:01 +0100 (CET)
X-AuditID: c1b4fb3c-b0fcabb0000007de-ba-45d5707535e1 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	9AB6C203D5
	for <sipping@ietf.org>; Fri, 16 Feb 2007 09:51:01 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 16 Feb 2007 09:51:01 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Fri, 16 Feb 2007 09:51:00 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD02886444@esealmw111>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C00915146D@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdNmH89O4T46iGoRjicZ3hYknVpawB8E6oQAH++leAABzLk8A==
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 16 Feb 2007 08:51:01.0304 (UTC)
	FILETIME=[9581AB80:01C751A7]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 731ea0e9f5725b67e634db1918f3b951
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hello Andrew,

Thanks for your comments.  I have a couple of brief responses below.

Best Regards,

..//Steve=20

-----Original Message-----
From: Andrew Allen [mailto:aallen@rim.com]=20
Sent: Friday, February 16, 2007 7:15 AM
To: Stephen Terrill (MI/EEM); sipping@ietf.org
Cc: Andrew Allen
Subject: RE: [Sipping] Service Identification in SIP


Some comments on Stephen's comments on the draft:
=20

-Moving onto the proposed solution draft itself, in addition for
providing for the identification of the "application" and the
"communication service", I have a few additional thoughts.

The draft seems to have the implication that all services (communication
services in other terminology) that are to be identified have to be
standardised within the IETF.  This is clearly an restriction that will
not be acceptable to the industry as a whole.  My guess is that if such
a statement is mandated by the IETF, it will just be words on paper
which would likely be ignored in practice by other standardisation
organisations.

[AA] This I think is an incorrect interpretation of the draft. My
understanding is that the only services according to the draft that
require some standardisation in IETF are those that by their very nature
cannot interoperate unless there is some common additional understanding
between the endpoints outside of what is currently specified in SIP. The
use of SIP option tags with Require header is proposed by the draft for
this purpose as such services create interoperability issues with SIP.

<Steve>
My comment applies even for these services
</Steve>
----------------------------------------------------

The draft also only recognises the case where all explicitly knowledge
of the service is required.  I think it would also be good to allow the
case where knowledge of the service would be helpful, but not explicitly
required.  This could be a case where the service to the end-user would
not be degraded if this was not supported.  This functionality seems to
be missing.

[AA] Service URIs are defined in the draft for these services where the
support by the other endpoint is not required for the session to be
sucessfully established.

The draft seems to indicate that that application servers in the network
are involved only at the request of the originating terminal.  This
seems a strange restriction.  For some network support that supports the
originating user, such as the one provided in the example "I would like
this call recorded for later retrieval", such mechanisms may make sense,
however as a general approach it seems a bit limited.  For example, in
3GPP networks there is the need to invoke application logic in the
originating side and terminating side, not only at the request of the
originating terminal, but as a consequence of the requested service.  If
there is an SIP message contains indication of the requested service, a
proxy can do this (or loose route to an AS that contains the necessary
service logic).=20

[AA] The draft supports both possibilities using the Route header for
originating network servers that need to be involved in the route and
based on the terminating networks routing policy for the Service URI in
the Request URI.

<Steve>
I don't see the advantage of having different solutions for the
originating network and the terminating network.  I don't think that the
originating logic is invoked only at the request of the originating
terminal.
</Steve>
--------------------------------------

 I note here that the proposed service URN does not contain the address
of a specific server containing the application logic.  An AS can look
at that, and loose route to an application server that does contain such
logic (If it is separated from the proxy), however I fail to see the
difference for that proxy if it looks at a feature tag, or it looks at
the service URN for this.  In both cases, it has to parse the message
for the data, and modify the routing information. =20

[AA]Route headers are the SIP mechanism for requesting intermediaries to
be includede on the route of the request. There is nothing in RFC
3840/RFC 3841 that addresses using feature tags in Accept-Contact header
for any routing decisions prior to determining which of the registered
contacts the request should be forked to. Accept-Contact is not a Route
header it is merely a filter that is applied to the set of valid
contacts. Using a parameter in the filter as a service identifier has a
potential significant impact on the use of Accept-Contact for its
original purpose.=20

<Steve>
I don't believe there is anthing that states a Proxy CAN NOT route to
applications or application logic based upon elements in the SIP
message, including what is in the Accept-Contact header (some people
have even proposed SDP ...).
</Steve>

Further, the route header could be removed in the originating network;
and it is still required to identify the service in the terminating
network.  As proxies may remove the route, it can't be used to transport
the requested service. =20

[AA] Proxies arbitarily removing Route headers is a BAD thing akin to a
bull in a china shop - almost certainly going to break something. RFC
3261 summarises what proxies should be doing with Route headers:
16.12 Summary of Proxy Route Processing

   In the absence of local policy to the contrary, the processing a
   proxy performs on a request containing a Route header field can be
   summarized in the following steps.

      1.  The proxy will inspect the Request-URI.  If it indicates a
          resource owned by this proxy, the proxy will replace it with
          the results of running a location service.  Otherwise, the
          proxy will not change the Request-URI.

      2.  The proxy will inspect the URI in the topmost Route header
          field value.  If it indicates this proxy, the proxy removes it
          from the Route header field (this route node has been
          reached).

      3.  The proxy will forward the request to the resource indicated
          by the URI in the topmost Route header field value or in the
          Request-URI if no Route header field is present.  The proxy
          determines the address, port and transport to use when
          forwarding the request by applying the procedures in [4] to
          that URI.

   If no strict-routing elements are encountered on the path of the
   request, the Request-URI will always indicate the target of the
   request.

<Steve>
The first sentance is interesting "In absence of local policy to the
contrary", indicating that there can be different policies.  I
understand that some proxies do remove this.  While I am on this - I
have lead to believe that R-URI parameters may also be removed by
intermediate proxies - is this correct?
</Steve>

[AA] Generally a UA wouldn't know the appropriate service identifiers
used in the terminating network or whether a Route header is appropriate
to be used in the terminating network. The draft defines the Service URI
for the purpose of identifying services in the terminating network and
service URIs can be obtained using Presence.

<Steve>
I don't get the above paragraph.  It suggest that the "appropriate
service identifier" used in the terminating network is not known by the
originating UE, but the Service URI is.  I thought that the service URI
is an encoding of the "appropriate service identifier".
</Steve>
-----------------------------------

>From this perspective it is not clear to me the advantage of using a
service URN in a route header compared to the use of a feature tag.

[AA] I hope the above responses clarify that somewhat. Also the Service
URN and Service URIs allow different service identifiers to be used in
the originating and terminating network for session compatiable services
instead of forcing everyone to deploy and understand the same service
and service identifier.

<Steve>
Nope, don't get it, but are information that is carried end-to-end ....
And as far as I see, could be used by intermediate proxies to impact
local routing policy ... Or ?
</Steve>

The draft seems to mandate GRUU - is this really so, is that really
necessary.  It seems strange to place such a dependency, it also starts
to move the GRUU as being the primary means of which to contact a user.

[AA] My understanding is that GRUU in the draft is used to address how a
request addressed to Service URI is forked only to those contacts that
support the service identified by the that Service URI. A contact would
seem to be fine so long as it had GRUU properties - no need to obtain a
GRUU if you already have one.

This implies that the originating user decides which terminal I should
receive my requested service on, even if I have two devices for that.
As such, I don't believe that there should be a dependency on GRUU.

[AA] If you have obtained a Service URI (probably through presence) for
that user then you can address the request to the Service URI which
represents all the terminals that support that service. However GRUU and
the Service Instance URIs created from GRUUs provide another mechanism
through which a user can obtain a URI to reach that user for the
required service.

Parts of the draft seems to rely on the use of presence.  While presence
could be a supportive mechanism, I doubt that we can mandate that the
number you want to dial has be available in presence as I assume not all
entities are available in presence (diallable network services, users in
the PSTN).  This would be strange to mandated.

[AA] PSTN gateways are presumably in the same domain as the terminating
proxy and therefore these PSTN gateways can be configured with the
appropriate service identifiers for telephony services and can include
this in either a Route header to route to the right application server
or can construct a valid Service URI. The real problem is how do users
in other networks know what the service identifier that's is used in the
terminating network is. The use of Service URIs and Presence provides a
solution to that problem.

<Steve>
Again, presence could be a useful mechanism, but I think it would be a
problem for the draft to rely it.  A user making a call to the PSTN may
not find that user in the presence archicture (and similar to services
like the 800 or 900 services).  (I should note, that you shouldn't make
any assumptions about the location of the PSTN gateway ... But I don't
think that changes anything). =20
</Steve>

Moving on from the draft I have seen the questioned raised about "why
can't we expand the power of SDP".  Such an approach would mandate that
SIP methods use SDP, I don't think that is appropriate.=20

[AA] I agree with you here

I hear the appeals that the mechanisms should not be abused, and that we
should not have 1000s of separate "communication services", and I fully
agree with this. I see a smaller number of communication services, not a
larger number; where those communication services are used to access
other end-user-services.

*flame away*

Regards,

Steve

---


-----Original Message-----
From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net]
Sent: Sunday, February 11, 2007 5:50 AM
To: sipping@ietf.org
Subject: Re: [Sipping] Service Identification in SIP

I've attempted to re-read the threads that Gonzalo referred to, and the
major limitation that I see is the lack of any concrete example of why
these identifiers are needed.  Probably the IMS people have discussed
this so many times that they've forgotten the detailed rationale behind
the requirement.  But I'm a net-head, (and most of the people here are
net-heads) and the SIP systems that I've dealt with show no evidence of
needing "service identification", they're pure proxies, and requests are
passed transparently end-to-end.

Could somebody please lay out an example showing the utility of service
identification?

   From: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>

   I think that everyone here agree that there is the possibility to use
a
   media in different ways, and that we need a distinction about what
the
   capabilities are and how they are used, whatever will be the
mechanism
   that at the end we will identify to do that.

Respectfully, I disagree.  We do need ways of identifying the
capabilities of SIP agents and transports, but I see no necessity of
creating specified bundles of capabailities, and then constraining
dialogs to require an initial agreement as to which bundle of
capabilities will be used.

One of the central features of Internet architecture is that agents
negotiate compatibility on a feature-by-feature basis; there are no
"profiles" which must be known on a network-wide basis.  (And I have
observed that this feature is of significant commercial value.)

In regard to identifying an "application" to process the dialog (and I
use "application" in a standard computing way), I see no reason that a
user cannot select URI-parameters to append to his AOR, and arrange for
his UA to route invites with those URI-parameters to the appropriate
application.  (And given the rich mechanisms for redirection in SIP, the
application need not even reside at a UA which is registered for the
AOR.)  These URI-parameters can be formally or informally standardized,
or established ad-hoc by each user.

In regard to PoC (as someone said, "a specialized floor-control
protocol") and similar features, SIP currently has mechanisms for
indicating support of features and requiring that a recipient support a
feature before accepting a call (the Supported and Requires headers, and
simply using the feature).  The only limitation I can see that might
need to be addressed is that there is no way to create an option tag for
a new feature that one has invented that is not yet thoroughly
standardized.

But I now discover RFC 3261 section 8.1.1.9:

   The option tags listed MUST only refer to extensions defined in
   standards-track RFCs.  This is to prevent servers from insisting that
   clients implement non-standard, vendor-defined features in order to
   receive service.  Extensions defined by experimental and
   informational RFCs are explicitly excluded from usage with the
   Supported header field in a request, since they too are often used to
   document vendor-defined extensions.

So I see that as a fundamental architectural decision, SIP does not
permit sending INVITEs that require specialized recipients, excepting if
the specialization is standards-track.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential
information, privileged material (including material protected by the
solicitor-client or other applicable privileges), or constitute
non-public information. Any use of this information by anyone other than
the intended recipient is prohibited. If you have received this
transmission in error, please immediately reply to the sender and delete
this information from your system. Use, dissemination, distribution, or
reproduction of this transmission by unintended recipients is not
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 16 06:22:27 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HI18x-0004R8-88; Fri, 16 Feb 2007 06:20:47 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HI18w-0004Qs-5M
	for sipping@ietf.org; Fri, 16 Feb 2007 06:20:46 -0500
Received: from zrtps0kn.nortel.com ([47.140.192.55])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HI18u-0005XL-TT
	for sipping@ietf.org; Fri, 16 Feb 2007 06:20:46 -0500
Received: from zrc2hxm2.corp.nortel.com (zrc2hxm2.corp.nortel.com
	[47.103.123.73])
	by zrtps0kn.nortel.com (Switch-2.2.6/Switch-2.2.0) with ESMTP id
	l1GBKWv29154; Fri, 16 Feb 2007 06:20:32 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [SIPPING] SDP
	offer/answer	inearly	media:draft-sawada-sipping-sip-offeranswer-01.txt
Date: Fri, 16 Feb 2007 05:20:30 -0600
Message-ID: <6FC4416DDE56C44DA0AEE67BC7CA437112C55A3C@zrc2hxm2.corp.nortel.com>
In-Reply-To: <7374777208BDC7449D5620EF94232567020F1D85@esealmw113.eemea.ericsson.se>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [SIPPING] SDP
	offer/answer	inearly	media:draft-sawada-sipping-sip-offeranswer-01.txt
Thread-index: AcdQiTcxnI4a5jjpS4ygifv8DmpI4wAa2+eAADHwLHA=
From: "Brian Stucker" <bstucker@nortel.com>
To: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>,
	<Dale.Worley@comcast.net>, <sipping@ietf.org>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a7d6aff76b15f3f56fcb94490e1052e4
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Well... Sorta. It just says it's no longer authorized. You have to get
the timing straight to switch between two streams.

Regards,
Brian=20

> -----Original Message-----
> From: Christer Holmberg (JO/LMF)=20
> [mailto:christer.holmberg@ericsson.com]=20
> Sent: Thursday, February 15, 2007 8:31 PM
> To: Dale.Worley@comcast.net; sipping@ietf.org
> Subject: RE: [SIPPING] SDP offer/answer inearly=20
> media:draft-sawada-sipping-sip-offeranswer-01.txt
>=20
>=20
> Hi,=20
>=20
> >[JRE] I don't think 180 means that. It means play ringback=20
> tone until=20
> >media starts to arrive.
> >=20
> >It doesn't mean that, either.  It means that the UAS that=20
> was reached=20
> >is declaring that it is notifying a user.  Whether or how the UAC=20
> >wishes to notify its user of this fact is its business.
> >=20
> >Currently, SIP has no defined means for indicating "Stop playing the=20
> >ringback and start playing this media stream."
>=20
> In TISPAN the usage of P-Early-Media header solves that, at=20
> least to some extent.
>=20
> Regards,
>=20
> Christer
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 16 10:56:48 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HI5QQ-0002lU-Fo; Fri, 16 Feb 2007 10:55:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HI5QP-0002lC-9i
	for sipping@ietf.org; Fri, 16 Feb 2007 10:55:05 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HI5QG-000600-4P
	for sipping@ietf.org; Fri, 16 Feb 2007 10:55:05 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	8665321263; Fri, 16 Feb 2007 16:54:53 +0100 (CET)
X-AuditID: c1b4fb3e-aeed1bb0000007e1-ee-45d5d3cdaeff 
Received: from esealmw129.eemea.ericsson.se (unknown [153.88.254.124])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	59EC720228; Fri, 16 Feb 2007 16:54:53 +0100 (CET)
Received: from esealmw127.eemea.ericsson.se ([153.88.254.171]) by
	esealmw129.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 16 Feb 2007 16:54:53 +0100
Received: from mail.lmf.ericsson.se ([131.160.11.50]) by
	esealmw127.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 16 Feb 2007 16:54:52 +0100
Received: from nomadiclab.lmf.ericsson.se (nomadiclab.lmf.ericsson.se
	[131.160.33.3])
	by mail.lmf.ericsson.se (Postfix) with ESMTP id 8F511246F;
	Fri, 16 Feb 2007 17:54:52 +0200 (EET)
Received: from nomadiclab.lmf.ericsson.se (localhost [127.0.0.1])
	by nomadiclab.lmf.ericsson.se (Postfix) with ESMTP id 40D824DB55;
	Fri, 16 Feb 2007 17:54:52 +0200 (EET)
Received: from localhost.localdomain (localhost [IPv6:::1])
	by nomadiclab.lmf.ericsson.se (Postfix) with ESMTP id 6E7AB4DA95;
	Fri, 16 Feb 2007 17:54:50 +0200 (EET)
Subject: RE: [Sipping] Service Identification in SIP
From: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>
To: Andrew Allen <aallen@rim.com>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C009151456@XCH47YKF.rim.net>
References: <61968779B8AC4C4BAB421D4C12F008C009151456@XCH47YKF.rim.net>
Content-Type: text/plain
Date: Fri, 16 Feb 2007 17:55:33 +0200
Message-Id: <1171641333.3800.34.camel@localhost.localdomain>
Mime-Version: 1.0
X-Mailer: Evolution 2.2.3 (2.2.3-4.fc4) 
Content-Transfer-Encoding: 7bit
X-Virus-Scanned: ClamAV using ClamSMTP
X-OriginalArrivalTime: 16 Feb 2007 15:54:52.0742 (UTC)
	FILETIME=[CBD32660:01C751E2]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: f0ea5880a0890be2408609376fa519aa
Cc: Paul Kyzivat <pkyzivat@cisco.com>, Gonzalo.Camarillo@ericsson.com,
	sipping@ietf.org, mary.barnes@nortel.com, ravishankar.shiroor@wipro.com
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

On Thu, 2007-02-15 at 18:35 -0500, Andrew Allen wrote:
> I agree with Salvatore that IM is a better example of where service
> identification may have a role in expressing a preference for a
> particular application. Certainly there are different emoticons and
> other value added features proprietary to different messaging
> applications so having the ability to express preferences to use a
> preferred application seems to me to have value. I think that email is a
> bad counter example as with email the email client interfaces to the
> email server not directly to the other email client.

I am convincing myself that a good way to interpret service, or if you
want "communication service" (in 3GPP terminology) is just as a context 
for the communication. 
So a Communication Service Identifier should not override or hide the
media capabilities but just specify the context or better the preferred
context where the user would like use them.
Whatever will be the mechanism will be chosen, I really think it should
possible for the user for example ask



About the discussion below, I want clarify my opinion, I am not
religiously for CallPref and for the way 3GPP is going to use it.
I share the same concern about the interoperability, and I agree with
Andrew about the interest and the opportunity for both 3GPP and Internet

My intent is just to understand better all the technical aspects.
For example I am curios about the possibility highlighted by Paul
previously in this same thread of considering the ASs as a First class
UASs and using call-pref to select them.

At same time I am also curious about the mechanism proposed by Jonathan.
It is an interesting proposal, but to tell you the truth, 
at the present stage in both the proposal call-pref and Jonathan
mechanism I can identify more or less the same number of issues that
need a lot of discussion and work before reach a robust stage.
For example the Jonathan proposal is based on the louse routing that is
something that has not yet reached a large consensus in the sip
community, and also the re-ordering is something that need a lot of work
and clarification. But as I said it is an interesting proposal.


/sal




> 
> Where I disagree with Salvatore is that the fact that the use of
> callerprefs doesn't support well, if at all, the inclusion of
> intermediate servers into the call path is not a problem for 3GPP as
> well. From my viewpoint 3GPP needs to make use of Callerprefs in the
> same way as was intended in RFC 3840/3841 and not extend this with an
> entire bunch of semantics that were not defined for Accept-Contact. To
> do otherwise is only likely to create problems with services that work
> in the internet but do not work with mobile and other devices in IMS
> networks. Also service identification needs to work in both IMS networks
> and non-IMS networks so that the the same services can be deployed in
> both environments. The fact that IMS filter criteria enables routing on
> just about anything (Using iFC I could route based on my favourite
> number appearing in a Call-ID or Cseq etc) doesn't mean that it's Ok or
> a good idea to simply add a new token anywhere you can and route based
> on that. Services should not depend on the use of filter criteria and
> the IMS architecture and routing decisions should be based on the SIP
> routing mechanisms that use the contents of the Route header and
> Request-URI for routing decisions. IMHO Accept-Contact is not a routing
> header per say but simply allows a filter to be specified that filters
> the list of registered contacts in order to determine which contacts of
> the AOR to fork the request to. Building additional semantics on top of
> this filter potentially significantly restricts the use of this filter
> mechanism as originally intended.
> 
> I think it is in the interest of both 3GPP and the internet that a
> general mechanism that works also in a pure SIP environment is
> developed.
> 
> Andrew
> 
> 
> -----Original Message-----
> From: Salvatore Loreo (JO/LMF) [mailto:salvatore.loreto@ericsson.com] 
> Sent: Friday, February 09, 2007 9:39 AM
> To: Paul Kyzivat
> Cc: mary.barnes@nortel.com; sipping@ietf.org;
> ravishankar.shiroor@wipro.com; Gonzalo.Camarillo@ericsson.com
> Subject: Re: [Sipping] Service Identification in SIP
> 
> Hi,
> 
> In my opinion we should also consider the user expectation.
> I fully agree with you about the mail service, but if I have look at the
> Instant Messaging service where at present there are different and
> incompatible programs, then maybe the necessity of a identification for
> the specific IM service become a little bit more clear.
> 
> 
> I think that everyone here agree that there is the possibility to use a
> media in different ways, and that we need a distinction about what the
> capabilities are and how they are used, whatever will be the mechanism
> that at the end we will identify to do that.
> What we do need is a way to identify a preference.
> 
> then what I don't completely understand, surely it is my fault, that a
> combination of capabilities and ways to use those capabilities it is not
> itself a capability, perhaps something we can call a *macro-capability*
> Even if I agree on the opportunity of don't create a real Tower of Babel
> and avoid . So definitely this macro-capability should be well defined
> and standardized.
> 
> 
> Of course we need also discuss the opportunity to use the collerpref
> mechanism to address this capability.
> I agree with Paul that "callerprefs doesn't support well, if at all, is
> the inclusion of intermediate servers into the call path."
> This of course is not a problem from 3GPP point of view, because in a
> IMS network it is present the IFC mechanism that resolve this
> "callerpref" problem. 
> But I do think we need a general mechanism that works also in a pure SIP
> environment.
> 
> 
> 
> /sal
> 
> 
> 
> 
> 
> 
> On Fri, 2007-02-09 at 08:26 -0500, Paul Kyzivat wrote:
> > 
> > ravishankar.shiroor@wipro.com wrote:
> > > Hi Paul,
> > > 
> > > IMO, service identification is different from identifying the tool
> that
> > > should be used for executing that service on the end device. In the
> > > analogy that you have given, e-mail is the service. Eudora/gmail/...
> Are
> > > the tools that are used to provide the service.
> > 
> > The comment below was about identifying the application in the end 
> > system, not the protocol.
> > 
> > In the case of email, the "service" is determined by the protocol and 
> > the address/port of the destination. Nobody has suggested adding
> headers 
> > to mail messages to identify the service/application in the end system
> 
> > that should process the mail message. But here we see proposals to add
> a 
> > header to a sip message to identify the service/application in the
> UAS.
> > 
> > 	Paul
> > 
> > > Regards,
> > > Ravi.
> > > 
> > > 
> > > --
> > >  
> > > Ravishankar. G. Shiroor
> > > Wipro Technologies, Bangalore.
> > >  
> > > ravishankar.shiroor@wipro.com
> > > --
> > >  
> > > 
> > >> -----Original Message-----
> > >> From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
> > >> Sent: Wednesday, February 07, 2007 10:43 PM
> > >> To: Salvatore Loreto (JO/LMF)
> > >> Cc: RaviShankar Shiroor (WT01 - Product Engineering 
> > >> Solutions); sipping@ietf.org; Gonzalo.Camarillo@ericsson.com; 
> > >> mary.barnes@nortel.com
> > >> Subject: Re: [Sipping] Service Identification in SIP
> > >>
> > >> I still don't get the intent of identifying the application. 
> > >> It seems to me this is evil - akin to sending an email 
> > >> message and identifying that the recipient must use Outlook 
> > >> to receive it, not Thunderbird, Eudora, gmail, or whatever.
> > >>
> > >> I also wish somebody could give better examples of a 
> > >> communication service. Chess is too artificial to be 
> > >> believable, and IMO PoC is only a service because of the way 
> > >> the implementation has been done. 
> > >> Conceptually PoC could be nothing but a special floor control 
> > >> protocol accompanying a voice session. In that case none of 
> > >> the Service mechanism would be needed.
> > >>
> > >> 	Paul
> > >>
> > >> Salvatore Loreto (JO/LMF) wrote:
> > >>> hi Ravi,
> > >>>
> > >>> in IMS a Service is a "type of communication defined by a service 
> > >>> definition that specifies the rules and procedures and 
> > >> allowed media, 
> > >>> if any, for a specific type of communication and that 
> > >> utilises the IMS 
> > >>> enabler. An IMS communication service defines what is 
> > >> possible within 
> > >>> a single SIP session or standalone transaction."
> > >>> So a Service implies some combination of capabilities and 
> > >> ways to use 
> > >>> those capabilities.
> > >>> If you want the Service specifies the context for a communication.
> > >>>
> > >>> Instead the Application define the specific communication 
> > >> (or if you 
> > >>> want application) that you want run in the specific context. So
> you 
> > >>> should have several different applications designed to run on a 
> > >>> specific environment or that have different performance or 
> > >> behavior in 
> > >>> different context (running on top of different service).
> > >>>
> > >>> Lets go to Jonathan's example of the Chess.
> > >>> In his example the Chess is a service, for IMS instead is an 
> > >>> application that you can run on top of different service.
> > >>> For example you can run Chess application on top of PoC 
> > >> service or on 
> > >>> top of the "standard" service. So let say you run Chess on 
> > >> top of PoC 
> > >>> service, then you can use only voice in the way that PoC
> specifies.
> > >>>
> > >>> /sal
> > >>>
> > >>>
> > >>>
> > >>>
> > >>> On Thu, 2007-02-01 at 16:39 +0530, 
> > >> ravishankar.shiroor@wipro.com wrote:
> > >>>> Hello,
> > >>>>
> > >>>> IMO, Service Identification is definitely needed and 
> > >> should be taken 
> > >>>> up as a WG as proposed below.
> > >>>>
> > >>>> Further, I think Service Identification scope should be as per
> the 
> > >>>> requirement in Jonathan's draft, as I still do not understand the
> 
> > >>>> need for having a Communication Service Identifier and another 
> > >>>> Application Identifier as given in the 
> > >> 3gpp-ics-requirements document.
> > >>>> Regards,
> > >>>> Ravi.
> > >>>>
> > >>>> --
> > >>>>  
> > >>>> Ravishankar. G. Shiroor
> > >>>> Wipro Technologies, Bangalore.
> > >>>>  
> > >>>> ravishankar.shiroor@wipro.com
> > >>>> --
> > >>>>  
> > >>>>
> > >>>>> -----Original Message-----
> > >>>>> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
> > >>>>> Sent: Wednesday, January 31, 2007 8:30 PM
> > >>>>> To: sipping
> > >>>>> Cc: Mary Barnes
> > >>>>> Subject: [Sipping] Service Identification in SIP
> > >>>>>
> > >>>>> Folks,
> > >>>>>
> > >>>>> during the last year, there have been a number of threads on the
> 
> > >>>>> mailing list discussing service identification in SIP. In 
> > >>>>> particular, the following thread discussed 3GPP 
> > >> requirements on this 
> > >>>>> area:
> > >>>>>
> > >>>>>
> > >> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
> > >>>>> At a later point, those 3GPP requirements were documented in the
> 
> > >>>>> following (expired by now) Internet Draft, which was 
> > >> presented in a 
> > >>>>> face-to-face IETF meeting.
> > >>>>>
> > >>>>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
> > >>>>> cs-requirements-00.txt
> > >>>>>
> > >>>>> Lately, the following draft has been submitted to the archives:
> > >>>>>
> > >>>>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
> > >>>>> rvice-identification-00.txt
> > >>>>>
> > >>>>> This draft proposes a solution that intends to meet known 
> > >>>>> requirements in this area.
> > >>>>>
> > >>>>> Given the amount of discussions on issues related to service 
> > >>>>> identification we have seen during the last year, we 
> > >> believe this is 
> > >>>>> a fairly important topic for the SIPPING WG to tackle. At some 
> > >>>>> point, we would like to have a milestone with its 
> > >> associated WG item 
> > >>>>> dealing with this issue.
> > >>>>>
> > >>>>> Therefore, we would like to encourage people to read this 
> > >> draft and 
> > >>>>> send comments to the list. Additionally, if you believe that
> this 
> > >>>>> work is (or is not) important, we would like to hear that as
> well.
> > >>>>>
> > >>>>> Cheers,
> > >>>>>
> > >>>>> Gonzalo
> > >>>>> SIPPING co-chair
> > >>>>>
> > >>>>>
> > >>>>> _______________________________________________
> > >>>>> Sipping mailing list  
> > >> https://www1.ietf.org/mailman/listinfo/sipping
> > >>>>> This list is for NEW development of the application of SIP Use 
> > >>>>> sip-implementors@cs.columbia.edu for questions on current sip
> Use 
> > >>>>> sip@ietf.org for new developments of core SIP
> > >>>>>
> > >>>> _______________________________________________
> > >>>> Sipping mailing list  
> > >> https://www1.ietf.org/mailman/listinfo/sipping
> > >>>> This list is for NEW development of the application of SIP Use 
> > >>>> sip-implementors@cs.columbia.edu for questions on current sip Use
> 
> > >>>> sip@ietf.org for new developments of core SIP
> > >>>
> > >>> _______________________________________________
> > >>> Sipping mailing list
> https://www1.ietf.org/mailman/listinfo/sipping
> > >>> This list is for NEW development of the application of SIP Use 
> > >>> sip-implementors@cs.columbia.edu for questions on current sip Use 
> > >>> sip@ietf.org for new developments of core SIP
> > >>>
> > > 
> > > 
> > > The information contained in this electronic message and any
> attachments to this message are intended for the exclusive use of the
> addressee(s) and may contain proprietary, confidential or privileged
> information. If you are not the intended recipient, you should not
> disseminate, distribute or copy this e-mail. Please notify the sender
> immediately and destroy all copies of this message and any attachments. 
> > > 
> > > WARNING: Computer viruses can be transmitted via email. The
> recipient should check this email and any attachments for the presence
> of viruses. The company accepts no liability for any damage caused by
> any virus transmitted by this email.
> > >  
> > > www.wipro.com
> > > 
> 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 
> 
> 
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential information, privileged material (including material protected by the solicitor-client or other applicable privileges), or constitute non-public information. Any use of this information by anyone other than the intended recipient is prohibited. If you have received this transmission in error, please immediately reply to the sender and delete this information from your system. Use, dissemination, distribution, or reproduction of this transmission by unintended recipients is not authorized and may be unlawful.


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 16 11:03:03 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HI5Xw-0001If-5W; Fri, 16 Feb 2007 11:02:52 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HI5Xv-0001Ew-6y
	for sipping@ietf.org; Fri, 16 Feb 2007 11:02:51 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HI5Xp-0007cp-3f
	for sipping@ietf.org; Fri, 16 Feb 2007 11:02:51 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	863CF205E9; Fri, 16 Feb 2007 17:02:42 +0100 (CET)
X-AuditID: c1b4fb3c-b1fccbb0000007de-80-45d5d5a265d7 
Received: from esealmw129.eemea.ericsson.se (unknown [153.88.254.124])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	5F2D42048B; Fri, 16 Feb 2007 17:02:42 +0100 (CET)
Received: from esealmw127.eemea.ericsson.se ([153.88.254.171]) by
	esealmw129.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 16 Feb 2007 17:02:42 +0100
Received: from mail.lmf.ericsson.se ([131.160.11.50]) by
	esealmw127.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 16 Feb 2007 17:02:41 +0100
Received: from nomadiclab.lmf.ericsson.se (nomadiclab.lmf.ericsson.se
	[131.160.33.3])
	by mail.lmf.ericsson.se (Postfix) with ESMTP id 757A5246F;
	Fri, 16 Feb 2007 18:02:41 +0200 (EET)
Received: from nomadiclab.lmf.ericsson.se (localhost [127.0.0.1])
	by nomadiclab.lmf.ericsson.se (Postfix) with ESMTP id 267644DB55;
	Fri, 16 Feb 2007 18:02:41 +0200 (EET)
Received: from localhost.localdomain (localhost [IPv6:::1])
	by nomadiclab.lmf.ericsson.se (Postfix) with ESMTP id EB5FE4DA95;
	Fri, 16 Feb 2007 18:02:39 +0200 (EET)
Subject: RE: [Sipping] Service Identification in SIP
From: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>
To: Henry Sinnreich <hsinnrei@adobe.com>
In-Reply-To: <24CCCC428EFEA2469BF046DB3C7A8D2249F0AA@namail5.corp.adobe.com>
References: <61968779B8AC4C4BAB421D4C12F008C00915140A@XCH47YKF.rim.net>
	<24CCCC428EFEA2469BF046DB3C7A8D2249F0AA@namail5.corp.adobe.com>
Content-Type: text/plain
Date: Fri, 16 Feb 2007 18:03:23 +0200
Message-Id: <1171641803.3800.38.camel@localhost.localdomain>
Mime-Version: 1.0
X-Mailer: Evolution 2.2.3 (2.2.3-4.fc4) 
Content-Transfer-Encoding: 7bit
X-Virus-Scanned: ClamAV using ClamSMTP
X-OriginalArrivalTime: 16 Feb 2007 16:02:41.0655 (UTC)
	FILETIME=[E3519870:01C751E3]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8a961490db2a74c7613bf0201229f176
Cc: ravishankar.shiroor@wipro.com, Paul Kyzivat <pkyzivat@cisco.com>,
	Gonzalo.Camarillo@ericsson.com, mary.barnes@nortel.com,
	Andrew Allen <aallen@rim.com>, sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi Henry,

thank you for point out this two drafts, I didn't know the dccp one, it
is interesting that there is a similar idea going on also at transport
level. I'll read the draft immediately.

/sal



On Thu, 2007-02-15 at 18:03 -0800, Henry Sinnreich wrote:
> Andrew,
> 
> It is interesting to note that there is similar thinking going on in
> other WGs, see for example:
> 
> http://www.ietf.org/internet-drafts/draft-fairhurst-dccp-serv-codes-02.t
> xt 
> 
> and
> 
> http://www.ietf.org/internet-drafts/draft-andreasen-sipping-rfc3603bis-0
> 2.txt 
> 
> Along with your correct observation:
> 
> > with an increasing number of competing service providers in the market
> 
>  One has to wonder about the scalability and complexity all levels of
> defining so many services for so many providers, not only for SIP but
> also in the transport area. 
> 
> Coming back to scalability, even plain services could have promotions,
> coupons "free video for frequent flyers", etc. that interestingly enough
> may not be a good idea to trumpet to the competition or get the
> agreement first from the competition. Perish the thought...
> 
> It is also correct that 
> 
> > this argument will be settled by market competition and is not really 
> > the role of engineering to settle.  
> 
> Internet engineers may however share the discomfort that service
> identification facilitates the following on the Internet:
> 
> - Possible discrimination of unwanted services by the competition,
> - Restricting the freedom of development for edge devices and apps,
> - Charging for the two above, to top it off.
> 
> Thanks again to all for the thoughtful contributions to this discussion.
> 
> Henry
> 
> 
> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com] 
> Sent: Thursday, February 15, 2007 2:17 PM
> To: Henry Sinnreich; Paul Kyzivat; Salvatore Loreto (JO/LMF)
> Cc: Gonzalo.Camarillo@ericsson.com; sipping@ietf.org;
> mary.barnes@nortel.com; ravishankar.shiroor@wipro.com; Andrew Allen
> Subject: RE: [Sipping] Service Identification in SIP
> 
> 
>  
> IN response to Henry:
> 
> RFC 4485 also states:
> 
>    Session initiation also depends on the ability of the called party to
>    have enough information about the session itself to make a decision
>    on whether to join.  That information includes data about the caller,
>    the purpose for the invitation, and parameters of the session itself.
>    For this reason, SIP includes this kind of information.
> 
> IMHO "the purpose for the invitation" also includes information about
> what type of service or application this session initiation request is
> related to and SDP doesn't provide sufficent information to definitively
> determine this.
> 
> I think we can all debate the advantages and disadvantages of building
> intelligence in the network as opposed to distributing it to the
> endpoints and I feel that I am also generally on the distributed side of
> that particular argument however SIP allows for both models and many
> services/applications utilise both intelligence in the endpoint and
> intelligence in Application Servers in the network. A prime example is
> Presence where Presence can be provided with SIP using only SIP
> endpoints however many practical implementations make use of network
> based Presence Servers and Resource List Servers in order to create a
> scalable Presence Service and significant IETF work has been done on
> creating SIP based and other related mechanisms (e.g XCAP) to support
> Presence using Presence Servers and Resource List Servers. Being purist
> philosophical about moving all intelligence to the very edge doesn't
> always result in the most optimal solution as the scalability advantages
> of having all the intelligence in the endpoints has to be considered
> along with the scalability disadvantages of having all the signalling
> transported to the endpoints. In limited and high cost bandwidth access
> networks like cellular the scalability disadvantages of having all
> signaling transported to the endpoint can sometimes outweigh the
> advantages of distributed intelligence. 
> 
> My understanding is that SIP doesn't take any position on the business
> models used for communication service provision and IETF has been
> cooperating closely with those who are using SIP for carrier based
> services. Some people may not think that the carrier centric service
> model is the way to go in the future but with an increasing number of
> competing service providers in the market this argument will be settled
> by market competition and is not really the role of engineering to
> settle. 
> 
> Also don't think that failing to develop or delaying an IETF based
> solution will stop 3GPP from developing their own solution. There are
> already proposals being agreed within 3GPP based on Accept-Contact
> header and feature tags to implement this in what I believe is a way
> that potentially breaks interoperability between IMS and other SIP
> networks. This is because ithis proposalt requires pre knowledge of
> these service identifiers and their use in all networks for
> communication without providing a mechanism to obtain them and also
> through its use of Accept-Contact potentially signiifcantly limits the
> flexibility use of that header for its intended purpose. However RFC
> 3427 doesn't require that new feature tags be discussed in SIPPING even
> when the proposed usage is required for interoperability. So basically
> we have a choice between working on an open internet friendly solution
> to service identification or allowing by default a 3GPP specific
> solution that potentially will create interoperability issues between
> internet users and about 2 Billion Mobile users and several hundred
> million fixed carrier device users. Jonthan's proposal addresses these
> issues.
> 
> Something else from RFC 4485:
> 
>    Proxies can ignore bodies:  In order for proxies to scale well, they
>       must be able to operate with minimal message processing.  SIP has
>       been engineered so that proxies can always ignore bodies.
>       Extensions SHOULD NOT require proxies to examine bodies.
> 
> As already stated I don't believe that SDP always provides sufficent
> information to definitively determine the application or service that
> the session initiation request is related to and even when it does this
> requires that a Proxy examine the body in order to determine whether the
> request needs to be routed to some intermediate server that performs
> service logic or make other service based routing decisions. I think it
> is far preferable to have some kind of service identifier that
> specifcally identifies the service logic processing needed included in
> an appropriate place in the request as proposed by Jonathan and make
> route decisions based on that than to have proxies attempt to do this
> based on examing SDP or based on the content type. As well as being more
> efficient in terms of proxy processing this avoids routing failures due
> to encrypted bodies or unexpected multipart MIME content types or new
> alternative content types not yet understood by the proxy causing the
> service routing to fail.
> 
> So you can count me in the supporting the work on service identification
> based on Jonathan's draft column.
> 
> Andrew
> 
> 
> 
> -----Original Message-----
> From: Henry Sinnreich [mailto:hsinnrei@adobe.com] 
> Sent: Wednesday, February 07, 2007 12:34 PM
> To: Paul Kyzivat; Salvatore Loreto (JO/LMF)
> Cc: Gonzalo.Camarillo@ericsson.com; sipping@ietf.org;
> mary.barnes@nortel.com; ravishankar.shiroor@wipro.com
> Subject: RE: [Sipping] Service Identification in SIP
> 
> > I still don't get the intent of identifying the application. It seems
> to > me this is evil - akin to sending an email message and identifying
> that > the recipient must use Outlook to receive it, not Thunderbird,
> Eudora, 
> > gmail, or whatever.
> 
> This goes back to the old discussion about the IN versus the "dumb
> network" and we know how this has played out.
> 
> Placing intelligence (AKA services) in the network is just "dumb design"
> IMHO for all the good reasons, but makes no sense to discuss on this
> list.
> 
> SIP was designed as a rendezvous protocol for multimedia communications
> on the Internet and not to replicate ISDN, etc., as has been so well
> documented in RFC 4485:
> 
>    "SIP is not meant to be used as a strict Public Switched Telephone
>    Network (PSTN) signaling replacement.  It is not a superset of the
>    Integrated Services Digital Network (ISDN) User Part (ISUP).
>    Although it can support gatewaying of PSTN signaling and can provide
>    many features present in the PSTN, the mere existence of a feature or
>    capability in the PSTN is not a justification for its inclusion in
>    SIP."
> 
> Which makes me wonder how many folks have read RFC 4485 and taken it to
> heart or may have chosen to forget it.
> 
> Thanks, Henry
> 
> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
> Sent: Wednesday, February 07, 2007 11:13 AM
> To: Salvatore Loreto (JO/LMF)
> Cc: mary.barnes@nortel.com; sipping@ietf.org;
> ravishankar.shiroor@wipro.com; Gonzalo.Camarillo@ericsson.com
> Subject: Re: [Sipping] Service Identification in SIP
> 
> I still don't get the intent of identifying the application. It seems to
> 
> me this is evil - akin to sending an email message and identifying that
> the recipient must use Outlook to receive it, not Thunderbird, Eudora,
> gmail, or whatever.
> 
> I also wish somebody could give better examples of a communication
> service. Chess is too artificial to be believable, and IMO PoC is only a
> 
> service because of the way the implementation has been done. 
> Conceptually PoC could be nothing but a special floor control protocol
> accompanying a voice session. In that case none of the Service mechanism
> 
> would be needed.
> 
> 	Paul
> 
> Salvatore Loreto (JO/LMF) wrote:
> > hi Ravi,
> > 
> > in IMS a Service is a "type of communication defined by a service 
> > definition that specifies the rules and procedures and allowed media,
> if
> > any, for a specific type of communication and that utilises the IMS 
> > enabler. An IMS communication service defines what is possible within
> a
> > single SIP session or standalone transaction."
> > So a Service implies some combination of capabilities and ways to use 
> > those capabilities.
> > If you want the Service specifies the context for a communication.
> > 
> > Instead the Application define the specific communication (or if you 
> > want application) that you want run in the specific context. So you 
> > should have several different applications designed to run on a
> specific
> > environment or that have different performance or behavior in
> different
> > context (running on top of different service).
> > 
> > Lets go to Jonathan's example of the Chess.
> > In his example the Chess is a service, for IMS instead is an
> application
> > that you can run on top of different service.
> > For example you can run Chess application on top of PoC service or on 
> > top of the "standard" service. So let say you run Chess on top of PoC 
> > service, then you can use only voice in the way that PoC specifies.
> > 
> > /sal
> > 
> > 
> > 
> > 
> > On Thu, 2007-02-01 at 16:39 +0530, ravishankar.shiroor@wipro.com
> wrote:
> >> Hello,
> >>
> >> IMO, Service Identification is definitely needed and should be taken
> up
> >> as a WG as proposed below.
> >>
> >> Further, I think Service Identification scope should be as per the 
> >> requirement in Jonathan's draft, as I still do not understand the
> need
> >> for having a Communication Service Identifier and another Application
> 
> >> Identifier as given in the 3gpp-ics-requirements document.
> >>
> >> Regards,
> >> Ravi.
> >>
> >> --
> >>  
> >> Ravishankar. G. Shiroor
> >> Wipro Technologies, Bangalore.
> >>  
> >> ravishankar.shiroor@wipro.com
> >> --
> >>  
> >>
> >>> -----Original Message-----
> >>> From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]
> >>> Sent: Wednesday, January 31, 2007 8:30 PM
> >>> To: sipping
> >>> Cc: Mary Barnes
> >>> Subject: [Sipping] Service Identification in SIP
> >>>
> >>> Folks,
> >>>
> >>> during the last year, there have been a number of threads on the 
> >>> mailing list discussing service identification in SIP. In 
> >>> particular, the following thread discussed 3GPP requirements on this
> 
> >>> area:
> >>>
> >>> http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html
> >>>
> >>> At a later point, those 3GPP requirements were documented in the 
> >>> following (expired by now) Internet Draft, which was presented in a 
> >>> face-to-face IETF meeting.
> >>>
> >>> http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-i
> >>> cs-requirements-00.txt
> >>>
> >>> Lately, the following draft has been submitted to the archives:
> >>>
> >>> http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-se
> >>> rvice-identification-00.txt
> >>>
> >>> This draft proposes a solution that intends to meet known 
> >>> requirements in this area.
> >>>
> >>> Given the amount of discussions on issues related to service 
> >>> identification we have seen during the last year, we believe this is
> 
> >>> a fairly important topic for the SIPPING WG to tackle. At some 
> >>> point, we would like to have a milestone with its associated WG item
> 
> >>> dealing with this issue.
> >>>
> >>> Therefore, we would like to encourage people to read this draft and 
> >>> send comments to the list. Additionally, if you believe that this 
> >>> work is (or is not) important, we would like to hear that as well.
> >>>
> >>> Cheers,
> >>>
> >>> Gonzalo
> >>> SIPPING co-chair
> >>>
> >>>
> >>> _______________________________________________
> >>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> >>> This list is for NEW development of the application of SIP Use 
> >>> sip-implementors@cs.columbia.edu for questions on current sip Use 
> >>> sip@ietf.org for new developments of core SIP
> >>>
> >> _______________________________________________
> >> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> >> This list is for NEW development of the application of SIP Use 
> >> sip-implementors@cs.columbia.edu for questions on current sip Use 
> >> sip@ietf.org for new developments of core SIP
> > 
> > 
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP Use 
> > sip-implementors@cs.columbia.edu for questions on current sip Use 
> > sip@ietf.org for new developments of core SIP
> > 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use
> sip-implementors@cs.columbia.edu for questions on current sip Use
> sip@ietf.org for new developments of core SIP
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use
> sip-implementors@cs.columbia.edu for questions on current sip Use
> sip@ietf.org for new developments of core SIP
> 
> 
> 
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential
> information, privileged material (including material protected by the
> solicitor-client or other applicable privileges), or constitute
> non-public information. Any use of this information by anyone other than
> the intended recipient is prohibited. If you have received this
> transmission in error, please immediately reply to the sender and delete
> this information from your system. Use, dissemination, distribution, or
> reproduction of this transmission by unintended recipients is not
> authorized and may be unlawful.


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 16 11:16:31 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HI5kv-0007pk-UU; Fri, 16 Feb 2007 11:16:17 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HI5ku-0007pW-2F
	for sipping@ietf.org; Fri, 16 Feb 2007 11:16:16 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HI5kn-0002YS-Hc
	for sipping@ietf.org; Fri, 16 Feb 2007 11:16:16 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	145CB2054B; Fri, 16 Feb 2007 17:16:09 +0100 (CET)
X-AuditID: c1b4fb3e-aeed1bb0000007e1-6a-45d5d8c8c7c6 
Received: from esealmw127.eemea.ericsson.se (unknown [153.88.254.122])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	EC5572005D; Fri, 16 Feb 2007 17:16:08 +0100 (CET)
Received: from eseldmw101.eemea.ericsson.se ([136.225.228.103]) by
	esealmw127.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 16 Feb 2007 17:16:08 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Fri, 16 Feb 2007 17:15:35 +0100
Message-ID: <61D8D34BB13CFE408D154529C120E0799A5B98@eseldmw101.eemea.ericsson.se>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdFSRsbKJWzf9g3QRu4qLbIswIyiwMkwoaw
From: "Atle Monrad \(GR/ETO\)" <atle.monrad@ericsson.com>
To: "Gonzalo Camarillo \(JO/LMF\)" <gonzalo.camarillo@ericsson.com>,
	"sipping" <sipping@ietf.org>
X-OriginalArrivalTime: 16 Feb 2007 16:16:08.0687 (UTC)
	FILETIME=[C458E7F0:01C751E5]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: d185fa790257f526fedfd5d01ed9c976
Cc: Mary Barnes <mary.barnes@nortel.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hello all

If this draft is intended to replace the expired
"draft-loreto-sipping-3gpp-ics-requirements-00.txt", it would be useful
to get a confirmation that this is the case in the draft. Further, as
the requirements from 3GPP have not changed much but merely have been
clarified, some mapping towards these requirements as described in
Loretos expired draft and a linkage towards this expired draft would be
useful aswell.=20

3GPP is still working on a solution to the requirements as listed in
Loretos expired draft, and the current draft would IMHO benefit from
addressing these reqirements more explicit before mapping the
requirements to a solution in order to confirm that the requirements are
covered.=20

On the interoperability, I would just like to add that during
discussions in 3GPP, interoperability has not been considered a problem
by the operators. On the contrary, the GSM Association has asked 3GPP to
finalise the work on communication service identifiers and sort out the
bits and pieces in the protocol ASAP.

Whether this new draft is useful or not depends on the answers to the
questions and concerns raised. So far the thread has been quite
confusing and I have seen more questions than answers.

br
Atle Monrad


-----Original Message-----
From: Gonzalo Camarillo [mailto:Gonzalo.Camarillo@ericsson.com]=20
Sent: 31. januar 2007 16:00
To: sipping
Cc: Mary Barnes
Subject: [Sipping] Service Identification in SIP

Folks,

during the last year, there have been a number of threads on the mailing
list discussing service identification in SIP. In particular, the
following thread discussed 3GPP requirements on this area:

http://www1.ietf.org/mail-archive/web/sipping/current/msg10562.html

At a later point, those 3GPP requirements were documented in the
following (expired by now) Internet Draft, which was presented in a
face-to-face IETF meeting.

http://www.watersprings.org/pub/id/draft-loreto-sipping-3gpp-ics-require
ments-00.txt

Lately, the following draft has been submitted to the archives:

http://www.ietf.org/internet-drafts/draft-rosenberg-sipping-service-iden
tification-00.txt

This draft proposes a solution that intends to meet known requirements
in this area.

Given the amount of discussions on issues related to service
identification we have seen during the last year, we believe this is a
fairly important topic for the SIPPING WG to tackle. At some point, we
would like to have a milestone with its associated WG item dealing with
this issue.

Therefore, we would like to encourage people to read this draft and send
comments to the list. Additionally, if you believe that this work is (or
is not) important, we would like to hear that as well.

Cheers,

Gonzalo
SIPPING co-chair


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 16 11:28:16 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HI5w0-0003Ji-4n; Fri, 16 Feb 2007 11:27:44 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HI5vz-0003Jb-1u
	for sipping@ietf.org; Fri, 16 Feb 2007 11:27:43 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HI5vs-000497-40
	for sipping@ietf.org; Fri, 16 Feb 2007 11:27:43 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	9C58620657; Fri, 16 Feb 2007 17:27:33 +0100 (CET)
X-AuditID: c1b4fb3c-b0fcabb0000007de-b8-45d5db7598c3 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	81F582048B; Fri, 16 Feb 2007 17:27:33 +0100 (CET)
Received: from esealmw126.eemea.ericsson.se ([153.88.254.174]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 16 Feb 2007 17:27:33 +0100
Received: from mail.lmf.ericsson.se ([131.160.11.50]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 16 Feb 2007 17:27:32 +0100
Received: from nomadiclab.lmf.ericsson.se (nomadiclab.lmf.ericsson.se
	[131.160.33.3])
	by mail.lmf.ericsson.se (Postfix) with ESMTP id E46A5251F;
	Fri, 16 Feb 2007 18:27:32 +0200 (EET)
Received: from nomadiclab.lmf.ericsson.se (localhost [127.0.0.1])
	by nomadiclab.lmf.ericsson.se (Postfix) with ESMTP id 740094DB55;
	Fri, 16 Feb 2007 18:27:32 +0200 (EET)
Received: from localhost.localdomain (localhost [IPv6:::1])
	by nomadiclab.lmf.ericsson.se (Postfix) with ESMTP id A3D134DA95;
	Fri, 16 Feb 2007 18:27:31 +0200 (EET)
Subject: RE: [Sipping] Service Identification in SIP
From: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>
To: Andrew Allen <aallen@rim.com>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C009151469@XCH47YKF.rim.net>
References: <61968779B8AC4C4BAB421D4C12F008C009151469@XCH47YKF.rim.net>
Content-Type: text/plain
Date: Fri, 16 Feb 2007 18:28:15 +0200
Message-Id: <1171643295.3800.43.camel@localhost.localdomain>
Mime-Version: 1.0
X-Mailer: Evolution 2.2.3 (2.2.3-4.fc4) 
Content-Transfer-Encoding: 7bit
X-Virus-Scanned: ClamAV using ClamSMTP
X-OriginalArrivalTime: 16 Feb 2007 16:27:33.0041 (UTC)
	FILETIME=[5C410A10:01C751E7]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 2a76bcd37b1c8a21336eb0a1ea6bbf48
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi Andrew,

your mail reinforce my idea that what we need, or what we are looking
for, is something that specify the ***context*** for a specific call.

/sal 


On Thu, 2007-02-15 at 22:26 -0500, Andrew Allen wrote:
> I'll have a crack at laying out examples showing the utility of service
> identification.
> 
> First a gaming example somewhat similar to Jonathan's. Take a network
> based multiplayer game (Chess, D&D, Quake, Pool, Roulette, virtual golf,
> etc.....). This game allows multiple users to participate in the game
> either as players or as observers. An audio stream is established so
> that players and observers can exchange comments on the game along with
> the media stream (either using MSRP or an application specific data
> transport that is used to exchange game commands and game position
> updates between the participants and the game server which also acts as
> a focus. The gaming session is established using SIP and SDP over a SIP
> network such as IMS that has Application Servers that perform telephony
> call features for basic voice only telephony and multimedia "rich"
> telephony calls. The multimedia telephony sessions also support combined
> audio with MSRP and other application specific data streams (such as
> whiteboards) in a call. Now gaming sessions don't require telephony
> service call feature handling in fact this isnt even desired however
> based on the SIP and SDP it isnt possible to determine whether an INVITE
> request is related to a telephony call or a gaming session and hence
> whether the request needs to be routed via an Application Server that
> performs telephony call features or not. In addition users may have
> multiple devices only one of which has the application that understands
> the game. Other devices that the user has registered against his AoR may
> handle audio for telephony and other voice applications but if Alice
> invites Bob to play her favourite game it is not very interesting if Bob
> accepts the gaming session on a device that can only accept the audio
> stream. Also even when the terminal that has the game application
> receives the INVITE the correct application in that terminal needs to be
> identified to accept the session. In addition service providers may
> charge different rates for telephony and for games. I also don't want to
> be charged telephony per minute call rates for playing my game.  
> 
> Here is another example that I think shows that service identification
> doesn't just apply to INVITE and sessions but also can apply to other
> requests such a MESSAGE as well. SIP MESSAGE method is being used for
> Instant messaging applications but it is also being used for many non
> Instant Messaging applications as well. These include as a replacement
> SIP transport for transporting encapsulated SMS (Short Message Service)
> contents (see 3GPP SMS over IP), as a transport for Group Advertisement
> documents for Push To Talk (see OMA PoC), as a notification transport
> for various Push Services (see OMA SIP Push). Each of these different
> service applications of SIP MESSAGE method requires routing to different
> Application Servers in the network, delivery to different applications
> in the terminal and the ability to be charged for separately. The only
> SIP header that differentiates each of these different useages of SIP
> MESSAGE is the Content-Type header but routing based on Content-Type is
> a very bad idea. Firstly the Content-Type doesn't identify the service
> specifically but some content related to a subset of the service. Also
> as the service evolves the Content-Types used may change or new ones may
> be added and this shouldn't require updating the routing rules of the
> proxies or the internal application invocation logic in endpoints. The
> possibility of Multipart MIME Content-Types makes parsing of
> Content-Types by proxies inefficient and complex and potentially
> impossible if the message body is encrypted. What is needed is a means
> compliant with standard SIP Routing mechanisms that allow SIP proxies to
> efficently route these different MESSAGE requests to the appropriate
> servers, endpoints and invoke applications within endpoints. I think
> Jonathan's draft is a good basis for doing this.
> 
> Another example using INVITE is IPTV. IPTV may be established using
> INVITE containing SDP with both Audio and Video streams. How does the
> network determine the difference between an IPTV related INVITE and one
> related to multimedia telephony that supports video? Certainly it is not
> desired that the IPTV INVITE invoke telephony call features in an
> Application Server and you probably don't want the IPTV INVITE being
> routed to your mobile phone (unless it's a mobile TV service). It is
> likely that an SDP offer for IPTV is for a unidirectional media stream
> but this is real convoluted to use as the basis for routing and
> charging. 
> 
> In adddition it is possible that an INVITE doesn't contain SDP at all
> and then there is no basis in SDP for determining the related service.
> 
> Andrew
> 
> 
> 
> -----Original Message-----
> From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net] 
> Sent: Saturday, February 10, 2007 11:50 PM
> To: sipping@ietf.org
> Subject: Re: [Sipping] Service Identification in SIP
> 
> I've attempted to re-read the threads that Gonzalo referred to, and the
> major limitation that I see is the lack of any concrete example of why
> these identifiers are needed.  Probably the IMS people have discussed
> this so many times that they've forgotten the detailed rationale behind
> the requirement.  But I'm a net-head, (and most of the people here are
> net-heads) and the SIP systems that I've dealt with show no evidence of
> needing "service identification", they're pure proxies, and requests are
> passed transparently end-to-end.
> 
> Could somebody please lay out an example showing the utility of service
> identification?
> 
>    From: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>
> 
>    I think that everyone here agree that there is the possibility to use
> a
>    media in different ways, and that we need a distinction about what
> the
>    capabilities are and how they are used, whatever will be the
> mechanism
>    that at the end we will identify to do that.
> 
> Respectfully, I disagree.  We do need ways of identifying the
> capabilities of SIP agents and transports, but I see no necessity of
> creating specified bundles of capabailities, and then constraining
> dialogs to require an initial agreement as to which bundle of
> capabilities will be used.
> 
> One of the central features of Internet architecture is that agents
> negotiate compatibility on a feature-by-feature basis; there are no
> "profiles" which must be known on a network-wide basis.  (And I have
> observed that this feature is of significant commercial value.)
> 
> In regard to identifying an "application" to process the dialog (and I
> use "application" in a standard computing way), I see no reason that a
> user cannot select URI-parameters to append to his AOR, and arrange for
> his UA to route invites with those URI-parameters to the appropriate
> application.  (And given the rich mechanisms for redirection in SIP, the
> application need not even reside at a UA which is registered for the
> AOR.)  These URI-parameters can be formally or informally standardized,
> or established ad-hoc by each user.
> 
> In regard to PoC (as someone said, "a specialized floor-control
> protocol") and similar features, SIP currently has mechanisms for
> indicating support of features and requiring that a recipient support a
> feature before accepting a call (the Supported and Requires headers, and
> simply using the feature).  The only limitation I can see that might
> need to be addressed is that there is no way to create an option tag for
> a new feature that one has invented that is not yet thoroughly
> standardized.
> 
> But I now discover RFC 3261 section 8.1.1.9:
> 
>    The option tags listed MUST only refer to extensions defined in
>    standards-track RFCs.  This is to prevent servers from insisting that
>    clients implement non-standard, vendor-defined features in order to
>    receive service.  Extensions defined by experimental and
>    informational RFCs are explicitly excluded from usage with the
>    Supported header field in a request, since they too are often used to
>    document vendor-defined extensions.
> 
> So I see that as a fundamental architectural decision, SIP does not
> permit sending INVITEs that require specialized recipients, excepting if
> the specialization is standards-track.
> 
> Dale
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use
> sip-implementors@cs.columbia.edu for questions on current sip Use
> sip@ietf.org for new developments of core SIP
> 
> 
> 
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential information, privileged material (including material protected by the solicitor-client or other applicable privileges), or constitute non-public information. Any use of this information by anyone other than the intended recipient is prohibited. If you have received this transmission in error, please immediately reply to the sender and delete this information from your system. Use, dissemination, distribution, or reproduction of this transmission by unintended recipients is not authorized and may be unlawful.
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 16 15:03:16 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HI9Gz-0007Iq-Uu; Fri, 16 Feb 2007 15:01:37 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HI9F0-0006el-IS
	for sipping@ietf.org; Fri, 16 Feb 2007 14:59:34 -0500
Received: from c243-111.rim.net ([216.9.243.111] helo=MHS01YKF.rim.net)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HI9Ez-0001mi-0q
	for sipping@ietf.org; Fri, 16 Feb 2007 14:59:34 -0500
Importance: normal
Priority: normal
Received: from ngw01ykf.rim.net ([10.102.108.30]) by MHS01YKF.rim.net with
	Microsoft SMTPSVC(6.0.3790.211); Fri, 16 Feb 2007 14:59:22 -0500
Received: from XCH20YKF.rim.net ([10.102.100.35]) by ngw01ykf.rim.net (SAVSMTP
	3.1.6.45) with SMTP id M2007021614592207444 ;
	Fri, 16 Feb 2007 14:59:22 -0500
Received: from XCH47YKF.rim.net ([10.64.31.217]) by XCH20YKF.rim.net with
	Microsoft SMTPSVC(5.0.2195.6713); Fri, 16 Feb 2007 14:59:22 -0500
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.181
Content-Class: urn:content-classes:message
MIME-Version: 1.0
Subject: RE: Device architecture? WAS: RE: [Sipping] Service
	Identificatio	ninSIP
Date: Fri, 16 Feb 2007 14:59:17 -0500
Message-ID: <61968779B8AC4C4BAB421D4C12F008C00915161E@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Device architecture? WAS: RE: [Sipping] Service
	Identificatio	ninSIP
thread-index: AcdRol1V3xs8EY8aTam56kl7wqTNiAATZT2Q
From: "Andrew Allen" <aallen@rim.com>
To: "Elwell, John" <john.elwell@siemens.com>,
	"Malleswara Rao Ankem" <malleshavn@gmail.com>
X-OriginalArrivalTime: 16 Feb 2007 19:59:22.0279 (UTC)
	FILETIME=[F38CDB70:01C75204]
X-Spam-Score: 0.1 (/)
X-Scan-Signature: ad5310d4a40550521c18cacd3b6e7111
X-Mailman-Approved-At: Fri, 16 Feb 2007 15:01:34 -0500
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1410705292=="
Errors-To: sipping-bounces@ietf.org

This is a multi-part message in MIME format.

--===============1410705292==
Content-Transfer-Encoding: 7bit
Content-Class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C75204.F13937DB"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C75204.F13937DB
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

=20
John
=20
I think whether an AS in the terminating network is considered a contact =
is a debatable point. Certainly it doesnt register feature tags using =
RFC 3840 mechanisms and (at least in IMS) they dont show up as =
registered contacts in the reg-event notifications. Also AS in the =
originating network (which may need to be involved in some communication =
services) are certainly not contacts for the resource identified by the =
Request-URI and Accept-Contact certainly has no defined role in routing =
to them.
=20
In addition 3GPP TS 23.228 v7.6.0 subcaluse 4.13.2 specifies 15 =
requirements for IMS Communication Service Identifier and only one of =
these is regarding matching to the correct contact:
=20
11. The communication service identifier shall be taken into account =
when selecting the correct UE(s), if multiple UEs are registered for the =
same Public User Identity(s).

=20
 Several of the other requirements are in my view totally incompatible =
with Accept-Contact being the transport for the identifier and indeed =
may limit the use of Accept-Contact for its intended purpose. This =
includes charging and policy control:
=20
8.   Based on operator policy S-CSCF and AS shall consider the IMS =
communication service identifier for online and offline charging, e.g. =
put appropriate data into call detailed records.

9.   The communication service identifier shall be capable of being an =
input into the policy control and charging rules.

Also:
=20
2.   Based on operator policy the S-CSCF or an AS shall be able to =
validate an IMS communication service identifier in a SIP request. This =
includes e.g. to check the syntactical correctness of a service =
identifier, and policing the usage of a communication service =
identifier.

10. It shall be possible to use the IMS communication service identifier =
as a means to authorise whether a subscriber is allowed to initiate or =
receive request for a communication service.

Another problem with UAs placing feature tags in Accept-Contact headers =
as Service Identifiers is how does the initiator obtains the service =
identifier of the destination party for the related service? The concept =
only works in the scenario where all networks and devices have a common =
understanding of the same services and the identifiers that identify =
them. As far as I am aware SIP doesnt provide a mechanism to discover =
proprietary non standard feature tags that are used by other devices in =
other domains and map them to services that they relate to. This means =
that if Accept-Contact is used the communication services are closed and =
can only be used by those who are part of the closed network, (unless of =
course IETF were to get into standardising services which I think is =
highly unlikely). If URIs are used as the container for the Service =
Identifier these can be obtained using Presence as proposed in =
Jonathan's draft.
=20
This helps fulfil the following requirements in 3GPP TS 23.228:
=20
12. The usage of communication service identifiers shall not adversely =
affect interoperability between IMS networks and interoperability with =
external SIP networks and CS networks. The behaviour of a network =
receiving the IMS requests without an IMS communication service =
identifier is a matter of operator policy. Usage of communication =
service identifiers shall not decrease the level of interoperability =
with networks and UEs that are unaware of the communication service =
identifier.

14. The usage of communication service identifiers shall not restrict =
the inherent capabilities of SIP.

=20
SDP is not good for routing decisions for service identification. =
Firstly SDP is intended to be end to end only. It may be encrypted and =
understanding the details of the service may require service or =
application level understanding of the SDP and the service which is =
knowledge that proxies should not need to have. Also I dont think that =
SDP can always uniquely identify a service in any case. In addition not =
all requests that are related to a service will contain SDP in any case.
=20
Andrew
________________________________

From: Elwell, John [mailto:john.elwell@siemens.com]=20
Sent: Friday, February 16, 2007 3:12 AM
To: Andrew Allen; Malleswara Rao Ankem
Cc: Andrew Allen; sipping@ietf.org
Subject: RE: Device architecture? WAS: RE: [Sipping] Service =
Identificatio ninSIP


Andrew,
=20
Well, an AS is just another contact (perhaps a permanently registered =
one, rather than one that uses SIP REGISTER), isn't it? So caller =
preferences can be used to choose an AS or any other contact that =
supports the desired capability.
=20
I agree caller preferences is not a suitable basis for charging. Just =
because I say I have a preference for a contact that supports video, =
doesn't mean that I will necessarily use video during that call. That =
might be determined during some mid-call session-re-negotiation. Surely =
charging has to be based on what is actually used, not what might be =
used, and you can't always determine this at INVITE time.
=20
So I am saying use caller preferences to select a preferred contact, use =
SDP to negotiate and re-negotiate the session with the contact that you =
end up with. I haven't heard any compelling arguments for anything else =
so far.
=20
John


________________________________

	From: Andrew Allen [mailto:aallen@rim.com]=20
	Sent: 16 February 2007 03:51
	To: Elwell, John; Malleswara Rao Ankem
	Cc: sipping@ietf.org; Andrew Allen
	Subject: RE: Device architecture? WAS: RE: [Sipping] Service =
Identificatio n inSIP
=09
=09
	=20
	John,
	=20
	If Service Identification was only about selecting the appropriate =
contacts from those registered agaisnt the AOR then RFC 3840/3841 would =
be fine however Service Identification is much more than that and may be =
used to route requests to application servers for service logic =
invocation, and used as the basis for charging. Accept-Contact is =
basically just a filter expression to filter a set of registered =
contacts. Including a parameter in a filter expression and using that as =
the basis for charging, authorisation and service invocation is going to =
cause problems.
	=20
	Andrew
	=20
________________________________

	From: Elwell, John [mailto:john.elwell@siemens.com]=20
	Sent: Thursday, February 15, 2007 6:16 AM
	To: Malleswara Rao Ankem
	Cc: sipping@ietf.org
	Subject: RE: Device architecture? WAS: RE: [Sipping] Service =
Identificatio n inSIP
=09
=09
	Mallesh,
	=20
	Yes, I agree that some means of selecting a preferred device might be =
needed, but that is indeed what RFC3841/3840 does. Why do you consider =
this overkill?
	=20
	John


________________________________

		From: Malleswara Rao Ankem [mailto:malleshavn@gmail.com]=20
		Sent: 15 February 2007 10:54
		To: Elwell, John
		Cc: G=F6ran Eriksson AP (KI/EAB); sipping@ietf.org
		Subject: Re: Device architecture? WAS: RE: [Sipping] Service =
Identificatio n in SIP
	=09
	=09
		Inorder to identify a endpoint that has the capabilities to serve a =
very specialized functionality, I guess it is simpler to name the =
functionality. Naming the functionality not only helps in decision =
making while routing the request to that specific endpoint but also =
helps the endpoint identify which specialized application on the =
endpoint can take up the responsibility. Instead of all the routing =
entities looking into the SDP/message bodies it should be easier/faster =
if they can look for a specific header.=20
		I thought feature tag mechanisms indicated in the RFC3841/3840 =
combination should have sufficed but implementing those RFCs seems to be =
a overkill for describing/selecting/routing requests to the specific =
endpoint and also choosing a service/feature/application on the =
endpoint.=20
		=20
		-Mallesh
		=20
		=20
		=20
		On 2/15/07, Elwell, John <john.elwell@siemens.com > wrote:=20

			G=F6ran,
		=09
			Your interpretation is different from what is stated in the Rosenberg =
draft.
			Anyway, let's not dwell on this particular use case, which may or may =
not be=20
			realistic. The important thing is whether the media being sent is =
suitable
			for general purpose devices or requires or could be used by some =
special
			class of application.
		=09
			If a medium is somehow specialised ( e.g., audio modulated for =
interpretation
			by a modem or fax machine), then the necessary information needs to =
be
			conveyed along with the m-line in the SDP. Perhaps even if it is
			human-intelligible voice but with a restricted repertoire of words =
that make=20
			it suitable for a particular class of application, maybe that =
information
			too should be conveyed with the m-line. Perhaps the case described in
			Jonathan's draft fits this latter category: I believe the medium is =
just=20
			regular voice that can be interpreted by a human, but is also =
constrained in
			its repertoire of words such that it can be used by a chess =
application.
		=09
			The important thing is that there be sufficient information in the =
SDP to=20
			allow appropriate application selection at the end device. I am not
			convinced by a need for information at the SIP level during call
			establishment, because of course SDP can be re-negotiated during the =
call
			( e.g., when the users decide to use a particular application) and =
what is
			indicated in SIP at call establishment time might no longer be valid =
later
			in the call.
		=09
			John
		=09
		=09
			> -----Original Message-----=20
			> From: G=F6ran Eriksson AP (KI/EAB)
			> [mailto:goran.ap.eriksson@ericsson.com]
			> Sent: 14 February 2007 23:34=20
			> To: Elwell, John
			> Cc: sipping@ietf.org
			> Subject: Device architecture? WAS: RE: [Sipping] Service=20
			> Identification in SIP
			>
			> My humble, perhaps faulty, understanding of the "chess game"
			> was not that voice commands were used to control the chess=20
			> itself, rather
			> that the chess game, while having some IP-based proprietary=20
			> gaming interaction protocol, also had a feature letting
			> the users talk to each others while playing. The scenario are=20
			> two, or more, users play a chess game of the (Inter)net, talkin
			> to eachother from time to time during the game (yeah, don't=20
			> ask me if I think people will love this!!).
			>
			> The purpose, as I understand it, was a)to show that the=20
			> 'voice component' in the chess game application is similar to that
			> in a stand alone POC-Only application when it comes to=20
			> exchange and transfer of voice samples and associated floor control =

			> and b) to point to differences, for instance when it comes to
			> supplementary services, which may not apply in the chess game case. =

			>
			> Perhaps it is important to be clear on the assumed user
			> device realisation, as far as I understand it:
			>
			>  The "chess game + voice" example is, as I understand it,
			> discussed under the assumption that voice and chess game is=20
			>  managed by custome designed appliction, or client, which is=20
			> a necessary/likely approach in a mobile device given the
			>  limited MMI (and also other factors in the mobile context).
			>
			>  In a PC, I could easily handle this as two different=20
			> applications/client, like when I use Skype/MSN to talk to my=20
			> friends while
			>  playing some turn based game, but again, the discussion is
			> held under the assumption that usability requirements=20
			>  leads to an application client integrating voice and chess
			> game events.
			>
			> The alternative device realisation You describe- different
			> clients managing different portions of what is negotiated in=20
			> the SDP or MSRP MIME attachment-
			> is, I think, a good illustration of the need to be precise=20
			> and explicit about the assumed device architecture when
			> discussing the "service identification" issue(s).=20
			>
			> Regards
			> G=F6ran
			>
			>
			>
			>
			> -----Original Message-----=20
			> From: Elwell, John [mailto: john.elwell@siemens.com =
<mailto:john.elwell@siemens.com> ]
			> Sent: den 14 februari 2007 11:57
			> To: Henry Sinnreich; sipping@ietf.org=20
			> Subject: RE: [Sipping] Service Identification in SIP
			>
			> Having followed this thread, I too find the chess example
			> very contrived. If chess moves are sent as voice commands,
			> then surely it would be acceptable for the normal telephone=20
			> application to answer the call, i.e., the remote player
			> listens to the move, performs this move on a conventional or
			> electronic chess board and speaks his next move. The calling
			> UA should not care whether the call is answered by a computer=20
			> application or a person.
			> Furthermore, even if the audio needs to consumed by a
			> particular chess application at the called UA, then other
			> media in the same session such as video or messaging might=20
			> not be. So any qualification of the nature of media to be
			> exchanged probably should be at the media level, not at the
			> session level or SIP level. As yet another variation, what if
			> the players set up a conventional telephone call and then=20
			> mutually agree to switch in their chess applications?
			>
			> Now, let's consider another case. I establish an MSRP
			> session. Within this MSRP session I can send messages with
			> various MIME attachments that are intended for consumption by=20
			> various applications. I should not need to know in advance
			> what applications I will use during the course of the MSRP
			> session. All I should put in the INVITE request is the fact
			> that I want to establish an MSRP session.=20
			>
			> John
			>
			> > -----Original Message-----
			> > From: Henry Sinnreich [mailto:hsinnrei@adobe.com ]
			> > Sent: 13 February 2007 18:19
			> > To: sipping@ietf.org
			> > Subject: RE: [Sipping] Service Identification in SIP=20
			> >
			> > It must be my lack of imagination for not being able to
			> come up with a=20
			> > credible usage scenario, given the following:
			> >
			> > - Is playing chess by voice a credible usage scenario? :-)=20
			> Especially
			> > since voice recognition is now so often found in PCs and in =
mobile=20
			> > phones as well. In other words, chess playing by voice may
			> also be an
			> > endpoint application, without any network involvement. Actually, =
if=20
			> > placed in the endpoint it works just as well on poor=20
			> bandwidth links
			> > by contrast to chess in the network that may require adequate
			> > broadband.
			> >
			> > - Is anyone's boss or spouse going to approve extra phone charges =

			> > above the flat rate for playing chess at a time when the gaming
			> > industry has pioneered free VoIP?
			> > Will such a service be taxed extra and if yes using by what
			> criteria?
			> > Gambling comes to my mind :-)=20
			> >
			> > - VoIP is reported in the press as a give-away and I have seen no
			> > reports of any VoIP provider recording a profit on VoIP. I
			> assume the
			> > losses are often hidden by bundling in triple play, etc.=20
			> >
			> > Given my lack of imagination, can anyone share a more
			> credible usage
			> > scenario for Service Identification in SIP?
			> >
			> > What's the use of a standard without a credible usage scenario?=20
			> >
			> > Thanks, Henry
			> >
			> >
			> > _______________________________________________
			> > Sipping mailing list   =
https://www1.ietf.org/mailman/listinfo/sipping =
<https://www1.ietf.org/mailman/listinfo/sipping>=20
			> > This list is for NEW development of the application of SIP Use
			> > sip-implementors@cs.columbia.edu for questions on current sip Use
			> > sip@ietf.org for new developments of core SIP=20
			> >
			>
			> _______________________________________________
			> Sipping mailing list   =
https://www1.ietf.org/mailman/listinfo/sipping =
<https://www1.ietf.org/mailman/listinfo/sipping>=20
			> This list is for NEW development of the application of SIP
			> Use sip-implementors@cs.columbia.edu for questions on current=20
			> sip Use sip@ietf.org for new developments of core SIP=20
			>
		=09
			_______________________________________________
			Sipping mailing list   https://www1.ietf.org/mailman/listinfo/sipping =
<https://www1.ietf.org/mailman/listinfo/sipping>=20
			This list is for NEW development of the application of SIP
			Use sip-implementors@cs.columbia.edu for questions on current sip=20
			Use sip@ietf.org for new developments of core SIP
		=09


=09
________________________________

	This transmission (including any attachments) may contain confidential =
information, privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.=20




---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential =
information, privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.

------_=_NextPart_001_01C75204.F13937DB
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; =
charset=3Diso-8859-1">
<META content=3D"MSHTML 6.00.2900.2963" name=3DGENERATOR></HEAD>
<BODY>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2>&nbsp;</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2>John</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff size=3D2>I =
think whether=20
an AS in the terminating network is considered a contact is a debatable =
point.=20
Certainly it doesnt register feature tags using RFC 3840 mechanisms and =
(at=20
least in IMS) they dont show up as registered contacts in the reg-event=20
notifications. Also AS in the originating network (which may need to be =
involved=20
in some communication services) are certainly not contacts for the =
resource=20
identified by the Request-URI and Accept-Contact certainly has no =
defined role=20
in routing to them.</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff =
size=3D2>In addition 3GPP=20
TS 23.228 v7.6.0 subcaluse 4.13.2 specifies 15 requirements for IMS=20
Communication Service Identifier and only one of these is regarding =
matching to=20
the correct contact:</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007>
<P class=3DB1=20
style=3D"MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: 23228CR0542R1 =
20051123T1735"><SPAN=20
lang=3DEN-GB><FONT size=3D2>11.<SPAN style=3D"mso-tab-count: 1"> =
</SPAN>The=20
communication service identifier shall be taken into account when =
selecting the=20
correct UE(s), if multiple UEs are registered for the same Public User=20
Identity(s).</FONT></SPAN></P></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff =
size=3D2>&nbsp;Several of=20
the other requirements are in my view totally incompatible with =
Accept-Contact=20
being the transport for the identifier and indeed may limit the use of=20
Accept-Contact for its intended purpose. This includes charging and =
policy=20
control:</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007>
<P class=3DB1=20
style=3D"MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: 23228CR0542R1 =
20051123T1735"><SPAN=20
lang=3DEN-GB><FONT size=3D2>8.<SPAN style=3D"mso-tab-count: =
1">&nbsp;&nbsp;=20
</SPAN>Based on operator policy S-CSCF and AS shall consider the IMS=20
communication service identifier for online and offline charging, e.g. =
put=20
appropriate data into call detailed records.</FONT></SPAN></P>
<P class=3DB1=20
style=3D"MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: 23228CR0542R1 =
20051123T1735"><SPAN=20
lang=3DEN-GB><FONT size=3D2>9.<SPAN style=3D"mso-tab-count: =
1">&nbsp;&nbsp; </SPAN>The=20
communication service identifier shall be capable of being an input into =
the=20
policy control and charging rules.</FONT></SPAN></P></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2>Also:</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007>
<P class=3DB1=20
style=3D"MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: 23228CR0542R1 =
20051123T1735"><SPAN=20
lang=3DEN-GB><FONT size=3D2>2.<SPAN style=3D"mso-tab-count: =
1">&nbsp;&nbsp;=20
</SPAN>Based on operator policy the S-CSCF or an AS shall be able to =
validate an=20
IMS communication service identifier in a SIP request. This includes =
e.g. to=20
check the syntactical correctness of a service identifier, and policing =
the=20
usage of a communication service =
identifier.</FONT></SPAN></SPAN></P></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007>
<P class=3DB1=20
style=3D"MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: 23228CR0542R1 =
20051123T1735"><SPAN=20
lang=3DEN-GB><FONT size=3D2>10.<SPAN style=3D"mso-tab-count: 1"> =
</SPAN>It shall be=20
possible to use the IMS communication service identifier as a means to =
authorise=20
whether a subscriber is allowed to initiate or receive request for a=20
communication service.</FONT></SPAN></P></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff =
size=3D2>Another problem=20
with UAs placing feature tags in Accept-Contact headers as Service =
Identifiers=20
is&nbsp;how does the initiator obtains the service identifier of the =
destination=20
party for the related service? The concept only works in the scenario =
where all=20
networks and devices have a common understanding of the same services =
and the=20
identifiers that identify them. As far as I am aware SIP doesnt provide =
a=20
mechanism to discover proprietary non standard feature tags that are =
used by=20
other devices in other domains and map them to services that they relate =
to.=20
This means that if Accept-Contact is used the communication services are =
closed=20
and can only be used by those who are part of the closed network, =
(unless of=20
course IETF were to get into standardising services which I think is =
highly=20
unlikely). If URIs are used as the container for the Service Identifier =
these=20
can be obtained using Presence as proposed in Jonathan's=20
draft.</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff =
size=3D2>This helps fulfil=20
the following requirements in 3GPP TS 23.228:</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007>
<P class=3DB1 style=3D"MARGIN: 0in 0in 9pt 28.4pt"><SPAN =
lang=3DEN-GB><FONT=20
size=3D2>12.<SPAN style=3D"mso-tab-count: 1"> </SPAN>The usage of =
communication=20
service identifiers shall not adversely affect interoperability between =
IMS=20
networks and interoperability with external SIP networks and CS =
networks. The=20
behaviour of a network receiving the IMS requests without an IMS =
communication=20
service identifier is a matter of operator policy. Usage of =
communication=20
service identifiers shall not decrease the level of interoperability =
with=20
networks and UEs that are unaware of the communication service=20
identifier.</FONT></SPAN></P>
<P class=3DB1=20
style=3D"MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: '23.228CR0567' =
20060309T1718"><SPAN=20
lang=3DEN-GB><FONT size=3D2>14.<SPAN style=3D"mso-tab-count: 1"> =
</SPAN>The usage of=20
communication service identifiers shall not restrict the inherent =
capabilities=20
of SIP.</FONT></SPAN></P></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff =
size=3D2>SDP is not good=20
for routing decisions for service identification. Firstly SDP is =
intended to be=20
end to end only. It may be encrypted and understanding&nbsp;the details =
of the=20
service may require service or application level understanding of the =
SDP and=20
the service which is knowledge that proxies should not need to have. =
Also I dont=20
think that SDP can always uniquely identify a service in any case. In =
addition=20
not all requests that are related to a service will contain SDP in any=20
case.</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2>Andrew</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr align=3Dleft>
<HR tabIndex=3D-1>
</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><FONT face=3DTahoma=20
size=3D2><B>From:</B> Elwell, John [mailto:john.elwell@siemens.com]=20
<BR><B>Sent:</B> Friday, February 16, 2007 3:12 AM<BR><B>To:</B> Andrew =
Allen;=20
Malleswara Rao Ankem<BR><B>Cc:</B> Andrew Allen;=20
sipping@ietf.org<BR><B>Subject:</B> RE: Device architecture? WAS: RE: =
[Sipping]=20
Service Identificatio ninSIP<BR></FONT><BR></DIV>
<DIV></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>Andrew,</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>Well, an AS is just another contact (perhaps a =
permanently=20
registered one, rather than one that uses SIP REGISTER), isn't it? So =
caller=20
preferences can be used to choose an AS or any other contact that =
supports the=20
desired capability.</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>I agree caller preferences&nbsp;is not a =
suitable basis for=20
charging. Just because I say I have a preference for a contact that =
supports=20
video, doesn't mean that I will necessarily use video during that call. =
That=20
might be determined during some mid-call session-re-negotiation. Surely =
charging=20
has to be based on what is actually used, not what might be used, and =
you can't=20
always determine this at INVITE time.</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>So I am saying use caller preferences to select =
a preferred=20
contact, use SDP to negotiate and re-negotiate the session with the =
contact that=20
you end up with. I haven't heard any compelling arguments for anything =
else so=20
far.</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>John</FONT></SPAN></DIV><BR>
<BLOCKQUOTE=20
style=3D"PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #0000ff 2px =
solid; MARGIN-RIGHT: 0px">
  <DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr align=3Dleft>
  <HR tabIndex=3D-1>
  <FONT face=3DTahoma size=3D2><B>From:</B> Andrew Allen =
[mailto:aallen@rim.com]=20
  <BR><B>Sent:</B> 16 February 2007 03:51<BR><B>To:</B> Elwell, John; =
Malleswara=20
  Rao Ankem<BR><B>Cc:</B> sipping@ietf.org; Andrew =
Allen<BR><B>Subject:</B> RE:=20
  Device architecture? WAS: RE: [Sipping] Service Identificatio n=20
  inSIP<BR></FONT><BR></DIV>
  <DIV></DIV>
  <DIV dir=3Dltr align=3Dleft><FONT face=3DArial color=3D#0000ff=20
  size=3D2></FONT>&nbsp;</DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2>John,</FONT></SPAN></DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2>If Service Identification was only about =
selecting the=20
  appropriate contacts from those registered agaisnt the AOR then RFC =
3840/3841=20
  would be fine however Service Identification is much more than that =
and may be=20
  used to route requests to application servers for service logic =
invocation,=20
  and used as the basis for charging. Accept-Contact is basically just a =
filter=20
  expression to filter a set of registered contacts. Including a =
parameter in a=20
  filter expression and using that as the basis for charging,=20
  authorisation&nbsp;and service invocation is going to cause=20
  problems.</FONT></SPAN></DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2>Andrew</FONT></SPAN></DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN =
class=3D927204303-16022007></SPAN>&nbsp;</DIV>
  <DIV dir=3Dltr align=3Dleft>
  <HR tabIndex=3D-1>
  </DIV>
  <DIV dir=3Dltr align=3Dleft><FONT face=3DTahoma size=3D2><B>From:</B> =
Elwell, John=20
  [mailto:john.elwell@siemens.com] <BR><B>Sent:</B> Thursday, February =
15, 2007=20
  6:16 AM<BR><B>To:</B> Malleswara Rao Ankem<BR><B>Cc:</B>=20
  sipping@ietf.org<BR><B>Subject:</B> RE: Device architecture? WAS: RE:=20
  [Sipping] Service Identificatio n inSIP<BR></FONT><BR></DIV>
  <DIV></DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2>Mallesh,</FONT></SPAN></DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2>Yes, I agree that some means of selecting a =
preferred=20
  device might be needed, but that is indeed what RFC3841/3840 does. Why =
do you=20
  consider this overkill?</FONT></SPAN></DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D820071411-15022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2>John</FONT></SPAN></DIV><BR>
  <BLOCKQUOTE=20
  style=3D"PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #0000ff 2px =
solid; MARGIN-RIGHT: 0px">
    <DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft>
    <HR tabIndex=3D-1>
    <FONT face=3DTahoma size=3D2><B>From:</B> Malleswara Rao Ankem=20
    [mailto:malleshavn@gmail.com] <BR><B>Sent:</B> 15 February 2007=20
    10:54<BR><B>To:</B> Elwell, John<BR><B>Cc:</B> G=F6ran Eriksson AP =
(KI/EAB);=20
    sipping@ietf.org<BR><B>Subject:</B> Re: Device architecture? WAS: =
RE:=20
    [Sipping] Service Identificatio n in SIP<BR></FONT><BR></DIV>
    <DIV></DIV>
    <DIV>Inorder to identify a endpoint that has the capabilities to =
serve a=20
    very specialized functionality, I guess it is simpler to name the=20
    functionality. Naming the functionality not only helps in decision =
making=20
    while routing the request to that specific endpoint but also helps =
the=20
    endpoint identify which specialized application on the endpoint can =
take up=20
    the responsibility. Instead of all the routing entities looking into =
the=20
    SDP/message bodies it should be easier/faster if they can look for a =

    specific header. </DIV>
    <DIV>I thought feature tag mechanisms indicated in the RFC3841/3840=20
    combination should have sufficed but implementing those RFCs seems =
to be a=20
    overkill for describing/selecting/routing requests&nbsp;to the =
specific=20
    endpoint and also choosing a service/feature/application on the =
endpoint.=20
    </DIV>
    <DIV>&nbsp;</DIV>
    <DIV>-Mallesh</DIV>
    <DIV>&nbsp;</DIV>
    <DIV>&nbsp;</DIV>
    <DIV>&nbsp;</DIV>
    <DIV><SPAN class=3Dgmail_quote>On 2/15/07, <B =
class=3Dgmail_sendername>Elwell,=20
    John</B> &lt;<A onclick=3D"return =
top.js.OpenExtLink(window,event,this)"=20
    href=3D"mailto:john.elwell@siemens.com" =
target=3D_blank>john.elwell@siemens.com=20
    </A>&gt; wrote:</SPAN>=20
    <BLOCKQUOTE class=3Dgmail_quote=20
    style=3D"PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: =
#ccc 1px solid">G=F6ran,<BR><BR>Your=20
      interpretation is different from what is stated in the Rosenberg=20
      draft.<BR>Anyway, let's not dwell on this particular use case, =
which may=20
      or may not be <BR>realistic. The important thing is whether the =
media=20
      being sent is suitable<BR>for general purpose devices or requires =
or could=20
      be used by some special<BR>class of application.<BR><BR>If a =
medium is=20
      somehow specialised ( e.g., audio modulated for =
interpretation<BR>by a=20
      modem or fax machine), then the necessary information needs to=20
      be<BR>conveyed along with the m-line in the SDP. Perhaps even if =
it=20
      is<BR>human-intelligible voice but with a restricted repertoire of =
words=20
      that make <BR>it suitable for a particular class of application, =
maybe=20
      that information<BR>too should be conveyed with the m-line. =
Perhaps the=20
      case described in<BR>Jonathan's draft fits this latter category: I =
believe=20
      the medium is just <BR>regular voice that can be interpreted by a =
human,=20
      but is also constrained in<BR>its repertoire of words such that it =
can be=20
      used by a chess application.<BR><BR>The important thing is that =
there be=20
      sufficient information in the SDP to <BR>allow appropriate =
application=20
      selection at the end device. I am not<BR>convinced by a need for=20
      information at the SIP level during call<BR>establishment, because =
of=20
      course SDP can be re-negotiated during the call<BR>( e.g., when =
the users=20
      decide to use a particular application) and what is<BR>indicated =
in SIP at=20
      call establishment time might no longer be valid later<BR>in the=20
      call.<BR><BR>John<BR><BR><BR>&gt; -----Original Message----- =
<BR>&gt;=20
      From: G=F6ran Eriksson AP (KI/EAB)<BR>&gt; [mailto:<A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:goran.ap.eriksson@ericsson.com"=20
      target=3D_blank>goran.ap.eriksson@ericsson.com</A>]<BR>&gt; Sent: =
14=20
      February 2007 23:34 <BR>&gt; To: Elwell, John<BR>&gt; Cc: <A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sipping@ietf.org" =
target=3D_blank>sipping@ietf.org</A><BR>&gt;=20
      Subject: Device architecture? WAS: RE: [Sipping] Service <BR>&gt;=20
      Identification in SIP<BR>&gt;<BR>&gt; My humble, perhaps faulty,=20
      understanding of the "chess game"<BR>&gt; was not that voice =
commands were=20
      used to control the chess <BR>&gt; itself, rather<BR>&gt; that the =
chess=20
      game, while having some IP-based proprietary <BR>&gt; gaming =
interaction=20
      protocol, also had a feature letting<BR>&gt; the users talk to =
each others=20
      while playing. The scenario are <BR>&gt; two, or more, users play =
a chess=20
      game of the (Inter)net, talkin<BR>&gt; to eachother from time to =
time=20
      during the game (yeah, don't <BR>&gt; ask me if I think people =
will love=20
      this!!).<BR>&gt;<BR>&gt; The purpose, as I understand it, was a)to =
show=20
      that the <BR>&gt; 'voice component' in the chess game application =
is=20
      similar to that<BR>&gt; in a stand alone POC-Only application when =
it=20
      comes to <BR>&gt; exchange and transfer of voice samples and =
associated=20
      floor control <BR>&gt; and b) to point to differences, for =
instance when=20
      it comes to<BR>&gt; supplementary services, which may not apply in =
the=20
      chess game case. <BR>&gt;<BR>&gt; Perhaps it is important to be =
clear on=20
      the assumed user<BR>&gt; device realisation, as far as I =
understand=20
      it:<BR>&gt;<BR>&gt;&nbsp;&nbsp;The "chess game + voice" example =
is, as I=20
      understand it,<BR>&gt; discussed under the assumption that voice =
and chess=20
      game is <BR>&gt;&nbsp;&nbsp;managed by custome designed =
appliction, or=20
      client, which is <BR>&gt; a necessary/likely approach in a mobile =
device=20
      given the<BR>&gt;&nbsp;&nbsp;limited MMI (and also other factors =
in the=20
      mobile context).<BR>&gt;<BR>&gt;&nbsp;&nbsp;In a PC, I could =
easily handle=20
      this as two different <BR>&gt; applications/client, like when I =
use=20
      Skype/MSN to talk to my <BR>&gt; friends =
while<BR>&gt;&nbsp;&nbsp;playing=20
      some turn based game, but again, the discussion is<BR>&gt; held =
under the=20
      assumption that usability requirements <BR>&gt;&nbsp;&nbsp;leads =
to an=20
      application client integrating voice and chess<BR>&gt; game=20
      events.<BR>&gt;<BR>&gt; The alternative device realisation You =
describe-=20
      different<BR>&gt; clients managing different portions of what is=20
      negotiated in <BR>&gt; the SDP or MSRP MIME attachment-<BR>&gt; =
is, I=20
      think, a good illustration of the need to be precise <BR>&gt; and =
explicit=20
      about the assumed device architecture when<BR>&gt; discussing the =
"service=20
      identification" issue(s). <BR>&gt;<BR>&gt; Regards<BR>&gt;=20
      G=F6ran<BR>&gt;<BR>&gt;<BR>&gt;<BR>&gt;<BR>&gt; -----Original =
Message-----=20
      <BR>&gt; From: Elwell, John [mailto:<A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:john.elwell@siemens.com" target=3D_blank>=20
      john.elwell@siemens.com</A>]<BR>&gt; Sent: den 14 februari 2007=20
      11:57<BR>&gt; To: Henry Sinnreich; <A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sipping@ietf.org" target=3D_blank>sipping@ietf.org =
</A><BR>&gt;=20
      Subject: RE: [Sipping] Service Identification in =
SIP<BR>&gt;<BR>&gt;=20
      Having followed this thread, I too find the chess example<BR>&gt; =
very=20
      contrived. If chess moves are sent as voice commands,<BR>&gt; then =
surely=20
      it would be acceptable for the normal telephone <BR>&gt; =
application to=20
      answer the call, i.e., the remote player<BR>&gt; listens to the =
move,=20
      performs this move on a conventional or<BR>&gt; electronic chess =
board and=20
      speaks his next move. The calling<BR>&gt; UA should not care =
whether the=20
      call is answered by a computer <BR>&gt; application or a =
person.<BR>&gt;=20
      Furthermore, even if the audio needs to consumed by a<BR>&gt; =
particular=20
      chess application at the called UA, then other<BR>&gt; media in =
the same=20
      session such as video or messaging might <BR>&gt; not be. So any=20
      qualification of the nature of media to be<BR>&gt; exchanged =
probably=20
      should be at the media level, not at the<BR>&gt; session level or =
SIP=20
      level. As yet another variation, what if<BR>&gt; the players set =
up a=20
      conventional telephone call and then <BR>&gt; mutually agree to =
switch in=20
      their chess applications?<BR>&gt;<BR>&gt; Now, let's consider =
another=20
      case. I establish an MSRP<BR>&gt; session. Within this MSRP =
session I can=20
      send messages with<BR>&gt; various MIME attachments that are =
intended for=20
      consumption by <BR>&gt; various applications. I should not need to =
know in=20
      advance<BR>&gt; what applications I will use during the course of =
the=20
      MSRP<BR>&gt; session. All I should put in the INVITE request is =
the=20
      fact<BR>&gt; that I want to establish an MSRP session. =
<BR>&gt;<BR>&gt;=20
      John<BR>&gt;<BR>&gt; &gt; -----Original Message-----<BR>&gt; &gt; =
From:=20
      Henry Sinnreich [mailto:<A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:hsinnrei@adobe.com" =
target=3D_blank>hsinnrei@adobe.com=20
      </A>]<BR>&gt; &gt; Sent: 13 February 2007 18:19<BR>&gt; &gt; To: =
<A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sipping@ietf.org" =
target=3D_blank>sipping@ietf.org</A><BR>&gt;=20
      &gt; Subject: RE: [Sipping] Service Identification in SIP <BR>&gt; =

      &gt;<BR>&gt; &gt; It must be my lack of imagination for not being =
able=20
      to<BR>&gt; come up with a <BR>&gt; &gt; credible usage scenario, =
given the=20
      following:<BR>&gt; &gt;<BR>&gt; &gt; - Is playing chess by voice a =

      credible usage scenario? :-) <BR>&gt; Especially<BR>&gt; &gt; =
since voice=20
      recognition is now so often found in PCs and in mobile <BR>&gt; =
&gt;=20
      phones as well. In other words, chess playing by voice may<BR>&gt; =
also be=20
      an<BR>&gt; &gt; endpoint application, without any network =
involvement.=20
      Actually, if <BR>&gt; &gt; placed in the endpoint it works just as =
well on=20
      poor <BR>&gt; bandwidth links<BR>&gt; &gt; by contrast to chess in =
the=20
      network that may require adequate<BR>&gt; &gt; broadband.<BR>&gt;=20
      &gt;<BR>&gt; &gt; - Is anyone's boss or spouse going to approve =
extra=20
      phone charges <BR>&gt; &gt; above the flat rate for playing chess =
at a=20
      time when the gaming<BR>&gt; &gt; industry has pioneered free=20
      VoIP?<BR>&gt; &gt; Will such a service be taxed extra and if yes =
using by=20
      what<BR>&gt; criteria?<BR>&gt; &gt; Gambling comes to my mind :-) =
<BR>&gt;=20
      &gt;<BR>&gt; &gt; - VoIP is reported in the press as a give-away =
and I=20
      have seen no<BR>&gt; &gt; reports of any VoIP provider recording a =
profit=20
      on VoIP. I<BR>&gt; assume the<BR>&gt; &gt; losses are often hidden =
by=20
      bundling in triple play, etc. <BR>&gt; &gt;<BR>&gt; &gt; Given my =
lack of=20
      imagination, can anyone share a more<BR>&gt; credible =
usage<BR>&gt; &gt;=20
      scenario for Service Identification in SIP?<BR>&gt; &gt;<BR>&gt; =
&gt;=20
      What's the use of a standard without a credible usage scenario? =
<BR>&gt;=20
      &gt;<BR>&gt; &gt; Thanks, Henry<BR>&gt; &gt;<BR>&gt; &gt;<BR>&gt; =
&gt;=20
      _______________________________________________<BR>&gt; &gt; =
Sipping=20
      mailing list&nbsp;&nbsp;<A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"https://www1.ietf.org/mailman/listinfo/sipping" =
target=3D_blank>=20
      https://www1.ietf.org/mailman/listinfo/sipping </A><BR>&gt; &gt; =
This list=20
      is for NEW development of the application of SIP Use<BR>&gt; &gt; =
<A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sip-implementors@cs.columbia.edu"=20
      target=3D_blank>sip-implementors@cs.columbia.edu</A> for questions =
on=20
      current sip Use<BR>&gt; &gt; <A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sip@ietf.org" target=3D_blank>sip@ietf.org</A> for =
new=20
      developments of core SIP <BR>&gt; &gt;<BR>&gt;<BR>&gt;=20
      _______________________________________________<BR>&gt; Sipping =
mailing=20
      list&nbsp;&nbsp;<A onclick=3D"return =
top.js.OpenExtLink(window,event,this)"=20
      href=3D"https://www1.ietf.org/mailman/listinfo/sipping" =
target=3D_blank>=20
      https://www1.ietf.org/mailman/listinfo/sipping</A><BR>&gt; This =
list is=20
      for NEW development of the application of SIP<BR>&gt; Use <A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sip-implementors@cs.columbia.edu"=20
      target=3D_blank>sip-implementors@cs.columbia.edu</A> for questions =
on=20
      current <BR>&gt; sip Use <A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sip@ietf.org" target=3D_blank>sip@ietf.org</A> for =
new=20
      developments of core SIP=20
      =
<BR>&gt;<BR><BR>_______________________________________________<BR>Sippin=
g=20
      mailing list&nbsp;&nbsp;<A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"https://www1.ietf.org/mailman/listinfo/sipping" =
target=3D_blank>=20
      https://www1.ietf.org/mailman/listinfo/sipping </A><BR>This list =
is for=20
      NEW development of the application of SIP<BR>Use <A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sip-implementors@cs.columbia.edu"=20
      target=3D_blank>sip-implementors@cs.columbia.edu </A>for questions =
on=20
      current sip <BR>Use <A=20
      onclick=3D"return top.js.OpenExtLink(window,event,this)"=20
      href=3D"mailto:sip@ietf.org" target=3D_blank>sip@ietf.org</A> for =
new=20
      developments of core SIP<BR></BLOCKQUOTE></DIV><BR></BLOCKQUOTE>
  <P>
  <HR>
  This transmission (including any attachments) may contain confidential =

  information, privileged material (including material protected by the=20
  solicitor-client or other applicable privileges), or constitute =
non-public=20
  information. Any use of this information by anyone other than the =
intended=20
  recipient is prohibited. If you have received this transmission in =
error,=20
  please immediately reply to the sender and delete this information =
from your=20
  system. Use, dissemination, distribution, or reproduction of this =
transmission=20
  by unintended recipients is not authorized and may be unlawful.=20
</BLOCKQUOTE><p><hr>This transmission (including any attachments) may =
contain confidential information, privileged material (including =
material protected by the solicitor-client or other applicable =
privileges), or constitute non-public information. Any use of this =
information by anyone other than the intended recipient is prohibited. =
If you have received this transmission in error, please immediately =
reply to the sender and delete this information from your system. Use, =
dissemination, distribution, or reproduction of this transmission by =
unintended recipients is not authorized and may be =
unlawful.</BODY></HTML>

------_=_NextPart_001_01C75204.F13937DB--


--===============1410705292==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1410705292==--




From sipping-bounces@ietf.org Fri Feb 16 16:40:40 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HIAoF-0003KC-Vo; Fri, 16 Feb 2007 16:40:03 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HIAoE-0003EY-LK
	for sipping@ietf.org; Fri, 16 Feb 2007 16:40:02 -0500
Received: from c243-111.rim.net ([216.9.243.111] helo=MHS02YKF.rim.net)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HIAoD-00049e-Sq
	for sipping@ietf.org; Fri, 16 Feb 2007 16:40:02 -0500
Importance: normal
Priority: normal
Received: from ngw02ykf.rim.net ([10.102.108.31]) by MHS02YKF.rim.net with
	Microsoft SMTPSVC(6.0.3790.211); Fri, 16 Feb 2007 16:39:51 -0500
Received: from XCH20YKF.rim.net ([10.102.100.35]) by ngw02ykf.rim.net (SAVSMTP
	3.1.6.45) with SMTP id M2007021616394711183 ;
	Fri, 16 Feb 2007 16:39:47 -0500
Received: from XCH47YKF.rim.net ([10.64.31.217]) by XCH20YKF.rim.net with
	Microsoft SMTPSVC(5.0.2195.6713); Fri, 16 Feb 2007 16:39:47 -0500
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.181
Content-class: urn:content-classes:message
MIME-Version: 1.0
Subject: RE: Device architecture? WAS: RE: [Sipping] Service
	Identificatio	ninSIP
Date: Fri, 16 Feb 2007 16:39:41 -0500
Message-ID: <61968779B8AC4C4BAB421D4C12F008C009151675@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Device architecture? WAS: RE: [Sipping] Service
	Identificatio	ninSIP
Thread-Index: AcdRol1V3xs8EY8aTam56kl7wqTNiAAcIs8Q
From: "Andrew Allen" <aallen@rim.com>
To: "Elwell, John" <john.elwell@siemens.com>,
	"Malleswara Rao Ankem" <malleshavn@gmail.com>
X-OriginalArrivalTime: 16 Feb 2007 21:39:47.0739 (UTC)
	FILETIME=[FB0116B0:01C75212]
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 93ea548d5fde6eb89af31f1faab96344
Cc: Andrew Allen <aallen@rim.com>, sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1717018118=="
Errors-To: sipping-bounces@ietf.org

This is a multi-part message in MIME format.

--===============1717018118==
Content-Transfer-Encoding: 7bit
Content-class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C75212.FABA4493"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C75212.FABA4493
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable

=20
John
=20
I think whether an AS in the terminating network is considered a contact
is a debatable point. Certainly it doesnt register feature tags using
RFC 3840 mechanisms and (at least in IMS) they dont show up as
registered contacts in the reg-event notifications. Also AS in the
originating network (which may need to be involved in some communication
services) are certainly not contacts for the resource identified by the
Request-URI and Accept-Contact certainly has no defined role in routing
to them.
=20
In addition 3GPP TS 23.228 v7.6.0 subcaluse 4.13.2 specifies 15
requirements for IMS Communication Service Identifier and only one of
these is regarding matching to the correct contact:
=20
11. The communication service identifier shall be taken into account
when selecting the correct UE(s), if multiple UEs are registered for the
same Public User Identity(s).

=20
 Several of the other requirements are in my view totally incompatible
with Accept-Contact being the transport for the identifier and indeed
may limit the use of Accept-Contact for its intended purpose. This
includes charging and policy control:
=20
8.   Based on operator policy S-CSCF and AS shall consider the IMS
communication service identifier for online and offline charging, e.g.
put appropriate data into call detailed records.

9.   The communication service identifier shall be capable of being an
input into the policy control and charging rules.

Also:
=20
2.   Based on operator policy the S-CSCF or an AS shall be able to
validate an IMS communication service identifier in a SIP request. This
includes e.g. to check the syntactical correctness of a service
identifier, and policing the usage of a communication service
identifier.

10. It shall be possible to use the IMS communication service identifier
as a means to authorise whether a subscriber is allowed to initiate or
receive request for a communication service.

Another problem with UAs placing feature tags in Accept-Contact headers
as Service Identifiers is how does the initiator obtains the service
identifier of the destination party for the related service? The concept
only works in the scenario where all networks and devices have a common
understanding of the same services and the identifiers that identify
them. As far as I am aware SIP doesnt provide a mechanism to discover
proprietary non standard feature tags that are used by other devices in
other domains and map them to services that they relate to. This means
that if Accept-Contact is used the communication services are closed and
can only be used by those who are part of the closed network, (unless of
course IETF were to get into standardising services which I think is
highly unlikely). If URIs are used as the container for the Service
Identifier these can be obtained using Presence as proposed in
Jonathan's draft.
=20
This helps fulfil the following requirements in 3GPP TS 23.228:
=20
12. The usage of communication service identifiers shall not adversely
affect interoperability between IMS networks and interoperability with
external SIP networks and CS networks. The behaviour of a network
receiving the IMS requests without an IMS communication service
identifier is a matter of operator policy. Usage of communication
service identifiers shall not decrease the level of interoperability
with networks and UEs that are unaware of the communication service
identifier.

14. The usage of communication service identifiers shall not restrict
the inherent capabilities of SIP.

=20
SDP is not good for routing decisions for service identification.
Firstly SDP is intended to be end to end only. It may be encrypted and
understanding the details of the service may require service or
application level understanding of the SDP and the service which is
knowledge that proxies should not need to have. Also I dont think that
SDP can always uniquely identify a service in any case. In addition not
all requests that are related to a service will contain SDP in any case.
=20
Andrew

________________________________

From: Elwell, John [mailto:john.elwell@siemens.com]=20
Sent: Friday, February 16, 2007 3:12 AM
To: Andrew Allen; Malleswara Rao Ankem
Cc: Andrew Allen; sipping@ietf.org
Subject: RE: Device architecture? WAS: RE: [Sipping] Service
Identificatio ninSIP


Andrew,
=20
Well, an AS is just another contact (perhaps a permanently registered
one, rather than one that uses SIP REGISTER), isn't it? So caller
preferences can be used to choose an AS or any other contact that
supports the desired capability.
=20
I agree caller preferences is not a suitable basis for charging. Just
because I say I have a preference for a contact that supports video,
doesn't mean that I will necessarily use video during that call. That
might be determined during some mid-call session-re-negotiation. Surely
charging has to be based on what is actually used, not what might be
used, and you can't always determine this at INVITE time.
=20
So I am saying use caller preferences to select a preferred contact, use
SDP to negotiate and re-negotiate the session with the contact that you
end up with. I haven't heard any compelling arguments for anything else
so far.
=20
John


________________________________

	From: Andrew Allen [mailto:aallen@rim.com]=20
	Sent: 16 February 2007 03:51
	To: Elwell, John; Malleswara Rao Ankem
	Cc: sipping@ietf.org; Andrew Allen
	Subject: RE: Device architecture? WAS: RE: [Sipping] Service
Identificatio n inSIP
=09
=09
	=20
	John,
	=20
	If Service Identification was only about selecting the
appropriate contacts from those registered agaisnt the AOR then RFC
3840/3841 would be fine however Service Identification is much more than
that and may be used to route requests to application servers for
service logic invocation, and used as the basis for charging.
Accept-Contact is basically just a filter expression to filter a set of
registered contacts. Including a parameter in a filter expression and
using that as the basis for charging, authorisation and service
invocation is going to cause problems.
	=20
	Andrew
	=20
________________________________





---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential =
information, privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.

------_=_NextPart_001_01C75212.FABA4493
Content-Type: text/html;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; =
charset=3Dus-ascii">
<META content=3D"MSHTML 6.00.2900.2963" name=3DGENERATOR></HEAD>
<BODY>
<DIV dir=3Dltr align=3Dleft><FONT face=3DArial color=3D#0000ff size=3D2>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2>John</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff size=3D2>I =
think whether=20
an AS in the terminating network is considered a contact is a debatable =
point.=20
Certainly it doesnt register feature tags using RFC 3840 mechanisms and =
(at=20
least in IMS) they dont show up as registered contacts in the reg-event=20
notifications. Also AS in the originating network (which may need to be =
involved=20
in some communication services) are certainly not contacts for the =
resource=20
identified by the Request-URI and Accept-Contact certainly has no =
defined role=20
in routing to them.</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff =
size=3D2>In addition 3GPP=20
TS 23.228 v7.6.0 subcaluse 4.13.2 specifies 15 requirements for IMS=20
Communication Service Identifier and only one of these is regarding =
matching to=20
the correct contact:</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007>
<P class=3DB1=20
style=3D"MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: 23228CR0542R1 =
20051123T1735"><SPAN=20
lang=3DEN-GB><FONT size=3D2>11.<SPAN style=3D"mso-tab-count: 1"> =
</SPAN>The=20
communication service identifier shall be taken into account when =
selecting the=20
correct UE(s), if multiple UEs are registered for the same Public User=20
Identity(s).</FONT></SPAN></P></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff =
size=3D2>&nbsp;Several of=20
the other requirements are in my view totally incompatible with =
Accept-Contact=20
being the transport for the identifier and indeed may limit the use of=20
Accept-Contact for its intended purpose. This includes charging and =
policy=20
control:</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007>
<P class=3DB1=20
style=3D"MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: 23228CR0542R1 =
20051123T1735"><SPAN=20
lang=3DEN-GB><FONT size=3D2>8.<SPAN style=3D"mso-tab-count: =
1">&nbsp;&nbsp;=20
</SPAN>Based on operator policy S-CSCF and AS shall consider the IMS=20
communication service identifier for online and offline charging, e.g. =
put=20
appropriate data into call detailed records.</FONT></SPAN></P>
<P class=3DB1=20
style=3D"MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: 23228CR0542R1 =
20051123T1735"><SPAN=20
lang=3DEN-GB><FONT size=3D2>9.<SPAN style=3D"mso-tab-count: =
1">&nbsp;&nbsp; </SPAN>The=20
communication service identifier shall be capable of being an input into =
the=20
policy control and charging rules.</FONT></SPAN></P></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2>Also:</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007>
<P class=3DB1=20
style=3D"MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: 23228CR0542R1 =
20051123T1735"><SPAN=20
lang=3DEN-GB><FONT size=3D2>2.<SPAN style=3D"mso-tab-count: =
1">&nbsp;&nbsp;=20
</SPAN>Based on operator policy the S-CSCF or an AS shall be able to =
validate an=20
IMS communication service identifier in a SIP request. This includes =
e.g. to=20
check the syntactical correctness of a service identifier, and policing =
the=20
usage of a communication service =
identifier.</FONT></SPAN></SPAN></P></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007>
<P class=3DB1=20
style=3D"MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: 23228CR0542R1 =
20051123T1735"><SPAN=20
lang=3DEN-GB><FONT size=3D2>10.<SPAN style=3D"mso-tab-count: 1"> =
</SPAN>It shall be=20
possible to use the IMS communication service identifier as a means to =
authorise=20
whether a subscriber is allowed to initiate or receive request for a=20
communication service.</FONT></SPAN></P></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff =
size=3D2>Another problem=20
with UAs placing feature tags in Accept-Contact headers as Service =
Identifiers=20
is&nbsp;how does the initiator obtains the service identifier of the =
destination=20
party for the related service? The concept only works in the scenario =
where all=20
networks and devices have a common understanding of the same services =
and the=20
identifiers that identify them. As far as I am aware SIP doesnt provide =
a=20
mechanism to discover proprietary non standard feature tags that are =
used by=20
other devices in other domains and map them to services that they relate =
to.=20
This means that if Accept-Contact is used the communication services are =
closed=20
and can only be used by those who are part of the closed network, =
(unless of=20
course IETF were to get into standardising services which I think is =
highly=20
unlikely). If URIs are used as the container for the Service Identifier =
these=20
can be obtained using Presence as proposed in Jonathan's=20
draft.</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff =
size=3D2>This helps fulfil=20
the following requirements in 3GPP TS 23.228:</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007>
<P class=3DB1 style=3D"MARGIN: 0in 0in 9pt 28.4pt"><SPAN =
lang=3DEN-GB><FONT=20
size=3D2>12.<SPAN style=3D"mso-tab-count: 1"> </SPAN>The usage of =
communication=20
service identifiers shall not adversely affect interoperability between =
IMS=20
networks and interoperability with external SIP networks and CS =
networks. The=20
behaviour of a network receiving the IMS requests without an IMS =
communication=20
service identifier is a matter of operator policy. Usage of =
communication=20
service identifiers shall not decrease the level of interoperability =
with=20
networks and UEs that are unaware of the communication service=20
identifier.</FONT></SPAN></P>
<P class=3DB1=20
style=3D"MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: '23.228CR0567' =
20060309T1718"><SPAN=20
lang=3DEN-GB><FONT size=3D2>14.<SPAN style=3D"mso-tab-count: 1"> =
</SPAN>The usage of=20
communication service identifiers shall not restrict the inherent =
capabilities=20
of SIP.</FONT></SPAN></P></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff =
size=3D2>SDP is not good=20
for routing decisions for service identification. Firstly SDP is =
intended to be=20
end to end only. It may be encrypted and understanding&nbsp;the details =
of the=20
service may require service or application level understanding of the =
SDP and=20
the service which is knowledge that proxies should not need to have. =
Also I dont=20
think that SDP can always uniquely identify a service in any case. In =
addition=20
not all requests that are related to a service will contain SDP in any=20
case.</FONT></SPAN></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><SPAN=20
class=3D778002917-16022007><FONT face=3DArial color=3D#0000ff=20
size=3D2>Andrew</FONT></SPAN></DIV></FONT></DIV><BR>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr align=3Dleft>
<HR tabIndex=3D-1>
<FONT face=3DTahoma size=3D2><B>From:</B> Elwell, John=20
[mailto:john.elwell@siemens.com] <BR><B>Sent:</B> Friday, February 16, =
2007 3:12=20
AM<BR><B>To:</B> Andrew Allen; Malleswara Rao Ankem<BR><B>Cc:</B> Andrew =
Allen;=20
sipping@ietf.org<BR><B>Subject:</B> RE: Device architecture? WAS: RE: =
[Sipping]=20
Service Identificatio ninSIP<BR></FONT><BR></DIV>
<DIV></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>Andrew,</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>Well, an AS is just another contact (perhaps a =
permanently=20
registered one, rather than one that uses SIP REGISTER), isn't it? So =
caller=20
preferences can be used to choose an AS or any other contact that =
supports the=20
desired capability.</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>I agree caller preferences&nbsp;is not a =
suitable basis for=20
charging. Just because I say I have a preference for a contact that =
supports=20
video, doesn't mean that I will necessarily use video during that call. =
That=20
might be determined during some mid-call session-re-negotiation. Surely =
charging=20
has to be based on what is actually used, not what might be used, and =
you can't=20
always determine this at INVITE time.</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>So I am saying use caller preferences to select =
a preferred=20
contact, use SDP to negotiate and re-negotiate the session with the =
contact that=20
you end up with. I haven't heard any compelling arguments for anything =
else so=20
far.</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D700265807-16022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>John</FONT></SPAN></DIV><BR>
<BLOCKQUOTE=20
style=3D"PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #0000ff 2px =
solid; MARGIN-RIGHT: 0px">
  <DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr align=3Dleft>
  <HR tabIndex=3D-1>
  <FONT face=3DTahoma size=3D2><B>From:</B> Andrew Allen =
[mailto:aallen@rim.com]=20
  <BR><B>Sent:</B> 16 February 2007 03:51<BR><B>To:</B> Elwell, John; =
Malleswara=20
  Rao Ankem<BR><B>Cc:</B> sipping@ietf.org; Andrew =
Allen<BR><B>Subject:</B> RE:=20
  Device architecture? WAS: RE: [Sipping] Service Identificatio n=20
  inSIP<BR></FONT><BR></DIV>
  <DIV></DIV>
  <DIV dir=3Dltr align=3Dleft><FONT face=3DArial color=3D#0000ff=20
  size=3D2></FONT>&nbsp;</DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2>John,</FONT></SPAN></DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2>If Service Identification was only about =
selecting the=20
  appropriate contacts from those registered agaisnt the AOR then RFC =
3840/3841=20
  would be fine however Service Identification is much more than that =
and may be=20
  used to route requests to application servers for service logic =
invocation,=20
  and used as the basis for charging. Accept-Contact is basically just a =
filter=20
  expression to filter a set of registered contacts. Including a =
parameter in a=20
  filter expression and using that as the basis for charging,=20
  authorisation&nbsp;and service invocation is going to cause=20
  problems.</FONT></SPAN></DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN class=3D927204303-16022007><FONT =
face=3DArial=20
  color=3D#0000ff size=3D2>Andrew</FONT></SPAN></DIV>
  <DIV dir=3Dltr align=3Dleft><SPAN =
class=3D927204303-16022007></SPAN>&nbsp;</DIV>
  <DIV dir=3Dltr align=3Dleft>
  <HR tabIndex=3D-1>
  </DIV></BLOCKQUOTE><p><hr>This transmission (including any =
attachments) may contain confidential information, privileged material =
(including material protected by the solicitor-client or other =
applicable privileges), or constitute non-public information. Any use of =
this information by anyone other than the intended recipient is =
prohibited. If you have received this transmission in error, please =
immediately reply to the sender and delete this information from your =
system. Use, dissemination, distribution, or reproduction of this =
transmission by unintended recipients is not authorized and may be =
unlawful.</BODY></HTML>

------_=_NextPart_001_01C75212.FABA4493--


--===============1717018118==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1717018118==--




From sipping-bounces@ietf.org Fri Feb 16 18:53:09 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HICrQ-0005TS-2Q; Fri, 16 Feb 2007 18:51:28 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HICrO-0005TF-Ug
	for sipping@ietf.org; Fri, 16 Feb 2007 18:51:26 -0500
Received: from c243-111.rim.net ([216.9.243.111] helo=MHS01YKF.rim.net)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HICrO-0005qz-BB
	for sipping@ietf.org; Fri, 16 Feb 2007 18:51:26 -0500
Importance: normal
Priority: normal
Received: from ngw02ykf.rim.net ([10.102.108.31]) by MHS01YKF.rim.net with
	Microsoft SMTPSVC(6.0.3790.211); Fri, 16 Feb 2007 18:51:15 -0500
Received: from XCH20YKF.rim.net ([10.102.100.35]) by ngw02ykf.rim.net (SAVSMTP
	3.1.6.45) with SMTP id M2007021618511517479 ;
	Fri, 16 Feb 2007 18:51:15 -0500
Received: from XCH47YKF.rim.net ([10.64.31.217]) by XCH20YKF.rim.net with
	Microsoft SMTPSVC(5.0.2195.6713); Fri, 16 Feb 2007 18:51:15 -0500
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.181
Content-Class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Fri, 16 Feb 2007 18:51:14 -0500
Message-ID: <61968779B8AC4C4BAB421D4C12F008C0091516B6@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
thread-index: AcdNmH89O4T46iGoRjicZ3hYknVpawB8E6oQAH++leAABzLk8AAYIixw
From: "Andrew Allen" <aallen@rim.com>
To: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>,
	<sipping@ietf.org>
X-OriginalArrivalTime: 16 Feb 2007 23:51:15.0727 (UTC)
	FILETIME=[589C89F0:01C75225]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 96e0f8497f38c15fbfc8f6f315bcdecb
Cc: Andrew Allen <aallen@rim.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Hi Steve,

My responses inline

Andrew=20

-----Original Message-----
From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]=20
Sent: Friday, February 16, 2007 3:51 AM
To: sipping@ietf.org
Subject: RE: [Sipping] Service Identification in SIP

Hello Andrew,

Thanks for your comments.  I have a couple of brief responses below.

Best Regards,

..//Steve=20

-----Original Message-----
From: Andrew Allen [mailto:aallen@rim.com]
Sent: Friday, February 16, 2007 7:15 AM
To: Stephen Terrill (MI/EEM); sipping@ietf.org
Cc: Andrew Allen
Subject: RE: [Sipping] Service Identification in SIP


Some comments on Stephen's comments on the draft:
=20

-Moving onto the proposed solution draft itself, in addition for
providing for the identification of the "application" and the
"communication service", I have a few additional thoughts.

The draft seems to have the implication that all services (communication
services in other terminology) that are to be identified have to be
standardised within the IETF.  This is clearly an restriction that will
not be acceptable to the industry as a whole.  My guess is that if such
a statement is mandated by the IETF, it will just be words on paper
which would likely be ignored in practice by other standardisation
organisations.

[AA] This I think is an incorrect interpretation of the draft. My
understanding is that the only services according to the draft that
require some standardisation in IETF are those that by their very nature
cannot interoperate unless there is some common additional understanding
between the endpoints outside of what is currently specified in SIP. The
use of SIP option tags with Require header is proposed by the draft for
this purpose as such services create interoperability issues with SIP.

<Steve>
My comment applies even for these services
</Steve>


<AA>=20
Well then we are in big trouble because what you are proposing is that
its OK for bodies outside of IETF to create non interoperable extensions
for the behavour of SIP entities. This will result in the fragmentation
of SIP into different flavors that don't interoperate. This is the
balkanization of SIP that Paul talks about.even
<AA>=20

----------------------------------------------------

The draft also only recognises the case where all explicitly knowledge
of the service is required.  I think it would also be good to allow the
case where knowledge of the service would be helpful, but not explicitly
required.  This could be a case where the service to the end-user would
not be degraded if this was not supported.  This functionality seems to
be missing.

[AA] Service URIs are defined in the draft for these services where the
support by the other endpoint is not required for the session to be
sucessfully established.

The draft seems to indicate that that application servers in the network
are involved only at the request of the originating terminal.  This
seems a strange restriction.  For some network support that supports the
originating user, such as the one provided in the example "I would like
this call recorded for later retrieval", such mechanisms may make sense,
however as a general approach it seems a bit limited.  For example, in
3GPP networks there is the need to invoke application logic in the
originating side and terminating side, not only at the request of the
originating terminal, but as a consequence of the requested service.  If
there is an SIP message contains indication of the requested service, a
proxy can do this (or loose route to an AS that contains the necessary
service logic).=20

[AA] The draft supports both possibilities using the Route header for
originating network servers that need to be involved in the route and
based on the terminating networks routing policy for the Service URI in
the Request URI.

<Steve>
I don't see the advantage of having different solutions for the
originating network and the terminating network.  I don't think that the
originating logic is invoked only at the request of the originating
terminal.
</Steve>

<AA>=20
Unless IETF is willing to standardise all services and mandate that all
SIP entities understand these (which will never happen) then we cannot
assume that all networks and all endpoints understand and use the same
identifiers. It is possible through configuration for an endpoint to
obtain the identifiers used by its own service provider but there is no
way to ensure that other networks use or understand the same identifiers
for the equivalent or similar service that is compatible at the session
layer. Therefore we need to support the possibility to have different
identifiers in different networks in order to not violate the TS 23.228
requirement:

12.	The usage of communication service identifiers shall not
adversely affect interoperability between IMS networks and
interoperability with external SIP networks and CS networks. The
behaviour of a network receiving the IMS requests without an IMS
communication service identifier is a matter of operator policy. Usage
of communication service identifiers shall not decrease the level of
interoperability with networks and UEs that are unaware of the
communication service identifier.

If the originating service logic isn't invoked based on an identifier
placed in the request by the originating terminal on what basis do you
think it is invoked? How does the use of Accept-Contact have advantages
in this case?
<AA>=20

--------------------------------------

 I note here that the proposed service URN does not contain the address
of a specific server containing the application logic.  An AS can look
at that, and loose route to an application server that does contain such
logic (If it is separated from the proxy), however I fail to see the
difference for that proxy if it looks at a feature tag, or it looks at
the service URN for this.  In both cases, it has to parse the message
for the data, and modify the routing information. =20

[AA]Route headers are the SIP mechanism for requesting intermediaries to
be includede on the route of the request. There is nothing in RFC
3840/RFC 3841 that addresses using feature tags in Accept-Contact header
for any routing decisions prior to determining which of the registered
contacts the request should be forked to. Accept-Contact is not a Route
header it is merely a filter that is applied to the set of valid
contacts. Using a parameter in the filter as a service identifier has a
potential significant impact on the use of Accept-Contact for its
original purpose.=20

<Steve>
I don't believe there is anthing that states a Proxy CAN NOT route to
applications or application logic based upon elements in the SIP
message, including what is in the Accept-Contact header (some people
have even proposed SDP ...).
</Steve>

<AA>=20
As I had stated in an earlier email on this subject a proxy can do a lot
of things (including routing based on your favorite number appearing in
a call-Id) but that doesn't mean they are a good idea or a good design
decision. If you want to ensure you don't break things then you need to
be careful to use SIP headers with the semantics intended for them
otherwise problems can result if extra semantics are put upon existing
headers.
<AA>=20

Further, the route header could be removed in the originating network;
and it is still required to identify the service in the terminating
network.  As proxies may remove the route, it can't be used to transport
the requested service. =20

[AA] Proxies arbitarily removing Route headers is a BAD thing akin to a
bull in a china shop - almost certainly going to break something. RFC
3261 summarises what proxies should be doing with Route headers:
16.12 Summary of Proxy Route Processing

   In the absence of local policy to the contrary, the processing a
   proxy performs on a request containing a Route header field can be
   summarized in the following steps.

      1.  The proxy will inspect the Request-URI.  If it indicates a
          resource owned by this proxy, the proxy will replace it with
          the results of running a location service.  Otherwise, the
          proxy will not change the Request-URI.

      2.  The proxy will inspect the URI in the topmost Route header
          field value.  If it indicates this proxy, the proxy removes it
          from the Route header field (this route node has been
          reached).

      3.  The proxy will forward the request to the resource indicated
          by the URI in the topmost Route header field value or in the
          Request-URI if no Route header field is present.  The proxy
          determines the address, port and transport to use when
          forwarding the request by applying the procedures in [4] to
          that URI.

   If no strict-routing elements are encountered on the path of the
   request, the Request-URI will always indicate the target of the
   request.

<Steve>
The first sentance is interesting "In absence of local policy to the
contrary", indicating that there can be different policies.  I
understand that some proxies do remove this.  While I am on this - I
have lead to believe that R-URI parameters may also be removed by
intermediate proxies - is this correct?
</Steve>

<AA>=20
My understanding is that local policy may determine that additional
nodes need to be routed to in addition to the Route contained in the
request. However deleting existing Route headers other than the Route
header containing the address of the proxy as part of the loose routing
procedures is liable to break things and is not part of RFC 3261 routing
(loose routing).
<AA>=20


[AA] Generally a UA wouldn't know the appropriate service identifiers
used in the terminating network or whether a Route header is appropriate
to be used in the terminating network. The draft defines the Service URI
for the purpose of identifying services in the terminating network and
service URIs can be obtained using Presence.

<Steve>
I don't get the above paragraph.  It suggest that the "appropriate
service identifier" used in the terminating network is not known by the
originating UE, but the Service URI is.  I thought that the service URI
is an encoding of the "appropriate service identifier".
</Steve>

<AA>=20
Yes but in the general case the UA doesn't need to know what the service
identifier is it obtains through Presence a Service URI that contains
the Service identifier as a URI parameter. The great thing about using
Presence to obtain a Service URI is that the originating UA doesn't even
need to know anything about service identifiers at all for this to work.
It just uses Presence to find the URI to use for the service which can
be indicated to the user via an indication such as a graphic icon (e,g a
phone for telephony, smiley face for IM, etc) in the presence document.
<AA>=20

-----------------------------------

>From this perspective it is not clear to me the advantage of using a
service URN in a route header compared to the use of a feature tag.

[AA] I hope the above responses clarify that somewhat. Also the Service
URN and Service URIs allow different service identifiers to be used in
the originating and terminating network for session compatiable services
instead of forcing everyone to deploy and understand the same service
and service identifier.

<Steve>
Nope, don't get it, but are information that is carried end-to-end ....
And as far as I see, could be used by intermediate proxies to impact
local routing policy ... Or ?
</Steve>

<AA>=20
I don't understand this response either. The key point is that UAs in
corporate networks and in the internet are not going to deploy 3GPP
flavor services and use the same 3GPP service identifiers or even
possibly any service identifiers at all. The Accept-Contact based
proposal assumes that all endpoints understand the service identifiers
to be used for each service. This concept is not generally applicable
and create interoperability problems.
<AA>=20

The draft seems to mandate GRUU - is this really so, is that really
necessary.  It seems strange to place such a dependency, it also starts
to move the GRUU as being the primary means of which to contact a user.

[AA] My understanding is that GRUU in the draft is used to address how a
request addressed to Service URI is forked only to those contacts that
support the service identified by the that Service URI. A contact would
seem to be fine so long as it had GRUU properties - no need to obtain a
GRUU if you already have one.

This implies that the originating user decides which terminal I should
receive my requested service on, even if I have two devices for that.
As such, I don't believe that there should be a dependency on GRUU.

[AA] If you have obtained a Service URI (probably through presence) for
that user then you can address the request to the Service URI which
represents all the terminals that support that service. However GRUU and
the Service Instance URIs created from GRUUs provide another mechanism
through which a user can obtain a URI to reach that user for the
required service.

Parts of the draft seems to rely on the use of presence.  While presence
could be a supportive mechanism, I doubt that we can mandate that the
number you want to dial has be available in presence as I assume not all
entities are available in presence (diallable network services, users in
the PSTN).  This would be strange to mandated.

[AA] PSTN gateways are presumably in the same domain as the terminating
proxy and therefore these PSTN gateways can be configured with the
appropriate service identifiers for telephony services and can include
this in either a Route header to route to the right application server
or can construct a valid Service URI. The real problem is how do users
in other networks know what the service identifier that's is used in the
terminating network is. The use of Service URIs and Presence provides a
solution to that problem.

<Steve>
Again, presence could be a useful mechanism, but I think it would be a
problem for the draft to rely it.  A user making a call to the PSTN may
not find that user in the presence archicture (and similar to services
like the 800 or 900 services).  (I should note, that you shouldn't make
any assumptions about the location of the PSTN gateway ... But I don't
think that changes anything). =20
</Steve>

<AA>=20
How does a PSTN gateway in another domain from the terminating network
know the service identifier for telephony used in the terminating
network with the Accept-Contact header based proposal?=20
<AA>=20

Moving on from the draft I have seen the questioned raised about "why
can't we expand the power of SDP".  Such an approach would mandate that
SIP methods use SDP, I don't think that is appropriate.=20

[AA] I agree with you here

I hear the appeals that the mechanisms should not be abused, and that we
should not have 1000s of separate "communication services", and I fully
agree with this. I see a smaller number of communication services, not a
larger number; where those communication services are used to access
other end-user-services.

*flame away*

Regards,

Steve

---



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential =
information, privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sat Feb 17 12:02:01 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HISuW-0005nq-3S; Sat, 17 Feb 2007 11:59:44 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HISuV-0005nl-6B
	for sipping@ietf.org; Sat, 17 Feb 2007 11:59:43 -0500
Received: from sj-iport-6.cisco.com ([171.71.176.117])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HISuQ-0007nT-5k
	for sipping@ietf.org; Sat, 17 Feb 2007 11:59:43 -0500
Received: from sj-dkim-2.cisco.com ([171.71.179.186])
	by sj-iport-6.cisco.com with ESMTP; 17 Feb 2007 08:59:33 -0800
X-IronPort-AV: i="4.14,185,1170662400"; 
	d="scan'208"; a="113885900:sNHT64387161"
Received: from sj-core-4.cisco.com (sj-core-4.cisco.com [171.68.223.138])
	by sj-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1HGxVM9016066; 
	Sat, 17 Feb 2007 08:59:31 -0800
Received: from xbh-sjc-221.amer.cisco.com (xbh-sjc-221.cisco.com
	[128.107.191.63])
	by sj-core-4.cisco.com (8.12.10/8.12.6) with ESMTP id l1HGxKnN003410;
	Sat, 17 Feb 2007 08:59:31 -0800 (PST)
Received: from xfe-sjc-211.amer.cisco.com ([171.70.151.174]) by
	xbh-sjc-221.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Sat, 17 Feb 2007 08:59:29 -0800
Received: from [192.168.1.113] ([10.21.96.33]) by xfe-sjc-211.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Sat, 17 Feb 2007 08:59:28 -0800
Message-ID: <45D65EE8.7010501@cisco.com>
Date: Fri, 16 Feb 2007 20:48:24 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: Henning Schulzrinne <hgs@cs.columbia.edu>
Subject: Re: [Sipping] Service Identification in SIP
References: <45C0AEF2.8010101@ericsson.com>	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>	<1170835642.29763.54.camel@n95.nomadiclab.com>	<532B18E13CF9E64380EF5FDDE265E071B61978@blr-m2-msg.wipro.com>	<1171033470.3464.115.camel@localhost.localdomain>
	<FE059A45-A940-4CAE-9B9C-19EFE99F1D66@cs.columbia.edu>
In-Reply-To: <FE059A45-A940-4CAE-9B9C-19EFE99F1D66@cs.columbia.edu>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 17 Feb 2007 16:59:28.0932 (UTC)
	FILETIME=[FCA08E40:01C752B4]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=14042; t=1171731571;
	x=1172595571; c=relaxed/simple; s=sjdkim2002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20; bh=t//rbIWtbst5qE+lt99IfLeNfdzxARsAjL4NSbwqGfE=;
	b=1K9cWBqgxjl+IQjNCxVefiIAPwvq+5+XeRVEYCeuk7aTUCN7GjMtIRO01G3svK1bQWm52pIa
	7f6R+tq9PH5zc6Otno7njRtYj3uxNqeQldaiskhLolnXllz7GJIb7bxX;
Authentication-Results: sj-dkim-2; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim2002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 7698d1420ecbbce1995432e99bb6d1a1
Cc: sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I think its a fair point that the example is contrived. Really, the main 
use case being targeted is IMS PoC. So, let me talk a bit about that.

One would think that PoC is just a regular SIP call that has floor 
control added to it as another media type, and so one could indicate 
that a call is a PoC call just because of the presence of this new media 
type in the SDP. Sadly, I wish that were true. If it was, we would not 
need a mechanism here.

Unfortunately, PoC has layered a whole bunch of additional behaviors 
into the way SIP requests are handled. Dale has indicated one of them - 
that a 200 OK is returned on the terminating side priod to the call 
actually being completed. PoC also *requires* the insertion of servers 
into the SIP signaling path to make the system work. Consequently, 
though the messages that PoC uses *look* like SIP messages with a new 
media type, there is a hidden 'extension' in place that changes the 
fundamental semantics of INVITE. The fact that this extension is not 
visible as a specific header in a well defined way does not change the 
fact that PoC CHANGES THE BEHAVIOR OF SIP AND IS THUS A SIP EXTENSION.

To convince yourself of this, consider what would happen if a PoC client 
calls a regular SIP phone. The regular SIP phone would see the PoC floor 
control stream and reject it as unknown. If PoC were just SIP as we've 
specified it, then everything would be fine, and the call would proceed 
but without floor control.

But alas, it is not so.

Because PoC makes all kinds of assumptions about the behaviors of its 
participants, and what the real meaning is for various SIP messages, the 
entire thing just won't work when a call terminates on a non-PoC device. 
This is because PoC is, effectively, a fairly significant and 
non-backwards-compatible extension to SIP.

So, my proposal is, if you are going to go and DO something like that, 
you better use the SIP techniques that we've defined to deal with it. 
And when you extend SIP in a way that doesn't work with a SIP endpoint 
that doesn't understand an extension, you use a Require header. And that 
is exactly what I'm proposing that 3gpp should do. Use Require because 
you have made a non-backwards-compatible extension to SIP.

The general problem here is that there are other places where people are 
layering additional 'context', as Loreto has called it, ontop of SIP 
signaling. This is a worrisome trend because this context is subtle and 
not explicitly signaled. Its just assumed. And that is dangerous, since 
it leads to the kind of balkanization and non-interoperability that 
myself, Paul, Dale and others have complained about in this thread.

The current 3gpp specs, instead of using Require to signal this 
extension, put a tag into the Accept-Contact header field. I find this 
extremely worrisome, because it is really just masking the reality of 
what has happened. If you've got a situration where a SIP call *cannot 
be processed* without going through application servers, then this is 
not something that is signaled with Accept-Contact. Its signaled with 
Require or Proxy-Require as appropriate.

Thanks,
-Jonathan R.

Henning Schulzrinne wrote:

> I have to admit that I find a discussion that uses a single example a  
> bit peculiar, particularly since the example is rather contrived.  (Why 
> would you use voice, plus speech-to-text presumably, to talk to  a chess 
> application?)
> 
> In a networking sense, the issue seems to be that we're using voice  as 
> a transport protocol to carry different distinct applications. (In  the 
> old days, a modem would have been such an application.)
> 
> I'll repeat what I've said before: SIP doesn't exactly suffer from a  
> lack of complexity and mechanisms. At this point, before adding more  
> machinery, we might want to ask as to whether there is truly a broad- 
> based need for justifying additional machinery.
> 
> With regards to service URNs, I happen to believe that they are very  
> reasonable as "modern" replacements for tel URIs, essentially  extending 
> the concept of 800 numbers. But that's a different problem  from what I 
> understand this discussion to be about. (In that usage,  service URNs 
> would be mapped to SIP AORs based on geography, time-of- day and other 
> factors.)
> 
> Henning
> 
> 
> On Feb 9, 2007, at 10:04 AM, Salvatore Loreto (JO/LMF) wrote:
> 
>> Hi Ravi,
>>
>> if you consider the Jonathan draft, he says that
>> "...chess game that utilizes voice commands to move pieces.  When  one 
>> player says "Queen to D7", the software on their phone  recognizes 
>> this and moves the piece.  At the same time, the user's  voice is sent 
>> to the other player, where it is both rendered to the  user, and 
>> interpreted locally in order to move the piece."
>> and he suggests that the following urn
>>
>>    urn:service:games:voice-chess
>>
>> could be used to identify the voice chess application.
>>
>> I guess that with voice he supposes audio, but nowadays we can also  
>> have voice using PoC. Then we need a way to specify which kind of  
>> voice we want: a standard full-duplex or a PoC. Maybe the chess  game 
>> works well with, or it is designed for a standard audio but  not for a 
>> PoC audio. So we need to address also the voice  preference we want. I 
>> like your proposal, having something  hierarchical was something I was 
>> thinking too I don't know if it is  your idea, but with the service 
>> urn you are proposing:  Urn:service:voice.poc.chess I imaging you are 
>> asking to play a  Chess game, that work utilizing voice commands, but 
>> specifically  voice over PoC. This is also compliant to the 
>> draft-ietf-ecrit- service-urn (see section 3) "Declaration of 
>> syntactic structure:  The URN consists of a hierarchical service 
>> identifier, with a  sequence of labels separated by periods.  The 
>> left-most label is  the most significant one and is called 'top-level 
>> service', while  names to the right are called 'sub-services'.  The 
>> set of allowable  characters is the same as that for domain names [1] 
>> and a subset of  the labels allowed in [5].  Labels are 
>> case-insensitive and MUST be  specified in all lower-case.  For any 
>> given service URN, service- identifiers can be removed right-to-left 
>> and the resulting URN is  still valid, referring a more generic 
>> service.  In other words, if  a service 'x.y.z' exists, the URNs 'x' 
>> and 'x.y' are also valid  service URNs." Regards sal
>> On Fri, 2007-02-09 at 13:44 +0530, ravishankar.shiroor@wipro.com  wrote:
>>
>>> Hi Salvatore, As an end user, poc is also an application that I  use 
>>> and chess is also an application that I can use. The way I see  it, 
>>> in your example, it will be running chess application on top  of PoC 
>>> application. So the service urn for this would be something  like: 
>>> Urn:service:voice.poc.chess Regards, Ravi. -- Ravishankar.  G. 
>>> Shiroor Wipro Technologies, Bangalore.  ravishankar.shiroor@wipro.com 
>>> -- > -----Original Message----- >  From: Salvatore Loreto (JO/LMF) >  
>>> [mailto:salvatore.loreto@ericsson.com] > Sent: Wednesday, February  
>>> 07, 2007 1:37 PM > To: RaviShankar Shiroor (WT01 - Product  
>>> Engineering Solutions) > Cc: Gonzalo.Camarillo@ericsson.com;  
>>> sipping@ietf.org; > mary.barnes@nortel.com > Subject: RE:  [Sipping] 
>>> Service Identification in SIP > > hi Ravi, > > in IMS a  Service is a 
>>> "type of communication defined by a > service  definition that 
>>> specifies the rules and procedures > and allowed  media, if any, for 
>>> a specific type of > communication and that  utilises the IMS 
>>> enabler. An IMS > communication service defines  what is possible 
>>> within a > single SIP session or standalone  transaction." > So a 
>>> Service implies some combination of  capabilities and > ways to use 
>>> those capabilities. > If you want  the Service specifies the context 
>>> for a communication. > > Instead  the Application define the specific 
>>> communication (or > if you  want application) that you want run in 
>>> the specific > context. So  you should have several different 
>>> applications > designed to run  on a specific environment or that 
>>> have > different performance or  behavior in different context > 
>>> (running on top of different  service). > > Lets go to Jonathan's 
>>> example of the Chess. > In his  example the Chess is a service, for 
>>> IMS instead is an >  application that you can run on top of different 
>>> service. > For  example you can run Chess application on top of PoC > 
>>> service or  on top of the "standard" service. So let say you > run 
>>> Chess on  top of PoC service, then you can use only voice > in the 
>>> way that  PoC specifies. > > /sal > > > > > On Thu, 2007-02-01 at 
>>> 16:39  +0530, > ravishankar.shiroor@wipro.com wrote: > > Hello, > > > 
>>> >  IMO, Service Identification is definitely needed and should > be  
>>> taken > > up as a WG as proposed below. > > > > Further, I think  
>>> Service Identification scope should be as per the > > requirement  in 
>>> Jonathan's draft, as I still do not > understand the need > >  for 
>>> having a Communication Service Identifier and another >  Application 
>>> > > Identifier as given in the 3gpp-ics-requirements  document. > > > 
>>> > Regards, > > Ravi. > > > > -- > > > >  Ravishankar. G. Shiroor > > 
>>> Wipro Technologies, Bangalore. > > > >  ravishankar.shiroor@wipro.com 
>>> > > -- > > > > > > > -----Original  Message----- > > > From: Gonzalo 
>>> Camarillo  [mailto:Gonzalo.Camarillo@ericsson.com] > > > Sent: 
>>> Wednesday,  January 31, 2007 8:30 PM > > > To: sipping > > > Cc: Mary 
>>> Barnes >  > > Subject: [Sipping] Service Identification in SIP > > > 
>>> > > >  Folks, > > > > > > during the last year, there have been a 
>>> number  of threads on the > > > mailing list discussing service  
>>> identification in SIP. In > > > particular, the following thread  
>>> discussed 3GPP > requirements on this > > > area: > > > > > > >  
>>> http://www1.ietf.org/mail-archive/web/sipping/current/ msg10562.html 
>>> > > > > > > At a later point, those 3GPP  requirements were 
>>> documented in the > > > following (expired by  now) Internet Draft, 
>>> which was > presented in a > > > face-to-face  IETF meeting. > > > > 
>>> > > http://www.watersprings.org/pub/id/draft- loreto-sipping-3gpp-i > 
>>> > > cs-requirements-00.txt > > > > > >  Lately, the following draft 
>>> has been submitted to the archives: >  > > > > > 
>>> http://www.ietf.org/internet-drafts/draft-rosenberg- sipping-se > > > 
>>> rvice-identification-00.txt > > > > > > This  draft proposes a 
>>> solution that intends to meet known > > >  requirements in this area. 
>>> > > > > > > Given the amount of  discussions on issues related to 
>>> service > > > identification we  have seen during the last year, we > 
>>> believe this is > > > a  fairly important topic for the SIPPING WG to 
>>> tackle. At some > > >  point, we would like to have a milestone with 
>>> its > associated WG  item > > > dealing with this issue. > > > > > > 
>>> Therefore, we  would like to encourage people to read this > draft 
>>> and > > > send  comments to the list. Additionally, if you believe 
>>> that this > > >  work is (or is not) important, we would like to hear 
>>> that as well.  > > > > > > Cheers, > > > > > > Gonzalo > > > SIPPING 
>>> co-chair > >  > > > > > > > 
>>> _______________________________________________ > >  > Sipping 
>>> mailing list > https://www1.ietf.org/mailman/listinfo/ sipping > > > 
>>> This list is for NEW development of the application  of SIP Use  > > 
>>> > sip-implementors@cs.columbia.edu for questions  on current sip Use 
>>> > > > sip@ietf.org for new developments of core  SIP > > > > > > > 
>>> _______________________________________________  > > Sipping mailing 
>>> list https://www1.ietf.org/mailman/listinfo/ sipping > > This list is 
>>> for NEW development of the application of  SIP Use > > 
>>> sip-implementors@cs.columbia.edu for questions on  current sip Use > 
>>> > sip@ietf.org for new developments of core SIP  > > The information 
>>> contained in this electronic message and any  attachments to this 
>>> message are intended for the exclusive use of  the addressee(s) and 
>>> may contain proprietary, confidential or  privileged information. If 
>>> you are not the intended recipient, you  should not disseminate, 
>>> distribute or copy this e-mail. Please  notify the sender immediately 
>>> and destroy all copies of this  message and any attachments. WARNING: 
>>> Computer viruses can be  transmitted via email. The recipient should 
>>> check this email and  any attachments for the presence of viruses. 
>>> The company accepts  no liability for any damage caused by any virus 
>>> transmitted by  this email. www.wipro.com
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP
>> Use sip-implementors@cs.columbia.edu for questions on current sip
>> Use sip@ietf.org for new developments of core SIP
> 
> 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sat Feb 17 12:27:03 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HITKi-0003yC-DQ; Sat, 17 Feb 2007 12:26:48 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HITKg-0003xn-MP
	for sipping@ietf.org; Sat, 17 Feb 2007 12:26:46 -0500
Received: from serrano.cc.columbia.edu ([128.59.29.6])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HITKe-0003rS-Ck
	for sipping@ietf.org; Sat, 17 Feb 2007 12:26:46 -0500
Received: from [192.168.0.41] (pool-141-153-194-152.mad.east.verizon.net
	[141.153.194.152]) (user=hgs10 mech=PLAIN bits=0)
	by serrano.cc.columbia.edu (8.13.7/8.13.6) with ESMTP id l1HHQcqH001428
	(version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NOT);
	Sat, 17 Feb 2007 12:26:38 -0500 (EST)
In-Reply-To: <45D65EE8.7010501@cisco.com>
References: <45C0AEF2.8010101@ericsson.com>	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>	<1170835642.29763.54.camel@n95.nomadiclab.com>	<532B18E13CF9E64380EF5FDDE265E071B61978@blr-m2-msg.wipro.com>	<1171033470.3464.115.camel@localhost.localdomain>
	<FE059A45-A940-4CAE-9B9C-19EFE99F1D66@cs.columbia.edu>
	<45D65EE8.7010501@cisco.com>
Mime-Version: 1.0 (Apple Message framework v752.2)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <BF1308AC-4B8B-4E43-B9E3-78CA178EE3C0@cs.columbia.edu>
Content-Transfer-Encoding: 7bit
From: Henning Schulzrinne <hgs@cs.columbia.edu>
Subject: Re: [Sipping] Service Identification in SIP
Date: Sat, 17 Feb 2007 12:26:34 -0500
To: Jonathan Rosenberg <jdrosen@cisco.com>
X-Mailer: Apple Mail (2.752.2)
X-No-Spam-Score: Local
X-Scanned-By: MIMEDefang 2.48 on 128.59.29.6
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 386e0819b1192672467565a524848168
Cc: sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I think we're in general agreement; in particular, Require seems like  
the right approach. (service URNs do not, in my opinion).

Stepping back, I don't think we should generalize this problem since  
it essentially "blesses" a particular design approach that is  
undesirable. The right approach, from a SIP perspective, for IMS PoC  
would have been to come up with a solution that is backwards- 
compatible. If that's not possible, getting "branded" with Require  
seems like the right penalty.

In general, we have probably two requirements:

(1) When contacting an AOR, a UAC that needs a special UAS to work  
properly needs to get routed to that UAS only. Callee caps and caller  
preferences seem to be the right tools for that, as it involves call  
routing and capability matching.

(2) If, despite doing (1), a SIP request accidentally reaches a UAS  
that can't do anything useful with it, Requires minimizes the damage  
(and allows a proxy to serially-fork to something else that might  
have a chance of working).

Henning

On Feb 16, 2007, at 8:48 PM, Jonathan Rosenberg wrote:

> I think its a fair point that the example is contrived. Really, the  
> main use case being targeted is IMS PoC. So, let me talk a bit  
> about that.
>
> One would think that PoC is just a regular SIP call that has floor  
> control added to it as another media type, and so one could  
> indicate that a call is a PoC call just because of the presence of  
> this new media type in the SDP. Sadly, I wish that were true. If it  
> was, we would not need a mechanism here.
>
> Unfortunately, PoC has layered a whole bunch of additional  
> behaviors into the way SIP requests are handled. Dale has indicated  
> one of them - that a 200 OK is returned on the terminating side  
> priod to the call actually being completed. PoC also *requires* the  
> insertion of servers into the SIP signaling path to make the system  
> work. Consequently, though the messages that PoC uses *look* like  
> SIP messages with a new media type, there is a hidden 'extension'  
> in place that changes the fundamental semantics of INVITE. The fact  
> that this extension is not visible as a specific header in a well  
> defined way does not change the fact that PoC CHANGES THE BEHAVIOR  
> OF SIP AND IS THUS A SIP EXTENSION.
>
> To convince yourself of this, consider what would happen if a PoC  
> client calls a regular SIP phone. The regular SIP phone would see  
> the PoC floor control stream and reject it as unknown. If PoC were  
> just SIP as we've specified it, then everything would be fine, and  
> the call would proceed but without floor control.
>
> But alas, it is not so.
>
> Because PoC makes all kinds of assumptions about the behaviors of  
> its participants, and what the real meaning is for various SIP  
> messages, the entire thing just won't work when a call terminates  
> on a non-PoC device. This is because PoC is, effectively, a fairly  
> significant and non-backwards-compatible extension to SIP.
>
> So, my proposal is, if you are going to go and DO something like  
> that, you better use the SIP techniques that we've defined to deal  
> with it. And when you extend SIP in a way that doesn't work with a  
> SIP endpoint that doesn't understand an extension, you use a  
> Require header. And that is exactly what I'm proposing that 3gpp  
> should do. Use Require because you have made a non-backwards- 
> compatible extension to SIP.
>
> The general problem here is that there are other places where  
> people are layering additional 'context', as Loreto has called it,  
> ontop of SIP signaling. This is a worrisome trend because this  
> context is subtle and not explicitly signaled. Its just assumed.  
> And that is dangerous, since it leads to the kind of balkanization  
> and non-interoperability that myself, Paul, Dale and others have  
> complained about in this thread.
>
> The current 3gpp specs, instead of using Require to signal this  
> extension, put a tag into the Accept-Contact header field. I find  
> this extremely worrisome, because it is really just masking the  
> reality of what has happened. If you've got a situration where a  
> SIP call *cannot be processed* without going through application  
> servers, then this is not something that is signaled with Accept- 
> Contact. Its signaled with Require or Proxy-Require as appropriate.
>
> Thanks,
> -Jonathan R.
>
> Henning Schulzrinne wrote:
>


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sat Feb 17 14:24:22 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HIV8y-0006rQ-FQ; Sat, 17 Feb 2007 14:22:48 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HIV8w-0006rK-NG
	for sipping@ietf.org; Sat, 17 Feb 2007 14:22:46 -0500
Received: from lohi.tutpro.com ([192.98.100.2] helo=tutpro.com)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HIV8r-00083f-92
	for sipping@ietf.org; Sat, 17 Feb 2007 14:22:46 -0500
Received: from localhost (localhost [127.0.0.1])
	by tutpro.com (Postfix) with ESMTP id DB92A1EC643;
	Sat, 17 Feb 2007 21:22:34 +0200 (EET)
Received: from tutpro.com ([127.0.0.1])
	by localhost (tutpro.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id eTaOJRHthpGB; Sat, 17 Feb 2007 21:22:19 +0200 (EET)
Received: from rautu (unknown [124.6.202.202])
	by tutpro.com (Postfix) with ESMTP;
	Sat, 17 Feb 2007 21:22:19 +0200 (EET)
Received: by rautu (Postfix, from userid 1000)
	id CEDA6F14B5; Sat, 17 Feb 2007 21:21:52 +0200 (EET)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <17879.21968.773179.929052@rautu.test.fi>
Date: Sat, 17 Feb 2007 21:21:52 +0200
From: Juha Heinanen <jh@tutpro.com>
To: Jonathan Rosenberg <jdrosen@cisco.com>
Subject: Re: [Sipping] Service Identification in SIP
In-Reply-To: <45D65EE8.7010501@cisco.com>
References: <45C0AEF2.8010101@ericsson.com>
	<532B18E13CF9E64380EF5FDDE265E071B26B1F@blr-m2-msg.wipro.com>
	<1170835642.29763.54.camel@n95.nomadiclab.com>
	<532B18E13CF9E64380EF5FDDE265E071B61978@blr-m2-msg.wipro.com>
	<1171033470.3464.115.camel@localhost.localdomain>
	<FE059A45-A940-4CAE-9B9C-19EFE99F1D66@cs.columbia.edu>
	<45D65EE8.7010501@cisco.com>
X-Mailer: VM 7.19 under Emacs 21.4.1
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 7a6398bf8aaeabc7a7bb696b6b0a2aad
Cc: sipping <sipping@ietf.org>, Henning Schulzrinne <hgs@cs.columbia.edu>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Jonathan Rosenberg writes:

 > So, my proposal is, if you are going to go and DO something like that, 
 > you better use the SIP techniques that we've defined to deal with it. 
 > And when you extend SIP in a way that doesn't work with a SIP endpoint 
 > that doesn't understand an extension, you use a Require header. And that 
 > is exactly what I'm proposing that 3gpp should do. Use Require because 
 > you have made a non-backwards-compatible extension to SIP.

i have said many times, let 3gpp people shoot their feet as long as they
don't bother ietf mailing lists with their incompatibilities.

-- juha

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sat Feb 17 14:57:29 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HIVg9-0006o4-NE; Sat, 17 Feb 2007 14:57:05 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HIVg7-0006nz-Lg
	for sipping@ietf.org; Sat, 17 Feb 2007 14:57:03 -0500
Received: from ihemail1.lucent.com ([135.245.0.33])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HIVg4-0008S7-9X
	for sipping@ietf.org; Sat, 17 Feb 2007 14:57:03 -0500
Received: from ilexp02.ndc.lucent.com (h135-3-39-2.lucent.com [135.3.39.2])
	by ihemail1.lucent.com (8.13.8/IER-o) with ESMTP id l1HJupUi002124;
	Sat, 17 Feb 2007 13:56:51 -0600 (CST)
Received: from DEEXP01.de.lucent.com ([135.248.187.65]) by
	ilexp02.ndc.lucent.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Sat, 17 Feb 2007 13:56:51 -0600
Received: from DEEXC1U01.de.lucent.com ([135.248.187.28]) by
	DEEXP01.de.lucent.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Sat, 17 Feb 2007 20:56:49 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP - proxy types
Date: Sat, 17 Feb 2007 20:56:49 +0100
Message-ID: <5D1A7985295922448D5550C94DE29180CEE89E@DEEXC1U01.de.lucent.com>
In-Reply-To: <7374777208BDC7449D5620EF9423256702047C66@esealmw113.eemea.ericsson.se>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP - proxy types
Thread-Index: AcdKpsbFDrW3FiTgTaqoCBJtt1RnoQEDfmXQAOE92aA=
References: <1170835642.29763.54.camel@n95.nomadiclab.com>
	<7374777208BDC7449D5620EF9423256702047C66@esealmw113.eemea.ericsson.se>
From: "Drage, Keith \(Keith\)" <drage@alcatel-lucent.com>
To: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>,
	<sipping@ietf.org>
X-OriginalArrivalTime: 17 Feb 2007 19:56:49.0441 (UTC)
	FILETIME=[C2DD6110:01C752CD]
X-Scanned-By: MIMEDefang 2.57 on 135.245.2.33
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 50a516d93fd399dc60588708fd9a3002
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

>From the proceedings of the IETF#67 for SIP:

"Consensus noted that we do need some sort of solution for the general =20
problem of delivery of URI parameters across contact-routing
operations."

"Question: Adopt this draft as baseline text for a WG effort, should =20
the ADs approve the milestone? Consensus to adopt noted by chairs. To-=20
do: Chairs to work with AD to set new milestone."

Now I admit that the SIP chairs have been somewhat dilatory in updating
the milestones from the last meeting, but the proceedings went to the
list - and nobody said these were wrong.

Keith


=20

> -----Original Message-----
> From: Christer Holmberg (JO/LMF)=20
> [mailto:christer.holmberg@ericsson.com]=20
> Sent: 13 February 2007 10:23
> To: sipping@ietf.org
> Subject: [Sipping] Service Identification in SIP - proxy types
>=20
>=20
> Hi,
>=20
> First, the document talks about outbound proxy, home proxy,=20
> originating proxy and terminating proxy, but there is no real=20
> definition/explanation/reference to what those proxies really are.
>=20
> Second, the draft also uses Jonathan's loose route draft.=20
> Have we decided to work on that draft? I remember there was=20
> quite much discussion at one point, but what happened after=20
> that? In any case, a few words WHY the mechanism is needed=20
> would also be useful.
>=20
> Regards,
>=20
> Christer
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sat Feb 17 19:40:43 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HIa46-0001Fd-TP; Sat, 17 Feb 2007 19:38:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HIa45-0001FS-9q
	for sipping@ietf.org; Sat, 17 Feb 2007 19:38:05 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HIa3v-0007pC-24
	for sipping@ietf.org; Sat, 17 Feb 2007 19:38:05 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	F0D7E2044F; Sun, 18 Feb 2007 01:37:39 +0100 (CET)
X-AuditID: c1b4fb3c-affc8bb0000007de-ec-45d79fd346ff 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	CD5CD203F5; Sun, 18 Feb 2007 01:37:39 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Sun, 18 Feb 2007 01:37:39 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Sun, 18 Feb 2007 01:37:38 +0100
Message-ID: <B0AF4CCDCE7F4D439EAC01F3331B1329265BE2@esealmw111>
In-Reply-To: <45D65EE8.7010501@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdSta5H5No71HxLQf6Yt9fqCeVjkgAIHnyw
From: =?iso-8859-1?Q?G=F6ran_Eriksson_AP_=28KI/EAB=29?=
	<goran.ap.eriksson@ericsson.com>
To: "Jonathan Rosenberg" <jdrosen@cisco.com>
X-OriginalArrivalTime: 18 Feb 2007 00:37:39.0129 (UTC)
	FILETIME=[FE0FBA90:01C752F4]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 9f79b8e383fd3af2b1b5b1d0910f6094
Cc: sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Jonathan,

You wrote:

	The general problem here is that there are other places where people =
are layering additional 'context', as Loreto=20
	has called it, ontop of SIP signaling. This is a worrisome trend =
because this context is subtle and not explicitly signaled.=20
	Its just assumed. And that is dangerous, since it leads to the kind of =
balkanization and non-interoperability that myself,=20
	Paul, Dale and others have complained about in this thread.

I understand You to mean that what is worrisome is not that "people" =
thinks that handling "contexts" is needed but rather the way to do it.
Have I understood Your correctly?

I think this Service instance URI is a good idea; the URN I would like =
to understand the consequences of a bit better. The Service Id draft =
state that=20
service URN's should be used for "well known services" and in the =
Service Id draft, "SMS" as a well known service- does this mean "SMS" =
must be defined
in a standard track RFC?=20

Perhaps You could expand on this one a bit?

Finally, I would to say that I sympathize with the view that the PoC =
realisation on SIP is a very special case and would qualify for Require.

Thanks
-G=F6ran


-----Original Message-----
From: Jonathan Rosenberg [mailto:jdrosen@cisco.com]=20
Sent: den 17 februari 2007 02:48
To: Henning Schulzrinne
Cc: sipping
Subject: Re: [Sipping] Service Identification in SIP

I think its a fair point that the example is contrived. Really, the main =
use case being targeted is IMS PoC. So, let me talk a bit about that.

One would think that PoC is just a regular SIP call that has floor =
control added to it as another media type, and so one could indicate =
that a call is a PoC call just because of the presence of this new media =
type in the SDP. Sadly, I wish that were true. If it was, we would not =
need a mechanism here.

Unfortunately, PoC has layered a whole bunch of additional behaviors =
into the way SIP requests are handled. Dale has indicated one of them - =
that a 200 OK is returned on the terminating side priod to the call =
actually being completed. PoC also *requires* the insertion of servers =
into the SIP signaling path to make the system work. Consequently, =
though the messages that PoC uses *look* like SIP messages with a new =
media type, there is a hidden 'extension' in place that changes the =
fundamental semantics of INVITE. The fact that this extension is not =
visible as a specific header in a well defined way does not change the =
fact that PoC CHANGES THE BEHAVIOR OF SIP AND IS THUS A SIP EXTENSION.

To convince yourself of this, consider what would happen if a PoC client =
calls a regular SIP phone. The regular SIP phone would see the PoC floor =
control stream and reject it as unknown. If PoC were just SIP as we've =
specified it, then everything would be fine, and the call would proceed =
but without floor control.

But alas, it is not so.

Because PoC makes all kinds of assumptions about the behaviors of its =
participants, and what the real meaning is for various SIP messages, the =
entire thing just won't work when a call terminates on a non-PoC device. =

This is because PoC is, effectively, a fairly significant and =
non-backwards-compatible extension to SIP.

So, my proposal is, if you are going to go and DO something like that, =
you better use the SIP techniques that we've defined to deal with it.=20
And when you extend SIP in a way that doesn't work with a SIP endpoint =
that doesn't understand an extension, you use a Require header. And that =
is exactly what I'm proposing that 3gpp should do. Use Require because =
you have made a non-backwards-compatible extension to SIP.

The general problem here is that there are other places where people are =
layering additional 'context', as Loreto has called it, ontop of SIP =
signaling. This is a worrisome trend because this context is subtle and =
not explicitly signaled. Its just assumed. And that is dangerous, since =
it leads to the kind of balkanization and non-interoperability that =
myself, Paul, Dale and others have complained about in this thread.

The current 3gpp specs, instead of using Require to signal this =
extension, put a tag into the Accept-Contact header field. I find this =
extremely worrisome, because it is really just masking the reality of =
what has happened. If you've got a situration where a SIP call *cannot =
be processed* without going through application servers, then this is =
not something that is signaled with Accept-Contact. Its signaled with =
Require or Proxy-Require as appropriate.

Thanks,
-Jonathan R.

Henning Schulzrinne wrote:

> I have to admit that I find a discussion that uses a single example a=20
> bit peculiar, particularly since the example is rather contrived. =20
> (Why would you use voice, plus speech-to-text presumably, to talk to =20
> a chess
> application?)
>=20
> In a networking sense, the issue seems to be that we're using voice =20
> as a transport protocol to carry different distinct applications. (In  =

> the old days, a modem would have been such an application.)
>=20
> I'll repeat what I've said before: SIP doesn't exactly suffer from a=20
> lack of complexity and mechanisms. At this point, before adding more=20
> machinery, we might want to ask as to whether there is truly a broad-=20
> based need for justifying additional machinery.
>=20
> With regards to service URNs, I happen to believe that they are very=20
> reasonable as "modern" replacements for tel URIs, essentially =20
> extending the concept of 800 numbers. But that's a different problem =20
> from what I understand this discussion to be about. (In that usage, =20
> service URNs would be mapped to SIP AORs based on geography, time-of-=20
> day and other
> factors.)
>=20
> Henning
>=20
>=20
> On Feb 9, 2007, at 10:04 AM, Salvatore Loreto (JO/LMF) wrote:
>=20
>> Hi Ravi,
>>
>> if you consider the Jonathan draft, he says that "...chess game that=20
>> utilizes voice commands to move pieces.  When  one player says "Queen =

>> to D7", the software on their phone  recognizes this and moves the=20
>> piece.  At the same time, the user's  voice is sent to the other=20
>> player, where it is both rendered to the  user, and interpreted=20
>> locally in order to move the piece."
>> and he suggests that the following urn
>>
>>    urn:service:games:voice-chess
>>
>> could be used to identify the voice chess application.
>>
>> I guess that with voice he supposes audio, but nowadays we can also=20
>> have voice using PoC. Then we need a way to specify which kind of=20
>> voice we want: a standard full-duplex or a PoC. Maybe the chess  game =

>> works well with, or it is designed for a standard audio but  not for=20
>> a PoC audio. So we need to address also the voice  preference we=20
>> want. I like your proposal, having something  hierarchical was=20
>> something I was thinking too I don't know if it is  your idea, but=20
>> with the service urn you are proposing:  Urn:service:voice.poc.chess=20
>> I imaging you are asking to play a  Chess game, that work utilizing=20
>> voice commands, but specifically  voice over PoC. This is also=20
>> compliant to the
>> draft-ietf-ecrit- service-urn (see section 3) "Declaration of=20
>> syntactic structure:  The URN consists of a hierarchical service=20
>> identifier, with a  sequence of labels separated by periods.  The=20
>> left-most label is  the most significant one and is called 'top-level =

>> service', while  names to the right are called 'sub-services'.  The=20
>> set of allowable  characters is the same as that for domain names [1] =

>> and a subset of  the labels allowed in [5].  Labels are=20
>> case-insensitive and MUST be  specified in all lower-case.  For any=20
>> given service URN, service- identifiers can be removed right-to-left=20
>> and the resulting URN is  still valid, referring a more generic=20
>> service.  In other words, if  a service 'x.y.z' exists, the URNs 'x'
>> and 'x.y' are also valid  service URNs." Regards sal On Fri,=20
>> 2007-02-09 at 13:44 +0530, ravishankar.shiroor@wipro.com  wrote:
>>
>>> Hi Salvatore, As an end user, poc is also an application that I  use =

>>> and chess is also an application that I can use. The way I see  it,=20
>>> in your example, it will be running chess application on top  of PoC =

>>> application. So the service urn for this would be something  like:
>>> Urn:service:voice.poc.chess Regards, Ravi. -- Ravishankar.  G.=20
>>> Shiroor Wipro Technologies, Bangalore. =20
>>> ravishankar.shiroor@wipro.com
>>> -- > -----Original Message----- >  From: Salvatore Loreto (JO/LMF) > =

>>> [mailto:salvatore.loreto@ericsson.com] > Sent: Wednesday, February=20
>>> 07, 2007 1:37 PM > To: RaviShankar Shiroor (WT01 - Product=20
>>> Engineering Solutions) > Cc: Gonzalo.Camarillo@ericsson.com;=20
>>> sipping@ietf.org; > mary.barnes@nortel.com > Subject: RE:  [Sipping] =

>>> Service Identification in SIP > > hi Ravi, > > in IMS a  Service is=20
>>> a "type of communication defined by a > service  definition that=20
>>> specifies the rules and procedures > and allowed  media, if any, for =

>>> a specific type of > communication and that  utilises the IMS=20
>>> enabler. An IMS > communication service defines  what is possible=20
>>> within a > single SIP session or standalone  transaction." > So a=20
>>> Service implies some combination of  capabilities and > ways to use=20
>>> those capabilities. > If you want  the Service specifies the context =

>>> for a communication. > > Instead  the Application define the=20
>>> specific communication (or > if you  want application) that you want =

>>> run in the specific > context. So  you should have several different =

>>> applications > designed to run  on a specific environment or that=20
>>> have > different performance or  behavior in different context >=20
>>> (running on top of different  service). > > Lets go to Jonathan's=20
>>> example of the Chess. > In his  example the Chess is a service, for=20
>>> IMS instead is an >  application that you can run on top of=20
>>> different service. > For  example you can run Chess application on=20
>>> top of PoC > service or  on top of the "standard" service. So let=20
>>> say you > run Chess on  top of PoC service, then you can use only=20
>>> voice > in the way that  PoC specifies. > > /sal > > > > > On Thu,=20
>>> 2007-02-01 at
>>> 16:39  +0530, > ravishankar.shiroor@wipro.com wrote: > > Hello, > >=20
>>> >
>>> >  IMO, Service Identification is definitely needed and should > be
>>> taken > > up as a WG as proposed below. > > > > Further, I think=20
>>> Service Identification scope should be as per the > > requirement =20
>>> in Jonathan's draft, as I still do not > understand the need > > =20
>>> for having a Communication Service Identifier and another > =20
>>> Application
>>> > > Identifier as given in the 3gpp-ics-requirements  document. > >=20
>>> > > >
>>> > Regards, > > Ravi. > > > > -- > > > >  Ravishankar. G. Shiroor > >
>>> Wipro Technologies, Bangalore. > > > > =20
>>> ravishankar.shiroor@wipro.com
>>> > > -- > > > > > > > -----Original  Message----- > > > From: Gonzalo
>>> Camarillo  [mailto:Gonzalo.Camarillo@ericsson.com] > > > Sent:=20
>>> Wednesday,  January 31, 2007 8:30 PM > > > To: sipping > > > Cc:=20
>>> Mary Barnes >  > > Subject: [Sipping] Service Identification in SIP=20
>>> > > >
>>> > > >  Folks, > > > > > > during the last year, there have been a
>>> number  of threads on the > > > mailing list discussing service=20
>>> identification in SIP. In > > > particular, the following thread=20
>>> discussed 3GPP > requirements on this > > > area: > > > > > > >=20
>>> http://www1.ietf.org/mail-archive/web/sipping/current/ msg10562.html
>>> > > > > > > At a later point, those 3GPP  requirements were
>>> documented in the > > > following (expired by  now) Internet Draft,=20
>>> which was > presented in a > > > face-to-face  IETF meeting. > > > >
>>> > > http://www.watersprings.org/pub/id/draft- loreto-sipping-3gpp-i=20
>>> > > > cs-requirements-00.txt > > > > > >  Lately, the following=20
>>> > > draft
>>> has been submitted to the archives: >  > > > > >
>>> http://www.ietf.org/internet-drafts/draft-rosenberg- sipping-se > >=20
>>> > rvice-identification-00.txt > > > > > > This  draft proposes a=20
>>> solution that intends to meet known > > >  requirements in this =
area.
>>> > > > > > > Given the amount of  discussions on issues related to
>>> service > > > identification we  have seen during the last year, we=20
>>> > believe this is > > > a  fairly important topic for the SIPPING WG =

>>> to tackle. At some > > >  point, we would like to have a milestone=20
>>> with its > associated WG  item > > > dealing with this issue. > > >=20
>>> > > > Therefore, we  would like to encourage people to read this >=20
>>> draft and > > > send  comments to the list. Additionally, if you=20
>>> believe that this > > >  work is (or is not) important, we would=20
>>> like to hear that as well.  > > > > > > Cheers, > > > > > > Gonzalo=20
>>> > > > SIPPING co-chair > >  > > > > > > >=20
>>> _______________________________________________ > >  > Sipping=20
>>> mailing list > https://www1.ietf.org/mailman/listinfo/ sipping > > > =

>>> This list is for NEW development of the application  of SIP Use  > >
>>> > sip-implementors@cs.columbia.edu for questions  on current sip Use
>>> > > > sip@ietf.org for new developments of core  SIP > > > > > > >
>>> _______________________________________________  > > Sipping mailing =

>>> list https://www1.ietf.org/mailman/listinfo/ sipping > > This list=20
>>> is for NEW development of the application of  SIP Use > >=20
>>> sip-implementors@cs.columbia.edu for questions on  current sip Use >
>>> > sip@ietf.org for new developments of core SIP  > > The information
>>> contained in this electronic message and any  attachments to this=20
>>> message are intended for the exclusive use of  the addressee(s) and=20
>>> may contain proprietary, confidential or  privileged information. If =

>>> you are not the intended recipient, you  should not disseminate,=20
>>> distribute or copy this e-mail. Please  notify the sender=20
>>> immediately and destroy all copies of this  message and any =
attachments. WARNING:
>>> Computer viruses can be  transmitted via email. The recipient should =

>>> check this email and  any attachments for the presence of viruses.
>>> The company accepts  no liability for any damage caused by any virus =

>>> transmitted by  this email. www.wipro.com
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP Use=20
>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
>> sip@ietf.org for new developments of core SIP
>=20
>=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20

--=20
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use =
sip-implementors@cs.columbia.edu for questions on current sip Use =
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sat Feb 17 22:18:10 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HIcX5-00022Z-5w; Sat, 17 Feb 2007 22:16:11 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HIcX3-00021s-PI
	for sipping@ietf.org; Sat, 17 Feb 2007 22:16:09 -0500
Received: from c243-111.rim.net ([216.9.243.111] helo=MHS02YKF.rim.net)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HIcX3-0006Kb-Ae
	for sipping@ietf.org; Sat, 17 Feb 2007 22:16:09 -0500
Importance: normal
Priority: normal
Received: from ngw02ykf.rim.net ([10.102.108.31]) by MHS02YKF.rim.net with
	Microsoft SMTPSVC(6.0.3790.211); Sat, 17 Feb 2007 22:15:58 -0500
Received: from XCH20YKF.rim.net ([10.102.100.35]) by ngw02ykf.rim.net (SAVSMTP
	3.1.6.45) with SMTP id M2007021722154915417 ;
	Sat, 17 Feb 2007 22:15:49 -0500
Received: from XCH47YKF.rim.net ([10.64.31.217]) by XCH20YKF.rim.net with
	Microsoft SMTPSVC(5.0.2195.6713); Sat, 17 Feb 2007 22:15:50 -0500
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.181
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Sat, 17 Feb 2007 22:15:48 -0500
Message-ID: <61968779B8AC4C4BAB421D4C12F008C0091516D0@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdSuNrFNKtKne4MQLm5PW07CCO4BgAR5wYA
From: "Andrew Allen" <aallen@rim.com>
To: "Henning Schulzrinne" <hgs@cs.columbia.edu>,
	"Jonathan Rosenberg" <jdrosen@cisco.com>
X-OriginalArrivalTime: 18 Feb 2007 03:15:50.0546 (UTC)
	FILETIME=[17630720:01C7530B]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 501044f827b673024f6a4cb1d46e67d2
Cc: sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi Henning,

First point on the use of Service URNs. I think the following properties
of a URN are important to this work:

a)       the guaranteed uniqueness - we need to ensure that service
identifiers are unique. URNs have an existing process for ensuring
uniqueness.

b)       the ability to use a URN as a URI - A UA can be pre-configured
or through some other means obtain a URN for a given service and then
this can be used in the Route header or the Request-URI in order to
contact Application Servers in the local domain when initiating a
communication related to the service.=20

c)       a hierarchical structure - in order to enhance the
interoperability URNs provide the possibility to perform best match
comparisons in order to determine what service should be provided.

e.g =20
If=20
   URN:service.comm.telephony.multimedia
Is present in a request but only
   URN:service.comm.telephony.basic=20
is supported then that can be considered a good match.=20


Some of these communication services may also be location based for
example the equivalent of 1-800-PIZZA could be
URN:service.comm.telephony.free.restaurant.pizza.

Linking this to the Service URN used in ecrit seems convenient but not
absolutely essential.=20


Secondly I don't have a problem with using Callee Caps and Caller
preferences to find the appropriate UAS however the requirements for
service identification are much more than just that. =20

This following requirements from 3GPP TS 23.228 do not I believe fit
well with the semantics of Accept-Contact:

2.   Based on operator policy the S-CSCF or an AS shall be able to
validate an IMS communication service identifier in a SIP request. This
includes e.g. to check the syntactical correctness of a service
identifier, and policing the usage of a communication service
identifier.

4.   It shall be possible for the S-CSCF to invoke appropriate service
logic based on the IMS communication service identifier contained in a
SIP request, e.g. route a SIP request containing a service identifier
based on initial filter criteria to the correct AS.

8.   Based on operator policy S-CSCF and AS shall consider the IMS
communication service identifier for online and offline charging, e.g.
put appropriate data into call detailed records.

9.   The communication service identifier shall be capable of being an
input into the policy control and charging rules.

10. It shall be possible to use the IMS communication service identifier
as a means to authorise whether a subscriber is allowed to initiate or
receive request for a communication service.

Note that the some of AS (Application Servers) that need to be invoked
based on requirement (4) may not have any relationship to the resource
identified by the Request-URI because they are Applications Servers in
the domain of the originating UAC.

Not only do these requirements not semantically fit with Accept-Contact
but the fact that a service feature tag appears somewhere in an
Accept-Contact cannot be taken as a definitive indication the initiating
user is only requesting that particular service as the expression may
involve other feature tags in a complex expression.

In addition there needs to be a mechanism that allows UAs to obtain the
appropriate service identifiers used by other UAs and in other domains.
I think Jonathan's proposal to use Service URIs and Service Instance
URIs with Presence creates a backward compatible way for UAs to obtain
these identifiers.

Andrew


-----Original Message-----
From: Henning Schulzrinne [mailto:hgs@cs.columbia.edu]=20
Sent: Saturday, February 17, 2007 12:27 PM
To: Jonathan Rosenberg
Cc: sipping
Subject: Re: [Sipping] Service Identification in SIP

I think we're in general agreement; in particular, Require seems like =20
the right approach. (service URNs do not, in my opinion).

Stepping back, I don't think we should generalize this problem since =20
it essentially "blesses" a particular design approach that is =20
undesirable. The right approach, from a SIP perspective, for IMS PoC =20
would have been to come up with a solution that is backwards-=20
compatible. If that's not possible, getting "branded" with Require =20
seems like the right penalty.

In general, we have probably two requirements:

(1) When contacting an AOR, a UAC that needs a special UAS to work =20
properly needs to get routed to that UAS only. Callee caps and caller =20
preferences seem to be the right tools for that, as it involves call =20
routing and capability matching.

(2) If, despite doing (1), a SIP request accidentally reaches a UAS =20
that can't do anything useful with it, Requires minimizes the damage =20
(and allows a proxy to serially-fork to something else that might =20
have a chance of working).

Henning

On Feb 16, 2007, at 8:48 PM, Jonathan Rosenberg wrote:

> I think its a fair point that the example is contrived. Really, the =20
> main use case being targeted is IMS PoC. So, let me talk a bit =20
> about that.
>
> One would think that PoC is just a regular SIP call that has floor =20
> control added to it as another media type, and so one could =20
> indicate that a call is a PoC call just because of the presence of =20
> this new media type in the SDP. Sadly, I wish that were true. If it =20
> was, we would not need a mechanism here.
>
> Unfortunately, PoC has layered a whole bunch of additional =20
> behaviors into the way SIP requests are handled. Dale has indicated =20
> one of them - that a 200 OK is returned on the terminating side =20
> priod to the call actually being completed. PoC also *requires* the =20
> insertion of servers into the SIP signaling path to make the system =20
> work. Consequently, though the messages that PoC uses *look* like =20
> SIP messages with a new media type, there is a hidden 'extension' =20
> in place that changes the fundamental semantics of INVITE. The fact =20
> that this extension is not visible as a specific header in a well =20
> defined way does not change the fact that PoC CHANGES THE BEHAVIOR =20
> OF SIP AND IS THUS A SIP EXTENSION.
>
> To convince yourself of this, consider what would happen if a PoC =20
> client calls a regular SIP phone. The regular SIP phone would see =20
> the PoC floor control stream and reject it as unknown. If PoC were =20
> just SIP as we've specified it, then everything would be fine, and =20
> the call would proceed but without floor control.
>
> But alas, it is not so.
>
> Because PoC makes all kinds of assumptions about the behaviors of =20
> its participants, and what the real meaning is for various SIP =20
> messages, the entire thing just won't work when a call terminates =20
> on a non-PoC device. This is because PoC is, effectively, a fairly =20
> significant and non-backwards-compatible extension to SIP.
>
> So, my proposal is, if you are going to go and DO something like =20
> that, you better use the SIP techniques that we've defined to deal =20
> with it. And when you extend SIP in a way that doesn't work with a =20
> SIP endpoint that doesn't understand an extension, you use a =20
> Require header. And that is exactly what I'm proposing that 3gpp =20
> should do. Use Require because you have made a non-backwards-=20
> compatible extension to SIP.
>
> The general problem here is that there are other places where =20
> people are layering additional 'context', as Loreto has called it, =20
> ontop of SIP signaling. This is a worrisome trend because this =20
> context is subtle and not explicitly signaled. Its just assumed. =20
> And that is dangerous, since it leads to the kind of balkanization =20
> and non-interoperability that myself, Paul, Dale and others have =20
> complained about in this thread.
>
> The current 3gpp specs, instead of using Require to signal this =20
> extension, put a tag into the Accept-Contact header field. I find =20
> this extremely worrisome, because it is really just masking the =20
> reality of what has happened. If you've got a situration where a =20
> SIP call *cannot be processed* without going through application =20
> servers, then this is not something that is signaled with Accept-=20
> Contact. Its signaled with Require or Proxy-Require as appropriate.
>
> Thanks,
> -Jonathan R.
>
> Henning Schulzrinne wrote:
>


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential =
information, privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sun Feb 18 10:38:44 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HIo5s-0005gR-Sb; Sun, 18 Feb 2007 10:36:52 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HIo5r-0005fH-BE
	for sipping@ietf.org; Sun, 18 Feb 2007 10:36:51 -0500
Received: from serrano.cc.columbia.edu ([128.59.29.6])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HIo5q-0000Av-S7
	for sipping@ietf.org; Sun, 18 Feb 2007 10:36:51 -0500
Received: from [192.168.0.41] (pool-141-153-194-152.mad.east.verizon.net
	[141.153.194.152]) (user=hgs10 mech=PLAIN bits=0)
	by serrano.cc.columbia.edu (8.13.7/8.13.6) with ESMTP id l1IFaYdR011296
	(version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NOT);
	Sun, 18 Feb 2007 10:36:35 -0500 (EST)
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C0091516D0@XCH47YKF.rim.net>
References: <61968779B8AC4C4BAB421D4C12F008C0091516D0@XCH47YKF.rim.net>
Mime-Version: 1.0 (Apple Message framework v752.2)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <42E60FD2-CA6C-45C4-A1E3-99C93A867A54@cs.columbia.edu>
Content-Transfer-Encoding: 7bit
From: Henning Schulzrinne <hgs@cs.columbia.edu>
Subject: Re: [Sipping] Service Identification in SIP
Date: Sun, 18 Feb 2007 10:36:29 -0500
To: Andrew Allen <aallen@rim.com>
X-Mailer: Apple Mail (2.752.2)
X-No-Spam-Score: Local
X-Scanned-By: MIMEDefang 2.48 on 128.59.29.6
X-Spam-Score: 0.0 (/)
X-Scan-Signature: ed68cc91cc637fea89623888898579ba
Cc: sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Andrew,

my concern is that I'm unsure that developing a very generic  
mechanism to address a problem we don't want to happen in the first  
case seems both misplaced effort and counterproductive. A robust and  
distributed allocation mechanism for identifiers makes sense if you  
expect hundreds of entities to independently need to allocate such  
identifiers. This is clearly undesirable in this case, based on the  
examples presented, so the IANA registration mechanism for Require  
tags seems to be both sufficient and sufficiently onerous to prevent  
gratuitous incompatibility.

I also believe that service URNs are wholly inappropriate for this  
particular purpose, even if you don't believe what I wrote above.

Henning

On Feb 17, 2007, at 10:15 PM, Andrew Allen wrote:

>
> Hi Henning,
>
> First point on the use of Service URNs. I think the following  
> properties
> of a URN are important to this work:
>
> a)       the guaranteed uniqueness - we need to ensure that service
> identifiers are unique. URNs have an existing process for ensuring
> uniqueness.
>
> b)       the ability to use a URN as a URI - A UA can be pre- 
> configured
> or through some other means obtain a URN for a given service and then
> this can be used in the Route header or the Request-URI in order to
> contact Application Servers in the local domain when initiating a
> communication related to the service.
>
> c)       a hierarchical structure - in order to enhance the
> interoperability URNs provide the possibility to perform best match
> comparisons in order to determine what service should be provided.
>
> e.g
> If
>    URN:service.comm.telephony.multimedia
> Is present in a request but only
>    URN:service.comm.telephony.basic
> is supported then that can be considered a good match.
>
>
> Some of these communication services may also be location based for
> example the equivalent of 1-800-PIZZA could be
> URN:service.comm.telephony.free.restaurant.pizza.
>
> Linking this to the Service URN used in ecrit seems convenient but not
> absolutely essential.
>
>
> Secondly I don't have a problem with using Callee Caps and Caller
> preferences to find the appropriate UAS however the requirements for
> service identification are much more than just that.
>
> This following requirements from 3GPP TS 23.228 do not I believe fit
> well with the semantics of Accept-Contact:
>
> 2.   Based on operator policy the S-CSCF or an AS shall be able to
> validate an IMS communication service identifier in a SIP request.  
> This
> includes e.g. to check the syntactical correctness of a service
> identifier, and policing the usage of a communication service
> identifier.
>
> 4.   It shall be possible for the S-CSCF to invoke appropriate service
> logic based on the IMS communication service identifier contained in a
> SIP request, e.g. route a SIP request containing a service identifier
> based on initial filter criteria to the correct AS.
>
> 8.   Based on operator policy S-CSCF and AS shall consider the IMS
> communication service identifier for online and offline charging, e.g.
> put appropriate data into call detailed records.
>
> 9.   The communication service identifier shall be capable of being an
> input into the policy control and charging rules.
>
> 10. It shall be possible to use the IMS communication service  
> identifier
> as a means to authorise whether a subscriber is allowed to initiate or
> receive request for a communication service.
>
> Note that the some of AS (Application Servers) that need to be invoked
> based on requirement (4) may not have any relationship to the resource
> identified by the Request-URI because they are Applications Servers in
> the domain of the originating UAC.
>
> Not only do these requirements not semantically fit with Accept- 
> Contact
> but the fact that a service feature tag appears somewhere in an
> Accept-Contact cannot be taken as a definitive indication the  
> initiating
> user is only requesting that particular service as the expression may
> involve other feature tags in a complex expression.
>
> In addition there needs to be a mechanism that allows UAs to obtain  
> the
> appropriate service identifiers used by other UAs and in other  
> domains.
> I think Jonathan's proposal to use Service URIs and Service Instance
> URIs with Presence creates a backward compatible way for UAs to obtain
> these identifiers.
>
> Andrew
>
>
> -----Original Message-----
> From: Henning Schulzrinne [mailto:hgs@cs.columbia.edu]
> Sent: Saturday, February 17, 2007 12:27 PM
> To: Jonathan Rosenberg
> Cc: sipping
> Subject: Re: [Sipping] Service Identification in SIP
>
> I think we're in general agreement; in particular, Require seems like
> the right approach. (service URNs do not, in my opinion).
>
> Stepping back, I don't think we should generalize this problem since
> it essentially "blesses" a particular design approach that is
> undesirable. The right approach, from a SIP perspective, for IMS PoC
> would have been to come up with a solution that is backwards-
> compatible. If that's not possible, getting "branded" with Require
> seems like the right penalty.
>
> In general, we have probably two requirements:
>
> (1) When contacting an AOR, a UAC that needs a special UAS to work
> properly needs to get routed to that UAS only. Callee caps and caller
> preferences seem to be the right tools for that, as it involves call
> routing and capability matching.
>
> (2) If, despite doing (1), a SIP request accidentally reaches a UAS
> that can't do anything useful with it, Requires minimizes the damage
> (and allows a proxy to serially-fork to something else that might
> have a chance of working).
>
> Henning
>
> On Feb 16, 2007, at 8:48 PM, Jonathan Rosenberg wrote:
>
>> I think its a fair point that the example is contrived. Really, the
>> main use case being targeted is IMS PoC. So, let me talk a bit
>> about that.
>>
>> One would think that PoC is just a regular SIP call that has floor
>> control added to it as another media type, and so one could
>> indicate that a call is a PoC call just because of the presence of
>> this new media type in the SDP. Sadly, I wish that were true. If it
>> was, we would not need a mechanism here.
>>
>> Unfortunately, PoC has layered a whole bunch of additional
>> behaviors into the way SIP requests are handled. Dale has indicated
>> one of them - that a 200 OK is returned on the terminating side
>> priod to the call actually being completed. PoC also *requires* the
>> insertion of servers into the SIP signaling path to make the system
>> work. Consequently, though the messages that PoC uses *look* like
>> SIP messages with a new media type, there is a hidden 'extension'
>> in place that changes the fundamental semantics of INVITE. The fact
>> that this extension is not visible as a specific header in a well
>> defined way does not change the fact that PoC CHANGES THE BEHAVIOR
>> OF SIP AND IS THUS A SIP EXTENSION.
>>
>> To convince yourself of this, consider what would happen if a PoC
>> client calls a regular SIP phone. The regular SIP phone would see
>> the PoC floor control stream and reject it as unknown. If PoC were
>> just SIP as we've specified it, then everything would be fine, and
>> the call would proceed but without floor control.
>>
>> But alas, it is not so.
>>
>> Because PoC makes all kinds of assumptions about the behaviors of
>> its participants, and what the real meaning is for various SIP
>> messages, the entire thing just won't work when a call terminates
>> on a non-PoC device. This is because PoC is, effectively, a fairly
>> significant and non-backwards-compatible extension to SIP.
>>
>> So, my proposal is, if you are going to go and DO something like
>> that, you better use the SIP techniques that we've defined to deal
>> with it. And when you extend SIP in a way that doesn't work with a
>> SIP endpoint that doesn't understand an extension, you use a
>> Require header. And that is exactly what I'm proposing that 3gpp
>> should do. Use Require because you have made a non-backwards-
>> compatible extension to SIP.
>>
>> The general problem here is that there are other places where
>> people are layering additional 'context', as Loreto has called it,
>> ontop of SIP signaling. This is a worrisome trend because this
>> context is subtle and not explicitly signaled. Its just assumed.
>> And that is dangerous, since it leads to the kind of balkanization
>> and non-interoperability that myself, Paul, Dale and others have
>> complained about in this thread.
>>
>> The current 3gpp specs, instead of using Require to signal this
>> extension, put a tag into the Accept-Contact header field. I find
>> this extremely worrisome, because it is really just masking the
>> reality of what has happened. If you've got a situration where a
>> SIP call *cannot be processed* without going through application
>> servers, then this is not something that is signaled with Accept-
>> Contact. Its signaled with Require or Proxy-Require as appropriate.
>>
>> Thanks,
>> -Jonathan R.
>>
>> Henning Schulzrinne wrote:
>>
>
>
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
>
>
>
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain  
> confidential information, privileged material (including material  
> protected by the solicitor-client or other applicable privileges),  
> or constitute non-public information. Any use of this information  
> by anyone other than the intended recipient is prohibited. If you  
> have received this transmission in error, please immediately reply  
> to the sender and delete this information from your system. Use,  
> dissemination, distribution, or reproduction of this transmission  
> by unintended recipients is not authorized and may be unlawful.
>
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 19 01:04:55 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJ1d4-0007Jx-Um; Mon, 19 Feb 2007 01:04:02 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJ1d3-0007Jq-UW
	for sipping@ietf.org; Mon, 19 Feb 2007 01:04:01 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HJ1d0-0005HO-2a
	for sipping@ietf.org; Mon, 19 Feb 2007 01:04:01 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	67B3120087; Mon, 19 Feb 2007 07:03:51 +0100 (CET)
X-AuditID: c1b4fb3e-b16d6bb0000007e1-ee-45d93dc7b185 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	44D5B2006D; Mon, 19 Feb 2007 07:03:51 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 19 Feb 2007 07:03:51 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP - Routing decissions
	based on Require/P-R
Date: Mon, 19 Feb 2007 07:03:50 +0100
Message-ID: <7374777208BDC7449D5620EF94232567021CE512@esealmw113.eemea.ericsson.se>
In-Reply-To: <45D65EE8.7010501@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP - Routing decissions
	based on Require/P-R
Thread-Index: AcdStY7gy91UY1w5R8ef+pYIv0iRvwBMjEog
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Jonathan Rosenberg" <jdrosen@cisco.com>,
	"Henning Schulzrinne" <hgs@cs.columbia.edu>
X-OriginalArrivalTime: 19 Feb 2007 06:03:51.0042 (UTC)
	FILETIME=[BA3E6620:01C753EB]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 83867a50fd8f547996ccdaf89af24437
Cc: sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi Jonathan,=20

>If you've got a situration where a SIP call *cannot be processed*
without=20
>going through application servers, then this is not something=20
>that is signaled with Accept-Contact. Its signaled with=20
>Require or Proxy-Require as appropriate.

Just for my clarification, where is it said that Require and
Proxy-Require are used, or at least more appropriate, for routing?=20

As far as I know, they are only used to require capabilities by the
nodes that receive a SIP request (as you said yourself earlier) - but
not to make routing decissions based upon.

Regards,

Christer



> Henning Schulzrinne wrote:
>=20
> > I have to admit that I find a discussion that uses a single=20
> example a=20
> > bit peculiar, particularly since the example is rather contrived. =20
> > (Why would you use voice, plus speech-to-text presumably,=20
> to talk to =20
> > a chess
> > application?)
> >=20
> > In a networking sense, the issue seems to be that we're=20
> using voice =20
> > as a transport protocol to carry different distinct=20
> applications. (In =20
> > the old days, a modem would have been such an application.)
> >=20
> > I'll repeat what I've said before: SIP doesn't exactly=20
> suffer from a=20
> > lack of complexity and mechanisms. At this point, before=20
> adding more=20
> > machinery, we might want to ask as to whether there is=20
> truly a broad-=20
> > based need for justifying additional machinery.
> >=20
> > With regards to service URNs, I happen to believe that they=20
> are very=20
> > reasonable as "modern" replacements for tel URIs, essentially =20
> > extending the concept of 800 numbers. But that's a=20
> different problem =20
> > from what I understand this discussion to be about. (In=20
> that usage, =20
> > service URNs would be mapped to SIP AORs based on=20
> geography, time-of-=20
> > day and other
> > factors.)
> >=20
> > Henning
> >=20
> >=20
> > On Feb 9, 2007, at 10:04 AM, Salvatore Loreto (JO/LMF) wrote:
> >=20
> >> Hi Ravi,
> >>
> >> if you consider the Jonathan draft, he says that "...chess=20
> game that=20
> >> utilizes voice commands to move pieces.  When  one player=20
> says "Queen=20
> >> to D7", the software on their phone  recognizes this and moves the=20
> >> piece.  At the same time, the user's  voice is sent to the other=20
> >> player, where it is both rendered to the  user, and interpreted=20
> >> locally in order to move the piece."
> >> and he suggests that the following urn
> >>
> >>    urn:service:games:voice-chess
> >>
> >> could be used to identify the voice chess application.
> >>
> >> I guess that with voice he supposes audio, but nowadays we=20
> can also=20
> >> have voice using PoC. Then we need a way to specify which kind of=20
> >> voice we want: a standard full-duplex or a PoC. Maybe the=20
> chess  game=20
> >> works well with, or it is designed for a standard audio=20
> but  not for=20
> >> a PoC audio. So we need to address also the voice  preference we=20
> >> want. I like your proposal, having something  hierarchical was=20
> >> something I was thinking too I don't know if it is  your idea, but=20
> >> with the service urn you are proposing: =20
> Urn:service:voice.poc.chess=20
> >> I imaging you are asking to play a  Chess game, that work=20
> utilizing=20
> >> voice commands, but specifically  voice over PoC. This is also=20
> >> compliant to the
> >> draft-ietf-ecrit- service-urn (see section 3) "Declaration of=20
> >> syntactic structure:  The URN consists of a hierarchical service=20
> >> identifier, with a  sequence of labels separated by periods.  The=20
> >> left-most label is  the most significant one and is called=20
> 'top-level=20
> >> service', while  names to the right are called=20
> 'sub-services'.  The=20
> >> set of allowable  characters is the same as that for=20
> domain names [1]=20
> >> and a subset of  the labels allowed in [5].  Labels are=20
> >> case-insensitive and MUST be  specified in all lower-case.=20
>  For any=20
> >> given service URN, service- identifiers can be removed=20
> right-to-left=20
> >> and the resulting URN is  still valid, referring a more generic=20
> >> service.  In other words, if  a service 'x.y.z' exists,=20
> the URNs 'x'
> >> and 'x.y' are also valid  service URNs." Regards sal On Fri,=20
> >> 2007-02-09 at 13:44 +0530, ravishankar.shiroor@wipro.com  wrote:
> >>
> >>> Hi Salvatore, As an end user, poc is also an application=20
> that I  use=20
> >>> and chess is also an application that I can use. The way=20
> I see  it,=20
> >>> in your example, it will be running chess application on=20
> top  of PoC=20
> >>> application. So the service urn for this would be something  like:
> >>> Urn:service:voice.poc.chess Regards, Ravi. -- Ravishankar.  G.=20
> >>> Shiroor Wipro Technologies, Bangalore. =20
> >>> ravishankar.shiroor@wipro.com
> >>> -- > -----Original Message----- >  From: Salvatore Loreto=20
> (JO/LMF) >=20
> >>> [mailto:salvatore.loreto@ericsson.com] > Sent: Wednesday,=20
> February=20
> >>> 07, 2007 1:37 PM > To: RaviShankar Shiroor (WT01 - Product=20
> >>> Engineering Solutions) > Cc: Gonzalo.Camarillo@ericsson.com;=20
> >>> sipping@ietf.org; > mary.barnes@nortel.com > Subject: RE:=20
>  [Sipping]=20
> >>> Service Identification in SIP > > hi Ravi, > > in IMS a =20
> Service is=20
> >>> a "type of communication defined by a > service  definition that=20
> >>> specifies the rules and procedures > and allowed  media,=20
> if any, for=20
> >>> a specific type of > communication and that  utilises the IMS=20
> >>> enabler. An IMS > communication service defines  what is possible=20
> >>> within a > single SIP session or standalone  transaction." > So a=20
> >>> Service implies some combination of  capabilities and >=20
> ways to use=20
> >>> those capabilities. > If you want  the Service specifies=20
> the context=20
> >>> for a communication. > > Instead  the Application define the=20
> >>> specific communication (or > if you  want application)=20
> that you want=20
> >>> run in the specific > context. So  you should have=20
> several different=20
> >>> applications > designed to run  on a specific environment or that=20
> >>> have > different performance or  behavior in different context >=20
> >>> (running on top of different  service). > > Lets go to Jonathan's=20
> >>> example of the Chess. > In his  example the Chess is a=20
> service, for=20
> >>> IMS instead is an >  application that you can run on top of=20
> >>> different service. > For  example you can run Chess=20
> application on=20
> >>> top of PoC > service or  on top of the "standard" service. So let=20
> >>> say you > run Chess on  top of PoC service, then you can use only=20
> >>> voice > in the way that  PoC specifies. > > /sal > > > >=20
> > On Thu,=20
> >>> 2007-02-01 at
> >>> 16:39  +0530, > ravishankar.shiroor@wipro.com wrote: > >=20
> Hello, > >=20
> >>> >
> >>> >  IMO, Service Identification is definitely needed and=20
> should > be
> >>> taken > > up as a WG as proposed below. > > > > Further, I think=20
> >>> Service Identification scope should be as per the > >=20
> requirement =20
> >>> in Jonathan's draft, as I still do not > understand the need > > =20
> >>> for having a Communication Service Identifier and another > =20
> >>> Application
> >>> > > Identifier as given in the 3gpp-ics-requirements =20
> document. > >=20
> >>> > > >
> >>> > Regards, > > Ravi. > > > > -- > > > >  Ravishankar. G.=20
> Shiroor > >
> >>> Wipro Technologies, Bangalore. > > > > =20
> >>> ravishankar.shiroor@wipro.com
> >>> > > -- > > > > > > > -----Original  Message----- > > >=20
> From: Gonzalo
> >>> Camarillo  [mailto:Gonzalo.Camarillo@ericsson.com] > > > Sent:=20
> >>> Wednesday,  January 31, 2007 8:30 PM > > > To: sipping > > > Cc:=20
> >>> Mary Barnes >  > > Subject: [Sipping] Service=20
> Identification in SIP=20
> >>> > > >
> >>> > > >  Folks, > > > > > > during the last year, there have been a
> >>> number  of threads on the > > > mailing list discussing service=20
> >>> identification in SIP. In > > > particular, the following thread=20
> >>> discussed 3GPP > requirements on this > > > area: > > > > > > >=20
> >>> http://www1.ietf.org/mail-archive/web/sipping/current/=20
> msg10562.html
> >>> > > > > > > At a later point, those 3GPP  requirements were
> >>> documented in the > > > following (expired by  now)=20
> Internet Draft,=20
> >>> which was > presented in a > > > face-to-face  IETF=20
> meeting. > > > >
> >>> > > http://www.watersprings.org/pub/id/draft-=20
> loreto-sipping-3gpp-i=20
> >>> > > > cs-requirements-00.txt > > > > > >  Lately, the following=20
> >>> > > draft
> >>> has been submitted to the archives: >  > > > > >
> >>> http://www.ietf.org/internet-drafts/draft-rosenberg-=20
> sipping-se > >=20
> >>> > rvice-identification-00.txt > > > > > > This  draft proposes a=20
> >>> solution that intends to meet known > > >  requirements=20
> in this area.
> >>> > > > > > > Given the amount of  discussions on issues related to
> >>> service > > > identification we  have seen during the=20
> last year, we=20
> >>> > believe this is > > > a  fairly important topic for the=20
> SIPPING WG=20
> >>> to tackle. At some > > >  point, we would like to have a=20
> milestone=20
> >>> with its > associated WG  item > > > dealing with this=20
> issue. > > >=20
> >>> > > > Therefore, we  would like to encourage people to=20
> read this >=20
> >>> draft and > > > send  comments to the list. Additionally, if you=20
> >>> believe that this > > >  work is (or is not) important, we would=20
> >>> like to hear that as well.  > > > > > > Cheers, > > > > >=20
> > Gonzalo=20
> >>> > > > SIPPING co-chair > >  > > > > > > >=20
> >>> _______________________________________________ > >  > Sipping=20
> >>> mailing list > https://www1.ietf.org/mailman/listinfo/=20
> sipping > > >=20
> >>> This list is for NEW development of the application  of=20
> SIP Use  > >
> >>> > sip-implementors@cs.columbia.edu for questions  on=20
> current sip Use
> >>> > > > sip@ietf.org for new developments of core  SIP > > > > > > >
> >>> _______________________________________________  > >=20
> Sipping mailing=20
> >>> list https://www1.ietf.org/mailman/listinfo/ sipping > >=20
> This list=20
> >>> is for NEW development of the application of  SIP Use > >=20
> >>> sip-implementors@cs.columbia.edu for questions on =20
> current sip Use >
> >>> > sip@ietf.org for new developments of core SIP  > > The=20
> information
> >>> contained in this electronic message and any  attachments to this=20
> >>> message are intended for the exclusive use of  the=20
> addressee(s) and=20
> >>> may contain proprietary, confidential or  privileged=20
> information. If=20
> >>> you are not the intended recipient, you  should not disseminate,=20
> >>> distribute or copy this e-mail. Please  notify the sender=20
> >>> immediately and destroy all copies of this  message and=20
> any attachments. WARNING:
> >>> Computer viruses can be  transmitted via email. The=20
> recipient should=20
> >>> check this email and  any attachments for the presence of viruses.
> >>> The company accepts  no liability for any damage caused=20
> by any virus=20
> >>> transmitted by  this email. www.wipro.com
> >>
> >> _______________________________________________
> >> Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> >> This list is for NEW development of the application of SIP Use=20
> >> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> >> sip@ietf.org for new developments of core SIP
> >=20
> >=20
> >=20
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP Use=20
> > sip-implementors@cs.columbia.edu for questions on current sip Use=20
> > sip@ietf.org for new developments of core SIP
> >=20
>=20
> --=20
> Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
> Cisco Fellow                                   Parsippany, NJ=20
> 07054-2711
> Cisco Systems
> jdrosen@cisco.com                              FAX:   (973) 952-5050
> http://www.jdrosen.net                         PHONE: (973) 952-5000
> http://www.cisco.com
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 19 03:35:41 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJ3yL-0001Ya-KG; Mon, 19 Feb 2007 03:34:09 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJ3yI-0001YU-CE
	for sipping@ietf.org; Mon, 19 Feb 2007 03:34:06 -0500
Received: from mailgate.siemenscomms.co.uk ([195.171.110.225]
	helo=bemg01.siemenscomms.co.uk)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HJ3yH-0001L1-As
	for sipping@ietf.org; Mon, 19 Feb 2007 03:34:06 -0500
Received: from ntht207e.uksgcs.siemenscomms.co.uk ([137.223.247.82])
	by siemenscomms.co.uk (PMDF V6.0-24 #40642)
	with ESMTP id <0JDP00I69BSGU3@siemenscomms.co.uk> for sipping@ietf.org;
	Mon, 19 Feb 2007 08:33:52 +0000 (GMT)
Received: by ntht207e.uksgcs.siemenscomms.co.uk with Internet Mail Service
	(5.5.2657.72)	id <FF5YPY5D>; Mon, 19 Feb 2007 08:33:53 +0000
Content-return: allowed
Date: Mon, 19 Feb 2007 08:33:52 +0000
From: "Elwell, John" <john.elwell@siemens.com>
Subject: RE: Device architecture? WAS: RE: [Sipping] Service	Identificatio
	ninSIP
To: Andrew Allen <aallen@rim.com>, Malleswara Rao Ankem <malleshavn@gmail.com>
Message-id: <50B1CBA96870A34799A506B2313F26670B0F8A1C@ntht201e.siemenscomms.co.uk>
MIME-version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 58b614506802734014829a093beb6879
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1630893377=="
Errors-To: sipping-bounces@ietf.org

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

--===============1630893377==
Content-return: allowed
Content-type: multipart/alternative;
	boundary="Boundary_(ID_8oK2oZ7N4SmW/DPn91EUDA)"

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

--Boundary_(ID_8oK2oZ7N4SmW/DPn91EUDA)
Content-type: text/plain
Content-transfer-encoding: 7BIT

Resending with some truncation (was too big to go through without moderator
approval).
#########################################################################
 
Andrew,
 
OK, so it seems you are talking about IMS Application Servers. I don't see
this mentioned in Jonathan's draft, the emphasis of which seems to be on
application selection at the terminating device. For example, in section 9,
4th paragraph it states:
"Firstly, they are meant specifically and exclusively for usage in cases
where some service logic must be present and matching on both the
originating and terminating sides in order for any type of reasonable
communications to exist."
This seems to preclude use by any network intermediaries. So when I talked
about an AS, I meant a terminating side entity. So while I definitely agree
that SDP is not suitable for use by intermediaries, it would be legitimate
for a terminating application to make use of it. Also caller preferences is
suitable for selecting between different terminating entities.
 
I have still not been convinced by arguments for having a service identifier
that can be used by intermediaries, but one possibility has emerged from
these discussions that I might be able to buy. It seems there might be some
need to convey some higher level "purpose" for a request that might
influence routing decisions at SIP intermediaries, in addition to being used
by the UAS. I think one example cited was push-to-talk (PTT), which might
want to avoid being subject to call forwarding that normal INVITE requests
might undergo. A user may be able to arrange that his local proxy retargets
or forwards requests under certain conditions. If a call were marked as
"PTT" or "chess", the user could arrange that such calls be subject to
different rules. So the called user (or an administration acting on the
called user's behalf) decides on the rules, which can take into account
whatever the calling user claims to be the purpose of the call. I am not
sure about the security implications here, because the caller might get to
learn what to do to get around retargeting imposed by the callee.
 
Furthermore, I am not convinced by the notion of authorisation or charging
being based on such a service identifier unless there are some appropriately
strong security mechanism to prevent misuses by the caller. Charging on the
basis of a service identifier is fine if the network really provides some
added value that it would not provide in the absence of that service
identifier. But given that network intermediaries cannot examine the SDP, it
it not clear exactly what added value might be provided. In the absence of
any value added by the network, uses would not use the service identifier
and therefore avoid higher charges (presumably finding some out of band
means to inform the called user of the wish to use a particular service).
This is where we still seem to lack compelling examples. 
 
John
 


  _____  

From: Andrew Allen [mailto:aallen@rim.com] 
Sent: 16 February 2007 19:59
To: Elwell, John; Malleswara Rao Ankem
Cc: sipping@ietf.org
Subject: RE: Device architecture? WAS: RE: [Sipping] Service Identificatio
ninSIP


 
John
 
I think whether an AS in the terminating network is considered a contact is
a debatable point. Certainly it doesnt register feature tags using RFC 3840
mechanisms and (at least in IMS) they dont show up as registered contacts in
the reg-event notifications. Also AS in the originating network (which may
need to be involved in some communication services) are certainly not
contacts for the resource identified by the Request-URI and Accept-Contact
certainly has no defined role in routing to them.
 
In addition 3GPP TS 23.228 v7.6.0 subcaluse 4.13.2 specifies 15 requirements
for IMS Communication Service Identifier and only one of these is regarding
matching to the correct contact:
 
11. The communication service identifier shall be taken into account when
selecting the correct UE(s), if multiple UEs are registered for the same
Public User Identity(s).

 
 Several of the other requirements are in my view totally incompatible with
Accept-Contact being the transport for the identifier and indeed may limit
the use of Accept-Contact for its intended purpose. This includes charging
and policy control:
 
8.   Based on operator policy S-CSCF and AS shall consider the IMS
communication service identifier for online and offline charging, e.g. put
appropriate data into call detailed records.

9.   The communication service identifier shall be capable of being an input
into the policy control and charging rules.

Also:
 
2.   Based on operator policy the S-CSCF or an AS shall be able to validate
an IMS communication service identifier in a SIP request. This includes e.g.
to check the syntactical correctness of a service identifier, and policing
the usage of a communication service identifier.

10. It shall be possible to use the IMS communication service identifier as
a means to authorise whether a subscriber is allowed to initiate or receive
request for a communication service.

Another problem with UAs placing feature tags in Accept-Contact headers as
Service Identifiers is how does the initiator obtains the service identifier
of the destination party for the related service? The concept only works in
the scenario where all networks and devices have a common understanding of
the same services and the identifiers that identify them. As far as I am
aware SIP doesnt provide a mechanism to discover proprietary non standard
feature tags that are used by other devices in other domains and map them to
services that they relate to. This means that if Accept-Contact is used the
communication services are closed and can only be used by those who are part
of the closed network, (unless of course IETF were to get into standardising
services which I think is highly unlikely). If URIs are used as the
container for the Service Identifier these can be obtained using Presence as
proposed in Jonathan's draft.
 
This helps fulfil the following requirements in 3GPP TS 23.228:
 
12. The usage of communication service identifiers shall not adversely
affect interoperability between IMS networks and interoperability with
external SIP networks and CS networks. The behaviour of a network receiving
the IMS requests without an IMS communication service identifier is a matter
of operator policy. Usage of communication service identifiers shall not
decrease the level of interoperability with networks and UEs that are
unaware of the communication service identifier.

14. The usage of communication service identifiers shall not restrict the
inherent capabilities of SIP.

 
SDP is not good for routing decisions for service identification. Firstly
SDP is intended to be end to end only. It may be encrypted and understanding
the details of the service may require service or application level
understanding of the SDP and the service which is knowledge that proxies
should not need to have. Also I dont think that SDP can always uniquely
identify a service in any case. In addition not all requests that are
related to a service will contain SDP in any case.
 
Andrew


--Boundary_(ID_8oK2oZ7N4SmW/DPn91EUDA)
Content-type: text/html
Content-transfer-encoding: 7BIT

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=us-ascii">


<META content="MSHTML 6.00.2900.3020" name=GENERATOR></HEAD>
<BODY>
<DIV dir=ltr align=left><SPAN class=341354907-19022007><FONT face=Arial 
color=#0000ff size=2>Resending with some truncation (was too big to go through 
without moderator approval).</FONT></SPAN></DIV>
<DIV dir=ltr align=left><SPAN class=341354907-19022007><FONT face=Arial 
color=#0000ff 
size=2>#########################################################################</FONT></SPAN></DIV>
<DIV dir=ltr align=left><SPAN class=341354907-19022007><FONT face=Arial 
color=#0000ff size=2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=ltr align=left><SPAN class=341354907-19022007><FONT face=Arial 
color=#0000ff size=2>Andrew,</FONT></SPAN></DIV>
<DIV dir=ltr align=left><SPAN class=341354907-19022007><FONT face=Arial 
color=#0000ff size=2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=ltr align=left><SPAN class=341354907-19022007><FONT face=Arial 
color=#0000ff size=2>OK, so it seems you are talking about IMS Application 
Servers. I don't see this mentioned in Jonathan's draft, the emphasis of which 
seems to be on application selection at the terminating device. For example, in 
section 9, 4th paragraph it states:</FONT></SPAN></DIV>
<DIV dir=ltr align=left><SPAN class=341354907-19022007><FONT face=Arial 
color=#0000ff size=2>"<SPAN 
style="FONT-SIZE: 11pt; FONT-FAMILY: Arial; mso-fareast-font-family: 'Times New Roman'; mso-bidi-font-family: 'Times New Roman'; mso-ansi-language: EN-US; mso-fareast-language: EN-US; mso-bidi-language: AR-SA"><FONT 
color=#000000>Firstly, they are meant specifically and exclusively for usage in 
cases where some service logic must be present and matching on both the 
originating and terminating sides in order for any type of reasonable 
communications to exist.</FONT></SPAN>"</FONT></SPAN></DIV>
<DIV dir=ltr align=left><SPAN class=341354907-19022007><FONT face=Arial 
color=#0000ff size=2>This seems to preclude use by any network intermediaries. 
So when I talked about an AS, I meant a terminating side entity. So while I 
definitely agree that SDP is not suitable for use by intermediaries, 
it&nbsp;would be&nbsp;legitimate for a terminating application to make use of 
it. Also caller preferences is suitable for selecting between 
different&nbsp;terminating entities.</FONT></SPAN></DIV>
<DIV dir=ltr align=left><SPAN class=341354907-19022007><FONT face=Arial 
color=#0000ff size=2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=ltr align=left><SPAN class=341354907-19022007><FONT face=Arial 
color=#0000ff size=2>I have still not been convinced by arguments for having a 
service identifier that can be used by intermediaries, but one possibility has 
</FONT></SPAN><SPAN class=341354907-19022007><FONT face=Arial color=#0000ff 
size=2>emerged from these discussions that I might be able to buy. It seems 
there might be some need to convey some higher level "purpose" for a request 
that might influence routing decisions&nbsp;at SIP intermediaries, in addition 
to being used by the UAS. I think one example cited was push-to-talk (PTT), 
which might want to avoid being subject to call forwarding that normal INVITE 
requests might undergo. A user may be able to arrange that his local proxy 
retargets or forwards requests under certain conditions. If a call were marked 
as "PTT" or "chess", the user could arrange that such calls be subject to 
different rules. So the called user (or an administration acting on the called 
user's behalf) decides on the rules, which can take into account whatever the 
calling user claims to be the purpose of the call. I am not sure about the 
security implications here, because the caller might get to learn what to do to 
get around retargeting imposed by the callee.</FONT></SPAN></DIV>
<DIV dir=ltr align=left><SPAN class=341354907-19022007><FONT face=Arial 
color=#0000ff size=2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=ltr align=left><SPAN class=341354907-19022007><FONT face=Arial 
color=#0000ff size=2>Furthermore, I am not convinced by the notion of 
authorisation or charging being based on such a service identifier unless there 
are some appropriately strong security mechanism to prevent misuses by the 
caller. Charging on the basis of a service identifier is fine if the network 
really provides some added value that it would not provide in the absence of 
that service identifier. But given that network intermediaries cannot examine 
the SDP, it it not clear exactly what added value might be provided. In the 
absence of any value added by the network, uses would not use the service 
identifier and therefore avoid higher charges (presumably finding some out of 
band means to inform the called user of the wish to use a particular service). 
This is where we still seem to lack compelling examples. </FONT></SPAN></DIV>
<DIV dir=ltr align=left><SPAN class=341354907-19022007><FONT face=Arial 
color=#0000ff size=2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=ltr align=left><SPAN class=341354907-19022007><FONT face=Arial 
color=#0000ff size=2>John</FONT></SPAN></DIV>
<DIV dir=ltr align=left><SPAN class=341354907-19022007><FONT face=Arial 
color=#0000ff size=2></FONT></SPAN>&nbsp;</DIV><BR>
<BLOCKQUOTE 
style="PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #0000ff 2px solid; MARGIN-RIGHT: 0px">
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left>
  <HR tabIndex=-1>
  <FONT face=Tahoma size=2><B>From:</B> Andrew Allen [mailto:aallen@rim.com] 
  <BR><B>Sent:</B> 16 February 2007 19:59<BR><B>To:</B> Elwell, John; Malleswara 
  Rao Ankem<BR><B>Cc:</B> sipping@ietf.org<BR><B>Subject:</B> RE: Device 
  architecture? WAS: RE: [Sipping] Service Identificatio 
  ninSIP<BR></FONT><BR></DIV>
  <DIV></DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff 
  size=2></FONT></SPAN>&nbsp;</DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff 
  size=2>John</FONT></SPAN></DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff 
  size=2></FONT></SPAN>&nbsp;</DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff size=2>I think whether 
  an AS in the terminating network is considered a contact is a debatable point. 
  Certainly it doesnt register feature tags using RFC 3840 mechanisms and (at 
  least in IMS) they dont show up as registered contacts in the reg-event 
  notifications. Also AS in the originating network (which may need to be 
  involved in some communication services) are certainly not contacts for the 
  resource identified by the Request-URI and Accept-Contact certainly has no 
  defined role in routing to them.</FONT></SPAN></DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff 
  size=2></FONT></SPAN>&nbsp;</DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff size=2>In addition 
  3GPP TS 23.228 v7.6.0 subcaluse 4.13.2 specifies 15 requirements for IMS 
  Communication Service Identifier and only one of these is regarding matching 
  to the correct contact:</FONT></SPAN></DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff 
  size=2></FONT></SPAN>&nbsp;</DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007>
  <P class=B1 
  style="MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: 23228CR0542R1 20051123T1735"><SPAN 
  lang=EN-GB><FONT size=2>11.<SPAN style="mso-tab-count: 1"> </SPAN>The 
  communication service identifier shall be taken into account when selecting 
  the correct UE(s), if multiple UEs are registered for the same Public User 
  Identity(s).</FONT></SPAN></P></SPAN></DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff 
  size=2></FONT></SPAN>&nbsp;</DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff size=2>&nbsp;Several 
  of the other requirements are in my view totally incompatible with 
  Accept-Contact being the transport for the identifier and indeed may limit the 
  use of Accept-Contact for its intended purpose. This includes charging and 
  policy control:</FONT></SPAN></DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff 
  size=2></FONT></SPAN>&nbsp;</DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007>
  <P class=B1 
  style="MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: 23228CR0542R1 20051123T1735"><SPAN 
  lang=EN-GB><FONT size=2>8.<SPAN style="mso-tab-count: 1">&nbsp;&nbsp; 
  </SPAN>Based on operator policy S-CSCF and AS shall consider the IMS 
  communication service identifier for online and offline charging, e.g. put 
  appropriate data into call detailed records.</FONT></SPAN></P>
  <P class=B1 
  style="MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: 23228CR0542R1 20051123T1735"><SPAN 
  lang=EN-GB><FONT size=2>9.<SPAN style="mso-tab-count: 1">&nbsp;&nbsp; 
  </SPAN>The communication service identifier shall be capable of being an input 
  into the policy control and charging rules.</FONT></SPAN></P></SPAN></DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff 
  size=2>Also:</FONT></SPAN></DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff 
  size=2></FONT></SPAN>&nbsp;</DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007>
  <P class=B1 
  style="MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: 23228CR0542R1 20051123T1735"><SPAN 
  lang=EN-GB><FONT size=2>2.<SPAN style="mso-tab-count: 1">&nbsp;&nbsp; 
  </SPAN>Based on operator policy the S-CSCF or an AS shall be able to validate 
  an IMS communication service identifier in a SIP request. This includes e.g. 
  to check the syntactical correctness of a service identifier, and policing the 
  usage of a communication service identifier.</FONT></SPAN></SPAN></P></DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007>
  <P class=B1 
  style="MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: 23228CR0542R1 20051123T1735"><SPAN 
  lang=EN-GB><FONT size=2>10.<SPAN style="mso-tab-count: 1"> </SPAN>It shall be 
  possible to use the IMS communication service identifier as a means to 
  authorise whether a subscriber is allowed to initiate or receive request for a 
  communication service.</FONT></SPAN></P></SPAN></DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff size=2>Another problem 
  with UAs placing feature tags in Accept-Contact headers as Service Identifiers 
  is&nbsp;how does the initiator obtains the service identifier of the 
  destination party for the related service? The concept only works in the 
  scenario where all networks and devices have a common understanding of the 
  same services and the identifiers that identify them. As far as I am aware SIP 
  doesnt provide a mechanism to discover proprietary non standard feature tags 
  that are used by other devices in other domains and map them to services that 
  they relate to. This means that if Accept-Contact is used the communication 
  services are closed and can only be used by those who are part of the closed 
  network, (unless of course IETF were to get into standardising services which 
  I think is highly unlikely). If URIs are used as the container for the Service 
  Identifier these can be obtained using Presence as proposed in Jonathan's 
  draft.</FONT></SPAN></DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff 
  size=2></FONT></SPAN>&nbsp;</DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff size=2>This helps 
  fulfil the following requirements in 3GPP TS 23.228:</FONT></SPAN></DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff 
  size=2></FONT></SPAN>&nbsp;</DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007>
  <P class=B1 style="MARGIN: 0in 0in 9pt 28.4pt"><SPAN lang=EN-GB><FONT 
  size=2>12.<SPAN style="mso-tab-count: 1"> </SPAN>The usage of communication 
  service identifiers shall not adversely affect interoperability between IMS 
  networks and interoperability with external SIP networks and CS networks. The 
  behaviour of a network receiving the IMS requests without an IMS communication 
  service identifier is a matter of operator policy. Usage of communication 
  service identifiers shall not decrease the level of interoperability with 
  networks and UEs that are unaware of the communication service 
  identifier.</FONT></SPAN></P>
  <P class=B1 
  style="MARGIN: 0in 0in 9pt 28.4pt; mso-prop-change: '23.228CR0567' 20060309T1718"><SPAN 
  lang=EN-GB><FONT size=2>14.<SPAN style="mso-tab-count: 1"> </SPAN>The usage of 
  communication service identifiers shall not restrict the inherent capabilities 
  of SIP.</FONT></SPAN></P></SPAN></DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff 
  size=2></FONT></SPAN>&nbsp;</DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff size=2>SDP is not good 
  for routing decisions for service identification. Firstly SDP is intended to 
  be end to end only. It may be encrypted and understanding&nbsp;the details of 
  the service may require service or application level understanding of the SDP 
  and the service which is knowledge that proxies should not need to have. Also 
  I dont think that SDP can always uniquely identify a service in any case. In 
  addition not all requests that are related to a service will contain SDP in 
  any case.</FONT></SPAN></DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff 
  size=2></FONT></SPAN>&nbsp;</DIV>
  <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><SPAN 
  class=778002917-16022007><FONT face=Arial color=#0000ff 
  size=2>Andrew</FONT></SPAN></DIV></BLOCKQUOTE></BODY></HTML>

--Boundary_(ID_8oK2oZ7N4SmW/DPn91EUDA)--


--===============1630893377==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1630893377==--




From sipping-bounces@ietf.org Mon Feb 19 06:02:07 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJ6Fu-00021P-8F; Mon, 19 Feb 2007 06:00:26 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJ6Fs-000211-R2
	for sipping@ietf.org; Mon, 19 Feb 2007 06:00:24 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HJ6Fk-0001Ap-Mh
	for sipping@ietf.org; Mon, 19 Feb 2007 06:00:24 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	12AC1205A7; Mon, 19 Feb 2007 12:00:14 +0100 (CET)
X-AuditID: c1b4fb3c-b07c9bb0000007de-df-45d9833d749c 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	DB47C20537; Mon, 19 Feb 2007 12:00:13 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 19 Feb 2007 12:00:13 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Mon, 19 Feb 2007 12:00:11 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD028BD258@esealmw111>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C0091516B6@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdNmH89O4T46iGoRjicZ3hYknVpawB8E6oQAH++leAABzLk8AAYIixwAIPObeA=
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: "Andrew Allen" <aallen@rim.com>,
	<sipping@ietf.org>
X-OriginalArrivalTime: 19 Feb 2007 11:00:13.0511 (UTC)
	FILETIME=[216BD570:01C75415]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 2bcdb6f0ba9636bf286b7e28ccb8bd9b
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi Andrew,

I see that in your assumptions, the originating and terminating networks
use different service identifiers, and I presume that this is for the
same service.  In my understanding, it is the same service identifier
that is used end-to-end (if not it is interworked).  That is why I do
not see the reliance on Presence being necessary.  In my opionion this
totally within the requirements of TS 23.228 (I think that different
service identifiers in different networks for the same service is
against TS 23.228).

This comes to the isssue of the PSTN gateway - I presume that you want
to go to the PSTN gateway if the service is telephony, and within that,
voice.  If we know that, then the we go to any PSTN gateway, and we
can't presume a particular network for a PSTN gateway.=20

Best Regards,

Steve.

-----Original Message-----
From: Andrew Allen [mailto:aallen@rim.com]=20
Sent: Saturday, February 17, 2007 12:51 AM
To: Stephen Terrill (MI/EEM); sipping@ietf.org
Cc: Andrew Allen
Subject: RE: [Sipping] Service Identification in SIP



Hi Steve,

My responses inline

Andrew=20

-----Original Message-----
From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
Sent: Friday, February 16, 2007 3:51 AM
To: sipping@ietf.org
Subject: RE: [Sipping] Service Identification in SIP

Hello Andrew,

Thanks for your comments.  I have a couple of brief responses below.

Best Regards,

..//Steve=20

-----Original Message-----
From: Andrew Allen [mailto:aallen@rim.com]
Sent: Friday, February 16, 2007 7:15 AM
To: Stephen Terrill (MI/EEM); sipping@ietf.org
Cc: Andrew Allen
Subject: RE: [Sipping] Service Identification in SIP


Some comments on Stephen's comments on the draft:
=20

-Moving onto the proposed solution draft itself, in addition for
providing for the identification of the "application" and the
"communication service", I have a few additional thoughts.

The draft seems to have the implication that all services (communication
services in other terminology) that are to be identified have to be
standardised within the IETF.  This is clearly an restriction that will
not be acceptable to the industry as a whole.  My guess is that if such
a statement is mandated by the IETF, it will just be words on paper
which would likely be ignored in practice by other standardisation
organisations.

[AA] This I think is an incorrect interpretation of the draft. My
understanding is that the only services according to the draft that
require some standardisation in IETF are those that by their very nature
cannot interoperate unless there is some common additional understanding
between the endpoints outside of what is currently specified in SIP. The
use of SIP option tags with Require header is proposed by the draft for
this purpose as such services create interoperability issues with SIP.

<Steve>
My comment applies even for these services </Steve>


<AA>
Well then we are in big trouble because what you are proposing is that
its OK for bodies outside of IETF to create non interoperable extensions
for the behavour of SIP entities. This will result in the fragmentation
of SIP into different flavors that don't interoperate. This is the
balkanization of SIP that Paul talks about.even <AA>=20

----------------------------------------------------

The draft also only recognises the case where all explicitly knowledge
of the service is required.  I think it would also be good to allow the
case where knowledge of the service would be helpful, but not explicitly
required.  This could be a case where the service to the end-user would
not be degraded if this was not supported.  This functionality seems to
be missing.

[AA] Service URIs are defined in the draft for these services where the
support by the other endpoint is not required for the session to be
sucessfully established.

The draft seems to indicate that that application servers in the network
are involved only at the request of the originating terminal.  This
seems a strange restriction.  For some network support that supports the
originating user, such as the one provided in the example "I would like
this call recorded for later retrieval", such mechanisms may make sense,
however as a general approach it seems a bit limited.  For example, in
3GPP networks there is the need to invoke application logic in the
originating side and terminating side, not only at the request of the
originating terminal, but as a consequence of the requested service.  If
there is an SIP message contains indication of the requested service, a
proxy can do this (or loose route to an AS that contains the necessary
service logic).=20

[AA] The draft supports both possibilities using the Route header for
originating network servers that need to be involved in the route and
based on the terminating networks routing policy for the Service URI in
the Request URI.

<Steve>
I don't see the advantage of having different solutions for the
originating network and the terminating network.  I don't think that the
originating logic is invoked only at the request of the originating
terminal.
</Steve>

<AA>
Unless IETF is willing to standardise all services and mandate that all
SIP entities understand these (which will never happen) then we cannot
assume that all networks and all endpoints understand and use the same
identifiers. It is possible through configuration for an endpoint to
obtain the identifiers used by its own service provider but there is no
way to ensure that other networks use or understand the same identifiers
for the equivalent or similar service that is compatible at the session
layer. Therefore we need to support the possibility to have different
identifiers in different networks in order to not violate the TS 23.228
requirement:

12.	The usage of communication service identifiers shall not
adversely affect interoperability between IMS networks and
interoperability with external SIP networks and CS networks. The
behaviour of a network receiving the IMS requests without an IMS
communication service identifier is a matter of operator policy. Usage
of communication service identifiers shall not decrease the level of
interoperability with networks and UEs that are unaware of the
communication service identifier.

If the originating service logic isn't invoked based on an identifier
placed in the request by the originating terminal on what basis do you
think it is invoked? How does the use of Accept-Contact have advantages
in this case?
<AA>=20

--------------------------------------

 I note here that the proposed service URN does not contain the address
of a specific server containing the application logic.  An AS can look
at that, and loose route to an application server that does contain such
logic (If it is separated from the proxy), however I fail to see the
difference for that proxy if it looks at a feature tag, or it looks at
the service URN for this.  In both cases, it has to parse the message
for the data, and modify the routing information. =20

[AA]Route headers are the SIP mechanism for requesting intermediaries to
be includede on the route of the request. There is nothing in RFC
3840/RFC 3841 that addresses using feature tags in Accept-Contact header
for any routing decisions prior to determining which of the registered
contacts the request should be forked to. Accept-Contact is not a Route
header it is merely a filter that is applied to the set of valid
contacts. Using a parameter in the filter as a service identifier has a
potential significant impact on the use of Accept-Contact for its
original purpose.=20

<Steve>
I don't believe there is anthing that states a Proxy CAN NOT route to
applications or application logic based upon elements in the SIP
message, including what is in the Accept-Contact header (some people
have even proposed SDP ...).
</Steve>

<AA>
As I had stated in an earlier email on this subject a proxy can do a lot
of things (including routing based on your favorite number appearing in
a call-Id) but that doesn't mean they are a good idea or a good design
decision. If you want to ensure you don't break things then you need to
be careful to use SIP headers with the semantics intended for them
otherwise problems can result if extra semantics are put upon existing
headers.
<AA>=20

Further, the route header could be removed in the originating network;
and it is still required to identify the service in the terminating
network.  As proxies may remove the route, it can't be used to transport
the requested service. =20

[AA] Proxies arbitarily removing Route headers is a BAD thing akin to a
bull in a china shop - almost certainly going to break something. RFC
3261 summarises what proxies should be doing with Route headers:
16.12 Summary of Proxy Route Processing

   In the absence of local policy to the contrary, the processing a
   proxy performs on a request containing a Route header field can be
   summarized in the following steps.

      1.  The proxy will inspect the Request-URI.  If it indicates a
          resource owned by this proxy, the proxy will replace it with
          the results of running a location service.  Otherwise, the
          proxy will not change the Request-URI.

      2.  The proxy will inspect the URI in the topmost Route header
          field value.  If it indicates this proxy, the proxy removes it
          from the Route header field (this route node has been
          reached).

      3.  The proxy will forward the request to the resource indicated
          by the URI in the topmost Route header field value or in the
          Request-URI if no Route header field is present.  The proxy
          determines the address, port and transport to use when
          forwarding the request by applying the procedures in [4] to
          that URI.

   If no strict-routing elements are encountered on the path of the
   request, the Request-URI will always indicate the target of the
   request.

<Steve>
The first sentance is interesting "In absence of local policy to the
contrary", indicating that there can be different policies.  I
understand that some proxies do remove this.  While I am on this - I
have lead to believe that R-URI parameters may also be removed by
intermediate proxies - is this correct?
</Steve>

<AA>
My understanding is that local policy may determine that additional
nodes need to be routed to in addition to the Route contained in the
request. However deleting existing Route headers other than the Route
header containing the address of the proxy as part of the loose routing
procedures is liable to break things and is not part of RFC 3261 routing
(loose routing).
<AA>=20


[AA] Generally a UA wouldn't know the appropriate service identifiers
used in the terminating network or whether a Route header is appropriate
to be used in the terminating network. The draft defines the Service URI
for the purpose of identifying services in the terminating network and
service URIs can be obtained using Presence.

<Steve>
I don't get the above paragraph.  It suggest that the "appropriate
service identifier" used in the terminating network is not known by the
originating UE, but the Service URI is.  I thought that the service URI
is an encoding of the "appropriate service identifier".
</Steve>

<AA>=20
Yes but in the general case the UA doesn't need to know what the service
identifier is it obtains through Presence a Service URI that contains
the Service identifier as a URI parameter. The great thing about using
Presence to obtain a Service URI is that the originating UA doesn't even
need to know anything about service identifiers at all for this to work.
It just uses Presence to find the URI to use for the service which can
be indicated to the user via an indication such as a graphic icon (e,g a
phone for telephony, smiley face for IM, etc) in the presence document.
<AA>=20

-----------------------------------

>From this perspective it is not clear to me the advantage of using a
service URN in a route header compared to the use of a feature tag.

[AA] I hope the above responses clarify that somewhat. Also the Service
URN and Service URIs allow different service identifiers to be used in
the originating and terminating network for session compatiable services
instead of forcing everyone to deploy and understand the same service
and service identifier.

<Steve>
Nope, don't get it, but are information that is carried end-to-end ....
And as far as I see, could be used by intermediate proxies to impact
local routing policy ... Or ?
</Steve>

<AA>=20
I don't understand this response either. The key point is that UAs in
corporate networks and in the internet are not going to deploy 3GPP
flavor services and use the same 3GPP service identifiers or even
possibly any service identifiers at all. The Accept-Contact based
proposal assumes that all endpoints understand the service identifiers
to be used for each service. This concept is not generally applicable
and create interoperability problems.
<AA>=20

The draft seems to mandate GRUU - is this really so, is that really
necessary.  It seems strange to place such a dependency, it also starts
to move the GRUU as being the primary means of which to contact a user.

[AA] My understanding is that GRUU in the draft is used to address how a
request addressed to Service URI is forked only to those contacts that
support the service identified by the that Service URI. A contact would
seem to be fine so long as it had GRUU properties - no need to obtain a
GRUU if you already have one.

This implies that the originating user decides which terminal I should
receive my requested service on, even if I have two devices for that.
As such, I don't believe that there should be a dependency on GRUU.

[AA] If you have obtained a Service URI (probably through presence) for
that user then you can address the request to the Service URI which
represents all the terminals that support that service. However GRUU and
the Service Instance URIs created from GRUUs provide another mechanism
through which a user can obtain a URI to reach that user for the
required service.

Parts of the draft seems to rely on the use of presence.  While presence
could be a supportive mechanism, I doubt that we can mandate that the
number you want to dial has be available in presence as I assume not all
entities are available in presence (diallable network services, users in
the PSTN).  This would be strange to mandated.

[AA] PSTN gateways are presumably in the same domain as the terminating
proxy and therefore these PSTN gateways can be configured with the
appropriate service identifiers for telephony services and can include
this in either a Route header to route to the right application server
or can construct a valid Service URI. The real problem is how do users
in other networks know what the service identifier that's is used in the
terminating network is. The use of Service URIs and Presence provides a
solution to that problem.

<Steve>
Again, presence could be a useful mechanism, but I think it would be a
problem for the draft to rely it.  A user making a call to the PSTN may
not find that user in the presence archicture (and similar to services
like the 800 or 900 services).  (I should note, that you shouldn't make
any assumptions about the location of the PSTN gateway ... But I don't
think that changes anything). =20
</Steve>

<AA>=20
How does a PSTN gateway in another domain from the terminating network
know the service identifier for telephony used in the terminating
network with the Accept-Contact header based proposal?=20
<AA>=20

Moving on from the draft I have seen the questioned raised about "why
can't we expand the power of SDP".  Such an approach would mandate that
SIP methods use SDP, I don't think that is appropriate.=20

[AA] I agree with you here

I hear the appeals that the mechanisms should not be abused, and that we
should not have 1000s of separate "communication services", and I fully
agree with this. I see a smaller number of communication services, not a
larger number; where those communication services are used to access
other end-user-services.

*flame away*

Regards,

Steve

---



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential
information, privileged material (including material protected by the
solicitor-client or other applicable privileges), or constitute
non-public information. Any use of this information by anyone other than
the intended recipient is prohibited. If you have received this
transmission in error, please immediately reply to the sender and delete
this information from your system. Use, dissemination, distribution, or
reproduction of this transmission by unintended recipients is not
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 19 06:07:33 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJ6Mn-00077v-6n; Mon, 19 Feb 2007 06:07:33 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJ6Ml-00077a-MZ
	for sipping@ietf.org; Mon, 19 Feb 2007 06:07:31 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HJ6Md-0002p5-Ma
	for sipping@ietf.org; Mon, 19 Feb 2007 06:07:31 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	31AE2200BD
	for <sipping@ietf.org>; Mon, 19 Feb 2007 12:07:23 +0100 (CET)
X-AuditID: c1b4fb3e-af6d2bb0000007e1-c3-45d984eb2b89 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	17320204CC
	for <sipping@ietf.org>; Mon, 19 Feb 2007 12:07:23 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 19 Feb 2007 12:07:22 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Mon, 19 Feb 2007 12:07:21 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD028BD26C@esealmw111>
In-Reply-To: <45D65EE8.7010501@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdSta1zIjJPiBIdRCKey9Fq0m/u7QBX7bWQ
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: "sipping" <sipping@ietf.org>
X-OriginalArrivalTime: 19 Feb 2007 11:07:22.0366 (UTC)
	FILETIME=[2109E9E0:01C75416]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 2d133cc328f58695161c98bb4f4dc213
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi,

Besides the discussion that PoC is "only the addition of floor control"
as there is more to it than that ...  The idea of being able to use a
Require header for some applications is not bad(though should not be
mandeated, i.e it should be possible to not use it as well while still
identifying the service).  I think that the implication that the use of
the the requires header then requires that the related specification
activity is constrained to the IETF is problematic for other forums - is
there a way around that?

Best Regards,

Steve.

-----Original Message-----
From: Jonathan Rosenberg [mailto:jdrosen@cisco.com]=20
Sent: Saturday, February 17, 2007 2:48 AM
To: Henning Schulzrinne
Cc: sipping
Subject: Re: [Sipping] Service Identification in SIP

I think its a fair point that the example is contrived. Really, the main
use case being targeted is IMS PoC. So, let me talk a bit about that.

One would think that PoC is just a regular SIP call that has floor
control added to it as another media type, and so one could indicate
that a call is a PoC call just because of the presence of this new media
type in the SDP. Sadly, I wish that were true. If it was, we would not
need a mechanism here.

Unfortunately, PoC has layered a whole bunch of additional behaviors
into the way SIP requests are handled. Dale has indicated one of them -
that a 200 OK is returned on the terminating side priod to the call
actually being completed. PoC also *requires* the insertion of servers
into the SIP signaling path to make the system work. Consequently,
though the messages that PoC uses *look* like SIP messages with a new
media type, there is a hidden 'extension' in place that changes the
fundamental semantics of INVITE. The fact that this extension is not
visible as a specific header in a well defined way does not change the
fact that PoC CHANGES THE BEHAVIOR OF SIP AND IS THUS A SIP EXTENSION.

To convince yourself of this, consider what would happen if a PoC client
calls a regular SIP phone. The regular SIP phone would see the PoC floor
control stream and reject it as unknown. If PoC were just SIP as we've
specified it, then everything would be fine, and the call would proceed
but without floor control.

But alas, it is not so.

Because PoC makes all kinds of assumptions about the behaviors of its
participants, and what the real meaning is for various SIP messages, the
entire thing just won't work when a call terminates on a non-PoC device.

This is because PoC is, effectively, a fairly significant and
non-backwards-compatible extension to SIP.

So, my proposal is, if you are going to go and DO something like that,
you better use the SIP techniques that we've defined to deal with it.=20
And when you extend SIP in a way that doesn't work with a SIP endpoint
that doesn't understand an extension, you use a Require header. And that
is exactly what I'm proposing that 3gpp should do. Use Require because
you have made a non-backwards-compatible extension to SIP.

The general problem here is that there are other places where people are
layering additional 'context', as Loreto has called it, ontop of SIP
signaling. This is a worrisome trend because this context is subtle and
not explicitly signaled. Its just assumed. And that is dangerous, since
it leads to the kind of balkanization and non-interoperability that
myself, Paul, Dale and others have complained about in this thread.

The current 3gpp specs, instead of using Require to signal this
extension, put a tag into the Accept-Contact header field. I find this
extremely worrisome, because it is really just masking the reality of
what has happened. If you've got a situration where a SIP call *cannot
be processed* without going through application servers, then this is
not something that is signaled with Accept-Contact. Its signaled with
Require or Proxy-Require as appropriate.

Thanks,
-Jonathan R.

Henning Schulzrinne wrote:

> I have to admit that I find a discussion that uses a single example a=20
> bit peculiar, particularly since the example is rather contrived. =20
> (Why would you use voice, plus speech-to-text presumably, to talk to =20
> a chess
> application?)
>=20
> In a networking sense, the issue seems to be that we're using voice =20
> as a transport protocol to carry different distinct applications. (In

> the old days, a modem would have been such an application.)
>=20
> I'll repeat what I've said before: SIP doesn't exactly suffer from a=20
> lack of complexity and mechanisms. At this point, before adding more=20
> machinery, we might want to ask as to whether there is truly a broad-=20
> based need for justifying additional machinery.
>=20
> With regards to service URNs, I happen to believe that they are very=20
> reasonable as "modern" replacements for tel URIs, essentially =20
> extending the concept of 800 numbers. But that's a different problem =20
> from what I understand this discussion to be about. (In that usage, =20
> service URNs would be mapped to SIP AORs based on geography, time-of-=20
> day and other
> factors.)
>=20
> Henning
>=20
>=20
> On Feb 9, 2007, at 10:04 AM, Salvatore Loreto (JO/LMF) wrote:
>=20
>> Hi Ravi,
>>
>> if you consider the Jonathan draft, he says that "...chess game that=20
>> utilizes voice commands to move pieces.  When  one player says "Queen

>> to D7", the software on their phone  recognizes this and moves the=20
>> piece.  At the same time, the user's  voice is sent to the other=20
>> player, where it is both rendered to the  user, and interpreted=20
>> locally in order to move the piece."
>> and he suggests that the following urn
>>
>>    urn:service:games:voice-chess
>>
>> could be used to identify the voice chess application.
>>
>> I guess that with voice he supposes audio, but nowadays we can also=20
>> have voice using PoC. Then we need a way to specify which kind of=20
>> voice we want: a standard full-duplex or a PoC. Maybe the chess  game

>> works well with, or it is designed for a standard audio but  not for=20
>> a PoC audio. So we need to address also the voice  preference we=20
>> want. I like your proposal, having something  hierarchical was=20
>> something I was thinking too I don't know if it is  your idea, but=20
>> with the service urn you are proposing:  Urn:service:voice.poc.chess=20
>> I imaging you are asking to play a  Chess game, that work utilizing=20
>> voice commands, but specifically  voice over PoC. This is also=20
>> compliant to the
>> draft-ietf-ecrit- service-urn (see section 3) "Declaration of=20
>> syntactic structure:  The URN consists of a hierarchical service=20
>> identifier, with a  sequence of labels separated by periods.  The=20
>> left-most label is  the most significant one and is called 'top-level

>> service', while  names to the right are called 'sub-services'.  The=20
>> set of allowable  characters is the same as that for domain names [1]

>> and a subset of  the labels allowed in [5].  Labels are=20
>> case-insensitive and MUST be  specified in all lower-case.  For any=20
>> given service URN, service- identifiers can be removed right-to-left=20
>> and the resulting URN is  still valid, referring a more generic=20
>> service.  In other words, if  a service 'x.y.z' exists, the URNs 'x'
>> and 'x.y' are also valid  service URNs." Regards sal On Fri,=20
>> 2007-02-09 at 13:44 +0530, ravishankar.shiroor@wipro.com  wrote:
>>
>>> Hi Salvatore, As an end user, poc is also an application that I  use

>>> and chess is also an application that I can use. The way I see  it,=20
>>> in your example, it will be running chess application on top  of PoC

>>> application. So the service urn for this would be something  like:
>>> Urn:service:voice.poc.chess Regards, Ravi. -- Ravishankar.  G.=20
>>> Shiroor Wipro Technologies, Bangalore. =20
>>> ravishankar.shiroor@wipro.com
>>> -- > -----Original Message----- >  From: Salvatore Loreto (JO/LMF) >

>>> [mailto:salvatore.loreto@ericsson.com] > Sent: Wednesday, February=20
>>> 07, 2007 1:37 PM > To: RaviShankar Shiroor (WT01 - Product=20
>>> Engineering Solutions) > Cc: Gonzalo.Camarillo@ericsson.com;=20
>>> sipping@ietf.org; > mary.barnes@nortel.com > Subject: RE:  [Sipping]

>>> Service Identification in SIP > > hi Ravi, > > in IMS a  Service is=20
>>> a "type of communication defined by a > service  definition that=20
>>> specifies the rules and procedures > and allowed  media, if any, for

>>> a specific type of > communication and that  utilises the IMS=20
>>> enabler. An IMS > communication service defines  what is possible=20
>>> within a > single SIP session or standalone  transaction." > So a=20
>>> Service implies some combination of  capabilities and > ways to use=20
>>> those capabilities. > If you want  the Service specifies the context

>>> for a communication. > > Instead  the Application define the=20
>>> specific communication (or > if you  want application) that you want

>>> run in the specific > context. So  you should have several different

>>> applications > designed to run  on a specific environment or that=20
>>> have > different performance or  behavior in different context >=20
>>> (running on top of different  service). > > Lets go to Jonathan's=20
>>> example of the Chess. > In his  example the Chess is a service, for=20
>>> IMS instead is an >  application that you can run on top of=20
>>> different service. > For  example you can run Chess application on=20
>>> top of PoC > service or  on top of the "standard" service. So let=20
>>> say you > run Chess on  top of PoC service, then you can use only=20
>>> voice > in the way that  PoC specifies. > > /sal > > > > > On Thu,=20
>>> 2007-02-01 at
>>> 16:39  +0530, > ravishankar.shiroor@wipro.com wrote: > > Hello, > >=20
>>> >
>>> >  IMO, Service Identification is definitely needed and should > be
>>> taken > > up as a WG as proposed below. > > > > Further, I think=20
>>> Service Identification scope should be as per the > > requirement =20
>>> in Jonathan's draft, as I still do not > understand the need > > =20
>>> for having a Communication Service Identifier and another > =20
>>> Application
>>> > > Identifier as given in the 3gpp-ics-requirements  document. > >=20
>>> > > >
>>> > Regards, > > Ravi. > > > > -- > > > >  Ravishankar. G. Shiroor > >
>>> Wipro Technologies, Bangalore. > > > > =20
>>> ravishankar.shiroor@wipro.com
>>> > > -- > > > > > > > -----Original  Message----- > > > From: Gonzalo
>>> Camarillo  [mailto:Gonzalo.Camarillo@ericsson.com] > > > Sent:=20
>>> Wednesday,  January 31, 2007 8:30 PM > > > To: sipping > > > Cc:=20
>>> Mary Barnes >  > > Subject: [Sipping] Service Identification in SIP=20
>>> > > >
>>> > > >  Folks, > > > > > > during the last year, there have been a
>>> number  of threads on the > > > mailing list discussing service=20
>>> identification in SIP. In > > > particular, the following thread=20
>>> discussed 3GPP > requirements on this > > > area: > > > > > > >=20
>>> http://www1.ietf.org/mail-archive/web/sipping/current/ msg10562.html
>>> > > > > > > At a later point, those 3GPP  requirements were
>>> documented in the > > > following (expired by  now) Internet Draft,=20
>>> which was > presented in a > > > face-to-face  IETF meeting. > > > >
>>> > > http://www.watersprings.org/pub/id/draft- loreto-sipping-3gpp-i=20
>>> > > > cs-requirements-00.txt > > > > > >  Lately, the following=20
>>> > > draft
>>> has been submitted to the archives: >  > > > > >
>>> http://www.ietf.org/internet-drafts/draft-rosenberg- sipping-se > >=20
>>> > rvice-identification-00.txt > > > > > > This  draft proposes a=20
>>> solution that intends to meet known > > >  requirements in this
area.
>>> > > > > > > Given the amount of  discussions on issues related to
>>> service > > > identification we  have seen during the last year, we=20
>>> > believe this is > > > a  fairly important topic for the SIPPING WG

>>> to tackle. At some > > >  point, we would like to have a milestone=20
>>> with its > associated WG  item > > > dealing with this issue. > > >=20
>>> > > > Therefore, we  would like to encourage people to read this >=20
>>> draft and > > > send  comments to the list. Additionally, if you=20
>>> believe that this > > >  work is (or is not) important, we would=20
>>> like to hear that as well.  > > > > > > Cheers, > > > > > > Gonzalo=20
>>> > > > SIPPING co-chair > >  > > > > > > >=20
>>> _______________________________________________ > >  > Sipping=20
>>> mailing list > https://www1.ietf.org/mailman/listinfo/ sipping > > >

>>> This list is for NEW development of the application  of SIP Use  > >
>>> > sip-implementors@cs.columbia.edu for questions  on current sip Use
>>> > > > sip@ietf.org for new developments of core  SIP > > > > > > >
>>> _______________________________________________  > > Sipping mailing

>>> list https://www1.ietf.org/mailman/listinfo/ sipping > > This list=20
>>> is for NEW development of the application of  SIP Use > >=20
>>> sip-implementors@cs.columbia.edu for questions on  current sip Use >
>>> > sip@ietf.org for new developments of core SIP  > > The information
>>> contained in this electronic message and any  attachments to this=20
>>> message are intended for the exclusive use of  the addressee(s) and=20
>>> may contain proprietary, confidential or  privileged information. If

>>> you are not the intended recipient, you  should not disseminate,=20
>>> distribute or copy this e-mail. Please  notify the sender=20
>>> immediately and destroy all copies of this  message and any
attachments. WARNING:
>>> Computer viruses can be  transmitted via email. The recipient should

>>> check this email and  any attachments for the presence of viruses.
>>> The company accepts  no liability for any damage caused by any virus

>>> transmitted by  this email. www.wipro.com
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP Use=20
>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
>> sip@ietf.org for new developments of core SIP
>=20
>=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20

--=20
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 19 08:34:48 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJ8eT-0006Cc-6D; Mon, 19 Feb 2007 08:33:57 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJ8eQ-0006CD-Rp
	for sipping@ietf.org; Mon, 19 Feb 2007 08:33:54 -0500
Received: from usaga01-in.huawei.com ([206.16.17.211])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HJ8eM-0003mb-Jd
	for sipping@ietf.org; Mon, 19 Feb 2007 08:33:54 -0500
Received: from huawei.com (usaga01-in [172.18.4.6])
	by usaga01-in.huawei.com (iPlanet Messaging Server 5.2 HotFix 1.25
	(built Mar
	3 2004)) with ESMTP id <0JDP00K7PPOBX0@usaga01-in.huawei.com> for
	sipping@ietf.org; Mon, 19 Feb 2007 05:33:47 -0800 (PST)
Received: from s73602 (cpe-72-190-0-23.tx.res.rr.com [72.190.0.23])
	by usaga01-in.huawei.com
	(iPlanet Messaging Server 5.2 HotFix 1.25 (built Mar  3 2004))
	with ESMTPA id <0JDP00KBDPO99N@usaga01-in.huawei.com> for
	sipping@ietf.org; Mon, 19 Feb 2007 05:33:47 -0800 (PST)
Date: Mon, 19 Feb 2007 07:33:25 -0600
From: Spencer Dawkins <spencer@mcsr-labs.org>
Subject: Re: [Sipping] Service Identification in SIP
To: sipping@ietf.org
Message-id: <009301c7542a$88d1f810$6501a8c0@china.huawei.com>
MIME-version: 1.0
X-MIMEOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
X-Mailer: Microsoft Outlook Express 6.00.2900.2869
Content-type: text/plain; format=flowed; charset=iso-8859-1;
	reply-type=original
Content-transfer-encoding: 7BIT
X-Priority: 3
X-MSMail-priority: Normal
References: <989763567E51AB48BE015DF7BB960CAD028BD258@esealmw111>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 9182cfff02fae4f1b6e9349e01d62f32
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi, Steve,

The distinction you're making here seems to capture the the biggest 
question - are we expecting service identifiers to be end-to-end, if not 
interworked, or are we expecting them to be different for the same service 
in originating and terminating networks.

Thanks,

Spencer

From: "Stephen Terrill (MI/EEM)" <stephen.terrill@ericsson.com>

Hi Andrew,

I see that in your assumptions, the originating and terminating networks
use different service identifiers, and I presume that this is for the
same service.  In my understanding, it is the same service identifier
that is used end-to-end (if not it is interworked).  That is why I do
not see the reliance on Presence being necessary.  In my opionion this
totally within the requirements of TS 23.228 (I think that different
service identifiers in different networks for the same service is
against TS 23.228). 



_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 19 09:07:38 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJ9A9-0003R4-8s; Mon, 19 Feb 2007 09:06:41 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJ9A7-0003Qy-Fq
	for sipping@ietf.org; Mon, 19 Feb 2007 09:06:39 -0500
Received: from jalapeno.cc.columbia.edu ([128.59.29.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HJ9A6-0008S0-9i
	for sipping@ietf.org; Mon, 19 Feb 2007 09:06:39 -0500
Received: from [192.168.0.41] (pool-141-153-194-152.mad.east.verizon.net
	[141.153.194.152]) (user=hgs10 mech=PLAIN bits=0)
	by jalapeno.cc.columbia.edu (8.13.7/8.13.6) with ESMTP id
	l1JE6UPF004418
	(version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NOT);
	Mon, 19 Feb 2007 09:06:30 -0500 (EST)
In-Reply-To: <7374777208BDC7449D5620EF94232567021CE512@esealmw113.eemea.ericsson.se>
References: <7374777208BDC7449D5620EF94232567021CE512@esealmw113.eemea.ericsson.se>
Mime-Version: 1.0 (Apple Message framework v752.2)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <6E0BD8DB-DF70-4CBE-AA8F-DB69CECF014E@cs.columbia.edu>
Content-Transfer-Encoding: 7bit
From: Henning Schulzrinne <hgs@cs.columbia.edu>
Subject: Re: [Sipping] Service Identification in SIP - Routing decissions
	based on Require/P-R
Date: Mon, 19 Feb 2007 09:06:27 -0500
To: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
X-Mailer: Apple Mail (2.752.2)
X-No-Spam-Score: Local
X-Scanned-By: MIMEDefang 2.48 on 128.59.29.5
X-Spam-Score: 0.0 (/)
X-Scan-Signature: ffa9dfbbe7cc58b3fa6b8ae3e57b0aa3
Cc: sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

RFC 3840, Section 10.15 ("Extensions") would be the mechanism to  
route according to capabilities.

On Feb 19, 2007, at 1:03 AM, Christer Holmberg ((JO/LMF)) wrote:

>
> Hi Jonathan,
>
>> If you've got a situration where a SIP call *cannot be processed*
> without
>> going through application servers, then this is not something
>> that is signaled with Accept-Contact. Its signaled with
>> Require or Proxy-Require as appropriate.
>
> Just for my clarification, where is it said that Require and
> Proxy-Require are used, or at least more appropriate, for routing?
>
> As far as I know, they are only used to require capabilities by the
> nodes that receive a SIP request (as you said yourself earlier) - but
> not to make routing decissions based upon.
>
> Regards,
>
> Christer
>


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 19 14:54:11 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJERh-0001CB-2v; Mon, 19 Feb 2007 14:45:09 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJERc-000192-U9
	for sipping@ietf.org; Mon, 19 Feb 2007 14:45:05 -0500
Received: from zrtps0kp.nortel.com ([47.140.192.56])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HJEOh-000222-4l
	for sipping@ietf.org; Mon, 19 Feb 2007 14:42:04 -0500
Received: from zcarhxs1.corp.nortel.com (zcarhxs1.corp.nortel.com
	[47.129.230.89])
	by zrtps0kp.nortel.com (Switch-2.2.6/Switch-2.2.0) with ESMTP id
	l1JJg1r09514
	for <sipping@ietf.org>; Mon, 19 Feb 2007 14:42:01 -0500 (EST)
Received: from [47.130.18.181] ([47.130.18.181] RDNS failed) by
	zcarhxs1.corp.nortel.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 19 Feb 2007 14:42:00 -0500
Message-ID: <45D9FD76.20007@nortel.com>
Date: Mon, 19 Feb 2007 14:41:42 -0500
From: "Tom-PT Taylor" <taylor@nortel.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: sipping@ietf.org
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 19 Feb 2007 19:42:00.0985 (UTC)
	FILETIME=[06214090:01C7545E]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 9182cfff02fae4f1b6e9349e01d62f32
Subject: [Sipping] [Fwd: I-D ACTION:draft-korhonen-mip6-service-00.txt]
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Service identifiers keep cropping up!

Internet-Drafts@ietf.org wrote:
> A New Internet-Draft is available from the on-line Internet-Drafts 
> directories.
> 
> 
> 	Title		: Service Selection for Mobile IPv6
> 	Author(s)	: J. Korhonen, U. Nilsson
> 	Filename	: draft-korhonen-mip6-service-00.txt
> 	Pages		: 8
> 	Date		: 2007-2-16
> 	
>    In some Mobile IPv6 deployments identifying the mobile node or the
>    mobility service subscriber is not enough to distinguish between
>    multiple services possibly provisioned to the said mobile node and
>    its mobility service subscription.  A capability to specify different
>    services in addition to the mobile node identity can be leveraged to
>    provide flexibility for mobility service providers on provisioning
>    multiple services to one mobility service subscription.  This
>    document describes a Service Selection Mobility Option for Mobile
>    IPv6 that is intended to assist home agents to make a specific
>    service selection for the mobility service subscription during the
>    binding update procedure.
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 20 04:53:19 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJReh-0007ne-7P; Tue, 20 Feb 2007 04:51:27 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJRef-0007nO-VL; Tue, 20 Feb 2007 04:51:25 -0500
Received: from mail128.messagelabs.com ([216.82.250.131])
	by ietf-mx.ietf.org with smtp (Exim 4.43)
	id 1HJRee-0008T2-Le; Tue, 20 Feb 2007 04:51:25 -0500
X-VirusChecked: Checked
X-Env-Sender: ranjit@motorola.com
X-Msg-Ref: server-12.tower-128.messagelabs.com!1171965083!7775749!1
X-StarScan-Version: 5.5.10.7.1; banners=-,-,-
X-Originating-IP: [129.188.136.8]
Received: (qmail 12235 invoked from network); 20 Feb 2007 09:51:23 -0000
Received: from motgate8.mot.com (HELO motgate8.mot.com) (129.188.136.8)
	by server-12.tower-128.messagelabs.com with SMTP;
	20 Feb 2007 09:51:23 -0000
Received: from il06exr04.mot.com (il06exr04.mot.com [129.188.137.134])
	by motgate8.mot.com (8.12.11/Motorola) with ESMTP id l1K9pNYB017894;
	Tue, 20 Feb 2007 02:51:23 -0700 (MST)
Received: from ZMY16EXM66.ds.mot.com (zmy16exm66.ap.mot.com [10.179.4.26])
	by il06exr04.mot.com (8.13.1/8.13.0) with ESMTP id l1K9pL0m012340;
	Tue, 20 Feb 2007 03:51:22 -0600 (CST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Date: Tue, 20 Feb 2007 17:51:19 +0800
Message-ID: <750BBC72E178114F9DC4872EBFF29A5B03E8E80D@ZMY16EXM66.ds.mot.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Knowing status of sent message
Thread-Index: AcdU1KvNPiuiYHveTTmH48rvxYZWtw==
X-Priority: 1
Priority: Urgent
Importance: high
From: "Avasarala Ranjit-A20990" <ranjit@motorola.com>
To: <sipping@ietf.org>, <sip@ietf.org>
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 538aad3a3c4f01d8b6a6477ca4248793
Cc: 
Subject: [Sipping] Knowing status of sent message
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1827483533=="
Errors-To: sipping-bounces@ietf.org

This is a multi-part message in MIME format.

--===============1827483533==
Content-class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C754D4.AC528304"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C754D4.AC528304
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable

Hi

Is there a mechanism in SIP to know the status of sent instant message
using MESSAGE?

Thanks

Regards
Ranjit


------_=_NextPart_001_01C754D4.AC528304
Content-Type: text/html;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3Dus-ascii">
<META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
6.5.7651.28">
<TITLE>Knowing status of sent message</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/rtf format -->

<P><FONT COLOR=3D"#0000FF" FACE=3D"Times New Roman">Hi</FONT>
</P>

<P><FONT COLOR=3D"#0000FF" FACE=3D"Times New Roman">Is there a mechanism =
in SIP to know the status of sent instant message using MESSAGE?</FONT>
</P>

<P><FONT COLOR=3D"#0000FF" FACE=3D"Times New Roman">Thanks</FONT>
</P>

<P><FONT COLOR=3D"#0000FF" FACE=3D"Monotype Corsiva">Regards</FONT>

<BR><FONT COLOR=3D"#0000FF" FACE=3D"Monotype Corsiva">Ranjit</FONT>
</P>

</BODY>
</HTML>
------_=_NextPart_001_01C754D4.AC528304--


--===============1827483533==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1827483533==--




From sipping-bounces@ietf.org Tue Feb 20 05:00:00 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJRmt-0006Id-MK; Tue, 20 Feb 2007 04:59:55 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJRmr-0006ID-8w; Tue, 20 Feb 2007 04:59:53 -0500
Received: from mtagate2.de.ibm.com ([195.212.29.151])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HJRmo-0004FW-NH; Tue, 20 Feb 2007 04:59:53 -0500
Received: from d12nrmr1607.megacenter.de.ibm.com
	(d12nrmr1607.megacenter.de.ibm.com [9.149.167.49])
	by mtagate2.de.ibm.com (8.13.8/8.13.8) with ESMTP id l1K9xojK074962;
	Tue, 20 Feb 2007 09:59:50 GMT
Received: from d12av02.megacenter.de.ibm.com (d12av02.megacenter.de.ibm.com
	[9.149.165.228])
	by d12nrmr1607.megacenter.de.ibm.com (8.13.8/8.13.8/NCO v8.2) with
	ESMTP id l1K9xnvf1810564; Tue, 20 Feb 2007 10:59:49 +0100
Received: from d12av02.megacenter.de.ibm.com (loopback [127.0.0.1])
	by d12av02.megacenter.de.ibm.com (8.12.11.20060308/8.13.3) with ESMTP
	id l1K9xnhp000884; Tue, 20 Feb 2007 10:59:49 +0100
Received: from d12ml102.megacenter.de.ibm.com (d12ml102.megacenter.de.ibm.com
	[9.149.166.138])
	by d12av02.megacenter.de.ibm.com (8.12.11.20060308/8.12.11) with ESMTP
	id l1K9xn0n000878; Tue, 20 Feb 2007 10:59:49 +0100
In-Reply-To: <750BBC72E178114F9DC4872EBFF29A5B03E8E80D@ZMY16EXM66.ds.mot.com>
To: "Avasarala Ranjit-A20990" <ranjit@motorola.com>
MIME-Version: 1.0
X-Mailer: Lotus Notes Release 7.0 HF144 February 01, 2006
From: Avshalom Houri <AVSHALOM@il.ibm.com>
Message-ID: <OF1670BB38.C0AE2E60-ONC2257288.003702A3-C2257288.00370DC9@il.ibm.com>
Date: Tue, 20 Feb 2007 11:59:47 +0200
X-MIMETrack: Serialize by Router on D12ML102/12/M/IBM(Release 7.0.2HF71 |
	November 3, 2006) at 20/02/2007 11:59:48,
	Serialize complete at 20/02/2007 11:59:48
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 34d35111647d654d033d58d318c0d21a
Cc: sip@ietf.org, sipping@ietf.org
Subject: [Sipping] Re: [Sip] Knowing status of sent message
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============0118005268=="
Errors-To: sipping-bounces@ietf.org

This is a multipart message in MIME format.
--===============0118005268==
Content-Type: multipart/alternative;
	boundary="=_alternative 00370CCAC2257288_="

This is a multipart message in MIME format.
--=_alternative 00370CCAC2257288_=
Content-Type: text/plain; charset="US-ASCII"

See: http://tools.ietf.org/html/draft-ietf-simple-imdn-03.

--Avshalom






"Avasarala Ranjit-A20990" <ranjit@motorola.com> 
20/02/2007 11:51

To
<sipping@ietf.org>, <sip@ietf.org>
cc

Subject
[Sip] Knowing status of sent message






Hi 
Is there a mechanism in SIP to know the status of sent instant message 
using MESSAGE? 
Thanks 
Regards 
Ranjit _______________________________________________
Sip mailing list  https://www1.ietf.org/mailman/listinfo/sip
This list is for NEW development of the core SIP Protocol
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sipping@ietf.org for new developments on the application of sip

--=_alternative 00370CCAC2257288_=
Content-Type: text/html; charset="US-ASCII"


<br><font size=2 face="sans-serif">See: http://tools.ietf.org/html/draft-ietf-simple-imdn-03.</font>
<br><font size=2 face="sans-serif"><br>
--Avshalom<br>
<br>
<br>
</font>
<br>
<br>
<br>
<table width=100%>
<tr valign=top>
<td width=40%><font size=1 face="sans-serif"><b>&quot;Avasarala Ranjit-A20990&quot;
&lt;ranjit@motorola.com&gt;</b> </font>
<p><font size=1 face="sans-serif">20/02/2007 11:51</font>
<td width=59%>
<table width=100%>
<tr valign=top>
<td>
<div align=right><font size=1 face="sans-serif">To</font></div>
<td><font size=1 face="sans-serif">&lt;sipping@ietf.org&gt;, &lt;sip@ietf.org&gt;</font>
<tr valign=top>
<td>
<div align=right><font size=1 face="sans-serif">cc</font></div>
<td>
<tr valign=top>
<td>
<div align=right><font size=1 face="sans-serif">Subject</font></div>
<td><font size=1 face="sans-serif">[Sip] Knowing status of sent message</font></table>
<br>
<table>
<tr valign=top>
<td>
<td></table>
<br></table>
<br>
<br>
<br><font size=3 color=blue face="Times New Roman">Hi</font><font size=3>
</font>
<p><font size=3 color=blue face="Times New Roman">Is there a mechanism
in SIP to know the status of sent instant message using MESSAGE?</font><font size=3>
</font>
<p><font size=3 color=blue face="Times New Roman">Thanks</font><font size=3>
</font>
<p><font size=3 color=blue face="Monotype Corsiva">Regards</font><font size=3>
</font><font size=3 color=blue face="Monotype Corsiva"><br>
Ranjit</font><font size=3> </font><tt><font size=2>_______________________________________________<br>
Sip mailing list &nbsp;https://www1.ietf.org/mailman/listinfo/sip<br>
This list is for NEW development of the core SIP Protocol<br>
Use sip-implementors@cs.columbia.edu for questions on current sip<br>
Use sipping@ietf.org for new developments on the application of sip</font></tt>
<p>
--=_alternative 00370CCAC2257288_=--


--===============0118005268==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============0118005268==--




From sipping-bounces@ietf.org Tue Feb 20 08:50:07 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJVMY-0001Ay-Rk; Tue, 20 Feb 2007 08:48:58 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJVMX-000187-AV
	for sipping@ietf.org; Tue, 20 Feb 2007 08:48:57 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HJVMP-0007f8-MV
	for sipping@ietf.org; Tue, 20 Feb 2007 08:48:57 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	1CD0D203F2
	for <sipping@ietf.org>; Tue, 20 Feb 2007 14:48:45 +0100 (CET)
X-AuditID: c1b4fb3e-b16d6bb0000007e1-cd-45dafc3dcfc5 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	0B3932008D
	for <sipping@ietf.org>; Tue, 20 Feb 2007 14:48:45 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 20 Feb 2007 14:48:44 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Tue, 20 Feb 2007 14:48:43 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD028BDC58@esealmw111>
In-Reply-To: <009301c7542a$88d1f810$6501a8c0@china.huawei.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdUKvwnE5/ZeTa0QY+RVL9QQXehSgAySiPg
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 20 Feb 2007 13:48:44.0547 (UTC)
	FILETIME=[D67B3D30:01C754F5]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 538aad3a3c4f01d8b6a6477ca4248793
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi Spencer,

I was a little suprised to understand that this was a difference of view
on this point.  Maybe it is confusing a little the User interface from
the "communication service" being identified.  In my understanding,
however, it is the same service identifer that should be used in the
originating network and terminating network - and indeed there was a
liaison from an operator community that see that want to use this for
identifiying the requested service between operators as well.  Of
course, nothing can prevent an operator from
inteworking/upgrading/downgrading in their own network ...

Best Regards,

Steve.

-----Original Message-----
From: Spencer Dawkins [mailto:spencer@mcsr-labs.org]=20
Sent: Monday, February 19, 2007 2:33 PM
To: sipping@ietf.org
Subject: Re: [Sipping] Service Identification in SIP

Hi, Steve,

The distinction you're making here seems to capture the the biggest
question - are we expecting service identifiers to be end-to-end, if not
interworked, or are we expecting them to be different for the same
service in originating and terminating networks.

Thanks,

Spencer

From: "Stephen Terrill (MI/EEM)" <stephen.terrill@ericsson.com>

Hi Andrew,

I see that in your assumptions, the originating and terminating networks
use different service identifiers, and I presume that this is for the
same service.  In my understanding, it is the same service identifier
that is used end-to-end (if not it is interworked).  That is why I do
not see the reliance on Presence being necessary.  In my opionion this
totally within the requirements of TS 23.228 (I think that different
service identifiers in different networks for the same service is
against TS 23.228).=20



_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 20 09:32:11 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJW1R-0004rZ-TU; Tue, 20 Feb 2007 09:31:13 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJW1Q-0004ls-25
	for sipping@ietf.org; Tue, 20 Feb 2007 09:31:12 -0500
Received: from sj-iport-2-in.cisco.com ([171.71.176.71]
	helo=sj-iport-2.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HJW1N-0008IP-No
	for sipping@ietf.org; Tue, 20 Feb 2007 09:31:12 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by sj-iport-2.cisco.com with ESMTP; 20 Feb 2007 06:31:08 -0800
X-IronPort-AV: i="4.14,196,1170662400"; 
	d="scan'208"; a="361776753:sNHT44334000"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1KEV8s4013995; 
	Tue, 20 Feb 2007 09:31:08 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1KEV4Vd003096; 
	Tue, 20 Feb 2007 09:31:08 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 20 Feb 2007 09:30:59 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 20 Feb 2007 09:30:58 -0500
Message-ID: <45DB0622.108@cisco.com>
Date: Tue, 20 Feb 2007 09:30:58 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Atle Monrad (GR/ETO)" <atle.monrad@ericsson.com>
Subject: Re: [Sipping] Service Identification in SIP
References: <61D8D34BB13CFE408D154529C120E0799A5B98@eseldmw101.eemea.ericsson.se>
In-Reply-To: <61D8D34BB13CFE408D154529C120E0799A5B98@eseldmw101.eemea.ericsson.se>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 20 Feb 2007 14:30:58.0856 (UTC)
	FILETIME=[BD0C1A80:01C754FB]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=626; t=1171981868;
	x=1172845868; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20
	|To:=20=22Atle=20Monrad=20(GR/ETO)=22=20<atle.monrad@ericsson.com>;
	bh=8BvpBCLroAW0VoRsvrNp5upT0+CpW9rf/Xup6Es/M7s=;
	b=sbeIMtn5npJrkDZbx6K2cTP1uq4yw45QVNKihDBV77Ijc+0GDO5wXP/phH01pTJiENw4cK1y
	Wu4ecXtqUBC2W281avQe38Aq724zBFGaWBdzyoZNYVcs4QBw8Q7HlFw9;
Authentication-Results: rtp-dkim-2; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: de4f315c9369b71d7dd5909b42224370
Cc: sipping <sipping@ietf.org>,
	"Gonzalo Camarillo \(JO/LMF\)" <gonzalo.camarillo@ericsson.com>,
	Mary Barnes <mary.barnes@nortel.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Atle Monrad (GR/ETO) wrote:

> On the interoperability, I would just like to add that during
> discussions in 3GPP, interoperability has not been considered a problem
> by the operators. On the contrary, the GSM Association has asked 3GPP to
> finalise the work on communication service identifiers and sort out the
> bits and pieces in the protocol ASAP.

Of course they don't find interoperability to be a problem! They 
constitute an oligopoly that is only concerned with interop among 
themselves. Interop with anyone else is presumably a non-goal.

The ietf has broader aspirations for interop.

	Paul

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 20 10:01:01 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJWTM-0006Ez-Aw; Tue, 20 Feb 2007 10:00:04 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJWTL-0006Dv-6t
	for sipping@ietf.org; Tue, 20 Feb 2007 10:00:03 -0500
Received: from sj-iport-1-in.cisco.com ([171.71.176.70]
	helo=sj-iport-1.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HJWTK-0006gM-5L
	for sipping@ietf.org; Tue, 20 Feb 2007 10:00:03 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by sj-iport-1.cisco.com with ESMTP; 20 Feb 2007 07:00:01 -0800
X-IronPort-AV: i="4.14,196,1170662400"; 
	d="scan'208"; a="763537018:sNHT73738548"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1KF00vY029741; 
	Tue, 20 Feb 2007 10:00:00 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1KExoOA016771; 
	Tue, 20 Feb 2007 09:59:58 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 20 Feb 2007 09:59:50 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 20 Feb 2007 09:59:49 -0500
Message-ID: <45DB0CE5.9090600@cisco.com>
Date: Tue, 20 Feb 2007 09:59:49 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Stephen Terrill (MI/EEM)" <stephen.terrill@ericsson.com>
Subject: Re: [Sipping] Service Identification in SIP
References: <989763567E51AB48BE015DF7BB960CAD028BD258@esealmw111>
In-Reply-To: <989763567E51AB48BE015DF7BB960CAD028BD258@esealmw111>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 20 Feb 2007 14:59:49.0697 (UTC)
	FILETIME=[C4B57B10:01C754FF]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=19686; t=1171983600;
	x=1172847600; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20
	|To:=20=22Stephen=20Terrill=20(MI/EEM)=22=20<stephen.terrill@ericsson.com
	>; bh=X98zU6MMKlk1aF3HIYEU2Jq2rgCa5KSiOI+lGjLpEHI=;
	b=UYF0wos9EukRADbYzf6PxdrlPuTUQONQ+tiwoH1MKIhE9yi9Vwt9SwrLk+4WOOsUL8cRav0+
	T0WtfHgGhwSCuYtxtPtIPHLkRHE6Q0GQcXz1Ztg10JriEAggiibb6QKX;
Authentication-Results: rtp-dkim-2; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 442d80051e0361a34f3560325c4a7092
Cc: Andrew Allen <aallen@rim.com>, sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Stephen Terrill (MI/EEM) wrote:
> Hi Andrew,
> 
> I see that in your assumptions, the originating and terminating networks
> use different service identifiers, and I presume that this is for the
> same service. 

What do you mean by "same service"?

If two organizations independently develop comparable functionality, is 
it the same service? If so, then having a common name for it cannot be 
expected.

Or is it only the same service if everyone who is going to interoperate 
has agreed to a common definition of the service? In that case, what 
forum is used to establish the agreement? And is the definition 
available to all who might want to support it? Or is the definition 
itself restricted?

IMO many of the most useful "services" are simply combinations of 
standard pieces, and don't require any agreement on a name, or any 
explicit agreement between interoperating parties. I would hate to see 
services that *can* be achieved this way being moved into the other 
category.

We already have things that do require agreements among the 
interoperating parties. Naturally sip itself is one, as is SDP. We have 
made those public standards. Codecs are another example, and they are 
mostly standard too, though I guess proprietary codecs are possible. 
MIME types are something else like this - that might be applicable to 
"services" using MESSAGE or MSRP.

Where something else must be standardized between the interoperating 
parties, then I hope it will follow the same kind of pattern. And I hope 
it won't be used if one of the other techniques can be used.

	Paul

> In my understanding, it is the same service identifier
> that is used end-to-end (if not it is interworked).  That is why I do
> not see the reliance on Presence being necessary.  In my opionion this
> totally within the requirements of TS 23.228 (I think that different
> service identifiers in different networks for the same service is
> against TS 23.228).
> 
> This comes to the isssue of the PSTN gateway - I presume that you want
> to go to the PSTN gateway if the service is telephony, and within that,
> voice.  If we know that, then the we go to any PSTN gateway, and we
> can't presume a particular network for a PSTN gateway. 
> 
> Best Regards,
> 
> Steve.
> 
> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com] 
> Sent: Saturday, February 17, 2007 12:51 AM
> To: Stephen Terrill (MI/EEM); sipping@ietf.org
> Cc: Andrew Allen
> Subject: RE: [Sipping] Service Identification in SIP
> 
> 
> 
> Hi Steve,
> 
> My responses inline
> 
> Andrew 
> 
> -----Original Message-----
> From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
> Sent: Friday, February 16, 2007 3:51 AM
> To: sipping@ietf.org
> Subject: RE: [Sipping] Service Identification in SIP
> 
> Hello Andrew,
> 
> Thanks for your comments.  I have a couple of brief responses below.
> 
> Best Regards,
> 
> ..//Steve 
> 
> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com]
> Sent: Friday, February 16, 2007 7:15 AM
> To: Stephen Terrill (MI/EEM); sipping@ietf.org
> Cc: Andrew Allen
> Subject: RE: [Sipping] Service Identification in SIP
> 
> 
> Some comments on Stephen's comments on the draft:
>  
> 
> -Moving onto the proposed solution draft itself, in addition for
> providing for the identification of the "application" and the
> "communication service", I have a few additional thoughts.
> 
> The draft seems to have the implication that all services (communication
> services in other terminology) that are to be identified have to be
> standardised within the IETF.  This is clearly an restriction that will
> not be acceptable to the industry as a whole.  My guess is that if such
> a statement is mandated by the IETF, it will just be words on paper
> which would likely be ignored in practice by other standardisation
> organisations.
> 
> [AA] This I think is an incorrect interpretation of the draft. My
> understanding is that the only services according to the draft that
> require some standardisation in IETF are those that by their very nature
> cannot interoperate unless there is some common additional understanding
> between the endpoints outside of what is currently specified in SIP. The
> use of SIP option tags with Require header is proposed by the draft for
> this purpose as such services create interoperability issues with SIP.
> 
> <Steve>
> My comment applies even for these services </Steve>
> 
> 
> <AA>
> Well then we are in big trouble because what you are proposing is that
> its OK for bodies outside of IETF to create non interoperable extensions
> for the behavour of SIP entities. This will result in the fragmentation
> of SIP into different flavors that don't interoperate. This is the
> balkanization of SIP that Paul talks about.even <AA> 
> 
> ----------------------------------------------------
> 
> The draft also only recognises the case where all explicitly knowledge
> of the service is required.  I think it would also be good to allow the
> case where knowledge of the service would be helpful, but not explicitly
> required.  This could be a case where the service to the end-user would
> not be degraded if this was not supported.  This functionality seems to
> be missing.
> 
> [AA] Service URIs are defined in the draft for these services where the
> support by the other endpoint is not required for the session to be
> sucessfully established.
> 
> The draft seems to indicate that that application servers in the network
> are involved only at the request of the originating terminal.  This
> seems a strange restriction.  For some network support that supports the
> originating user, such as the one provided in the example "I would like
> this call recorded for later retrieval", such mechanisms may make sense,
> however as a general approach it seems a bit limited.  For example, in
> 3GPP networks there is the need to invoke application logic in the
> originating side and terminating side, not only at the request of the
> originating terminal, but as a consequence of the requested service.  If
> there is an SIP message contains indication of the requested service, a
> proxy can do this (or loose route to an AS that contains the necessary
> service logic). 
> 
> [AA] The draft supports both possibilities using the Route header for
> originating network servers that need to be involved in the route and
> based on the terminating networks routing policy for the Service URI in
> the Request URI.
> 
> <Steve>
> I don't see the advantage of having different solutions for the
> originating network and the terminating network.  I don't think that the
> originating logic is invoked only at the request of the originating
> terminal.
> </Steve>
> 
> <AA>
> Unless IETF is willing to standardise all services and mandate that all
> SIP entities understand these (which will never happen) then we cannot
> assume that all networks and all endpoints understand and use the same
> identifiers. It is possible through configuration for an endpoint to
> obtain the identifiers used by its own service provider but there is no
> way to ensure that other networks use or understand the same identifiers
> for the equivalent or similar service that is compatible at the session
> layer. Therefore we need to support the possibility to have different
> identifiers in different networks in order to not violate the TS 23.228
> requirement:
> 
> 12.	The usage of communication service identifiers shall not
> adversely affect interoperability between IMS networks and
> interoperability with external SIP networks and CS networks. The
> behaviour of a network receiving the IMS requests without an IMS
> communication service identifier is a matter of operator policy. Usage
> of communication service identifiers shall not decrease the level of
> interoperability with networks and UEs that are unaware of the
> communication service identifier.
> 
> If the originating service logic isn't invoked based on an identifier
> placed in the request by the originating terminal on what basis do you
> think it is invoked? How does the use of Accept-Contact have advantages
> in this case?
> <AA> 
> 
> --------------------------------------
> 
>  I note here that the proposed service URN does not contain the address
> of a specific server containing the application logic.  An AS can look
> at that, and loose route to an application server that does contain such
> logic (If it is separated from the proxy), however I fail to see the
> difference for that proxy if it looks at a feature tag, or it looks at
> the service URN for this.  In both cases, it has to parse the message
> for the data, and modify the routing information.  
> 
> [AA]Route headers are the SIP mechanism for requesting intermediaries to
> be includede on the route of the request. There is nothing in RFC
> 3840/RFC 3841 that addresses using feature tags in Accept-Contact header
> for any routing decisions prior to determining which of the registered
> contacts the request should be forked to. Accept-Contact is not a Route
> header it is merely a filter that is applied to the set of valid
> contacts. Using a parameter in the filter as a service identifier has a
> potential significant impact on the use of Accept-Contact for its
> original purpose. 
> 
> <Steve>
> I don't believe there is anthing that states a Proxy CAN NOT route to
> applications or application logic based upon elements in the SIP
> message, including what is in the Accept-Contact header (some people
> have even proposed SDP ...).
> </Steve>
> 
> <AA>
> As I had stated in an earlier email on this subject a proxy can do a lot
> of things (including routing based on your favorite number appearing in
> a call-Id) but that doesn't mean they are a good idea or a good design
> decision. If you want to ensure you don't break things then you need to
> be careful to use SIP headers with the semantics intended for them
> otherwise problems can result if extra semantics are put upon existing
> headers.
> <AA> 
> 
> Further, the route header could be removed in the originating network;
> and it is still required to identify the service in the terminating
> network.  As proxies may remove the route, it can't be used to transport
> the requested service.  
> 
> [AA] Proxies arbitarily removing Route headers is a BAD thing akin to a
> bull in a china shop - almost certainly going to break something. RFC
> 3261 summarises what proxies should be doing with Route headers:
> 16.12 Summary of Proxy Route Processing
> 
>    In the absence of local policy to the contrary, the processing a
>    proxy performs on a request containing a Route header field can be
>    summarized in the following steps.
> 
>       1.  The proxy will inspect the Request-URI.  If it indicates a
>           resource owned by this proxy, the proxy will replace it with
>           the results of running a location service.  Otherwise, the
>           proxy will not change the Request-URI.
> 
>       2.  The proxy will inspect the URI in the topmost Route header
>           field value.  If it indicates this proxy, the proxy removes it
>           from the Route header field (this route node has been
>           reached).
> 
>       3.  The proxy will forward the request to the resource indicated
>           by the URI in the topmost Route header field value or in the
>           Request-URI if no Route header field is present.  The proxy
>           determines the address, port and transport to use when
>           forwarding the request by applying the procedures in [4] to
>           that URI.
> 
>    If no strict-routing elements are encountered on the path of the
>    request, the Request-URI will always indicate the target of the
>    request.
> 
> <Steve>
> The first sentance is interesting "In absence of local policy to the
> contrary", indicating that there can be different policies.  I
> understand that some proxies do remove this.  While I am on this - I
> have lead to believe that R-URI parameters may also be removed by
> intermediate proxies - is this correct?
> </Steve>
> 
> <AA>
> My understanding is that local policy may determine that additional
> nodes need to be routed to in addition to the Route contained in the
> request. However deleting existing Route headers other than the Route
> header containing the address of the proxy as part of the loose routing
> procedures is liable to break things and is not part of RFC 3261 routing
> (loose routing).
> <AA> 
> 
> 
> [AA] Generally a UA wouldn't know the appropriate service identifiers
> used in the terminating network or whether a Route header is appropriate
> to be used in the terminating network. The draft defines the Service URI
> for the purpose of identifying services in the terminating network and
> service URIs can be obtained using Presence.
> 
> <Steve>
> I don't get the above paragraph.  It suggest that the "appropriate
> service identifier" used in the terminating network is not known by the
> originating UE, but the Service URI is.  I thought that the service URI
> is an encoding of the "appropriate service identifier".
> </Steve>
> 
> <AA> 
> Yes but in the general case the UA doesn't need to know what the service
> identifier is it obtains through Presence a Service URI that contains
> the Service identifier as a URI parameter. The great thing about using
> Presence to obtain a Service URI is that the originating UA doesn't even
> need to know anything about service identifiers at all for this to work.
> It just uses Presence to find the URI to use for the service which can
> be indicated to the user via an indication such as a graphic icon (e,g a
> phone for telephony, smiley face for IM, etc) in the presence document.
> <AA> 
> 
> -----------------------------------
> 
>>From this perspective it is not clear to me the advantage of using a
> service URN in a route header compared to the use of a feature tag.
> 
> [AA] I hope the above responses clarify that somewhat. Also the Service
> URN and Service URIs allow different service identifiers to be used in
> the originating and terminating network for session compatiable services
> instead of forcing everyone to deploy and understand the same service
> and service identifier.
> 
> <Steve>
> Nope, don't get it, but are information that is carried end-to-end ....
> And as far as I see, could be used by intermediate proxies to impact
> local routing policy ... Or ?
> </Steve>
> 
> <AA> 
> I don't understand this response either. The key point is that UAs in
> corporate networks and in the internet are not going to deploy 3GPP
> flavor services and use the same 3GPP service identifiers or even
> possibly any service identifiers at all. The Accept-Contact based
> proposal assumes that all endpoints understand the service identifiers
> to be used for each service. This concept is not generally applicable
> and create interoperability problems.
> <AA> 
> 
> The draft seems to mandate GRUU - is this really so, is that really
> necessary.  It seems strange to place such a dependency, it also starts
> to move the GRUU as being the primary means of which to contact a user.
> 
> [AA] My understanding is that GRUU in the draft is used to address how a
> request addressed to Service URI is forked only to those contacts that
> support the service identified by the that Service URI. A contact would
> seem to be fine so long as it had GRUU properties - no need to obtain a
> GRUU if you already have one.
> 
> This implies that the originating user decides which terminal I should
> receive my requested service on, even if I have two devices for that.
> As such, I don't believe that there should be a dependency on GRUU.
> 
> [AA] If you have obtained a Service URI (probably through presence) for
> that user then you can address the request to the Service URI which
> represents all the terminals that support that service. However GRUU and
> the Service Instance URIs created from GRUUs provide another mechanism
> through which a user can obtain a URI to reach that user for the
> required service.
> 
> Parts of the draft seems to rely on the use of presence.  While presence
> could be a supportive mechanism, I doubt that we can mandate that the
> number you want to dial has be available in presence as I assume not all
> entities are available in presence (diallable network services, users in
> the PSTN).  This would be strange to mandated.
> 
> [AA] PSTN gateways are presumably in the same domain as the terminating
> proxy and therefore these PSTN gateways can be configured with the
> appropriate service identifiers for telephony services and can include
> this in either a Route header to route to the right application server
> or can construct a valid Service URI. The real problem is how do users
> in other networks know what the service identifier that's is used in the
> terminating network is. The use of Service URIs and Presence provides a
> solution to that problem.
> 
> <Steve>
> Again, presence could be a useful mechanism, but I think it would be a
> problem for the draft to rely it.  A user making a call to the PSTN may
> not find that user in the presence archicture (and similar to services
> like the 800 or 900 services).  (I should note, that you shouldn't make
> any assumptions about the location of the PSTN gateway ... But I don't
> think that changes anything).  
> </Steve>
> 
> <AA> 
> How does a PSTN gateway in another domain from the terminating network
> know the service identifier for telephony used in the terminating
> network with the Accept-Contact header based proposal? 
> <AA> 
> 
> Moving on from the draft I have seen the questioned raised about "why
> can't we expand the power of SDP".  Such an approach would mandate that
> SIP methods use SDP, I don't think that is appropriate. 
> 
> [AA] I agree with you here
> 
> I hear the appeals that the mechanisms should not be abused, and that we
> should not have 1000s of separate "communication services", and I fully
> agree with this. I see a smaller number of communication services, not a
> larger number; where those communication services are used to access
> other end-user-services.
> 
> *flame away*
> 
> Regards,
> 
> Steve
> 
> ---
> 
> 
> 
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential
> information, privileged material (including material protected by the
> solicitor-client or other applicable privileges), or constitute
> non-public information. Any use of this information by anyone other than
> the intended recipient is prohibited. If you have received this
> transmission in error, please immediately reply to the sender and delete
> this information from your system. Use, dissemination, distribution, or
> reproduction of this transmission by unintended recipients is not
> authorized and may be unlawful.
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 20 10:06:45 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJWZJ-0008Uf-MO; Tue, 20 Feb 2007 10:06:13 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJWZI-0008UZ-OA
	for sipping@ietf.org; Tue, 20 Feb 2007 10:06:12 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HJWZB-0007ly-Ky
	for sipping@ietf.org; Tue, 20 Feb 2007 10:06:12 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	E784E2069B; Tue, 20 Feb 2007 16:05:56 +0100 (CET)
X-AuditID: c1b4fb3c-af7c7bb0000007de-dd-45db0e541abc 
Received: from esealmw127.eemea.ericsson.se (unknown [153.88.254.122])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	D7D6D20645; Tue, 20 Feb 2007 16:05:56 +0100 (CET)
Received: from eseldmw101.eemea.ericsson.se ([136.225.228.103]) by
	esealmw127.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 20 Feb 2007 16:05:56 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Tue, 20 Feb 2007 16:05:56 +0100
Message-ID: <61D8D34BB13CFE408D154529C120E0799A6351@eseldmw101.eemea.ericsson.se>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdU+8Snbb12rvzXSHCMUBKxp9HYvwAAKXbg
From: "Atle Monrad \(GR/ETO\)" <atle.monrad@ericsson.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>
X-OriginalArrivalTime: 20 Feb 2007 15:05:56.0445 (UTC)
	FILETIME=[9F4EC8D0:01C75500]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 0ddefe323dd869ab027dbfff7eff0465
Cc: sipping <sipping@ietf.org>,
	"Gonzalo Camarillo \(JO/LMF\)" <gonzalo.camarillo@ericsson.com>,
	Mary Barnes <mary.barnes@nortel.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Paul

by the way, it would be useful to get this discussion on the sipping
list.

As a service is defined in terminals and possibly in application
servers, the interoperability on the service level are between the
terminals and the ASs that have implemented the service while
interoperability on the protocol level is something very different.
Please tell me which interop that is your concern
=20
I do not see any interoparability, if you do, it would be useful that
you share it.=20

Terminals may have different services and may not be interoperable in
all aspects, but isn't this just facts that is hard to get around? A PoC
terminal can only talk PoC to other PoC terminals for the PoC service
but it can communicate with other terminals aswell, right? =20

/atle

FYI, GSMA consists of 700 operators. When GSMA give their
recommendations, it would be useful to listen and at least give some
technical feeback and not this kind of political statement.=20


-----Original Message-----
From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
Sent: 20. februar 2007 15:31
To: Atle Monrad (GR/ETO)
Cc: Gonzalo Camarillo (JO/LMF); sipping; Mary Barnes
Subject: Re: [Sipping] Service Identification in SIP



Atle Monrad (GR/ETO) wrote:

> On the interoperability, I would just like to add that during=20
> discussions in 3GPP, interoperability has not been considered a=20
> problem by the operators. On the contrary, the GSM Association has=20
> asked 3GPP to finalise the work on communication service identifiers=20
> and sort out the bits and pieces in the protocol ASAP.

Of course they don't find interoperability to be a problem! They
constitute an oligopoly that is only concerned with interop among
themselves. Interop with anyone else is presumably a non-goal.

The ietf has broader aspirations for interop.

	Paul

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 20 11:19:39 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJXhy-00055Q-Sg; Tue, 20 Feb 2007 11:19:14 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJXhw-00055G-E9
	for sipping@ietf.org; Tue, 20 Feb 2007 11:19:13 -0500
Received: from sj-iport-6.cisco.com ([171.71.176.117])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HJXhv-00005n-DK
	for sipping@ietf.org; Tue, 20 Feb 2007 11:19:12 -0500
Received: from sj-dkim-6.cisco.com ([171.68.10.81])
	by sj-iport-6.cisco.com with ESMTP; 20 Feb 2007 08:19:10 -0800
X-IronPort-AV: i="4.14,197,1170662400"; 
	d="scan'208"; a="114500551:sNHT330834744"
Received: from sj-core-4.cisco.com (sj-core-4.cisco.com [171.68.223.138])
	by sj-dkim-6.cisco.com (8.12.11/8.12.11) with ESMTP id l1KGJACJ019629; 
	Tue, 20 Feb 2007 08:19:10 -0800
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by sj-core-4.cisco.com (8.12.10/8.12.6) with ESMTP id l1KGIqnP028887;
	Tue, 20 Feb 2007 08:19:10 -0800 (PST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 20 Feb 2007 11:18:43 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 20 Feb 2007 11:18:43 -0500
Message-ID: <45DB1F62.5060004@cisco.com>
Date: Tue, 20 Feb 2007 11:18:42 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Andrew Allen <aallen@rim.com>
Subject: Re: [Sipping] Service Identification in SIP
References: <61968779B8AC4C4BAB421D4C12F008C009151469@XCH47YKF.rim.net>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C009151469@XCH47YKF.rim.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 20 Feb 2007 16:18:43.0153 (UTC)
	FILETIME=[CA118C10:01C7550A]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=17249; t=1171988350;
	x=1172852350; c=relaxed/simple; s=sjdkim6002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20; bh=9vS1HdftZu8iHijXdoD0OlQoC9M3TO7rwZAgT7y35UU=;
	b=FH+MSOn6/CHfxLIjr/a0G80DP58M2d6Ofo9UA0iSUqNZ6jzc7mMNnTlwsPdR1UxPe5uJxHnN
	xa7paDHyOnN6NKCGyL9ODOKNVPhcICYCahNmLqTLLNn4U4WQ6Ah7aoPQ;
Authentication-Results: sj-dkim-6; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim6002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 2b3349545af520ba354ccdc9e1a03fc1
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Andrew Allen wrote:
> I'll have a crack at laying out examples showing the utility of service
> identification.

The examples are good. I think it is important not only to see how a 
service id could be used to solve these cases - it is also important to 
see if they could be addressed without service ids, using some existing 
mechanism.

> First a gaming example somewhat similar to Jonathan's. Take a network
> based multiplayer game (Chess, D&D, Quake, Pool, Roulette, virtual golf,
> etc.....). This game allows multiple users to participate in the game
> either as players or as observers. An audio stream is established so
> that players and observers can exchange comments on the game along with
> the media stream (either using MSRP or an application specific data
> transport that is used to exchange game commands and game position
> updates between the participants and the game server which also acts as
> a focus. The gaming session is established using SIP and SDP over a SIP
> network such as IMS that has Application Servers that perform telephony
> call features for basic voice only telephony and multimedia "rich"
> telephony calls. The multimedia telephony sessions also support combined
> audio with MSRP and other application specific data streams (such as
> whiteboards) in a call. Now gaming sessions don't require telephony
> service call feature handling in fact this isnt even desired however
> based on the SIP and SDP it isnt possible to determine whether an INVITE
> request is related to a telephony call or a gaming session and hence
> whether the request needs to be routed via an Application Server that
> performs telephony call features or not.

I'm not at all sure of the above.

For this example, the session will be between an UA and a game server 
that can be viewed as a conference focus. So how does that session get 
established? I can see three plausible ways:
- each player calls the server.
- the server calls a player, either based on out of band input
   (e.g. web access) or by another player sending a REFER to the
   game server

When a player calls the game server, originating "features" are still 
applicable. E.g. outbound call blocking.

When the game server calls the player's UA, terminating "features" are 
also probably still applicable. E.g. inbound call blocking is still 
applicable. Perhaps voicemail isn't, but that is hardly different in 
other cases where a caller doesn't want to be routed to voicemail, and 
should possible to handle via caller preferences.

In general, "mandatory" services at either the originating or 
terminating sides most likely need to be applied because if they aren't 
then it will open a target for a caller who wants to bypass those 
features. In the case of optional features, there ought to be some 
mechanism for callers to bypass them if they don't want them.

> In addition users may have
> multiple devices only one of which has the application that understands
> the game. Other devices that the user has registered against his AoR may
> handle audio for telephony and other voice applications but if Alice
> invites Bob to play her favourite game it is not very interesting if Bob
> accepts the gaming session on a device that can only accept the audio
> stream.

I definitely don't consider the service id to be the right solution for 
selecting a UA that supports the combination of media that a caller 
wants to use. Using service ids for this would imply that every 
meaningful combination of media be given a distinct service id. That is 
not a reasonable solution to the problem.

There are a variety of possible approaches to solving that problem. 
Callerprefs is a possible answer, though it provides no guarantees. 
There are also solutions involving SDP - indications of dependency among 
streams has been proposed. Its also possible to use preconditions to 
negotiate an acceptable combination of media.

This has been a problem just waiting for a critical mass of desire for a 
solution. Maybe this is the time to solve it.

> Also even when the terminal that has the game application
> receives the INVITE the correct application in that terminal needs to be
> identified to accept the session.

I'm not convinced of this one. In most cases this will be a no-brainer - 
there will only be one reasonable candidate. In cases where there are 
multiple candidates, it would be preferable to have the target device do 
so based on intrinsic attributes of the session, rather than artificial 
tags applied to it. (E.g. if MSRP is involved, use the IM application.) 
And if there is ambiguity and a suboptimal choice is made, the user 
ought to be able to change it without taking the call down.

> In addition service providers may
> charge different rates for telephony and for games. I also don't want to
> be charged telephony per minute call rates for playing my game.  

Well, the call will be to a game server, and it will involve some sort 
of media. The server may or may not be known to the SP.

If the game server is known to the SP, then it can adjust the billing as 
it wishes.

If the game server is unknown to the SP, then the SP would be stupid to 
reduce its billing based on a claim by the caller or callee that this is 
a game. If there are tangible differences in the resources provided by 
the SP (e.g. different bandwidth media streams) then billing can be 
adjusted based on that.

I don't see how a service id makes anything better here.

> Here is another example that I think shows that service identification
> doesn't just apply to INVITE and sessions but also can apply to other
> requests such a MESSAGE as well. SIP MESSAGE method is being used for
> Instant messaging applications but it is also being used for many non
> Instant Messaging applications as well. These include as a replacement
> SIP transport for transporting encapsulated SMS (Short Message Service)
> contents (see 3GPP SMS over IP), as a transport for Group Advertisement
> documents for Push To Talk (see OMA PoC), as a notification transport
> for various Push Services (see OMA SIP Push). Each of these different
> service applications of SIP MESSAGE method requires routing to different
> Application Servers in the network, delivery to different applications
> in the terminal and the ability to be charged for separately.

As far as handling by the receiving UA, it seems to me that the MIME 
type is perfectly adequate as a way to get the message properly 
processed, just as is done with attachments to email messages.

I'm not sure I understand where these need to be routed to an AS. 
Perhaps for store and forward SMS. Why is that any different than 
voicemail, where a server is the UA of last resort?

If something is needed, e.g. for routing to a B2BUA, it seems like 
callerprefs should work. If not, potentially a proxy could route based 
on the MIME type of the body.

So I'm not convinced you need service id to do this.

> The only
> SIP header that differentiates each of these different useages of SIP
> MESSAGE is the Content-Type header but routing based on Content-Type is
> a very bad idea. Firstly the Content-Type doesn't identify the service
> specifically but some content related to a subset of the service.

It works ok for email. And *especially* for MESSAGE, which is a 
datagram, it seems just right to me: please see that this is processed 
by something that understands the content. The type of the content can 
be narrowed as much as desired to achieve the desired selectivity.

> Also
> as the service evolves the Content-Types used may change or new ones may
> be added and this shouldn't require updating the routing rules of the
> proxies or the internal application invocation logic in endpoints. The

As MIME types change, the recipients need to understand them anyway.

And what is going to happen as the definition of a service changes? Will 
it get a new service id so you can be sure it is processed by something 
that understands it? Or will it keep the same service id but run the 
risk of being routed to an incompatible target?

I don't think the problem is substantially different whether you use 
MIME types of service ids.

> possibility of Multipart MIME Content-Types makes parsing of
> Content-Types by proxies inefficient and complex and potentially
> impossible if the message body is encrypted.

If you want a particular sort of intermediate server involved in a 
request then its indeed possible that selecting based on MIME type won't 
be ideal. Doing based on callerprefs, with the intermediate server being 
treated as a first class candidate target, may be an alternative.

> What is needed is a means
> compliant with standard SIP Routing mechanisms that allow SIP proxies to
> efficently route these different MESSAGE requests to the appropriate
> servers, endpoints and invoke applications within endpoints. I think
> Jonathan's draft is a good basis for doing this.

Using Route headers with URNs is indeed a possibility here. But how the 
originator arranges for such a header to be present when the request 
reaches the terminating side is hard for me to see. And how to ensure 
that the terminating side is capable of honoring that is also problematic.

> Another example using INVITE is IPTV. IPTV may be established using
> INVITE containing SDP with both Audio and Video streams. How does the
> network determine the difference between an IPTV related INVITE and one
> related to multimedia telephony that supports video? Certainly it is not
> desired that the IPTV INVITE invoke telephony call features in an
> Application Server and you probably don't want the IPTV INVITE being
> routed to your mobile phone (unless it's a mobile TV service). It is
> likely that an SDP offer for IPTV is for a unidirectional media stream
> but this is real convoluted to use as the basis for routing and
> charging. 

I think using differences in SDP to distinguish this is quite 
appropriate. It might be more than just unidirectional streams. For 
instance it might call for bandwidth and/or resolution that a phone 
won't support.

Nor do I find the bypassing of telephony services to be an issue. Either 
they are still meaningful, or it will be clear that they shouldn't be 
applied based on the nature of the call, without need for a service id.

As for charging, as I mentioned earlier, either the SP knows the source 
and can adjust the charging accordingly, or it doesn't know the source 
in which case it must charge based on the type of resources it is 
providing without knowing how they are being used.

> In adddition it is possible that an INVITE doesn't contain SDP at all
> and then there is no basis in SDP for determining the related service.

Well clearly an offerless invite means the caller desires the callee to 
choose the nature of the call. This doesn't present any new problems 
that aren't also present as a result of mid-call changes in the media 
used in a call. For instance, I could call you, and ask if you would 
like to see my latest video. Then switch the stream to hi def streaming 
video. This isn't much different than switching from voice to fax.

In part I think this demonstrates the folly of basing an architecture on 
the assumption that you can splice servers into a call at the beginning 
that will be appropriate for everything that might happen subsequently 
on that session. If you need servers in the middle of the call, then 
there had better be some way to introduce or change them mid-call 
if/when the nature of the call changes.

I also think that this again identifies the general problem of 
negotiating a mutually acceptable set of media during call 
establishment, as mentioned above. Again I don't see service ids as the 
right solution to this problem. The SDP capability work that Flemming is 
doing might be of some relevance here - it might permit sending an 
offerless invite with capabilities, so that UASs that can't live with 
the capabilities of the caller can refuse the call.

	Paul

> Andrew
> 
> 
> 
> -----Original Message-----
> From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net] 
> Sent: Saturday, February 10, 2007 11:50 PM
> To: sipping@ietf.org
> Subject: Re: [Sipping] Service Identification in SIP
> 
> I've attempted to re-read the threads that Gonzalo referred to, and the
> major limitation that I see is the lack of any concrete example of why
> these identifiers are needed.  Probably the IMS people have discussed
> this so many times that they've forgotten the detailed rationale behind
> the requirement.  But I'm a net-head, (and most of the people here are
> net-heads) and the SIP systems that I've dealt with show no evidence of
> needing "service identification", they're pure proxies, and requests are
> passed transparently end-to-end.
> 
> Could somebody please lay out an example showing the utility of service
> identification?
> 
>    From: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>
> 
>    I think that everyone here agree that there is the possibility to use
> a
>    media in different ways, and that we need a distinction about what
> the
>    capabilities are and how they are used, whatever will be the
> mechanism
>    that at the end we will identify to do that.
> 
> Respectfully, I disagree.  We do need ways of identifying the
> capabilities of SIP agents and transports, but I see no necessity of
> creating specified bundles of capabailities, and then constraining
> dialogs to require an initial agreement as to which bundle of
> capabilities will be used.
> 
> One of the central features of Internet architecture is that agents
> negotiate compatibility on a feature-by-feature basis; there are no
> "profiles" which must be known on a network-wide basis.  (And I have
> observed that this feature is of significant commercial value.)
> 
> In regard to identifying an "application" to process the dialog (and I
> use "application" in a standard computing way), I see no reason that a
> user cannot select URI-parameters to append to his AOR, and arrange for
> his UA to route invites with those URI-parameters to the appropriate
> application.  (And given the rich mechanisms for redirection in SIP, the
> application need not even reside at a UA which is registered for the
> AOR.)  These URI-parameters can be formally or informally standardized,
> or established ad-hoc by each user.
> 
> In regard to PoC (as someone said, "a specialized floor-control
> protocol") and similar features, SIP currently has mechanisms for
> indicating support of features and requiring that a recipient support a
> feature before accepting a call (the Supported and Requires headers, and
> simply using the feature).  The only limitation I can see that might
> need to be addressed is that there is no way to create an option tag for
> a new feature that one has invented that is not yet thoroughly
> standardized.
> 
> But I now discover RFC 3261 section 8.1.1.9:
> 
>    The option tags listed MUST only refer to extensions defined in
>    standards-track RFCs.  This is to prevent servers from insisting that
>    clients implement non-standard, vendor-defined features in order to
>    receive service.  Extensions defined by experimental and
>    informational RFCs are explicitly excluded from usage with the
>    Supported header field in a request, since they too are often used to
>    document vendor-defined extensions.
> 
> So I see that as a fundamental architectural decision, SIP does not
> permit sending INVITEs that require specialized recipients, excepting if
> the specialization is standards-track.
> 
> Dale
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use
> sip-implementors@cs.columbia.edu for questions on current sip Use
> sip@ietf.org for new developments of core SIP
> 
> 
> 
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential information, privileged material (including material protected by the solicitor-client or other applicable privileges), or constitute non-public information. Any use of this information by anyone other than the intended recipient is prohibited. If you have received this transmission in error, please immediately reply to the sender and delete this information from your system. Use, dissemination, distribution, or reproduction of this transmission by unintended recipients is not authorized and may be unlawful.
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 20 12:30:29 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJYo5-0002jX-EI; Tue, 20 Feb 2007 12:29:37 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJYo4-0002jR-1P
	for sipping@ietf.org; Tue, 20 Feb 2007 12:29:36 -0500
Received: from sj-iport-2-in.cisco.com ([171.71.176.71]
	helo=sj-iport-2.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HJYo2-0001BG-Uu
	for sipping@ietf.org; Tue, 20 Feb 2007 12:29:36 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by sj-iport-2.cisco.com with ESMTP; 20 Feb 2007 09:29:33 -0800
X-IronPort-AV: i="4.14,197,1170662400"; 
	d="scan'208"; a="361825810:sNHT84630404"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1KHTXI0015767; 
	Tue, 20 Feb 2007 12:29:33 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1KHTGVZ000351; 
	Tue, 20 Feb 2007 12:29:29 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 20 Feb 2007 12:29:20 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 20 Feb 2007 12:29:20 -0500
Message-ID: <45DB2FF0.6050105@cisco.com>
Date: Tue, 20 Feb 2007 12:29:20 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Andrew Allen <aallen@rim.com>
Subject: Re: Device architecture? WAS: RE: [Sipping] Service	Identificatio
	ninSIP
References: <61968779B8AC4C4BAB421D4C12F008C00915161E@XCH47YKF.rim.net>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C00915161E@XCH47YKF.rim.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 8bit
X-OriginalArrivalTime: 20 Feb 2007 17:29:20.0090 (UTC)
	FILETIME=[A77AD3A0:01C75514]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=24056; t=1171992573;
	x=1172856573; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20Device=20architecture?=20WAS=3A=20RE=3A=20[Sipping]=2
	0Service=09Identificatio=0A=20ninSIP |Sender:=20
	|To:=20Andrew=20Allen=20<aallen@rim.com>;
	bh=5cuJ6Xn3zOjDv6hKPWdmDzsDRwUslhvwmUUH/3v1NHc=;
	b=kwkHUNRimf+zgjng5B0chd6++iTnIN5CAXW531ebNnLXJHuOQH09Ry8qG6wTNGx9Y4STYpuN
	kW6nyUsEnoJT0lh6liKfJIYN8Hgqvxhu24D030beMD0F6BU197Re7+Mp;
Authentication-Results: rtp-dkim-1; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: d6c15d82a53e26283536b4a751453103
Cc: sipping@ietf.org, "Elwell, John" <john.elwell@siemens.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Andrew Allen wrote:
>  
> John
>  
> I think whether an AS in the terminating network is considered a contact 
> is a debatable point. Certainly it doesnt register feature tags using 
> RFC 3840 mechanisms and (at least in IMS) they dont show up as 
> registered contacts in the reg-event notifications.

I have long considered this a deficiency in IMS. The applicability of 
caller preferences to ASs is never discussed. IMO IMS needs to clarify 
this. In particular:
- should callerprefs be taken into account when the S-CSCF is
   deciding whether to route a request to a particular AS?
- when an AS receives a request from the S-CSCF, should it apply
   callerprefs before deciding if/how to process the request?

Take voicemail as an example. RFC 4596 discusses, among other things, 
how callerprefs can be used by a caller either avoid going to voicemail 
or to go directly to voicemail without attempting another UAS. These use 
cases work if the voicemail server is registered.

While I haven't found it specified in any IMS docs I have read, its my 
understanding that in IMS the intended way to handle VM is to route any 
call that *might* need to go to VM to a VM aware server. This server 
will then act as a proxy, and eventually the call will be routed to 
candidate registered target UAs. The VM aware proxy then monitors the 
call. If it isn't answered, then eventually the proxy will cancel those 
attempts and route the call to a VM server that was not registered as a 
candidate.

In such a scenario, the callerprefs capability to select only voicemail 
or reject voicemail won't work by default. It *might* be made to work if 
the VM aware proxy is also callerprefs aware and takes suitable actions. 
But it will be a bit odd, because in general it doesn't have any 
awareness of the actual registered candidates.

It is possible that the reason IMS feels it needs new features such as 
service id is because it has broken existing mechanisms.

	Paul

> Also AS in the 
> originating network (which may need to be involved in some communication 
> services) are certainly not contacts for the resource identified by the 
> Request-URI and Accept-Contact certainly has no defined role in routing 
> to them.
>  
> In addition 3GPP TS 23.228 v7.6.0 subcaluse 4.13.2 specifies 15 
> requirements for IMS Communication Service Identifier and only one of 
> these is regarding matching to the correct contact:
>  
> 
> 11. The communication service identifier shall be taken into account 
> when selecting the correct UE(s), if multiple UEs are registered for the 
> same Public User Identity(s).
> 
>  
>  Several of the other requirements are in my view totally incompatible 
> with Accept-Contact being the transport for the identifier and indeed 
> may limit the use of Accept-Contact for its intended purpose. This 
> includes charging and policy control:
>  
> 
> 8.   Based on operator policy S-CSCF and AS shall consider the IMS 
> communication service identifier for online and offline charging, e.g. 
> put appropriate data into call detailed records.
> 
> 9.   The communication service identifier shall be capable of being an 
> input into the policy control and charging rules.
> 
> Also:
>  
> 
> 2.   Based on operator policy the S-CSCF or an AS shall be able to 
> validate an IMS communication service identifier in a SIP request. This 
> includes e.g. to check the syntactical correctness of a service 
> identifier, and policing the usage of a communication service identifier.
> 
> 10. It shall be possible to use the IMS communication service identifier 
> as a means to authorise whether a subscriber is allowed to initiate or 
> receive request for a communication service.
> 
> Another problem with UAs placing feature tags in Accept-Contact headers 
> as Service Identifiers is how does the initiator obtains the service 
> identifier of the destination party for the related service? The concept 
> only works in the scenario where all networks and devices have a common 
> understanding of the same services and the identifiers that identify 
> them. As far as I am aware SIP doesnt provide a mechanism to discover 
> proprietary non standard feature tags that are used by other devices in 
> other domains and map them to services that they relate to. This means 
> that if Accept-Contact is used the communication services are closed and 
> can only be used by those who are part of the closed network, (unless of 
> course IETF were to get into standardising services which I think is 
> highly unlikely). If URIs are used as the container for the Service 
> Identifier these can be obtained using Presence as proposed in 
> Jonathan's draft.
>  
> This helps fulfil the following requirements in 3GPP TS 23.228:
>  
> 
> 12. The usage of communication service identifiers shall not adversely 
> affect interoperability between IMS networks and interoperability with 
> external SIP networks and CS networks. The behaviour of a network 
> receiving the IMS requests without an IMS communication service 
> identifier is a matter of operator policy. Usage of communication 
> service identifiers shall not decrease the level of interoperability 
> with networks and UEs that are unaware of the communication service 
> identifier.
> 
> 14. The usage of communication service identifiers shall not restrict 
> the inherent capabilities of SIP.
> 
>  
> SDP is not good for routing decisions for service identification. 
> Firstly SDP is intended to be end to end only. It may be encrypted and 
> understanding the details of the service may require service or 
> application level understanding of the SDP and the service which is 
> knowledge that proxies should not need to have. Also I dont think that 
> SDP can always uniquely identify a service in any case. In addition not 
> all requests that are related to a service will contain SDP in any case.
>  
> Andrew
> ------------------------------------------------------------------------
> *From:* Elwell, John [mailto:john.elwell@siemens.com]
> *Sent:* Friday, February 16, 2007 3:12 AM
> *To:* Andrew Allen; Malleswara Rao Ankem
> *Cc:* Andrew Allen; sipping@ietf.org
> *Subject:* RE: Device architecture? WAS: RE: [Sipping] Service 
> Identificatio ninSIP
> 
> Andrew,
>  
> Well, an AS is just another contact (perhaps a permanently registered 
> one, rather than one that uses SIP REGISTER), isn't it? So caller 
> preferences can be used to choose an AS or any other contact that 
> supports the desired capability.
>  
> I agree caller preferences is not a suitable basis for charging. Just 
> because I say I have a preference for a contact that supports video, 
> doesn't mean that I will necessarily use video during that call. That 
> might be determined during some mid-call session-re-negotiation. Surely 
> charging has to be based on what is actually used, not what might be 
> used, and you can't always determine this at INVITE time.
>  
> So I am saying use caller preferences to select a preferred contact, use 
> SDP to negotiate and re-negotiate the session with the contact that you 
> end up with. I haven't heard any compelling arguments for anything else 
> so far.
>  
> John
> 
>     ------------------------------------------------------------------------
>     *From:* Andrew Allen [mailto:aallen@rim.com]
>     *Sent:* 16 February 2007 03:51
>     *To:* Elwell, John; Malleswara Rao Ankem
>     *Cc:* sipping@ietf.org; Andrew Allen
>     *Subject:* RE: Device architecture? WAS: RE: [Sipping] Service
>     Identificatio n inSIP
> 
>      
>     John,
>      
>     If Service Identification was only about selecting the appropriate
>     contacts from those registered agaisnt the AOR then RFC 3840/3841
>     would be fine however Service Identification is much more than that
>     and may be used to route requests to application servers for service
>     logic invocation, and used as the basis for charging. Accept-Contact
>     is basically just a filter expression to filter a set of registered
>     contacts. Including a parameter in a filter expression and using
>     that as the basis for charging, authorisation and service invocation
>     is going to cause problems.
>      
>     Andrew
>      
>     ------------------------------------------------------------------------
>     *From:* Elwell, John [mailto:john.elwell@siemens.com]
>     *Sent:* Thursday, February 15, 2007 6:16 AM
>     *To:* Malleswara Rao Ankem
>     *Cc:* sipping@ietf.org
>     *Subject:* RE: Device architecture? WAS: RE: [Sipping] Service
>     Identificatio n inSIP
> 
>     Mallesh,
>      
>     Yes, I agree that some means of selecting a preferred device might
>     be needed, but that is indeed what RFC3841/3840 does. Why do you
>     consider this overkill?
>      
>     John
> 
>         ------------------------------------------------------------------------
>         *From:* Malleswara Rao Ankem [mailto:malleshavn@gmail.com]
>         *Sent:* 15 February 2007 10:54
>         *To:* Elwell, John
>         *Cc:* Göran Eriksson AP (KI/EAB); sipping@ietf.org
>         *Subject:* Re: Device architecture? WAS: RE: [Sipping] Service
>         Identificatio n in SIP
> 
>         Inorder to identify a endpoint that has the capabilities to
>         serve a very specialized functionality, I guess it is simpler to
>         name the functionality. Naming the functionality not only helps
>         in decision making while routing the request to that specific
>         endpoint but also helps the endpoint identify which specialized
>         application on the endpoint can take up the responsibility.
>         Instead of all the routing entities looking into the SDP/message
>         bodies it should be easier/faster if they can look for a
>         specific header.
>         I thought feature tag mechanisms indicated in the RFC3841/3840
>         combination should have sufficed but implementing those RFCs
>         seems to be a overkill for describing/selecting/routing
>         requests to the specific endpoint and also choosing a
>         service/feature/application on the endpoint.
>          
>         -Mallesh
>          
>          
>          
>         On 2/15/07, *Elwell, John* <john.elwell@siemens.com
>         <mailto:john.elwell@siemens.com>> wrote:
> 
>             Göran,
> 
>             Your interpretation is different from what is stated in the
>             Rosenberg draft.
>             Anyway, let's not dwell on this particular use case, which
>             may or may not be
>             realistic. The important thing is whether the media being
>             sent is suitable
>             for general purpose devices or requires or could be used by
>             some special
>             class of application.
> 
>             If a medium is somehow specialised ( e.g., audio modulated
>             for interpretation
>             by a modem or fax machine), then the necessary information
>             needs to be
>             conveyed along with the m-line in the SDP. Perhaps even if it is
>             human-intelligible voice but with a restricted repertoire of
>             words that make
>             it suitable for a particular class of application, maybe
>             that information
>             too should be conveyed with the m-line. Perhaps the case
>             described in
>             Jonathan's draft fits this latter category: I believe the
>             medium is just
>             regular voice that can be interpreted by a human, but is
>             also constrained in
>             its repertoire of words such that it can be used by a chess
>             application.
> 
>             The important thing is that there be sufficient information
>             in the SDP to
>             allow appropriate application selection at the end device. I
>             am not
>             convinced by a need for information at the SIP level during call
>             establishment, because of course SDP can be re-negotiated
>             during the call
>             ( e.g., when the users decide to use a particular
>             application) and what is
>             indicated in SIP at call establishment time might no longer
>             be valid later
>             in the call.
> 
>             John
> 
> 
>              > -----Original Message-----
>              > From: Göran Eriksson AP (KI/EAB)
>              > [mailto:goran.ap.eriksson@ericsson.com
>             <mailto:goran.ap.eriksson@ericsson.com>]
>              > Sent: 14 February 2007 23:34
>              > To: Elwell, John
>              > Cc: sipping@ietf.org <mailto:sipping@ietf.org>
>              > Subject: Device architecture? WAS: RE: [Sipping] Service
>              > Identification in SIP
>              >
>              > My humble, perhaps faulty, understanding of the "chess game"
>              > was not that voice commands were used to control the chess
>              > itself, rather
>              > that the chess game, while having some IP-based proprietary
>              > gaming interaction protocol, also had a feature letting
>              > the users talk to each others while playing. The scenario
>             are
>              > two, or more, users play a chess game of the (Inter)net,
>             talkin
>              > to eachother from time to time during the game (yeah, don't
>              > ask me if I think people will love this!!).
>              >
>              > The purpose, as I understand it, was a)to show that the
>              > 'voice component' in the chess game application is
>             similar to that
>              > in a stand alone POC-Only application when it comes to
>              > exchange and transfer of voice samples and associated
>             floor control
>              > and b) to point to differences, for instance when it comes to
>              > supplementary services, which may not apply in the chess
>             game case.
>              >
>              > Perhaps it is important to be clear on the assumed user
>              > device realisation, as far as I understand it:
>              >
>              >  The "chess game + voice" example is, as I understand it,
>              > discussed under the assumption that voice and chess game is
>              >  managed by custome designed appliction, or client, which is
>              > a necessary/likely approach in a mobile device given the
>              >  limited MMI (and also other factors in the mobile context).
>              >
>              >  In a PC, I could easily handle this as two different
>              > applications/client, like when I use Skype/MSN to talk to my
>              > friends while
>              >  playing some turn based game, but again, the discussion is
>              > held under the assumption that usability requirements
>              >  leads to an application client integrating voice and chess
>              > game events.
>              >
>              > The alternative device realisation You describe- different
>              > clients managing different portions of what is negotiated in
>              > the SDP or MSRP MIME attachment-
>              > is, I think, a good illustration of the need to be precise
>              > and explicit about the assumed device architecture when
>              > discussing the "service identification" issue(s).
>              >
>              > Regards
>              > Göran
>              >
>              >
>              >
>              >
>              > -----Original Message-----
>              > From: Elwell, John [mailto: john.elwell@siemens.com
>             <mailto:john.elwell@siemens.com>]
>              > Sent: den 14 februari 2007 11:57
>              > To: Henry Sinnreich; sipping@ietf.org
>             <mailto:sipping@ietf.org>
>              > Subject: RE: [Sipping] Service Identification in SIP
>              >
>              > Having followed this thread, I too find the chess example
>              > very contrived. If chess moves are sent as voice commands,
>              > then surely it would be acceptable for the normal telephone
>              > application to answer the call, i.e., the remote player
>              > listens to the move, performs this move on a conventional or
>              > electronic chess board and speaks his next move. The calling
>              > UA should not care whether the call is answered by a
>             computer
>              > application or a person.
>              > Furthermore, even if the audio needs to consumed by a
>              > particular chess application at the called UA, then other
>              > media in the same session such as video or messaging might
>              > not be. So any qualification of the nature of media to be
>              > exchanged probably should be at the media level, not at the
>              > session level or SIP level. As yet another variation, what if
>              > the players set up a conventional telephone call and then
>              > mutually agree to switch in their chess applications?
>              >
>              > Now, let's consider another case. I establish an MSRP
>              > session. Within this MSRP session I can send messages with
>              > various MIME attachments that are intended for
>             consumption by
>              > various applications. I should not need to know in advance
>              > what applications I will use during the course of the MSRP
>              > session. All I should put in the INVITE request is the fact
>              > that I want to establish an MSRP session.
>              >
>              > John
>              >
>              > > -----Original Message-----
>              > > From: Henry Sinnreich [mailto:hsinnrei@adobe.com
>             <mailto:hsinnrei@adobe.com>]
>              > > Sent: 13 February 2007 18:19
>              > > To: sipping@ietf.org <mailto:sipping@ietf.org>
>              > > Subject: RE: [Sipping] Service Identification in SIP
>              > >
>              > > It must be my lack of imagination for not being able to
>              > come up with a
>              > > credible usage scenario, given the following:
>              > >
>              > > - Is playing chess by voice a credible usage scenario? :-)
>              > Especially
>              > > since voice recognition is now so often found in PCs
>             and in mobile
>              > > phones as well. In other words, chess playing by voice may
>              > also be an
>              > > endpoint application, without any network involvement.
>             Actually, if
>              > > placed in the endpoint it works just as well on poor
>              > bandwidth links
>              > > by contrast to chess in the network that may require
>             adequate
>              > > broadband.
>              > >
>              > > - Is anyone's boss or spouse going to approve extra
>             phone charges
>              > > above the flat rate for playing chess at a time when
>             the gaming
>              > > industry has pioneered free VoIP?
>              > > Will such a service be taxed extra and if yes using by what
>              > criteria?
>              > > Gambling comes to my mind :-)
>              > >
>              > > - VoIP is reported in the press as a give-away and I
>             have seen no
>              > > reports of any VoIP provider recording a profit on VoIP. I
>              > assume the
>              > > losses are often hidden by bundling in triple play, etc.
>              > >
>              > > Given my lack of imagination, can anyone share a more
>              > credible usage
>              > > scenario for Service Identification in SIP?
>              > >
>              > > What's the use of a standard without a credible usage
>             scenario?
>              > >
>              > > Thanks, Henry
>              > >
>              > >
>              > > _______________________________________________
>              > > Sipping mailing list  
>             https://www1.ietf.org/mailman/listinfo/sipping
>             <https://www1.ietf.org/mailman/listinfo/sipping>
>              > > This list is for NEW development of the application of
>             SIP Use
>              > > sip-implementors@cs.columbia.edu
>             <mailto:sip-implementors@cs.columbia.edu> for questions on
>             current sip Use
>              > > sip@ietf.org <mailto:sip@ietf.org> for new developments
>             of core SIP
>              > >
>              >
>              > _______________________________________________
>              > Sipping mailing list  
>             https://www1.ietf.org/mailman/listinfo/sipping
>              > This list is for NEW development of the application of SIP
>              > Use sip-implementors@cs.columbia.edu
>             <mailto:sip-implementors@cs.columbia.edu> for questions on
>             current
>              > sip Use sip@ietf.org <mailto:sip@ietf.org> for new
>             developments of core SIP
>              >
> 
>             _______________________________________________
>             Sipping mailing list  
>             https://www1.ietf.org/mailman/listinfo/sipping
>             <https://www1.ietf.org/mailman/listinfo/sipping>
>             This list is for NEW development of the application of SIP
>             Use sip-implementors@cs.columbia.edu
>             <mailto:sip-implementors@cs.columbia.edu>for questions on
>             current sip
>             Use sip@ietf.org <mailto:sip@ietf.org> for new developments
>             of core SIP
> 
> 
>     ------------------------------------------------------------------------
>     This transmission (including any attachments) may contain
>     confidential information, privileged material (including material
>     protected by the solicitor-client or other applicable privileges),
>     or constitute non-public information. Any use of this information by
>     anyone other than the intended recipient is prohibited. If you have
>     received this transmission in error, please immediately reply to the
>     sender and delete this information from your system. Use,
>     dissemination, distribution, or reproduction of this transmission by
>     unintended recipients is not authorized and may be unlawful. 
> 
> ------------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential 
> information, privileged material (including material protected by the 
> solicitor-client or other applicable privileges), or constitute 
> non-public information. Any use of this information by anyone other than 
> the intended recipient is prohibited. If you have received this 
> transmission in error, please immediately reply to the sender and delete 
> this information from your system. Use, dissemination, distribution, or 
> reproduction of this transmission by unintended recipients is not 
> authorized and may be unlawful.
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 20 13:11:06 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJZRs-0006md-MR; Tue, 20 Feb 2007 13:10:44 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJZRr-0006mL-KZ
	for sipping@ietf.org; Tue, 20 Feb 2007 13:10:43 -0500
Received: from sj-iport-2-in.cisco.com ([171.71.176.71]
	helo=sj-iport-2.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HJZRp-0001SE-UB
	for sipping@ietf.org; Tue, 20 Feb 2007 13:10:43 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by sj-iport-2.cisco.com with ESMTP; 20 Feb 2007 10:10:40 -0800
X-IronPort-AV: i="4.14,197,1170662400"; 
	d="scan'208"; a="361837332:sNHT63468096"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1KIAe4i032765; 
	Tue, 20 Feb 2007 13:10:40 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1KIAYVv010966; 
	Tue, 20 Feb 2007 13:10:40 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 20 Feb 2007 13:10:34 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 20 Feb 2007 13:10:34 -0500
Message-ID: <45DB399A.9010108@cisco.com>
Date: Tue, 20 Feb 2007 13:10:34 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Elwell, John" <john.elwell@siemens.com>
Subject: Re: Device architecture? WAS: RE: [Sipping] Service	Identificatio
	ninSIP
References: <50B1CBA96870A34799A506B2313F26670B0F8A1C@ntht201e.siemenscomms.co.uk>
In-Reply-To: <50B1CBA96870A34799A506B2313F26670B0F8A1C@ntht201e.siemenscomms.co.uk>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 20 Feb 2007 18:10:34.0342 (UTC)
	FILETIME=[6A3FB860:01C7551A]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=7722; t=1171995040;
	x=1172859040; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20Device=20architecture?=20WAS=3A=20RE=3A=20[Sipping]=2
	0Service=09Identificatio=0A=20ninSIP |Sender:=20
	|To:=20=22Elwell,=20John=22=20<john.elwell@siemens.com>;
	bh=+xlxDMzAm5MBvdo2vNMRIUn5sDGd1Fta/In/8KMx970=;
	b=Fs+VLmQnvnpWtVLcgiYUhXHbZfbizQOYIls/tw0Erf2L+9DOhW0ptrscN86sNMfDoFIoVP4n
	F6gMZLrL2oiSaMy/UXxyBCavOGIKUulsxKWMheJ2CRqOm3j9IfIomQ/v;
Authentication-Results: rtp-dkim-1; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 03169bfe4792634a390035a01a6c6d2f
Cc: Andrew Allen <aallen@rim.com>, sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Elwell, John wrote:

> I have still not been convinced by arguments for having a service 
> identifier that can be used by intermediaries, but one possibility has 
> emerged from these discussions that I might be able to buy. It seems 
> there might be some need to convey some higher level "purpose" for a 
> request that might influence routing decisions at SIP intermediaries, in 
> addition to being used by the UAS. I think one example cited was 
> push-to-talk (PTT), which might want to avoid being subject to call 
> forwarding that normal INVITE requests might undergo. A user may be able 
> to arrange that his local proxy retargets or forwards requests under 
> certain conditions. If a call were marked as "PTT" or "chess", the user 
> could arrange that such calls be subject to different rules.

As soon as I discover you have done that, I will start to mark my calls 
as PTT any time I want to avoid having them forwarded, even though they 
aren't PTT calls.

For this, I think one might as well admit that you are opening the 
forwarding to caller preference, and simply use caller preferences for it.

	Paul

> So the 
> called user (or an administration acting on the called user's behalf) 
> decides on the rules, which can take into account whatever the calling 
> user claims to be the purpose of the call. I am not sure about the 
> security implications here, because the caller might get to learn what 
> to do to get around retargeting imposed by the callee.
>  
> Furthermore, I am not convinced by the notion of authorisation or 
> charging being based on such a service identifier unless there are some 
> appropriately strong security mechanism to prevent misuses by the 
> caller. Charging on the basis of a service identifier is fine if the 
> network really provides some added value that it would not provide in 
> the absence of that service identifier. But given that network 
> intermediaries cannot examine the SDP, it it not clear exactly what 
> added value might be provided. In the absence of any value added by the 
> network, uses would not use the service identifier and therefore avoid 
> higher charges (presumably finding some out of band means to inform the 
> called user of the wish to use a particular service). This is where we 
> still seem to lack compelling examples.
>  
> John
>  
> 
>     ------------------------------------------------------------------------
>     *From:* Andrew Allen [mailto:aallen@rim.com]
>     *Sent:* 16 February 2007 19:59
>     *To:* Elwell, John; Malleswara Rao Ankem
>     *Cc:* sipping@ietf.org
>     *Subject:* RE: Device architecture? WAS: RE: [Sipping] Service
>     Identificatio ninSIP
> 
>      
>     John
>      
>     I think whether an AS in the terminating network is considered a
>     contact is a debatable point. Certainly it doesnt register feature
>     tags using RFC 3840 mechanisms and (at least in IMS) they dont show
>     up as registered contacts in the reg-event notifications. Also AS in
>     the originating network (which may need to be involved in some
>     communication services) are certainly not contacts for the resource
>     identified by the Request-URI and Accept-Contact certainly has no
>     defined role in routing to them.
>      
>     In addition 3GPP TS 23.228 v7.6.0 subcaluse 4.13.2 specifies 15
>     requirements for IMS Communication Service Identifier and only one
>     of these is regarding matching to the correct contact:
>      
> 
>     11. The communication service identifier shall be taken into account
>     when selecting the correct UE(s), if multiple UEs are registered for
>     the same Public User Identity(s).
> 
>      
>      Several of the other requirements are in my view totally
>     incompatible with Accept-Contact being the transport for the
>     identifier and indeed may limit the use of Accept-Contact for its
>     intended purpose. This includes charging and policy control:
>      
> 
>     8.   Based on operator policy S-CSCF and AS shall consider the IMS
>     communication service identifier for online and offline charging,
>     e.g. put appropriate data into call detailed records.
> 
>     9.   The communication service identifier shall be capable of being
>     an input into the policy control and charging rules.
> 
>     Also:
>      
> 
>     2.   Based on operator policy the S-CSCF or an AS shall be able to
>     validate an IMS communication service identifier in a SIP request.
>     This includes e.g. to check the syntactical correctness of a service
>     identifier, and policing the usage of a communication service
>     identifier.
> 
>     10. It shall be possible to use the IMS communication service
>     identifier as a means to authorise whether a subscriber is allowed
>     to initiate or receive request for a communication service.
> 
>     Another problem with UAs placing feature tags in Accept-Contact
>     headers as Service Identifiers is how does the initiator obtains the
>     service identifier of the destination party for the related service?
>     The concept only works in the scenario where all networks and
>     devices have a common understanding of the same services and the
>     identifiers that identify them. As far as I am aware SIP doesnt
>     provide a mechanism to discover proprietary non standard feature
>     tags that are used by other devices in other domains and map them to
>     services that they relate to. This means that if Accept-Contact is
>     used the communication services are closed and can only be used by
>     those who are part of the closed network, (unless of course IETF
>     were to get into standardising services which I think is highly
>     unlikely). If URIs are used as the container for the Service
>     Identifier these can be obtained using Presence as proposed in
>     Jonathan's draft.
>      
>     This helps fulfil the following requirements in 3GPP TS 23.228:
>      
> 
>     12. The usage of communication service identifiers shall not
>     adversely affect interoperability between IMS networks and
>     interoperability with external SIP networks and CS networks. The
>     behaviour of a network receiving the IMS requests without an IMS
>     communication service identifier is a matter of operator policy.
>     Usage of communication service identifiers shall not decrease the
>     level of interoperability with networks and UEs that are unaware of
>     the communication service identifier.
> 
>     14. The usage of communication service identifiers shall not
>     restrict the inherent capabilities of SIP.
> 
>      
>     SDP is not good for routing decisions for service identification.
>     Firstly SDP is intended to be end to end only. It may be encrypted
>     and understanding the details of the service may require service or
>     application level understanding of the SDP and the service which is
>     knowledge that proxies should not need to have. Also I dont think
>     that SDP can always uniquely identify a service in any case. In
>     addition not all requests that are related to a service will contain
>     SDP in any case.
>      
>     Andrew
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 20 14:00:51 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJaE2-0007Kt-1X; Tue, 20 Feb 2007 14:00:30 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJaE0-0007Ki-6D
	for sipping@ietf.org; Tue, 20 Feb 2007 14:00:28 -0500
Received: from sccrmhc12.comcast.net ([204.127.200.82])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HJaDx-0004L0-Bd
	for sipping@ietf.org; Tue, 20 Feb 2007 14:00:28 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (sccrmhc12) with ESMTP
	id <2007022019002301200pr5dhe>; Tue, 20 Feb 2007 19:00:23 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1KJ0N6g022368
	for <sipping@ietf.org>; Tue, 20 Feb 2007 14:00:23 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1KJ0NnI022364;
	Tue, 20 Feb 2007 14:00:23 -0500
Date: Tue, 20 Feb 2007 14:00:23 -0500
Message-Id: <200702201900.l1KJ0NnI022364@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <989763567E51AB48BE015DF7BB960CAD028861AF@esealmw111>
	(stephen.terrill@ericsson.com)
Subject: Re: [Sipping] Service Identification in SIP
References: <989763567E51AB48BE015DF7BB960CAD028861AF@esealmw111>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 41c17b4b16d1eedaa8395c26e9a251c4
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>

   The draft also only recognises the case where all explicitly knowledge
   of the service is required.  I think it would also be good to allow the
   case where knowledge of the service would be helpful, but not explicitly
   required.  This could be a case where the service to the end-user would
   not be degraded if this was not supported.  This functionality seems to
   be missing.

And it seems to me that the ability to indicate support of a feature,
but not mandatory use, would greatly help prevent the "silo-ization"
problem.

   From: Jonathan Rosenberg <jdrosen@cisco.com>

   Unfortunately, PoC has layered a whole bunch of additional behaviors 
   into the way SIP requests are handled. Dale has indicated one of them - 
   that a 200 OK is returned on the terminating side priod to the call 
   actually being completed. PoC also *requires* the insertion of servers 
   into the SIP signaling path to make the system work. Consequently, 
   though the messages that PoC uses *look* like SIP messages with a new 
   media type, there is a hidden 'extension' in place that changes the 
   fundamental semantics of INVITE. The fact that this extension is not 
   visible as a specific header in a well defined way does not change the 
   fact that PoC CHANGES THE BEHAVIOR OF SIP AND IS THUS A SIP EXTENSION.

   To convince yourself of this, consider what would happen if a PoC client 
   calls a regular SIP phone. The regular SIP phone would see the PoC floor 
   control stream and reject it as unknown. If PoC were just SIP as we've 
   specified it, then everything would be fine, and the call would proceed 
   but without floor control.

   But alas, it is not so.

But why is it not so?  OK, I know nothing about PoC, but naively it
seems that a PoC [here I assume that PoC means "push-to-talk over
cellular"] UAC, if it did not see an immediate 200 would know that the
far end did not support PoC functionality and could fall back to
acting like an ordinary UAC.  (Indeed, it could quickly warn the user
that PoC is not in effect by playing a ring tone!)

Assuming that SIP-based PoC was designed with upward-compatibility in
mind, a PoC UA could reliably determine whether the far end supported
PoC on a call-by-call basis.

OTOH, if PoC changes the meanings of SIP messages so much that,
without a specified service context, they cannot be interpreted enough
to determine if PoC is in effect or not, um, well, what *were* they
thinking when they designed it?  And why haven't they fixed it?

Now, I suspect I've just revealed a large gap in my knowledge, but I
suspect that many other readers have the same gap, and it would
advance the discussion if you could reduce our ignorance.

   From: Henning Schulzrinne <hgs@cs.columbia.edu>

   A robust and distributed allocation mechanism for identifiers makes
   sense if you expect hundreds of entities to independently need to
   allocate such identifiers. This is clearly undesirable in this
   case, based on the examples presented, so the IANA registration
   mechanism for Require tags seems to be both sufficient and
   sufficiently onerous to prevent gratuitous incompatibility.

Naively, it seems that having more than a few defined services would
be useless in regard to the business of the cellular industry, as it
would balkanize the cellular PSTN, and customers would fall back to
using only the simplest modes of operation.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 20 14:01:33 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJaF2-0007qV-Vv; Tue, 20 Feb 2007 14:01:32 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJaF1-0007mE-Tu
	for sipping@ietf.org; Tue, 20 Feb 2007 14:01:31 -0500
Received: from sccrmhc14.comcast.net ([63.240.77.84])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HJaF0-0004Xo-Mn
	for sipping@ietf.org; Tue, 20 Feb 2007 14:01:31 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (sccrmhc14) with ESMTP
	id <2007022019012601400s8gp6e>; Tue, 20 Feb 2007 19:01:30 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1KJ1Q6g022452
	for <sipping@ietf.org>; Tue, 20 Feb 2007 14:01:26 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1KJ1Q6G022448;
	Tue, 20 Feb 2007 14:01:26 -0500
Date: Tue, 20 Feb 2007 14:01:26 -0500
Message-Id: <200702201901.l1KJ1Q6G022448@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <7374777208BDC7449D5620EF94232567021CE512@esealmw113.eemea.ericsson.se>
	(christer.holmberg@ericsson.com)
Subject: Re: [Sipping] Service Identification in SIP - Routing decissions
	based on Require/P-R
References: <7374777208BDC7449D5620EF94232567021CE512@esealmw113.eemea.ericsson.se>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 7a6398bf8aaeabc7a7bb696b6b0a2aad
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>

   Just for my clarification, where is it said that Require and
   Proxy-Require are used, or at least more appropriate, for routing?

   As far as I know, they are only used to require capabilities by the
   nodes that receive a SIP request (as you said yourself earlier) - but
   not to make routing decissions based upon.

Surely a proxy can make routing decisions based upon Require if it
desires to.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 20 15:51:34 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJbw8-0004qU-8P; Tue, 20 Feb 2007 15:50:08 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJbw2-0004nu-VJ; Tue, 20 Feb 2007 15:50:02 -0500
Received: from ns1.neustar.com ([2001:503:c779:1a::9c9a:108a])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HJbw2-00008W-KJ; Tue, 20 Feb 2007 15:50:02 -0500
Received: from stiedprstage1.ietf.org (stiedprstage1.va.neustar.com
	[10.31.47.10]) by ns1.neustar.com (Postfix) with ESMTP id 7965226EB6;
	Tue, 20 Feb 2007 20:50:02 +0000 (GMT)
Received: from ietf by stiedprstage1.ietf.org with local (Exim 4.43)
	id 1HJbw2-0007Ok-Br; Tue, 20 Feb 2007 15:50:02 -0500
Content-Type: Multipart/Mixed; Boundary="NextPart"
Mime-Version: 1.0
To: i-d-announce@ietf.org
From: Internet-Drafts@ietf.org
Message-Id: <E1HJbw2-0007Ok-Br@stiedprstage1.ietf.org>
Date: Tue, 20 Feb 2007 15:50:02 -0500
X-Spam-Score: -2.5 (--)
X-Scan-Signature: a2c12dacc0736f14d6b540e805505a86
Cc: sipping@ietf.org
Subject: [Sipping] I-D ACTION:draft-ietf-sipping-dialogusage-06.txt 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

--NextPart

A New Internet-Draft is available from the on-line Internet-Drafts 
directories.
This draft is a work item of the Session Initiation Proposal Investigation Working Group of the IETF.

	Title		: Multiple Dialog Usages in the Session Initiation Protocol
	Author(s)	: R. Sparks
	Filename	: draft-ietf-sipping-dialogusage-06.txt
	Pages		: 29
	Date		: 2007-2-20
	
Several methods in the Session Initiation Protocol can create an
   association between endpoints known as a dialog.  Some of these
   methods can also create a different, but related, association within
   an existing dialog.  These multiple associations, or dialog usages,
   require carefully coordinated processing as they have independent
   life-cycles, but share common dialog state.  Processing multiple
   dialog usages correctly is not completely understood.  What is
   understood is difficult to implement.
   This memo argues that multiple dialog usages should be avoided.  It
   discusses alternatives to their use and clarifies essential behavior
   for elements that cannot currently avoid them.

   This is an informative document and makes no normative statements of
   any kind.

A URL for this Internet-Draft is:
http://www.ietf.org/internet-drafts/draft-ietf-sipping-dialogusage-06.txt

To remove yourself from the I-D Announcement list, send a message to 
i-d-announce-request@ietf.org with the word unsubscribe in the body of 
the message. 
You can also visit https://www1.ietf.org/mailman/listinfo/I-D-announce 
to change your subscription settings.

Internet-Drafts are also available by anonymous FTP. Login with the 
username "anonymous" and a password of your e-mail address. After 
logging in, type "cd internet-drafts" and then 
"get draft-ietf-sipping-dialogusage-06.txt".

A list of Internet-Drafts directories can be found in
http://www.ietf.org/shadow.html 
or ftp://ftp.ietf.org/ietf/1shadow-sites.txt

Internet-Drafts can also be obtained by e-mail.

Send a message to:
	mailserv@ietf.org.
In the body type:
	"FILE /internet-drafts/draft-ietf-sipping-dialogusage-06.txt".
	
NOTE:	The mail server at ietf.org can return the document in
	MIME-encoded form by using the "mpack" utility.  To use this
	feature, insert the command "ENCODING mime" before the "FILE"
	command.  To decode the response(s), you will need "munpack" or
	a MIME-compliant mail reader.  Different MIME-compliant mail readers
	exhibit different behavior, especially when dealing with
	"multipart" MIME messages (i.e. documents which have been split
	up into multiple messages), so check your local documentation on
	how to manipulate these messages.

Below is the data which will enable a MIME compliant mail reader
implementation to automatically retrieve the ASCII version of the
Internet-Draft.

--NextPart
Content-Type: Multipart/Alternative; Boundary="OtherAccess"

--OtherAccess
Content-Type: Message/External-body; access-type="mail-server";
	server="mailserv@ietf.org"

Content-Type: text/plain
Content-ID: <2007-2-20102303.I-D@ietf.org>

ENCODING mime
FILE /internet-drafts/draft-ietf-sipping-dialogusage-06.txt

--OtherAccess
Content-Type: Message/External-body;
	name="draft-ietf-sipping-dialogusage-06.txt";
	site="ftp.ietf.org"; access-type="anon-ftp";
	directory="internet-drafts"

Content-Type: text/plain
Content-ID: <2007-2-20102303.I-D@ietf.org>


--OtherAccess--

--NextPart
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--NextPart--





From sipping-bounces@ietf.org Tue Feb 20 17:04:45 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJd5t-0003Rr-St; Tue, 20 Feb 2007 17:04:17 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJd5s-0003QX-LE
	for sipping@ietf.org; Tue, 20 Feb 2007 17:04:16 -0500
Received: from sj-iport-2-in.cisco.com ([171.71.176.71]
	helo=sj-iport-2.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HJd5q-0002pa-8z
	for sipping@ietf.org; Tue, 20 Feb 2007 17:04:16 -0500
Received: from sj-dkim-4.cisco.com ([171.71.179.196])
	by sj-iport-2.cisco.com with ESMTP; 20 Feb 2007 14:04:13 -0800
X-IronPort-AV: i="4.14,197,1170662400"; 
	d="scan'208"; a="361891993:sNHT53867380"
Received: from sj-core-5.cisco.com (sj-core-5.cisco.com [171.71.177.238])
	by sj-dkim-4.cisco.com (8.12.11/8.12.11) with ESMTP id l1KM4DT3020705; 
	Tue, 20 Feb 2007 14:04:13 -0800
Received: from dwingwxp ([10.32.240.194])
	by sj-core-5.cisco.com (8.12.10/8.12.6) with ESMTP id l1KM4CUw001549;
	Tue, 20 Feb 2007 14:04:12 -0800 (PST)
From: "Dan Wing" <dwing@cisco.com>
To: <ietf-rtpsec@imc.org>, <sipping@ietf.org>
Date: Tue, 20 Feb 2007 14:04:12 -0800
Message-ID: <053701c7553b$0e3b57b0$c2f0200a@amer.cisco.com>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.3028
Thread-Index: AcdVMo8LKyIIPWrYQMWlrDazeqrZZgACAC6Q
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=3077; t=1172009053;
	x=1172873053; c=relaxed/simple; s=sjdkim4002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=dwing@cisco.com;
	z=From:=20=22Dan=20Wing=22=20<dwing@cisco.com>
	|Subject:=20draft-wing-sipping-srtp-key-00.txt=20 |Sender:=20;
	bh=jXeaR6YmuCREt5bDrJ5aL7taUoPZdqAgnRy7K+VGWek=;
	b=MSA+zwl2x7/QdA9quBnNOWVCssSackYb2guV3CMbJkLVX8ANV1LCwcp6G6E09ezIdL1zGm5G
	adfXSAk+jOXAGe/xBoc9f9Ub8tvxViD6We3Xrv4dHF1MIwgIUN5xm+s4;
Authentication-Results: sj-dkim-4; header.From=dwing@cisco.com; dkim=pass (s
	ig from cisco.com/sjdkim4002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: bdc523f9a54890b8a30dd6fd53d5d024
Cc: 'Francois Audet' <audet@nortel.com>, "'Fries,
	Steffen'" <steffen.fries@siemens.com>
Subject: [Sipping] draft-wing-sipping-srtp-key-00.txt 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

This draft provides a way for cooperating endpoints to disclose their SRTP
keys to authorized elements in the network.  This allows deployment of SRTP
in environments where call recording is necessary to meet legal requirements
(stock brokers, banks) or business requirements (typical call centers where
"calls are monitored for quality purposes").

Comments welcome.

-d
 

> -----Original Message-----
> From: Internet-Drafts@ietf.org [mailto:Internet-Drafts@ietf.org] 
> Sent: Tuesday, February 20, 2007 12:50 PM
> To: i-d-announce@ietf.org
> Subject: I-D ACTION:draft-wing-sipping-srtp-key-00.txt 
> 
> A New Internet-Draft is available from the on-line Internet-Drafts 
> directories.
> 
> 
> 	Title		: Disclosing Secure RTP (SRTP) Session 
> Keys with a SIP Event Package
> 	Author(s)	: D. Wing, et al.
> 	Filename	: draft-wing-sipping-srtp-key-00.txt
> 	Pages		: 12
> 	Date		: 2007-2-20
> 	
>    Many Secure RTP (SRTP) key exchange mechanisms do not disclose the
>    SRTP session keys to intermediate SIP proxies.  However, these key
>    exchange mechanisms cannot be used In environments where 
> transcoding,
>    monitoring, or call recording are needed.  This document 
> specifies a
>    secure mechanism for a cooperating endpoint to disclose its SRTP
>    master keys to an authorized party.
> 
> 
> A URL for this Internet-Draft is:
> http://www.ietf.org/internet-drafts/draft-wing-sipping-srtp-key-00.txt
> 
> To remove yourself from the I-D Announcement list, send a message to 
> i-d-announce-request@ietf.org with the word unsubscribe in 
> the body of 
> the message. 
> You can also visit 
> https://www1.ietf.org/mailman/listinfo/I-D-announce 
> to change your subscription settings.
> 
> Internet-Drafts are also available by anonymous FTP. Login with the 
> username "anonymous" and a password of your e-mail address. After 
> logging in, type "cd internet-drafts" and then 
> "get draft-wing-sipping-srtp-key-00.txt".
> 
> A list of Internet-Drafts directories can be found in
> http://www.ietf.org/shadow.html 
> or ftp://ftp.ietf.org/ietf/1shadow-sites.txt
> 
> Internet-Drafts can also be obtained by e-mail.
> 
> Send a message to:
> 	mailserv@ietf.org.
> In the body type:
> 	"FILE /internet-drafts/draft-wing-sipping-srtp-key-00.txt".
> 	
> NOTE:	The mail server at ietf.org can return the document in
> 	MIME-encoded form by using the "mpack" utility.  To use this
> 	feature, insert the command "ENCODING mime" before the "FILE"
> 	command.  To decode the response(s), you will need "munpack" or
> 	a MIME-compliant mail reader.  Different MIME-compliant 
> mail readers
> 	exhibit different behavior, especially when dealing with
> 	"multipart" MIME messages (i.e. documents which have been split
> 	up into multiple messages), so check your local documentation on
> 	how to manipulate these messages.
> 
> Below is the data which will enable a MIME compliant mail reader
> implementation to automatically retrieve the ASCII version of the
> Internet-Draft.
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 20 17:30:39 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJdVA-0003mi-Gr; Tue, 20 Feb 2007 17:30:24 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJdV8-0003mE-Rb
	for sipping@ietf.org; Tue, 20 Feb 2007 17:30:22 -0500
Received: from sj-iport-1-in.cisco.com ([171.71.176.70]
	helo=sj-iport-1.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HJdV7-0000zf-DI
	for sipping@ietf.org; Tue, 20 Feb 2007 17:30:22 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by sj-iport-1.cisco.com with ESMTP; 20 Feb 2007 14:30:20 -0800
X-IronPort-AV: i="4.14,197,1170662400"; 
	d="scan'208"; a="763641251:sNHT709842688"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1KMUJ7M024860; 
	Tue, 20 Feb 2007 17:30:19 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1KMTvVj025978; 
	Tue, 20 Feb 2007 17:30:14 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 20 Feb 2007 17:30:07 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 20 Feb 2007 17:30:06 -0500
Message-ID: <45DB766E.6010001@cisco.com>
Date: Tue, 20 Feb 2007 17:30:06 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Atle Monrad (GR/ETO)" <atle.monrad@ericsson.com>
Subject: Re: [Sipping] Service Identification in SIP
References: <61D8D34BB13CFE408D154529C120E0799A6351@eseldmw101.eemea.ericsson.se>
In-Reply-To: <61D8D34BB13CFE408D154529C120E0799A6351@eseldmw101.eemea.ericsson.se>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 20 Feb 2007 22:30:06.0788 (UTC)
	FILETIME=[AC266840:01C7553E]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=3714; t=1172010619;
	x=1172874619; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20
	|To:=20=22Atle=20Monrad=20(GR/ETO)=22=20<atle.monrad@ericsson.com>;
	bh=nU6vnBuCR54wnE6oPFbFh6BZN+Fl239r0LfBlMKsF0k=;
	b=BRUaK1p8wI1aLB+fPGgw8KbHi0fNrluC68UjvUriPEasTRe+Bgkf2R0aJpJK7KqON/PKzA7A
	ncJLUiPh7Wt6lNJLdmo759JAS/L+rnvWcilnfvNXzRNEDOhKupOlBZlY;
Authentication-Results: rtp-dkim-1; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 386e0819b1192672467565a524848168
Cc: sipping <sipping@ietf.org>,
	"Gonzalo Camarillo \(JO/LMF\)" <gonzalo.camarillo@ericsson.com>,
	Mary Barnes <mary.barnes@nortel.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Atle Monrad (GR/ETO) wrote:
> Paul
> 
> by the way, it would be useful to get this discussion on the sipping
> list.

it is.

> As a service is defined in terminals and possibly in application
> servers, the interoperability on the service level are between the
> terminals and the ASs that have implemented the service while
> interoperability on the protocol level is something very different.
> Please tell me which interop that is your concern

I don't think it is specifically interop at the protocol level. But it 
is involved in the expected semantics involved in using sip. I don't 
know if that is the service level or what.

For instance somebody (I think Jonathan) already identified issues if 
PoC device called what turns out to be a vanilla sip audio UAS. At the 
protocol level everything is fine - the audio media stream has been 
accepted and anything else rejected. The callee expects an audio call. 
The calling PoC client, or maybe his PoC server, will explode or 
something I guess. This is an interoperability problem.

> I do not see any interoparability, if you do, it would be useful that
> you share it. 
> 
> Terminals may have different services and may not be interoperable in
> all aspects, but isn't this just facts that is hard to get around?

Well, I'm not concerned if my attempted IM call to you fails because you 
only have audio devices.

But if I take my vanilla sip audio/video client and make an audio/video 
call to you, who I know have an audio/video UAS, and my call is rejected 
because my UA didn't include the service ID that your provider requires 
for video calls, then I think there is an interop problem. If we don't 
support any common codecs then I will understand.

> A PoC
> terminal can only talk PoC to other PoC terminals for the PoC service
> but it can communicate with other terminals aswell, right?  

As Henning and others have pointed out, this is a fact based on the 
current approach to implementing PoC. But it isn't a desirable 
situation. There are other ways it could have been done that would have 
led to a more graceful handling of interop.

I gather that after you have failed to complete a call to me using PoC 
on your phone that you will switch it to regular audio mode and try 
again. It would be much more pleasing if your phone simply noticed that 
the response was not compatible with PoC and just switched to regular 
voice. In that case, PoC-specific servers that have Record-Routed would 
have to get out of the way somehow.

	Paul

> /atle
> 
> FYI, GSMA consists of 700 operators. When GSMA give their
> recommendations, it would be useful to listen and at least give some
> technical feeback and not this kind of political statement. 
> 
> 
> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
> Sent: 20. februar 2007 15:31
> To: Atle Monrad (GR/ETO)
> Cc: Gonzalo Camarillo (JO/LMF); sipping; Mary Barnes
> Subject: Re: [Sipping] Service Identification in SIP
> 
> 
> 
> Atle Monrad (GR/ETO) wrote:
> 
>> On the interoperability, I would just like to add that during 
>> discussions in 3GPP, interoperability has not been considered a 
>> problem by the operators. On the contrary, the GSM Association has 
>> asked 3GPP to finalise the work on communication service identifiers 
>> and sort out the bits and pieces in the protocol ASAP.
> 
> Of course they don't find interoperability to be a problem! They
> constitute an oligopoly that is only concerned with interop among
> themselves. Interop with anyone else is presumably a non-goal.
> 
> The ietf has broader aspirations for interop.
> 
> 	Paul
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 21 11:04:48 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJtvf-0005JV-0N; Wed, 21 Feb 2007 11:02:51 -0500
Received: from [10.90.34.44] (helo=chiedprmail1.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJtkJ-0008HH-6E; Wed, 21 Feb 2007 10:51:07 -0500
Received: from ns4.neustar.com ([156.154.24.139])
	by chiedprmail1.ietf.org with esmtp (Exim 4.43)
	id 1HJtkE-0007nD-Br; Wed, 21 Feb 2007 10:51:07 -0500
Received: from stiedprstage1.ietf.org (stiedprstage1.va.neustar.com
	[10.31.47.10]) by ns4.neustar.com (Postfix) with ESMTP id 63EBC2ACB3;
	Wed, 21 Feb 2007 15:50:02 +0000 (GMT)
Received: from ietf by stiedprstage1.ietf.org with local (Exim 4.43)
	id 1HJtjG-0008Sp-3M; Wed, 21 Feb 2007 10:50:02 -0500
Content-Type: Multipart/Mixed; Boundary="NextPart"
Mime-Version: 1.0
To: i-d-announce@ietf.org
From: Internet-Drafts@ietf.org
Message-Id: <E1HJtjG-0008Sp-3M@stiedprstage1.ietf.org>
Date: Wed, 21 Feb 2007 10:50:02 -0500
X-Spam-Score: -2.5 (--)
X-Scan-Signature: 10ba05e7e8a9aa6adb025f426bef3a30
Cc: sipping@ietf.org
Subject: [Sipping] I-D ACTION:draft-ietf-sipping-sbc-funcs-01.txt 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

--NextPart

A New Internet-Draft is available from the on-line Internet-Drafts 
directories.
This draft is a work item of the Session Initiation Proposal Investigation Working Group of the IETF.

	Title		: Requirements from SIP (Session Initiation Protocol) Session Border Control Deployments
	Author(s)	: J. Hautakorpi, et al.
	Filename	: draft-ietf-sipping-sbc-funcs-01.txt
	Pages		: 23
	Date		: 2007-2-21
	
This documents describes functions implemented in Session Initiation
   Protocol (SIP) intermediaries known as Session Border Controllers
   (SBCs).  The goal of this document is to describe the commonly
   provided functions of SBCs.  A special focus is given to those
   practices that are viewed to be in conflict with SIP architectural
   principles.  This document also explores the underlying requirements
   of network operators that have led to the use of these functions and
   practices in order to identify protocol requirements and determine
   whether those requirements are satisfied by existing specifications
   or additional standards work is required.

A URL for this Internet-Draft is:
http://www.ietf.org/internet-drafts/draft-ietf-sipping-sbc-funcs-01.txt

To remove yourself from the I-D Announcement list, send a message to 
i-d-announce-request@ietf.org with the word unsubscribe in the body of 
the message. 
You can also visit https://www1.ietf.org/mailman/listinfo/I-D-announce 
to change your subscription settings.

Internet-Drafts are also available by anonymous FTP. Login with the 
username "anonymous" and a password of your e-mail address. After 
logging in, type "cd internet-drafts" and then 
"get draft-ietf-sipping-sbc-funcs-01.txt".

A list of Internet-Drafts directories can be found in
http://www.ietf.org/shadow.html 
or ftp://ftp.ietf.org/ietf/1shadow-sites.txt

Internet-Drafts can also be obtained by e-mail.

Send a message to:
	mailserv@ietf.org.
In the body type:
	"FILE /internet-drafts/draft-ietf-sipping-sbc-funcs-01.txt".
	
NOTE:	The mail server at ietf.org can return the document in
	MIME-encoded form by using the "mpack" utility.  To use this
	feature, insert the command "ENCODING mime" before the "FILE"
	command.  To decode the response(s), you will need "munpack" or
	a MIME-compliant mail reader.  Different MIME-compliant mail readers
	exhibit different behavior, especially when dealing with
	"multipart" MIME messages (i.e. documents which have been split
	up into multiple messages), so check your local documentation on
	how to manipulate these messages.

Below is the data which will enable a MIME compliant mail reader
implementation to automatically retrieve the ASCII version of the
Internet-Draft.

--NextPart
Content-Type: Multipart/Alternative; Boundary="OtherAccess"

--OtherAccess
Content-Type: Message/External-body; access-type="mail-server";
	server="mailserv@ietf.org"

Content-Type: text/plain
Content-ID: <2007-2-21084608.I-D@ietf.org>

ENCODING mime
FILE /internet-drafts/draft-ietf-sipping-sbc-funcs-01.txt

--OtherAccess
Content-Type: Message/External-body;
	name="draft-ietf-sipping-sbc-funcs-01.txt"; site="ftp.ietf.org";
	access-type="anon-ftp"; directory="internet-drafts"

Content-Type: text/plain
Content-ID: <2007-2-21084608.I-D@ietf.org>


--OtherAccess--

--NextPart
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--NextPart--




From sipping-bounces@ietf.org Wed Feb 21 16:57:50 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJzSZ-00028r-RM; Wed, 21 Feb 2007 16:57:11 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJzSY-000200-29
	for sipping@ietf.org; Wed, 21 Feb 2007 16:57:10 -0500
Received: from usremg02.bea.com ([66.248.192.22])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HJzSX-00021Q-Hr
	for sipping@ietf.org; Wed, 21 Feb 2007 16:57:10 -0500
Received: from usremr01.bea.com (usremr01.bea.com [10.160.29.91])
	by usremg02.bea.com (Switch-3.2.2/Switch-3.2.2) with ESMTP id
	l1LLun36013833
	(version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=FAIL);
	Wed, 21 Feb 2007 13:56:49 -0800
Received: from repbex01.amer.bea.com (repbex01.bea.com [10.160.26.98])
	by usremr01.bea.com (Switch-3.2.2/Switch-3.2.2) with ESMTP id
	l1LLuVUp013088; Wed, 21 Feb 2007 13:56:47 -0800
Received: from 10.32.6.241 ([10.32.6.241]) by repbex01.amer.bea.com
	([10.160.26.98]) with Microsoft Exchange Server HTTP-DAV ; 
	Wed, 21 Feb 2007 21:55:09 +0000
User-Agent: Microsoft-Entourage/11.3.3.061214
Date: Wed, 21 Feb 2007 10:29:27 -0800
Subject: Re: [Sipping] Service Identification in SIP
From: Eric Burger <eburger@bea.com>
To: Henning Schulzrinne <hgs@cs.columbia.edu>,
	Jonathan Rosenberg <jdrosen@cisco.com>
Message-ID: <C201CF87.1FFA%eburger@bea.com>
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdV5jfHdmXZU8HZEdueygAWy4mm/w==
In-Reply-To: <BF1308AC-4B8B-4E43-B9E3-78CA178EE3C0@cs.columbia.edu>
Mime-version: 1.0
Content-type: text/plain;
	charset="US-ASCII"
Content-transfer-encoding: 7bit
X-PMX-Version: 4.7.1.128075, Antispam-Engine: 2.5.0.283055,
	Antispam-Data: 2007.2.21.120936
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 36c793b20164cfe75332aa66ddb21196
Cc: sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

How about this as a different way of saying "+1":

If PoC is broken, why are we breaking SIP to fix it?



On 2/17/07 12:26 PM, "Henning Schulzrinne" <hgs@cs.columbia.edu> wrote:

> I think we're in general agreement; in particular, Require seems like
> the right approach. (service URNs do not, in my opinion).
> 
> Stepping back, I don't think we should generalize this problem since
> it essentially "blesses" a particular design approach that is
> undesirable. The right approach, from a SIP perspective, for IMS PoC
> would have been to come up with a solution that is backwards-
> compatible. If that's not possible, getting "branded" with Require
> seems like the right penalty.
> 
> In general, we have probably two requirements:
> 
> (1) When contacting an AOR, a UAC that needs a special UAS to work
> properly needs to get routed to that UAS only. Callee caps and caller
> preferences seem to be the right tools for that, as it involves call
> routing and capability matching.
> 
> (2) If, despite doing (1), a SIP request accidentally reaches a UAS
> that can't do anything useful with it, Requires minimizes the damage
> (and allows a proxy to serially-fork to something else that might
> have a chance of working).
> 
> Henning
> 
> On Feb 16, 2007, at 8:48 PM, Jonathan Rosenberg wrote:
> 
>> I think its a fair point that the example is contrived. Really, the
>> main use case being targeted is IMS PoC. So, let me talk a bit
>> about that.
>> 
>> One would think that PoC is just a regular SIP call that has floor
>> control added to it as another media type, and so one could
>> indicate that a call is a PoC call just because of the presFrom sipping-bounces@ietf.org Wed Feb 21 16:57:50 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJzSZ-00028r-RM; Wed, 21 Feb 2007 16:57:11 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJzSY-000200-29
	for sipping@ietf.org; Wed, 21 Feb 2007 16:57:10 -0500
Received: from usremg02.bea.com ([66.248.192.22])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HJzSX-00021Q-Hr
	for sipping@ietf.org; Wed, 21 Feb 2007 16:57:10 -0500
Received: from usremr01.bea.com (usremr01.bea.com [10.160.29.91])
	by usremg02.bea.com (Switch-3.2.2/Switch-3.2.2) with ESMTP id
	l1LLun36013833
	(version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=FAIL);
	Wed, 21 Feb 2007 13:56:49 -0800
Received: from repbex01.amer.bea.com (repbex01.bea.com [10.160.26.98])
	by usremr01.bea.com (Switch-3.2.2/Switch-3.2.2) with ESMTP id
	l1LLuVUp013088; Wed, 21 Feb 2007 13:56:47 -0800
Received: from 10.32.6.241 ([10.32.6.241]) by repbex01.amer.bea.com
	([10.160.26.98]) with Microsoft Exchange Server HTTP-DAV ; 
	Wed, 21 Feb 2007 21:55:09 +0000
User-Agent: Microsoft-Entourage/11.3.3.061214
Date: Wed, 21 Feb 2007 10:29:27 -0800
Subject: Re: [Sipping] Service Identification in SIP
From: Eric Burger <eburger@bea.com>
To: Henning Schulzrinne <hgs@cs.columbia.edu>,
	Jonathan Rosenberg <jdrosen@cisco.com>
Message-ID: <C201CF87.1FFA%eburger@bea.com>
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdV5jfHdmXZU8HZEdueygAWy4mm/w==
In-Reply-To: <BF1308AC-4B8B-4E43-B9E3-78CA178EE3C0@cs.columbia.edu>
Mime-version: 1.0
Content-type: text/plain;
	charset="US-ASCII"
Content-transfer-encoding: 7bit
X-PMX-Version: 4.7.1.128075, Antispam-Engine: 2.5.0.283055,
	Antispam-Data: 2007.2.21.120936
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 36c793b20164cfe75332aa66ddb21196
Cc: sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

How about this as a different way of saying "+1":

If PoC is broken, why are we breaking SIP to fix it?



On 2/17/07 12:26 PM, "Henning Schulzrinne" <hgs@cs.columbia.edu> wrote:

> I think we're in general agreement; in particular, Require seems like
> the right approach. (service URNs do not, in my opinion).
> 
> Stepping back, I don't think we should generalize this problem since
> it essentially "blesses" a particular design approach that is
> undesirable. The right approach, from a SIP perspective, for IMS PoC
> would have been to come up with a solution that is backwards-
> compatible. If that's not possible, getting "branded" with Require
> seems like the right penalty.
> 
> In general, we have probably two requirements:
> 
> (1) When contacting an AOR, a UAC that needs a special UAS to work
> properly needs to get routed to that UAS only. Callee caps and caller
> preferences seem to be the right tools for that, as it involves call
> routing and capability matching.
> 
> (2) If, despite doing (1), a SIP request accidentally reaches a UAS
> that can't do anything useful with it, Requires minimizes the damage
> (and allows a proxy to serially-fork to something else that might
> have a chance of working).
> 
> Henning
> 
> On Feb 16, 2007, at 8:48 PM, Jonathan Rosenberg wrote:
> 
>> I think its a fair point that the example is contrived. Really, the
>> main use case being targeted is IMS PoC. So, let me talk a bit
>> about that.
>> 
>> One would think that PoC is just a regular SIP call that has floor
>> control added to it as another media type, and so one could
>> indicate that a call is a PoC call just because of the presence of
>> this new media type in the SDP. Sadly, I wish that were true. If it
>> was, we would not need a mechanism here.
>> 
>> Unfortunately, PoC has layered a whole bunch of additional
>> behaviors into the way SIP requests are handled. Dale has indicated
>> one of them - that a 200 OK is returned on the terminating side
>> priod to the call actually being completed. PoC also *requires* the
>> insertion of servers into the SIP signaling path to make the system
>> work. Consequently, though the messages that PoC uses *look* like
>> SIP messages with a new media type, there is a hidden 'extension'
>> in place that changes the fundamental semantics of INVITE. The fact
>> that this extension is not visible as a specific header in a well
>> defined way does not change the fact that PoC CHANGES THE BEHAVIOR
>> OF SIP AND IS THUS A SIP EXTENSION.
>> 
>> To convince yourself of this, consider what would happen if a PoC
>> client calls a regular SIP phone. The regular SIP phone would see
>> the PoC floor control stream and reject it as unknown. If PoC were
>> just SIP as we've specified it, then everything would be fine, and
>> the call would proceed but without floor control.
>> 
>> But alas, it is not so.
>> 
>> Because PoC makes all kinds of assumptions about the behaviors of
>> its participants, and what the real meaning is for various SIP
>> messages, the entire thing just won't work when a call terminates
>> on a non-PoC device. This is because PoC is, effectively, a fairly
>> significant and non-backwards-compatible extension to SIP.
>> 
>> So, my proposal is, if you are going to go and DO something like
>> that, you better use the SIP techniques that we've defined to deal
>> with it. And when you extend SIP in a way that doesn't work with a
>> SIP endpoint that doesn't understand an extension, you use a
>> Require header. And that is exactly what I'm proposing that 3gpp
>> should do. Use Require because you have made a non-backwards-
>> compatible extension to SIP.
>> 
>> The general problem here is that there are other places where
>> people are layering additional 'context', as Loreto has called it,
>> ontop of SIP signaling. This is a worrisome trend because this
>> context is subtle and not explicitly signaled. Its just assumed.
>> And that is dangerous, since it leads to the kind of balkanization
>> and non-interoperability that myself, Paul, Dale and others have
>> complained about in this thread.
>> 
>> The current 3gpp specs, instead of using Require to signal this
>> extension, put a tag into the Accept-Contact header field. I find
>> this extremely worrisome, because it is really just masking the
>> reality of what has happened. If you've got a situration where a
>> SIP call *cannot be processed* without going through application
>> servers, then this is not something that is signaled with Accept-
>> Contact. Its signaled with Require or Proxy-Require as appropriate.
>> 
>> Thanks,
>> -Jonathan R.
>> 
>> Henning Schulzrinne wrote:
>> 
> 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP

_______________________________________________________________________
Notice:  This email message, together with any attachments, may contain
information  of  BEA Systems,  Inc.,  its subsidiaries  and  affiliated
entities,  that may be confidential,  proprietary,  copyrighted  and/or
legally privileged, and is intended solely for the use of the individual
or entity named in this message. If you are not the intended recipient,
and have received this message in error, please immediately return this
by email and then delete it.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ence of
>> this new media type in the SDP. Sadly, I wish that were true. If it
>> was, we would not need a mechanism here.
>> 
>> Unfortunately, PoC has layered a whole bunch of additional
>> behaviors into the way SIP requests are handled. Dale has indicated
>> one of them - that a 200 OK is returned on the terminating side
>> priod to the call actually being completed. PoC also *requires* the
>> insertion of servers into the SIP signaling path to make the system
>> work. Consequently, though the messages that PoC uses *look* like
>> SIP messages with a new media type, there is a hidden 'extension'
>> in place that changes the fundamental semantics of INVITE. The fact
>> that this extension is not visible as a specific header in a well
>> defined way does not change the fact that PoC CHANGES THE BEHAVIOR
>> OF SIP AND IS THUS A SIP EXTENSION.
>> 
>> To convince yourself of this, consider what would happen if a PoC
>> client calls a regular SIP phone. The regular SIP phone would see
>> the PoC floor control stream and reject it as unknown. If PoC were
>> just SIP as we've specified it, then everything would be fine, and
>> the call would proceed but without floor control.
>> 
>> But alas, it is not so.
>> 
>> Because PoC makes all kinds of assumptions about the behaviors of
>> its participants, and what the real meaning is for various SIP
>> messages, the entire thing just won't work when a call terminates
>> on a non-PoC device. This is because PoC is, effectively, a fairly
>> significant and non-backwards-compatible extension to SIP.
>> 
>> So, my proposal is, if you are going to go and DO something like
>> that, you better use the SIP techniques that we've defined to deal
>> with it. And when you extend SIP in a way that doesn't work with a
>> SIP endpoint that doesn't understand an extension, you use a
>> Require header. And that is exactly what I'm proposing that 3gpp
>> should do. Use Require because you have made a non-backwards-
>> compatible extension to SIP.
>> 
>> The general problem here is that there are other places where
>> people are layering additional 'context', as Loreto has called it,
>> ontop of SIP signaling. This is a worrisome trend because this
>> context is subtle and not explicitly signaled. Its just assumed.
>> And that is dangerous, since it leads to the kind of balkanization
>> and non-interoperability that myself, Paul, Dale and others have
>> complained about in this thread.
>> 
>> The current 3gpp specs, instead of using Require to signal this
>> extension, put a tag into the Accept-Contact header field. I find
>> this extremely worrisome, because it is really just masking the
>> reality of what has happened. If you've got a situration where a
>> SIP call *cannot be processed* without going through application
>> servers, then this is not something that is signaled with Accept-
>> Contact. Its signaled with Require or Proxy-Require as appropriate.
>> 
>> Thanks,
>> -Jonathan R.
>> 
>> Henning Schulzrinne wrote:
>> 
> 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP

_______________________________________________________________________
Notice:  This email message, together with any attachments, may contain
information  of  BEA Systems,  Inc.,  its subsidiaries  and  affiliated
entities,  that may be confidential,  proprietary,  copyrighted  and/or
legally privileged, and is intended solely for the use of the individual
or entity named in this message. If you are not the intended recipient,
and have received this message in error, please immediately return this
by email and then delete it.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP

From sipping-bounces@ietf.org Wed Feb 21 16:57:50 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJzS1-0001bI-Uj; Wed, 21 Feb 2007 16:56:37 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJzS0-0001XC-8g
	for sipping@ietf.org; Wed, 21 Feb 2007 16:56:36 -0500
Received: from usremg02.bea.com ([66.248.192.22])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HJzRy-0001vv-Mn
	for sipping@ietf.org; Wed, 21 Feb 2007 16:56:36 -0500
Received: from usremr01.bea.com (usremr01.bea.com [10.160.29.91])
	by usremg02.bea.com (Switch-3.2.2/Switch-3.2.2) with ESMTP id
	l1LLuW1i013730
	(version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=FAIL)
	for <sipping@ietf.org>; Wed, 21 Feb 2007 13:56:33 -0800
Received: from repbex01.amer.bea.com (repbex01.bea.com [10.160.26.98])
	by usremr01.bea.com (Switch-3.2.2/Switch-3.2.2) with ESMTP id
	l1LLuVUP013088
	for <sipping@ietf.org>; Wed, 21 Feb 2007 13:56:31 -0800
Received: from 10.32.6.241 ([10.32.6.241]) by repbex01.amer.bea.com
	([10.160.26.98]) with Microsoft Exchange Server HTTP-DAV ; 
	Wed, 21 Feb 2007 21:55:09 +0000
User-Agent: Microsoft-Entourage/11.3.3.061214
Date: Wed, 21 Feb 2007 10:29:09 -0800
From: Eric Burger <eburger@bea.com>
To: <sipping@ietf.org>
Message-ID: <C201CF75.1FFA%eburger@bea.com>
Thread-Topic: Service Identification Model 
Thread-Index: AcdV5i0Ma5Tu/sHZEdueygAWy4mm/w==
Mime-version: 1.0
Content-type: text/plain;
	charset="US-ASCII"
Content-transfer-encoding: 7bit
X-PMX-Version: 4.7.1.128075, Antispam-Engine: 2.5.0.283055,
	Antispam-Data: 2007.2.21.120936
X-Spam-Score: 0.0 (/)
X-Scan-Signature: c83ccb5cc10e751496398f1233ca9c3a
Subject: [Sipping] Service Identification Model 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Folks have been pestering me for a while to look at the Service
Identification issue.  I finally have gotten around to it.


I guess I have a fundamental problem with the principal assertion of the
service-identification draft.  Namely, the assertion that there is a single
SIP Layer, separate and distinct from two services.  Here is Figure 2 from
the draft:

                           +---------------------------------+
                           |                                 |
                           | +-------------+ +-------------+ |
                           | |      UI     | |      UI     | |
                           | +-------------+ +-------------+ |
                           | +-------------+ +-------------+ |
                           | |             | |             | |
                           | |  Service 1  | |  Service 2  | |
                           | |             | |             | |
                           | +-------------+ +-------------+ |
                           | +-----------------------------+ |
                           | |                             | |
                           | |             SIP             | |
                           | |            Layer            | |
                           | |                             | |
                           | +-----------------------------+ |
                           |                                 |
                           +---------------------------------+

                                     Physical Device

I would offer this drawing confuses a common SIP Stack with a common User
Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use case:

> ... In addition users may have
> multiple devicietf.org for new developments of core SIP

From sipping-bounces@ietf.org Wed Feb 21 16:57:50 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJzS1-0001bI-Uj; Wed, 21 Feb 2007 16:56:37 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJzS0-0001XC-8g
	for sipping@ietf.org; Wed, 21 Feb 2007 16:56:36 -0500
Received: from usremg02.bea.com ([66.248.192.22])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HJzRy-0001vv-Mn
	for sipping@ietf.org; Wed, 21 Feb 2007 16:56:36 -0500
Received: from usremr01.bea.com (usremr01.bea.com [10.160.29.91])
	by usremg02.bea.com (Switch-3.2.2/Switch-3.2.2) with ESMTP id
	l1LLuW1i013730
	(version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=FAIL)
	for <sipping@ietf.org>; Wed, 21 Feb 2007 13:56:33 -0800
Received: from repbex01.amer.bea.com (repbex01.bea.com [10.160.26.98])
	by usremr01.bea.com (Switch-3.2.2/Switch-3.2.2) with ESMTP id
	l1LLuVUP013088
	for <sipping@ietf.org>; Wed, 21 Feb 2007 13:56:31 -0800
Received: from 10.32.6.241 ([10.32.6.241]) by repbex01.amer.bea.com
	([10.160.26.98]) with Microsoft Exchange Server HTTP-DAV ; 
	Wed, 21 Feb 2007 21:55:09 +0000
User-Agent: Microsoft-Entourage/11.3.3.061214
Date: Wed, 21 Feb 2007 10:29:09 -0800
From: Eric Burger <eburger@bea.com>
To: <sipping@ietf.org>
Message-ID: <C201CF75.1FFA%eburger@bea.com>
Thread-Topic: Service Identification Model 
Thread-Index: AcdV5i0Ma5Tu/sHZEdueygAWy4mm/w==
Mime-version: 1.0
Content-type: text/plain;
	charset="US-ASCII"
Content-transfer-encoding: 7bit
X-PMX-Version: 4.7.1.128075, Antispam-Engine: 2.5.0.283055,
	Antispam-Data: 2007.2.21.120936
X-Spam-Score: 0.0 (/)
X-Scan-Signature: c83ccb5cc10e751496398f1233ca9c3a
Subject: [Sipping] Service Identification Model 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Folks have been pestering me for a while to look at the Service
Identification issue.  I finally have gotten around to it.


I guess I have a fundamental problem with the principal assertion of the
service-identification draft.  Namely, the assertion that there is a single
SIP Layer, separate and distinct from two services.  Here is Figure 2 from
the draft:

                           +---------------------------------+
                           |                                 |
                           | +-------------+ +-------------+ |
                           | |      UI     | |      UI     | |
                           | +-------------+ +-------------+ |
                           | +-------------+ +-------------+ |
                           | |             | |             | |
                           | |  Service 1  | |  Service 2  | |
                           | |             | |             | |
                           | +-------------+ +-------------+ |
                           | +-----------------------------+ |
                           | |                             | |
                           | |             SIP             | |
                           | |            Layer            | |
                           | |                             | |
                           | +-----------------------------+ |
                           |                                 |
                           +---------------------------------+

                                     Physical Device

I would offer this drawing confuses a common SIP Stack with a common User
Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use case:

> ... In addition users may have
> multiple devices only one of which has the application that understands
> the game. Other devices that the user has registered against his AoR may
> handle audio for telephony and other voice applications but if Alice
> invites Bob to play her favourite game it is not very interesting if Bob
> accepts the gaming session on a device that can only accept the audio
> stream. Also even when the terminal that has the game application
> receives the INVITE the correct application in that terminal needs to be
> identified to accept the session.

I would offer that this is the clinching argument that the request is, in
fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a User
Agent.

Same argument for PoC versus a regular voice call: One is looking for a PoC
User Agent, NOT the PoC service at a given User Agent.

Thus, the picture really looks like the following:

                           +---------------------------------+
                           |                                 |
                           | +-------------+ +-------------+ |
                           | |    UI       | |    UI       | |
                           | +-------------+ +-------------+ |
                           | +-------------+ +-------------+ |
                           | |             | |             | |
                           | |    UAS 1    | |    UAS 2    | |
                           | |             | |             | |
                           | +-------------+ +-------------+ |
                           | +-----------------------------+ |
                           | |                             | |
                           | |             SIP             | |
                           | |            Stack            | |
                           | |                             | |
                           | +-----------------------------+ |
                           |                                 |
                           +---------------------------------+

                                     Physical Device


Another argument for different User Agents is the whole "route based on SDP"
argument.  As the author of RFC 3458, I think I was first to make the
assertion that knowing the media type did NOT mean you know the application
to use.  I don't want your phone or proxy or B2BUA to guess that because I
have Chess Media, I want the Chess Application; I want to talk with your
Chess Application.

Now, is the Chess Application a Preference, Requirement, or the thing that I
am addressing?  The answer drives one towards using caller prefs, the
Require header, or the Request-URI.

I would offer that it is not a preference: I really want the chess
application; I probably do not want to talk to you.

That leaves us with the Require header or a Request-URI.  My leanings are
not surprisingly towards the RFC 4240 approach.  Frankly, if you are
building an IMS (as an example) that can handle 10,000,000 subscribers, you
should be able to handle 100,000,000 registrations.  That order of magnitude
is way easier than the 10,000 to 10,000,000 jump.  I don't buy the
"registering each service is too much of a burden on the operator" argument,
because it is not that much of a burden, and the network actually does the
right thing.  Better yet, all of the corner cases work.  Best of all, the
IETF doesn't have to build an IANA Registry and process a thousand PS
documents to register urn:service:myco:widget.

I suppose I could live with Require:, only if you are way too lazy to
register each application instance.  It does semantically work, because if I
want to do PoC and you do not do PoC, I'll get a 4xx response, to which I
could try a regular call.

_______________________________________________________________________
Notice:  This email message, together with any attachments, may contain
information  of  BEA Systems,  Inc.,  its subsidiaries  and  affiliated
entities,  that may be confidential,  proprietary,  copyrighted  and/or
legally privileged, and is intended solely for the use of the individual
or entity named in this mees only one of which has the application that understands
> the game. Other devices that the user has registered against his AoR may
> handle audio for telephony and other voice applications but if Alice
> invites Bob to play her favourite game it is not very interesting if Bob
> accepts the gaming session on a device that can only accept the audio
> stream. Also even when the terminal that has the game application
> receives the INVITE the correct application in that terminal needs to be
> identified to accept the session.

I would offer that this is the clinching argument that the request is, in
fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a User
Agent.

Same argument for PoC versus a regular voice call: One is looking for a PoC
User Agent, NOT the PoC service at a given User Agent.

Thus, the picture really looks like the following:

                           +---------------------------------+
                           |                                 |
                           | +-------------+ +-------------+ |
                           | |    UI       | |    UI       | |
                           | +-------------+ +-------------+ |
                           | +-------------+ +-------------+ |
                           | |             | |             | |
                           | |    UAS 1    | |    UAS 2    | |
                           | |             | |             | |
                           | +-------------+ +-------------+ |
                           | +-----------------------------+ |
                           | |                             | |
                           | |             SIP             | |
                           | |            Stack            | |
                           | |                             | |
                           | +-----------------------------+ |
                           |                                 |
                           +---------------------------------+

                                     Physical Device


Another argument for different User Agents is the whole "route based on SDP"
argument.  As the author of RFC 3458, I think I was first to make the
assertion that knowing the media type did NOT mean you know the application
to use.  I don't want your phone or proxy or B2BUA to guess that because I
have Chess Media, I want the Chess Application; I want to talk with your
Chess Application.

Now, is the Chess Application a Preference, Requirement, or the thing that I
am addressing?  The answer drives one towards using caller prefs, the
Require header, or the Request-URI.

I would offer that it is not a preference: I really want the chess
application; I probably do not want to talk to you.

That leaves us with the Require header or a Request-URI.  My leanings are
not surprisingly towards the RFC 4240 approach.  Frankly, if you are
building an IMS (as an example) that can handle 10,000,000 subscribers, you
should be able to handle 100,000,000 registrations.  That order of magnitude
is way easier than the 10,000 to 10,000,000 jump.  I don't buy the
"registering each service is too much of a burden on the operator" argument,
because it is not that much of a burden, and the network actually does the
right thing.  Better yet, all of the corner cases work.  Best of all, the
IETF doesn't have to build an IANA Registry and process a thousand PS
documents to register urn:service:myco:widget.

I suppose I could live with Require:, only if you are way too lazy to
register each application instance.  It does semantically work, because if I
want to do PoC and you do not do PoC, I'll get a 4xx response, to which I
could try a regular call.

_______________________________________________________________________
Notice:  This email message, together with any attachments, may contain
information  of  BEA Systems,  Inc.,  its subsidiaries  and  affiliated
entities,  that may be confidential,  proprietary,  copyrighted  and/or
legally privileged, and is intended solely for the use of the individual
or entity named in this message. If you are not the intended recipient,
and have received this message in error, please immediately return this
by email and then delete it.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP





ssage. If you are not the intended recipient,
and have received this message in error, please immediately return this
by email and then delete it.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP





From sipping-bounces@ietf.org Wed Feb 21 17:12:57 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HJzhR-0006vm-GJ; Wed, 21 Feb 2007 17:12:33 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HJzhQ-0006vT-Eo
	for sipping@ietf.org; Wed, 21 Feb 2007 17:12:32 -0500
Received: from dsl001-129-069.dfw1.dsl.speakeasy.net ([72.1.129.69]
	helo=vicuna.estacado.net) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HJzhP-0003t7-PR
	for sipping@ietf.org; Wed, 21 Feb 2007 17:12:32 -0500
Received: from [172.17.2.60] ([172.17.2.60]) (authenticated bits=0)
	by vicuna.estacado.net (8.13.8/8.13.8) with ESMTP id l1LMCKBn090731
	(version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO);
	Wed, 21 Feb 2007 16:12:20 -0600 (CST)
	(envelope-from bcampen@estacado.net)
In-Reply-To: <C201CF75.1FFA%eburger@bea.com>
References: <C201CF75.1FFA%eburger@bea.com>
Mime-Version: 1.0 (Apple Message framework v752.3)
Message-Id: <E0E7620E-5E8E-4382-83AB-A4C6BECCCDEC@estacado.net>
From: Byron Campen <bcampen@estacado.net>
Subject: Re: [Sipping] Service Identification Model 
Date: Wed, 21 Feb 2007 16:12:11 -0600
To: Eric Burger <eburger@bea.com>
X-Mailer: Apple Mail (2.752.3)
X-Spam-Score: 0.1 (/)
X-Scan-Signature: bacfc6c7290e34d410f9bc22b825ce96
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1011460632=="
Errors-To: sipping-bounces@ietf.org


--===============1011460632==
Content-Type: multipart/signed; micalg=sha1; boundary=Apple-Mail-2-515209983;
	protocol="application/pkcs7-signature"


--Apple-Mail-2-515209983
Content-Transfer-Encoding: 7bit
Content-Type: text/plain;
	charset=US-ASCII;
	delsp=yes;
	format=flowed

	Yes! _This_ makes sense to me. This seems like the proper way to  
think about and solve this problem.

Best regards,
Byron Campen

> Folks have been pestering me for a while to look at the Service
> Identification issue.  I finally have gotten around to it.
>
>
> I guess I have a fundamental problem with the principal assertion  
> of the
> service-identification draft.  Namely, the assertion that there is  
> a single
> SIP Layer, separate and distinct from two services.  Here is Figure  
> 2 from
> the draft:
>
>                            +---------------------------------+
>                            |                                 |
>                            | +-------------+ +-------------+ |
>                            | |      UI     | |      UI     | |
>                            | +-------------+ +-------------+ |
>                            | +-------------+ +-------------+ |
>                            | |             | |             | |
>                            | |  Service 1  | |  Service 2  | |
>                            | |             | |             | |
>                            | +-------------+ +-------------+ |
>                            | +-----------------------------+ |
>                            | |                             | |
>                            | |             SIP             | |
>                            | |            Layer            | |
>                            | |                             | |
>                            | +-----------------------------+ |
>                            |                                 |
>                            +---------------------------------+
>
>                                      Physical Device
>
> I would offer this drawing confuses a common SIP Stack with a  
> common User
> Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use case:
>
>> ... In addition users may have
>> multiple devices only one of which has the application that  
>> understands
>> the game. Other devices that the user has registered against his  
>> AoR may
>> handle audio for telephony and other voice applications but if Alice
>> invites Bob to play her favourite game it is not very interesting  
>> if Bob
>> accepts the gaming session on a device that can only accept the audio
>> stream. Also even when the terminal that has the game application
>> receives the INVITE the correct application in that terminal needs  
>> to be
>> identified to accept the session.
>
> I would offer that this is the clinching argument that the request  
> is, in
> fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a  
> User
> Agent.
>
> Same argument for PoC versus a regular voice call: One is looking  
> for a PoC
> User Agent, NOT the PoC service at a given User Agent.
>
> Thus, the picture really looks like the following:
>
>                            +---------------------------------+
>                            |                                 |
>                            | +-------------+ +-------------+ |
>                            | |    UI       | |    UI       | |
>                            | +-------------+ +-------------+ |
>                            | +-------------+ +-------------+ |
>                            | |             | |             | |
>                            | |    UAS 1    | |    UAS 2    | |
>                            | |             | |             | |
>                            | +-------------+ +-------------+ |
>                            | +-----------------------------+ |
>                            | |                             | |
>                            | |             SIP             | |
>                            | |            Stack            | |
>                            | |                             | |
>                            | +-----------------------------+ |
>                            |                                 |
>                            +---------------------------------+
>
>                                      Physical Device
>
>
> Another argument for different User Agents is the whole "route  
> based on SDP"
> argument.  As the author of RFC 3458, I think I was first to make the
> assertion that knowing the media type did NOT mean you know the  
> application
> to use.  I don't want your phone or proxy or B2BUA to guess that  
> because I
> have Chess Media, I want the Chess Application; I want to talk with  
> your
> Chess Application.
>
> Now, is the Chess Application a Preference, Requirement, or the  
> thing that I
> am addressing?  The answer drives one towards using caller prefs, the
> Require header, or the Request-URI.
>
> I would offer that it is not a preference: I really want the chess
> application; I probably do not want to talk to you.
>
> That leaves us with the Require header or a Request-URI.  My  
> leanings are
> not surprisingly towards the RFC 4240 approach.  Frankly, if you are
> building an IMS (as an example) that can handle 10,000,000  
> subscribers, you
> should be able to handle 100,000,000 registrations.  That order of  
> magnitude
> is way easier than the 10,000 to 10,000,000 jump.  I don't buy the
> "registering each service is too much of a burden on the operator"  
> argument,
> because it is not that much of a burden, and the network actually  
> does the
> right thing.  Better yet, all of the corner cases work.  Best of  
> all, the
> IETF doesn't have to build an IANA Registry and process a thousand PS
> documents to register urn:service:myco:widget.
>
> I suppose I could live with Require:, only if you are way too lazy to
> register each application instance.  It does semantically work,  
> because if I
> want to do PoC and you do not do PoC, I'll get a 4xx response, to  
> which I
> could try a regular call.
>
> ______________________________________________________________________ 
> _
> Notice:  This email message, together with any attachments, may  
> contain
> information  of  BEA Systems,  Inc.,  its subsidiaries  and   
> affiliated
> entities,  that may be confidential,  proprietary,  copyrighted   
> and/or
> legally privileged, and is intended solely for the use of the  
> individual
> or entity named in this message. If you are not the intended  
> recipient,
> and have received this message in error, please immediately return  
> this
> by email and then delete it.
>
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP


--Apple-Mail-2-515209983
Content-Transfer-Encoding: base64
Content-Type: application/pkcs7-signature;
	name=smime.p7s
Content-Disposition: attachment;
	filename=smime.p7s
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--Apple-Mail-2-515209983--


--===============1011460632==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1011460632==--




From sipping-bounces@ietf.org Wed Feb 21 17:35:05 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HK02w-0003zJ-BE; Wed, 21 Feb 2007 17:34:46 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HK02t-0003t9-NO
	for sipping@ietf.org; Wed, 21 Feb 2007 17:34:43 -0500
Received: from zcars04f.nortel.com ([47.129.242.57])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HK02s-0007IU-D8
	for sipping@ietf.org; Wed, 21 Feb 2007 17:34:43 -0500
Received: from zrc2hxm2.corp.nortel.com (zrc2hxm2.corp.nortel.com
	[47.103.123.73])
	by zcars04f.nortel.com (Switch-2.2.6/Switch-2.2.0) with ESMTP id
	l1LMYcZ17454; Wed, 21 Feb 2007 17:34:38 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Wed, 21 Feb 2007 16:34:07 -0600
Message-ID: <6FC4416DDE56C44DA0AEE67BC7CA437112DE3701@zrc2hxm2.corp.nortel.com>
In-Reply-To: <45D65EE8.7010501@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-index: AcdStZFsSh8MPLqkR629LMNz6k4jqwDUYqMg
From: "Brian Stucker" <bstucker@nortel.com>
To: "Jonathan Rosenberg" <jdrosen@cisco.com>,
	"Henning Schulzrinne" <hgs@cs.columbia.edu>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 0ddefe323dd869ab027dbfff7eff0465
Cc: sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Jonathan,

You've got it all wrong.=20

PoC is not a non-backwards-compatable extension to SIP, it's a backwards
non-compatable extension to SIP.

The introduction of functions within the IMS architecture that
differentiate and interwork between 3GPP IMS SIP and IETF SIP (being
that, last I checked SIP was still an IETF RFC) at peering points shows
(to me) that solving IMS problems using the mechanisms that already
exist in IETF drafts (although the solution may not be apparent at the
time) is not always a top priority for a variety of reasons.=20

A commonly heard complaint/stated reason for this is the lack of the
IETF working groups to meet the milestones for solutions set by 3GPP.=20

Since we not only must agree upon a technical solution, but find it
desirable to convince 3GPP and others to use the solution recommended by
the IETF; getting agreement on producing a stable, referencable document
within the IETF within the 3GPP R7 timeline is critical. Otherwise, R7
will go out the door with what it has today for service identification,
and we'll wind up with an even bigger mess on our hands.

Regards,
Brian
=20

> -----Original Message-----
> From: Jonathan Rosenberg [mailto:jdrosen@cisco.com]=20
> Sent: Friday, February 16, 2007 7:48 PM
> To: Henning Schulzrinne
> Cc: sipping
> Subject: Re: [Sipping] Service Identification in SIP
>=20
> I think its a fair point that the example is contrived.=20
> Really, the main use case being targeted is IMS PoC. So, let=20
> me talk a bit about that.
>=20
[SNIP]
>=20
> But alas, it is not so.
>=20
> Because PoC makes all kinds of assumptions about the=20
> behaviors of its participants, and what the real meaning is=20
> for various SIP messages, the entire thing just won't work=20
> when a call terminates on a non-PoC device.=20
> This is because PoC is, effectively, a fairly significant and=20
> non-backwards-compatible extension to SIP.
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 21 22:06:25 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HK4GY-0003sP-OZ; Wed, 21 Feb 2007 22:05:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HK4GX-0003sK-Ty
	for sipping@ietf.org; Wed, 21 Feb 2007 22:05:05 -0500
Received: from c243-111.rim.net ([216.9.243.111] helo=MHS02YKF.rim.net)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HK4GV-0000Mu-6e
	for sipping@ietf.org; Wed, 21 Feb 2007 22:05:05 -0500
Importance: normal
Priority: normal
Received: from ngw01ykf.rim.net ([10.102.108.30]) by MHS02YKF.rim.net with
	Microsoft SMTPSVC(6.0.3790.211); Wed, 21 Feb 2007 22:04:56 -0500
Received: from XCH21YKF.rim.net ([10.102.100.36]) by ngw01ykf.rim.net (SAVSMTP
	3.1.6.45) with SMTP id M2007022122045328904 ;
	Wed, 21 Feb 2007 22:04:53 -0500
Received: from XCH47YKF.rim.net ([10.64.31.217]) by XCH21YKF.rim.net with
	Microsoft SMTPSVC(5.0.2195.6713); Wed, 21 Feb 2007 22:04:53 -0500
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.181
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Wed, 21 Feb 2007 22:04:51 -0500
Message-ID: <61968779B8AC4C4BAB421D4C12F008C009152056@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdNmH89O4T46iGoRjicZ3hYknVpawB8E6oQAH++leAABzLk8AAYIixwAIPObeAAhKg1EA==
From: "Andrew Allen" <aallen@rim.com>
To: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>,
	<sipping@ietf.org>
X-OriginalArrivalTime: 22 Feb 2007 03:04:53.0749 (UTC)
	FILETIME=[398EC250:01C7562E]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: ea36de7a5e28e9b4461c8d685f4e97f1
Cc: Andrew Allen <aallen@rim.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi Steve

In my view this is one of the interoperability issues with your
proposal. I think the assumption that all networks will deploy exactly
the same services with the same service identifiers is unrealistic and
also undesirable. Are corporate PBX vendors going to implement the same
services as defined by 3GPP or OMA? The possibility to have the
flexibility to have different service identifiers for session equivalent
services is in my view essential to any IETF standard and to do
otherwise is to create a situation where basic services like telephony
or messaging cannot interoperate between users in different networks and
users in the internet without the use of inter working gateway. SIP and
SIMPLE were setup specifically to create protocols that would enable
interoperability with IP networks without the need for gateways.=20

If 3GPP network operators choose to deploy the same services and use
common identifiers that is fine but there are whole other worlds out
there as well and service identifiers should not stand in the way of
interoperability between all these different SIP users. If the IETF
solution supports the ability for different networks to use different
identifiers that doesn't violate 3GPP requirements and doesn't prevent
3GPP using the same identifiers in different networks. I don't see how
that is against TS 23.228.

Andrew


-----Original Message-----
From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]=20
Sent: Monday, February 19, 2007 6:00 AM
To: Andrew Allen; sipping@ietf.org
Subject: RE: [Sipping] Service Identification in SIP

Hi Andrew,

I see that in your assumptions, the originating and terminating networks
use different service identifiers, and I presume that this is for the
same service.  In my understanding, it is the same service identifier
that is used end-to-end (if not it is interworked).  That is why I do
not see the reliance on Presence being necessary.  In my opionion this
totally within the requirements of TS 23.228 (I think that different
service identifiers in different networks for the same service is
against TS 23.228).

This comes to the isssue of the PSTN gateway - I presume that you want
to go to the PSTN gateway if the service is telephony, and within that,
voice.  If we know that, then the we go to any PSTN gateway, and we
can't presume a particular network for a PSTN gateway.=20

Best Regards,

Steve.

-----Original Message-----
From: Andrew Allen [mailto:aallen@rim.com]=20
Sent: Saturday, February 17, 2007 12:51 AM
To: Stephen Terrill (MI/EEM); sipping@ietf.org
Cc: Andrew Allen
Subject: RE: [Sipping] Service Identification in SIP



Hi Steve,

My responses inline

Andrew=20

-----Original Message-----
From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
Sent: Friday, February 16, 2007 3:51 AM
To: sipping@ietf.org
Subject: RE: [Sipping] Service Identification in SIP

Hello Andrew,

Thanks for your comments.  I have a couple of brief responses below.

Best Regards,

..//Steve=20

-----Original Message-----
From: Andrew Allen [mailto:aallen@rim.com]
Sent: Friday, February 16, 2007 7:15 AM
To: Stephen Terrill (MI/EEM); sipping@ietf.org
Cc: Andrew Allen
Subject: RE: [Sipping] Service Identification in SIP


Some comments on Stephen's comments on the draft:
=20

-Moving onto the proposed solution draft itself, in addition for
providing for the identification of the "application" and the
"communication service", I have a few additional thoughts.

The draft seems to have the implication that all services (communication
services in other terminology) that are to be identified have to be
standardised within the IETF.  This is clearly an restriction that will
not be acceptable to the industry as a whole.  My guess is that if such
a statement is mandated by the IETF, it will just be words on paper
which would likely be ignored in practice by other standardisation
organisations.

[AA] This I think is an incorrect interpretation of the draft. My
understanding is that the only services according to the draft that
require some standardisation in IETF are those that by their very nature
cannot interoperate unless there is some common additional understanding
between the endpoints outside of what is currently specified in SIP. The
use of SIP option tags with Require header is proposed by the draft for
this purpose as such services create interoperability issues with SIP.

<Steve>
My comment applies even for these services </Steve>


<AA>
Well then we are in big trouble because what you are proposing is that
its OK for bodies outside of IETF to create non interoperable extensions
for the behavour of SIP entities. This will result in the fragmentation
of SIP into different flavors that don't interoperate. This is the
balkanization of SIP that Paul talks about.even <AA>=20

----------------------------------------------------

The draft also only recognises the case where all explicitly knowledge
of the service is required.  I think it would also be good to allow the
case where knowledge of the service would be helpful, but not explicitly
required.  This could be a case where the service to the end-user would
not be degraded if this was not supported.  This functionality seems to
be missing.

[AA] Service URIs are defined in the draft for these services where the
support by the other endpoint is not required for the session to be
sucessfully established.

The draft seems to indicate that that application servers in the network
are involved only at the request of the originating terminal.  This
seems a strange restriction.  For some network support that supports the
originating user, such as the one provided in the example "I would like
this call recorded for later retrieval", such mechanisms may make sense,
however as a general approach it seems a bit limited.  For example, in
3GPP networks there is the need to invoke application logic in the
originating side and terminating side, not only at the request of the
originating terminal, but as a consequence of the requested service.  If
there is an SIP message contains indication of the requested service, a
proxy can do this (or loose route to an AS that contains the necessary
service logic).=20

[AA] The draft supports both possibilities using the Route header for
originating network servers that need to be involved in the route and
based on the terminating networks routing policy for the Service URI in
the Request URI.

<Steve>
I don't see the advantage of having different solutions for the
originating network and the terminating network.  I don't think that the
originating logic is invoked only at the request of the originating
terminal.
</Steve>

<AA>
Unless IETF is willing to standardise all services and mandate that all
SIP entities understand these (which will never happen) then we cannot
assume that all networks and all endpoints understand and use the same
identifiers. It is possible through configuration for an endpoint to
obtain the identifiers used by its own service provider but there is no
way to ensure that other networks use or understand the same identifiers
for the equivalent or similar service that is compatible at the session
layer. Therefore we need to support the possibility to have different
identifiers in different networks in order to not violate the TS 23.228
requirement:

12.	The usage of communication service identifiers shall not
adversely affect interoperability between IMS networks and
interoperability with external SIP networks and CS networks. The
behaviour of a network receiving the IMS requests without an IMS
communication service identifier is a matter of operator policy. Usage
of communication service identifiers shall not decrease the level of
interoperability with networks and UEs that are unaware of the
communication service identifier.

If the originating service logic isn't invoked based on an identifier
placed in the request by the originating terminal on what basis do you
think it is invoked? How does the use of Accept-Contact have advantages
in this case?
<AA>=20

--------------------------------------

 I note here that the proposed service URN does not contain the address
of a specific server containing the application logic.  An AS can look
at that, and loose route to an application server that does contain such
logic (If it is separated from the proxy), however I fail to see the
difference for that proxy if it looks at a feature tag, or it looks at
the service URN for this.  In both cases, it has to parse the message
for the data, and modify the routing information. =20

[AA]Route headers are the SIP mechanism for requesting intermediaries to
be includede on the route of the request. There is nothing in RFC
3840/RFC 3841 that addresses using feature tags in Accept-Contact header
for any routing decisions prior to determining which of the registered
contacts the request should be forked to. Accept-Contact is not a Route
header it is merely a filter that is applied to the set of valid
contacts. Using a parameter in the filter as a service identifier has a
potential significant impact on the use of Accept-Contact for its
original purpose.=20

<Steve>
I don't believe there is anthing that states a Proxy CAN NOT route to
applications or application logic based upon elements in the SIP
message, including what is in the Accept-Contact header (some people
have even proposed SDP ...).
</Steve>

<AA>
As I had stated in an earlier email on this subject a proxy can do a lot
of things (including routing based on your favorite number appearing in
a call-Id) but that doesn't mean they are a good idea or a good design
decision. If you want to ensure you don't break things then you need to
be careful to use SIP headers with the semantics intended for them
otherwise problems can result if extra semantics are put upon existing
headers.
<AA>=20

Further, the route header could be removed in the originating network;
and it is still required to identify the service in the terminating
network.  As proxies may remove the route, it can't be used to transport
the requested service. =20

[AA] Proxies arbitarily removing Route headers is a BAD thing akin to a
bull in a china shop - almost certainly going to break something. RFC
3261 summarises what proxies should be doing with Route headers:
16.12 Summary of Proxy Route Processing

   In the absence of local policy to the contrary, the processing a
   proxy performs on a request containing a Route header field can be
   summarized in the following steps.

      1.  The proxy will inspect the Request-URI.  If it indicates a
          resource owned by this proxy, the proxy will replace it with
          the results of running a location service.  Otherwise, the
          proxy will not change the Request-URI.

      2.  The proxy will inspect the URI in the topmost Route header
          field value.  If it indicates this proxy, the proxy removes it
          from the Route header field (this route node has been
          reached).

      3.  The proxy will forward the request to the resource indicated
          by the URI in the topmost Route header field value or in the
          Request-URI if no Route header field is present.  The proxy
          determines the address, port and transport to use when
          forwarding the request by applying the procedures in [4] to
          that URI.

   If no strict-routing elements are encountered on the path of the
   request, the Request-URI will always indicate the target of the
   request.

<Steve>
The first sentance is interesting "In absence of local policy to the
contrary", indicating that there can be different policies.  I
understand that some proxies do remove this.  While I am on this - I
have lead to believe that R-URI parameters may also be removed by
intermediate proxies - is this correct?
</Steve>

<AA>
My understanding is that local policy may determine that additional
nodes need to be routed to in addition to the Route contained in the
request. However deleting existing Route headers other than the Route
header containing the address of the proxy as part of the loose routing
procedures is liable to break things and is not part of RFC 3261 routing
(loose routing).
<AA>=20


[AA] Generally a UA wouldn't know the appropriate service identifiers
used in the terminating network or whether a Route header is appropriate
to be used in the terminating network. The draft defines the Service URI
for the purpose of identifying services in the terminating network and
service URIs can be obtained using Presence.

<Steve>
I don't get the above paragraph.  It suggest that the "appropriate
service identifier" used in the terminating network is not known by the
originating UE, but the Service URI is.  I thought that the service URI
is an encoding of the "appropriate service identifier".
</Steve>

<AA>=20
Yes but in the general case the UA doesn't need to know what the service
identifier is it obtains through Presence a Service URI that contains
the Service identifier as a URI parameter. The great thing about using
Presence to obtain a Service URI is that the originating UA doesn't even
need to know anything about service identifiers at all for this to work.
It just uses Presence to find the URI to use for the service which can
be indicated to the user via an indication such as a graphic icon (e,g a
phone for telephony, smiley face for IM, etc) in the presence document.
<AA>=20

-----------------------------------

>From this perspective it is not clear to me the advantage of using a
service URN in a route header compared to the use of a feature tag.

[AA] I hope the above responses clarify that somewhat. Also the Service
URN and Service URIs allow different service identifiers to be used in
the originating and terminating network for session compatiable services
instead of forcing everyone to deploy and understand the same service
and service identifier.

<Steve>
Nope, don't get it, but are information that is carried end-to-end ....
And as far as I see, could be used by intermediate proxies to impact
local routing policy ... Or ?
</Steve>

<AA>=20
I don't understand this response either. The key point is that UAs in
corporate networks and in the internet are not going to deploy 3GPP
flavor services and use the same 3GPP service identifiers or even
possibly any service identifiers at all. The Accept-Contact based
proposal assumes that all endpoints understand the service identifiers
to be used for each service. This concept is not generally applicable
and create interoperability problems.
<AA>=20

The draft seems to mandate GRUU - is this really so, is that really
necessary.  It seems strange to place such a dependency, it also starts
to move the GRUU as being the primary means of which to contact a user.

[AA] My understanding is that GRUU in the draft is used to address how a
request addressed to Service URI is forked only to those contacts that
support the service identified by the that Service URI. A contact would
seem to be fine so long as it had GRUU properties - no need to obtain a
GRUU if you already have one.

This implies that the originating user decides which terminal I should
receive my requested service on, even if I have two devices for that.
As such, I don't believe that there should be a dependency on GRUU.

[AA] If you have obtained a Service URI (probably through presence) for
that user then you can address the request to the Service URI which
represents all the terminals that support that service. However GRUU and
the Service Instance URIs created from GRUUs provide another mechanism
through which a user can obtain a URI to reach that user for the
required service.

Parts of the draft seems to rely on the use of presence.  While presence
could be a supportive mechanism, I doubt that we can mandate that the
number you want to dial has be available in presence as I assume not all
entities are available in presence (diallable network services, users in
the PSTN).  This would be strange to mandated.

[AA] PSTN gateways are presumably in the same domain as the terminating
proxy and therefore these PSTN gateways can be configured with the
appropriate service identifiers for telephony services and can include
this in either a Route header to route to the right application server
or can construct a valid Service URI. The real problem is how do users
in other networks know what the service identifier that's is used in the
terminating network is. The use of Service URIs and Presence provides a
solution to that problem.

<Steve>
Again, presence could be a useful mechanism, but I think it would be a
problem for the draft to rely it.  A user making a call to the PSTN may
not find that user in the presence archicture (and similar to services
like the 800 or 900 services).  (I should note, that you shouldn't make
any assumptions about the location of the PSTN gateway ... But I don't
think that changes anything). =20
</Steve>

<AA>=20
How does a PSTN gateway in another domain from the terminating network
know the service identifier for telephony used in the terminating
network with the Accept-Contact header based proposal?=20
<AA>=20

Moving on from the draft I have seen the questioned raised about "why
can't we expand the power of SDP".  Such an approach would mandate that
SIP methods use SDP, I don't think that is appropriate.=20

[AA] I agree with you here

I hear the appeals that the mechanisms should not be abused, and that we
should not have 1000s of separate "communication services", and I fully
agree with this. I see a smaller number of communication services, not a
larger number; where those communication services are used to access
other end-user-services.

*flame away*

Regards,

Steve

---



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential
information, privileged material (including material protected by the
solicitor-client or other applicable privileges), or constitute
non-public information. Any use of this information by anyone other than
the intended recipient is prohibited. If you have received this
transmission in error, please immediately reply to the sender and delete
this information from your system. Use, dissemination, distribution, or
reproduction of this transmission by unintended recipients is not
authorized and may be unlawful.



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential =
information, privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 21 22:12:45 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HK4Ns-00086l-O2; Wed, 21 Feb 2007 22:12:40 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HK4Nr-00086Y-3D
	for sipping@ietf.org; Wed, 21 Feb 2007 22:12:39 -0500
Received: from c243-111.rim.net ([216.9.243.111] helo=MHS01YKF.rim.net)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HK4Np-00015G-Pw
	for sipping@ietf.org; Wed, 21 Feb 2007 22:12:39 -0500
Importance: normal
Priority: normal
Received: from ngw01ykf.rim.net ([10.102.108.30]) by MHS01YKF.rim.net with
	Microsoft SMTPSVC(6.0.3790.211); Wed, 21 Feb 2007 22:12:37 -0500
Received: from XCH21YKF.rim.net ([10.102.100.36]) by ngw01ykf.rim.net (SAVSMTP
	3.1.6.45) with SMTP id M2007022122123729119 ;
	Wed, 21 Feb 2007 22:12:37 -0500
Received: from XCH47YKF.rim.net ([10.64.31.217]) by XCH21YKF.rim.net with
	Microsoft SMTPSVC(5.0.2195.6713); Wed, 21 Feb 2007 22:12:37 -0500
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.181
Content-Class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Wed, 21 Feb 2007 22:12:36 -0500
Message-ID: <61968779B8AC4C4BAB421D4C12F008C009152059@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
thread-index: AcdUKvwnE5/ZeTa0QY+RVL9QQXehSgAySiPgAE60jzA=
From: "Andrew Allen" <aallen@rim.com>
To: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>,
	<sipping@ietf.org>, <spencer@mcsr-labs.org>
X-OriginalArrivalTime: 22 Feb 2007 03:12:37.0509 (UTC)
	FILETIME=[4DFAEB50:01C7562F]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 73734d43604d52d23b3eba644a169745
Cc: Andrew Allen <aallen@rim.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Steve, Spencer


There is nothing that prevents operators from agreeing to use the same
identifiers for the same or similar services in their networks. However
the technical solution should not impose that as a constraint.

Andrew=20

-----Original Message-----
From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]=20
Sent: Tuesday, February 20, 2007 8:49 AM
To: sipping@ietf.org
Subject: RE: [Sipping] Service Identification in SIP

Hi Spencer,

I was a little suprised to understand that this was a difference of view
on this point.  Maybe it is confusing a little the User interface from
the "communication service" being identified.  In my understanding,
however, it is the same service identifer that should be used in the
originating network and terminating network - and indeed there was a
liaison from an operator community that see that want to use this for
identifiying the requested service between operators as well.  Of
course, nothing can prevent an operator from
inteworking/upgrading/downgrading in their own network ...

Best Regards,

Steve.

-----Original Message-----
From: Spencer Dawkins [mailto:spencer@mcsr-labs.org]=20
Sent: Monday, February 19, 2007 2:33 PM
To: sipping@ietf.org
Subject: Re: [Sipping] Service Identification in SIP

Hi, Steve,

The distinction you're making here seems to capture the the biggest
question - are we expecting service identifiers to be end-to-end, if not
interworked, or are we expecting them to be different for the same
service in originating and terminating networks.

Thanks,

Spencer

From: "Stephen Terrill (MI/EEM)" <stephen.terrill@ericsson.com>

Hi Andrew,

I see that in your assumptions, the originating and terminating networks
use different service identifiers, and I presume that this is for the
same service.  In my understanding, it is the same service identifier
that is used end-to-end (if not it is interworked).  That is why I do
not see the reliance on Presence being necessary.  In my opionion this
totally within the requirements of TS 23.228 (I think that different
service identifiers in different networks for the same service is
against TS 23.228).=20



_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential =
information, privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 21 23:52:49 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HK5vI-0007MV-SS; Wed, 21 Feb 2007 23:51:16 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HK5vI-0007MK-7u
	for sipping@ietf.org; Wed, 21 Feb 2007 23:51:16 -0500
Received: from c243-111.rim.net ([216.9.243.111] helo=MHS02YKF.rim.net)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HK5vH-0005yK-S9
	for sipping@ietf.org; Wed, 21 Feb 2007 23:51:16 -0500
Importance: normal
Priority: normal
Received: from ngw01ykf.rim.net ([10.102.108.30]) by MHS02YKF.rim.net with
	Microsoft SMTPSVC(6.0.3790.211); Wed, 21 Feb 2007 23:51:05 -0500
Received: from XCH20YKF.rim.net ([10.102.100.35]) by ngw01ykf.rim.net (SAVSMTP
	3.1.6.45) with SMTP id M2007022123510531476 ;
	Wed, 21 Feb 2007 23:51:05 -0500
Received: from XCH47YKF.rim.net ([10.64.31.217]) by XCH20YKF.rim.net with
	Microsoft SMTPSVC(5.0.2195.6713); Wed, 21 Feb 2007 23:51:05 -0500
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.181
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Wed, 21 Feb 2007 23:47:55 -0500
Message-ID: <61968779B8AC4C4BAB421D4C12F008C00915205C@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdVIXQIVywboKJuQdq0BksnGf7wPwBFSy+A
From: "Andrew Allen" <aallen@rim.com>
To: <Dale.Worley@comcast.net>,
	<sipping@ietf.org>
X-OriginalArrivalTime: 22 Feb 2007 04:51:05.0099 (UTC)
	FILETIME=[0F2D9DB0:01C7563D]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: dbb8771284c7a36189745aa720dc20ab
Cc: Andrew Allen <aallen@rim.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Dale,

PoC is a not simply a peer to peer protocol. It is very network server
centric with a network PTT Server that controls the floor for the
clients using a proprietary media control protocol that uses on RTCP.
The PTT server acts as master in a master-slave relationship with the
client. PoC sessions can be established with two or more PTT Servers in
the path of the signaling.=20

PTT also has a particular user experience because it mimics the half
duplex walkie talkie radio experience. The user pushes a button and
expects to be connected to the called party and then speak almost
immediately. Negotiating "down" to a regular full duplex VoIP call
without floor control would be completely unexpected. There are even
optimizations that grant the caller the right to speak before the called
party has answered in order to provide the user that almost
instantaneous Push Talk experience.

Some more information on some of this can be obtained by reading:

http://www.ietf.org/internet-drafts/draft-allen-sipping-poc-p-answer-sta
te-header-04.txt


Andrew

-----Original Message-----
From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net]=20
Sent: Tuesday, February 20, 2007 2:00 PM
To: sipping@ietf.org
Subject: Re: [Sipping] Service Identification in SIP

   From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>

   The draft also only recognises the case where all explicitly
knowledge
   of the service is required.  I think it would also be good to allow
the
   case where knowledge of the service would be helpful, but not
explicitly
   required.  This could be a case where the service to the end-user
would
   not be degraded if this was not supported.  This functionality seems
to
   be missing.

And it seems to me that the ability to indicate support of a feature,
but not mandatory use, would greatly help prevent the "silo-ization"
problem.

   From: Jonathan Rosenberg <jdrosen@cisco.com>

   Unfortunately, PoC has layered a whole bunch of additional behaviors=20
   into the way SIP requests are handled. Dale has indicated one of them
-=20
   that a 200 OK is returned on the terminating side priod to the call=20
   actually being completed. PoC also *requires* the insertion of
servers=20
   into the SIP signaling path to make the system work. Consequently,=20
   though the messages that PoC uses *look* like SIP messages with a new

   media type, there is a hidden 'extension' in place that changes the=20
   fundamental semantics of INVITE. The fact that this extension is not=20
   visible as a specific header in a well defined way does not change
the=20
   fact that PoC CHANGES THE BEHAVIOR OF SIP AND IS THUS A SIP
EXTENSION.

   To convince yourself of this, consider what would happen if a PoC
client=20
   calls a regular SIP phone. The regular SIP phone would see the PoC
floor=20
   control stream and reject it as unknown. If PoC were just SIP as
we've=20
   specified it, then everything would be fine, and the call would
proceed=20
   but without floor control.

   But alas, it is not so.

But why is it not so?  OK, I know nothing about PoC, but naively it
seems that a PoC [here I assume that PoC means "push-to-talk over
cellular"] UAC, if it did not see an immediate 200 would know that the
far end did not support PoC functionality and could fall back to
acting like an ordinary UAC.  (Indeed, it could quickly warn the user
that PoC is not in effect by playing a ring tone!)

Assuming that SIP-based PoC was designed with upward-compatibility in
mind, a PoC UA could reliably determine whether the far end supported
PoC on a call-by-call basis.

OTOH, if PoC changes the meanings of SIP messages so much that,
without a specified service context, they cannot be interpreted enough
to determine if PoC is in effect or not, um, well, what *were* they
thinking when they designed it?  And why haven't they fixed it?

Now, I suspect I've just revealed a large gap in my knowledge, but I
suspect that many other readers have the same gap, and it would
advance the discussion if you could reduce our ignorance.

   From: Henning Schulzrinne <hgs@cs.columbia.edu>

   A robust and distributed allocation mechanism for identifiers makes
   sense if you expect hundreds of entities to independently need to
   allocate such identifiers. This is clearly undesirable in this
   case, based on the examples presented, so the IANA registration
   mechanism for Require tags seems to be both sufficient and
   sufficiently onerous to prevent gratuitous incompatibility.

Naively, it seems that having more than a few defined services would
be useless in regard to the business of the cellular industry, as it
would balkanize the cellular PSTN, and customers would fall back to
using only the simplest modes of operation.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential =
information, privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 22 00:49:45 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HK6pW-0001ja-7y; Thu, 22 Feb 2007 00:49:22 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HK6pU-0001iX-5O
	for sipping@ietf.org; Thu, 22 Feb 2007 00:49:20 -0500
Received: from c243-111.rim.net ([216.9.243.111] helo=MHS01YKF.rim.net)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HK6pT-0007CW-OG
	for sipping@ietf.org; Thu, 22 Feb 2007 00:49:20 -0500
Importance: normal
Priority: normal
Received: from ngw02ykf.rim.net ([10.102.108.31]) by MHS01YKF.rim.net with
	Microsoft SMTPSVC(6.0.3790.211); Thu, 22 Feb 2007 00:49:11 -0500
Received: from XCH20YKF.rim.net ([10.102.100.35]) by ngw02ykf.rim.net (SAVSMTP
	3.1.6.45) with SMTP id M2007022200491029580 ;
	Thu, 22 Feb 2007 00:49:10 -0500
Received: from XCH47YKF.rim.net ([10.64.31.217]) by XCH20YKF.rim.net with
	Microsoft SMTPSVC(5.0.2195.6713); Thu, 22 Feb 2007 00:49:11 -0500
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.181
Content-Class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification Model 
Date: Thu, 22 Feb 2007 00:49:05 -0500
Message-ID: <61968779B8AC4C4BAB421D4C12F008C009152060@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification Model 
thread-index: AcdV5i0Ma5Tu/sHZEdueygAWy4mm/wAV9tzA
From: "Andrew Allen" <aallen@rim.com>
To: "Eric Burger" <eburger@bea.com>,
	<sipping@ietf.org>
X-OriginalArrivalTime: 22 Feb 2007 05:49:11.0200 (UTC)
	FILETIME=[2D0E6A00:01C75645]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a92270ba83d7ead10c5001bb42ec3221
Cc: Andrew Allen <aallen@rim.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi Eric,

It seems your prefer the approach of using URIs to identify services
which is part of Jonathan's draft as is the use of Require for those
services which simply do not interoperate unless the UAS understands the
service.

Where I am concerned is that you seem to be suggesting that the UA
separately register an AoR for each service it supports using REGISTER.
Certainly for cellular that wouldn't be acceptable as 10 or more
services could easily be supported by a single device. Registrations
create a significant load in the IMS network because its not just a SIP
REGISTER but also requires a lot of interactions with other network
elements such as the HSS and Notifications for reg-events etc. Also
having an order of magnitude increase in multiple registrations is going
to be a big issue for the avalanche restart problem for example when a
cell site goes down and comes back into service.

I think Jonathan's approach gives us the possibility for the registrar
to create the Service URIs based on the feature parameters included in
the Contact of the REGISTER.=20

I think the argument over whether the model is one UA with multiple
services or multiple UAs each representing a separate service is
somewhat academic. As I understand a UA can register multiple AoRs and
can have other AoRs configured or otherwise provided by the service
provider.=20

Andrew

-----Original Message-----
From: Eric Burger [mailto:eburger@bea.com]=20
Sent: Wednesday, February 21, 2007 1:29 PM
To: sipping@ietf.org
Subject: [Sipping] Service Identification Model=20

Folks have been pestering me for a while to look at the Service
Identification issue.  I finally have gotten around to it.


I guess I have a fundamental problem with the principal assertion of the
service-identification draft.  Namely, the assertion that there is a
single
SIP Layer, separate and distinct from two services.  Here is Figure 2
from
the draft:

                           +---------------------------------+
                           |                                 |
                           | +-------------+ +-------------+ |
                           | |      UI     | |      UI     | |
                           | +-------------+ +-------------+ |
                           | +-------------+ +-------------+ |
                           | |             | |             | |
                           | |  Service 1  | |  Service 2  | |
                           | |             | |             | |
                           | +-------------+ +-------------+ |
                           | +-----------------------------+ |
                           | |                             | |
                           | |             SIP             | |
                           | |            Layer            | |
                           | |                             | |
                           | +-----------------------------+ |
                           |                                 |
                           +---------------------------------+

                                     Physical Device

I would offer this drawing confuses a common SIP Stack with a common
User
Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use case:

> ... In addition users may have
> multiple devices only one of which has the application that
understands
> the game. Other devices that the user has registered against his AoR
may
> handle audio for telephony and other voice applications but if Alice
> invites Bob to play her favourite game it is not very interesting if
Bob
> accepts the gaming session on a device that can only accept the audio
> stream. Also even when the terminal that has the game application
> receives the INVITE the correct application in that terminal needs to
be
> identified to accept the session.

I would offer that this is the clinching argument that the request is,
in
fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a User
Agent.

Same argument for PoC versus a regular voice call: One is looking for a
PoC
User Agent, NOT the PoC service at a given User Agent.

Thus, the picture really looks like the following:

                           +---------------------------------+
                           |                                 |
                           | +-------------+ +-------------+ |
                           | |    UI       | |    UI       | |
                           | +-------------+ +-------------+ |
                           | +-------------+ +-------------+ |
                           | |             | |             | |
                           | |    UAS 1    | |    UAS 2    | |
                           | |             | |             | |
                           | +-------------+ +-------------+ |
                           | +-----------------------------+ |
                           | |                             | |
                           | |             SIP             | |
                           | |            Stack            | |
                           | |                             | |
                           | +-----------------------------+ |
                           |                                 |
                           +---------------------------------+

                                     Physical Device


Another argument for different User Agents is the whole "route based on
SDP"
argument.  As the author of RFC 3458, I think I was first to make the
assertion that knowing the media type did NOT mean you know the
application
to use.  I don't want your phone or proxy or B2BUA to guess that because
I
have Chess Media, I want the Chess Application; I want to talk with your
Chess Application.

Now, is the Chess Application a Preference, Requirement, or the thing
that I
am addressing?  The answer drives one towards using caller prefs, the
Require header, or the Request-URI.

I would offer that it is not a preference: I really want the chess
application; I probably do not want to talk to you.

That leaves us with the Require header or a Request-URI.  My leanings
are
not surprisingly towards the RFC 4240 approach.  Frankly, if you are
building an IMS (as an example) that can handle 10,000,000 subscribers,
you
should be able to handle 100,000,000 registrations.  That order of
magnitude
is way easier than the 10,000 to 10,000,000 jump.  I don't buy the
"registering each service is too much of a burden on the operator"
argument,
because it is not that much of a burden, and the network actually does
the
right thing.  Better yet, all of the corner cases work.  Best of all,
the
IETF doesn't have to build an IANA Registry and process a thousand PS
documents to register urn:service:myco:widget.

I suppose I could live with Require:, only if you are way too lazy to
register each application instance.  It does semantically work, because
if I
want to do PoC and you do not do PoC, I'll get a 4xx response, to which
I
could try a regular call.

_______________________________________________________________________
Notice:  This email message, together with any attachments, may contain
information  of  BEA Systems,  Inc.,  its subsidiaries  and  affiliated
entities,  that may be confidential,  proprietary,  copyrighted  and/or
legally privileged, and is intended solely for the use of the individual
or entity named in this message. If you are not the intended recipient,
and have received this message in error, please immediately return this
by email and then delete it.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential =
information, privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 22 03:22:43 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HK9Dj-0001BL-A8; Thu, 22 Feb 2007 03:22:31 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HK9Dh-0001BA-58
	for sipping@ietf.org; Thu, 22 Feb 2007 03:22:29 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HK9Da-00044v-9C
	for sipping@ietf.org; Thu, 22 Feb 2007 03:22:29 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	8C37820FA0
	for <sipping@ietf.org>; Thu, 22 Feb 2007 09:22:17 +0100 (CET)
X-AuditID: c1b4fb3e-b1ed7bb0000007e1-bd-45dd52b92f5e 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	6CADE2052F
	for <sipping@ietf.org>; Thu, 22 Feb 2007 09:22:17 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 09:22:17 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification Model 
Date: Thu, 22 Feb 2007 09:22:15 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD028F09A4@esealmw111>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C009152060@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification Model 
Thread-Index: AcdV5i0Ma5Tu/sHZEdueygAWy4mm/wAV9tzAAAZvQkA=
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 22 Feb 2007 08:22:17.0353 (UTC)
	FILETIME=[906E1F90:01C7565A]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 311e798ce51dbeacf5cdfcc8e9fda21b
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi,

Isn't relying on a different AoR record for each service a rather
backwards approach?  Today, there have been upto three numbers on the
phones representing different services (Telephony; fax and Data) - which
means business cards which means different numbers for the different
ways I want to contact a person.  In myunderstanding, that would apply
here as a different AoR record for Telephony; PoC and Messaging.  I
think it is much more preferable to have one AoR; and then to be able to
indicate the requested service.

Steve.

-----Original Message-----
From: Andrew Allen [mailto:aallen@rim.com]=20
Sent: Thursday, February 22, 2007 6:49 AM
To: Eric Burger; sipping@ietf.org
Cc: Andrew Allen
Subject: RE: [Sipping] Service Identification Model=20


Hi Eric,

It seems your prefer the approach of using URIs to identify services
which is part of Jonathan's draft as is the use of Require for those
services which simply do not interoperate unless the UAS understands the
service.

Where I am concerned is that you seem to be suggesting that the UA
separately register an AoR for each service it supports using REGISTER.
Certainly for cellular that wouldn't be acceptable as 10 or more
services could easily be supported by a single device. Registrations
create a significant load in the IMS network because its not just a SIP
REGISTER but also requires a lot of interactions with other network
elements such as the HSS and Notifications for reg-events etc. Also
having an order of magnitude increase in multiple regFrom sipping-bounces@ietf.org Thu Feb 22 03:22:43 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HK9Dj-0001BL-A8; Thu, 22 Feb 2007 03:22:31 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HK9Dh-0001BA-58
	for sipping@ietf.org; Thu, 22 Feb 2007 03:22:29 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HK9Da-00044v-9C
	for sipping@ietf.org; Thu, 22 Feb 2007 03:22:29 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	8C37820FA0
	for <sipping@ietf.org>; Thu, 22 Feb 2007 09:22:17 +0100 (CET)
X-AuditID: c1b4fb3e-b1ed7bb0000007e1-bd-45dd52b92f5e 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	6CADE2052F
	for <sipping@ietf.org>; Thu, 22 Feb 2007 09:22:17 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 09:22:17 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification Model 
Date: Thu, 22 Feb 2007 09:22:15 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD028F09A4@esealmw111>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C009152060@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification Model 
Thread-Index: AcdV5i0Ma5Tu/sHZEdueygAWy4mm/wAV9tzAAAZvQkA=
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 22 Feb 2007 08:22:17.0353 (UTC)
	FILETIME=[906E1F90:01C7565A]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 311e798ce51dbeacf5cdfcc8e9fda21b
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi,

Isn't relying on a different AoR record for each service a rather
backwards approach?  Today, there have been upto three numbers on the
phones representing different services (Telephony; fax and Data) - which
means business cards which means different numbers for the different
ways I want to contact a person.  In myunderstanding, that would apply
here as a different AoR record for Telephony; PoC and Messaging.  I
think it is much more preferable to have one AoR; and then to be able to
indicate the requested service.

Steve.

-----Original Message-----
From: Andrew Allen [mailto:aallen@rim.com]=20
Sent: Thursday, February 22, 2007 6:49 AM
To: Eric Burger; sipping@ietf.org
Cc: Andrew Allen
Subject: RE: [Sipping] Service Identification Model=20


Hi Eric,

It seems your prefer the approach of using URIs to identify services
which is part of Jonathan's draft as is the use of Require for those
services which simply do not interoperate unless the UAS understands the
service.

Where I am concerned is that you seem to be suggesting that the UA
separately register an AoR for each service it supports using REGISTER.
Certainly for cellular that wouldn't be acceptable as 10 or more
services could easily be supported by a single device. Registrations
create a significant load in the IMS network because its not just a SIP
REGISTER but also requires a lot of interactions with other network
elements such as the HSS and Notifications for reg-events etc. Also
having an order of magnitude increase in multiple registrations is going
to be a big issue for the avalanche restart problem for example when a
cell site goes down and comes back into service.

I think Jonathan's approach gives us the possibility for the registrar
to create the Service URIs based on the feature parameters included in
the Contact of the REGISTER.=20

I think the argument over whether the model is one UA with multiple
services or multiple UAs each representing a separate service is
somewhat academic. As I understand a UA can register multiple AoRs and
can have other AoRs configured or otherwise provided by the service
provider.=20

Andrew

-----Original Message-----
From: Eric Burger [mailto:eburger@bea.com]
Sent: Wednesday, February 21, 2007 1:29 PM
To: sipping@ietf.org
Subject: [Sipping] Service Identification Model=20

Folks have been pestering me for a while to look at the Service
Identification issue.  I finally have gotten around to it.


I guess I have a fundamental problem with the principal assertion of the
service-identification draft.  Namely, the assertion that there is a
single SIP Layer, separate and distinct from two services.  Here is
Figure 2 from the draft:

                           +---------------------------------+
                           |                                 |
                           | +-------------+ +-------------+ |
                           | |      UI     | |      UI     | |
                           | +-------------+ +-------------+ |
                           | +-------------+ +-------------+ |
                           | |             | |             | |
                           | |  Service 1  | |  Service 2  | |
                           | |             | |             | |
                           | +-------------+ +-------------+ |
                           | +-----------------------------+ |
                           | |                             | |
                           | |             SIP             | |
                           | |            Layer            | |
                           | |                             | |
                           | +-----------------------------+ |
                           |                                 |
                           +---------------------------------+

                                     Physical Device

I would offer this drawing confuses a common SIP Stack with a common
User Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use
case:

> ... In addition users may have
> multiple devices only one of which has the application that
understands
> the game. Other devices that the user has registered against his AoR
may
> handle audio for telephony and other voice applications but if Alice=20
> invites Bob to play her favourite game it is not very interesting if
Bob
> accepts the gaming session on a device that can only accept the audio=20
> stream. Also even when the terminal that has the game application=20
> receives the INVITE the correct application in that terminal needs to
be
> identified to accept the session.

I would offer that this is the clinching argument that the request is,
in fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a
User Agent.

Same argument for PoC versus a regular voice call: One is looking for a
PoC User Agent, NOT the PoC service at a given User Agent.

Thus, the picture really looks like the following:

                           +---------------------------------+
                           |                                 |
                           | +-------------+ +-------------+ |
                           | |    UI       | |    UI       | |
                           | +-------------+ +-------------+ |
                           | +-------------+ +-------------+ |
                           | |             | |             | |
                           | |    UAS 1    | |    UAS 2    | |
                           | |             | |             | |
                           | +-------------+ +-------------+ |
                           | +-------------------istrations is going
to be a big issue for the avalanche restart problem for example when a
cell site goes down and comes back into service.

I think Jonathan's approach gives us the possibility for the registrar
to create the Service URIs based on the feature parameters included in
the Contact of the REGISTER.=20

I think the argument over whether the model is one UA with multiple
services or multiple UAs each representing a separate service is
somewhat academic. As I understand a UA can register multiple AoRs and
can have other AoRs configured or otherwise provided by the service
provider.=20

Andrew

-----Original Message-----
From: Eric Burger [mailto:eburger@bea.com]
Sent: Wednesday, February 21, 2007 1:29 PM
To: sipping@ietf.org
Subject: [Sipping] Service Identification Model=20

Folks have been pestering me for a while to look at the Service
Identification issue.  I finally have gotten around to it.


I guess I have a fundamental problem with the principal assertion of the
service-identification draft.  Namely, the assertion that there is a
single SIP Layer, separate and distinct from two services.  Here is
Figure 2 from the draft:

                           +---------------------------------+
                           |                                 |
                           | +-------------+ +-------------+ |
                           | |      UI     | |      UI     | |
                           | +-------------+ +-------------+ |
                           | +-------------+ +-------------+ |
                           | |             | |             | |
                           | |  Service 1  | |  Service 2  | |
                           | |             | |             | |
                           | +-------------+ +-------------+ |
                           | +-----------------------------+ |
                           | |                             | |
                           | |             SIP             | |
                           | |            Layer            | |
                           | |                             | |
                           | +-----------------------------+ |
                           |                                 |
                           +---------------------------------+

                                     Physical Device

I would offer this drawing confuses a common SIP Stack with a common
User Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use
case:

> ... In addition users may have
> multiple devices only one of which has the application that
understands
> the game. Other devices that the user has registered against his AoR
may
> handle audio for telephony and other voice applications but if Alice=20
> invites Bob to play her favourite game it is not very interesting if
Bob
> accepts the gaming session on a device that can only accept the audio=20
> stream. Also even when the terminal that has the game application=20
> receives the INVITE the correct application in that terminal needs to
be
> identified to accept the session.

I would offer that this is the clinching argument that the request is,
in fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a
User Agent.

Same argument for PoC versus a regular voice call: One is looking for a
PoC User Agent, NOT the PoC service at a given User Agent.

Thus, the picture really looks like the following:

                           +---------------------------------+
                           |                                 |
                           | +-------------+ +-------------+ |
                           | |    UI       | |    UI       | |
                           | +-------------+ +-------------+ |
                           | +-------------+ +-------------+ |
                           | |             | |             | |
                           | |    UAS 1    | |    UAS 2    | |
                           | |             | |             | |
                           | +-------------+ +-------------+ |
                           | +-----------------------------+ |
                           | |                             | |
                           | |             SIP             | |
                           | |            Stack            | |
                           | |                             | |
                           | +-----------------------------+ |
                           |                                 |
                           +---------------------------------+

                                     Physical Device


Another argument for different User Agents is the whole "route based on
SDP"
argument.  As the author of RFC 3458, I think I was first to make the
assertion that knowing the media type did NOT mean you know the
application to use.  I don't want your phone or proxy or B2BUA to guess
that because I have Chess Media, I want the Chess Application; I want to
talk with your Chess Application.

Now, is the Chess Application a Preference, Requirement, or the thing
that I am addressing?  The answer drives one towards using caller prefs,
the Require header, or the Request-URI.

I would offer that it is not a preference: I really want the chess
application; I probably do not want to talk to you.

That leaves us with the Require header or a Request-URI.  My leanings
are not surprisingly towards the RFC 4240 approach.  Frankly, if you are
building an IMS (as an example) that can handle 10,000,000 subscribers,
you should be able to handle 100,000,000 registrations.  That order of
magnitude is way easier than the 10,000 to 10,000,000 jump.  I don't buy
the "registering each service is too much of a burden on the operator"
argument,
because it is not that much of a burden, and the network actually does
the right thing.  Better yet, all of the corner cases work.  Best of
all, the IETF doesn't have to build an IANA Registry and process a
thousand PS documents to register urn:service:myco:widget.

I suppose I could live with Require:, only if you are way too lazy to
register each application instance.  It does semantically work, because
if I want to do PoC and you do not do PoC, I'll get a 4xx response, to
which I could try a regular call.

_______________________________________________________________________
Notice:  This email message, together with any attachments, may contain
information  of  BEA Systems,  Inc.,  its subsidiaries  and  affiliated
entities,  that may be confidential,  proprietary,  copyrighted  and/or
legally privileged, and is intended solely for the use of the individual
or entity named in this message. If you are not the intended recipient,
and have received this message in error, please immediately return this
by email and then delete it.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential
information, privileged material (including material protected by the
solicitor-client or other applicable privileges), or constitute
non-public information. Any use of this information by anyone other than
the intended recipient is prohibited. If you have received this
transmission in error, please immediately reply to the sender and delete
this information from your system. Use, dissemination, distribution, or
reproduction of this transmission by unintended recipients is not
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the----------+ |
                           | |                             | |
                           | |             SIP             | |
                           | |            Stack            | |
                           | |                             | |
                           | +-----------------------------+ |
                           |                                 |
                           +---------------------------------+

                                     Physical Device


Another argument for different User Agents is the whole "route based on
SDP"
argument.  As the author of RFC 3458, I think I was first to make the
assertion that knowing the media type did NOT mean you know the
application to use.  I don't want your phone or proxy or B2BUA to guess
that because I have Chess Media, I want the Chess Application; I want to
talk with your Chess Application.

Now, is the Chess Application a Preference, Requirement, or the thing
that I am addressing?  The answer drives one towards using caller prefs,
the Require header, or the Request-URI.

I would offer that it is not a preference: I really want the chess
application; I probably do not want to talk to you.

That leaves us with the Require header or a Request-URI.  My leanings
are not surprisingly towards the RFC 4240 approach.  Frankly, if you are
building an IMS (as an example) that can handle 10,000,000 subscribers,
you should be able to handle 100,000,000 registrations.  That order of
magnitude is way easier than the 10,000 to 10,000,000 jump.  I don't buy
the "registering each service is too much of a burden on the operator"
argument,
because it is not that much of a burden, and the network actually does
the right thing.  Better yet, all of the corner cases work.  Best of
all, the IETF doesn't have to build an IANA Registry and process a
thousand PS documents to register urn:service:myco:widget.

I suppose I could live with Require:, only if you are way too lazy to
register each application instance.  It does semantically work, because
if I want to do PoC and you do not do PoC, I'll get a 4xx response, to
which I could try a regular call.

_______________________________________________________________________
Notice:  This email message, together with any attachments, may contain
information  of  BEA Systems,  Inc.,  its subsidiaries  and  affiliated
entities,  that may be confidential,  proprietary,  copyrighted  and/or
legally privileged, and is intended solely for the use of the individual
or entity named in this message. If you are not the intended recipient,
and have received this message in error, please immediately return this
by email and then delete it.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential
information, privileged material (including material protected by the
solicitor-client or other applicable privileges), or constitute
non-public information. Any use of this information by anyone other than
the intended recipient is prohibited. If you have received this
transmission in error, please immediately reply to the sender and delete
this information from your system. Use, dissemination, distribution, or
reproduction of this transmission by unintended recipients is not
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP

From sipping-bounces@ietf.org Thu Feb 22 03:22:43 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HK9Dm-0001CS-6j; Thu, 22 Feb 2007 03:22:34 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HK9Dj-0001BV-Qo
	for sipping@ietf.org; Thu, 22 Feb 2007 03:22:31 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HK9De-00045f-M6
	for sipping@ietf.org; Thu, 22 Feb 2007 03:22:31 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	0E0EE20670
	for <sipping@ietf.org>; Thu, 22 Feb 2007 09:22:20 +0100 (CET)
X-AuditID: c1b4fb3c-ae7c5bb0000007de-89-45dd52bcb66a 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	01B2C203D6
	for <sipping@ietf.org>; Thu, 22 Feb 2007 09:22:20 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 09:22:19 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Thu, 22 Feb 2007 09:22:18 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD028F09A5@esealmw111>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C009152056@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdNmH89O4T46iGoRjicZ3hYknVpawB8E6oQAH++leAABzLk8AAYIixwAIPObeAAhKg1EAAMiGmQ
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 22 Feb 2007 08:22:19.0931 (UTC)
	FILETIME=[91F77EB0:01C7565A]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8de8c36679aaa17b008853e74231c885
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi Andrew,

I have the opposite view.  I think that if the originating network and
the terminating network are going to interoperate, then they will have
the same understanding of the service, hence in my view the same
"identifier".  To have otherwise is problematic and complicated (the
operators will have to have a higher admininstrative effort in order to
understand how they should interconnect).

The PBX case is more complicated than can be responded to in a single
sentence - as there are many ways in which the PBXs are connected
together.  In one case - the PBXs are just treated as a terminal with
many users; and the service that the operator provides is really the
same basic service as is provided to other users (with maybe more
subscriber specific logic).  In other cases, the operator just provides
routing and connecting support, and here the service provided by the
operator is different; but I would assume that the peer PBXs had better
have a common understanding of the service (which they would if it was
the same vendor), or again there will be inter-operatability issues.

Regarding the URI parameter proposal, don't we have proxies that remove
additional R-URIs - it seems a natural thing to remove it also in e.g.
fowarding services etc ...

Regards,

Steve.

-----Original Message-----
From: Andrew Allen [mailto:aallen@rim.com application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP

From sipping-bounces@ietf.org Thu Feb 22 03:22:43 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HK9Dm-0001CS-6j; Thu, 22 Feb 2007 03:22:34 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HK9Dj-0001BV-Qo
	for sipping@ietf.org; Thu, 22 Feb 2007 03:22:31 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HK9De-00045f-M6
	for sipping@ietf.org; Thu, 22 Feb 2007 03:22:31 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	0E0EE20670
	for <sipping@ietf.org>; Thu, 22 Feb 2007 09:22:20 +0100 (CET)
X-AuditID: c1b4fb3c-ae7c5bb0000007de-89-45dd52bcb66a 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	01B2C203D6
	for <sipping@ietf.org>; Thu, 22 Feb 2007 09:22:20 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 09:22:19 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Thu, 22 Feb 2007 09:22:18 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD028F09A5@esealmw111>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C009152056@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdNmH89O4T46iGoRjicZ3hYknVpawB8E6oQAH++leAABzLk8AAYIixwAIPObeAAhKg1EAAMiGmQ
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 22 Feb 2007 08:22:19.0931 (UTC)
	FILETIME=[91F77EB0:01C7565A]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8de8c36679aaa17b008853e74231c885
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi Andrew,

I have the opposite view.  I think that if the originating network and
the terminating network are going to interoperate, then they will have
the same understanding of the service, hence in my view the same
"identifier".  To have otherwise is problematic and complicated (the
operators will have to have a higher admininstrative effort in order to
understand how they should interconnect).

The PBX case is more complicated than can be responded to in a single
sentence - as there are many ways in which the PBXs are connected
together.  In one case - the PBXs are just treated as a terminal with
many users; and the service that the operator provides is really the
same basic service as is provided to other users (with maybe more
subscriber specific logic).  In other cases, the operator just provides
routing and connecting support, and here the service provided by the
operator is different; but I would assume that the peer PBXs had better
have a common understanding of the service (which they would if it was
the same vendor), or again there will be inter-operatability issues.

Regarding the URI parameter proposal, don't we have proxies that remove
additional R-URIs - it seems a natural thing to remove it also in e.g.
fowarding services etc ...

Regards,

Steve.

-----Original Message-----
From: Andrew Allen [mailto:aallen@rim.com]=20
Sent: Thursday, February 22, 2007 4:05 AM
To: Stephen Terrill (MI/EEM); sipping@ietf.org
Cc: Andrew Allen
Subject: RE: [Sipping] Service Identification in SIP


Hi Steve

In my view this is one of the interoperability issues with your
proposal. I think the assumption that all networks will deploy exactly
the same services with the same service identifiers is unrealistic and
also undesirable. Are corporate PBX vendors going to implement the same
services as defined by 3GPP or OMA? The possibility to have the
flexibility to have different service identifiers for session equivalent
services is in my view essential to any IETF standard and to do
otherwise is to create a situation where basic services like telephony
or messaging cannot interoperate between users in different networks and
users in the internet without the use of inter working gateway. SIP and
SIMPLE were setup specifically to create protocols that would enable
interoperability with IP networks without the need for gateways.=20

If 3GPP network operators choose to deploy the same services and use
common identifiers that is fine but there are whole other worlds out
there as well and service identifiers should not stand in the way of
interoperability between all these different SIP users. If the IETF
solution supports the ability for different networks to use different
identifiers that doesn't violate 3GPP requirements and doesn't prevent
3GPP using the same identifiers in different networks. I don't see how
that is against TS 23.228.

Andrew


-----Original Message-----
From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
Sent: Monday, February 19, 2007 6:00 AM
To: Andrew Allen; sipping@ietf.org
Subject: RE: [Sipping] Service Identification in SIP

Hi Andrew,

I see that in your assumptions, the originating and terminating networks
use different service identifiers, and I presume that this is for the
same service.  In my understanding, it is the same service identifier
that is used end-to-end (if not it is interworked).  That is why I do
not see the reliance on Presence being necessary.  In my opionion this
totally within the requirements of TS 23.228 (I think that different
service identifiers in different networks for the same service is
against TS 23.228).

This comes to the isssue of the PSTN gateway - I presume that you want
to go to the PSTN gateway if the service is telephony, and within that,
voice.  If we know that, then the we go to any PSTN gateway, and we
can't presume a particular network for a PSTN gateway.=20

Best Regards,

Steve.

-----Original Message-----
From: Andrew Allen [mailto:aallen@rim.com]
Sent: Saturday, February 17, 2007 12:51 AM
To: Stephen Terrill (MI/EEM); sipping@ietf.org
Cc: Andrew Allen
Subject: RE: [Sipping] Service Identification in SIP



Hi Steve,

My responses inline

Andrew=20

-----Original Message-----
From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
Sent: Friday, February 16, 2007 3:51 AM
To: sipping@ietf.org
Subject: RE: [Sipping] Service Identification in SIP

Hello Andrew,

Thanks for your comments.  I have a couple of brief responses below.

Best Regards,

..//Steve=20

-----Original Message-----
From: Andrew Allen [mailto:aallen@rim.com]
Sent: Friday, February 16, 2007 7:15 AM
To: Stephen Terrill (MI/EEM); sipping@ietf.org
Cc: Andrew Allen
Subject: RE: [Sipping] Service Identification in SIP


Some comments on Stephen's comments on the draft:
=20

-Moving onto the proposed solution draft itself, in addition for
providing for the identification of the "application" and the
"communication service", I have a few additional thoughts.

The draft seems to have the implication that all services (communication
services in other terminology) that are to be identified have to be
standardised within the IETF.  This is clearly an restriction that will
not be acceptable to the industry as a whole.  My guess is that if such
a statement is mandated by the IETF, it will just be words on paper
which would likely be ignored in practice by other standardisation
organisations.

[AA] This I thi]=20
Sent: Thursday, February 22, 2007 4:05 AM
To: Stephen Terrill (MI/EEM); sipping@ietf.org
Cc: Andrew Allen
Subject: RE: [Sipping] Service Identification in SIP


Hi Steve

In my view this is one of the interoperability issues with your
proposal. I think the assumption that all networks will deploy exactly
the same services with the same service identifiers is unrealistic and
also undesirable. Are corporate PBX vendors going to implement the same
services as defined by 3GPP or OMA? The possibility to have the
flexibility to have different service identifiers for session equivalent
services is in my view essential to any IETF standard and to do
otherwise is to create a situation where basic services like telephony
or messaging cannot interoperate between users in different networks and
users in the internet without the use of inter working gateway. SIP and
SIMPLE were setup specifically to create protocols that would enable
interoperability with IP networks without the need for gateways.=20

If 3GPP network operators choose to deploy the same services and use
common identifiers that is fine but there are whole other worlds out
there as well and service identifiers should not stand in the way of
interoperability between all these different SIP users. If the IETF
solution supports the ability for different networks to use different
identifiers that doesn't violate 3GPP requirements and doesn't prevent
3GPP using the same identifiers in different networks. I don't see how
that is against TS 23.228.

Andrew


-----Original Message-----
From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
Sent: Monday, February 19, 2007 6:00 AM
To: Andrew Allen; sipping@ietf.org
Subject: RE: [Sipping] Service Identification in SIP

Hi Andrew,

I see that in your assumptions, the originating and terminating networks
use different service identifiers, and I presume that this is for the
same service.  In my understanding, it is the same service identifier
that is used end-to-end (if not it is interworked).  That is why I do
not see the reliance on Presence being necessary.  In my opionion this
totally within the requirements of TS 23.228 (I think that different
service identifiers in different networks for the same service is
against TS 23.228).

This comes to the isssue of the PSTN gateway - I presume that you want
to go to the PSTN gateway if the service is telephony, and within that,
voice.  If we know that, then the we go to any PSTN gateway, and we
can't presume a particular network for a PSTN gateway.=20

Best Regards,

Steve.

-----Original Message-----
From: Andrew Allen [mailto:aallen@rim.com]
Sent: Saturday, February 17, 2007 12:51 AM
To: Stephen Terrill (MI/EEM); sipping@ietf.org
Cc: Andrew Allen
Subject: RE: [Sipping] Service Identification in SIP



Hi Steve,

My responses inline

Andrew=20

-----Original Message-----
From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
Sent: Friday, February 16, 2007 3:51 AM
To: sipping@ietf.org
Subject: RE: [Sipping] Service Identification in SIP

Hello Andrew,

Thanks for your comments.  I have a couple of brief responses below.

Best Regards,

..//Steve=20

-----Original Message-----
From: Andrew Allen [mailto:aallen@rim.com]
Sent: Friday, February 16, 2007 7:15 AM
To: Stephen Terrill (MI/EEM); sipping@ietf.org
Cc: Andrew Allen
Subject: RE: [Sipping] Service Identification in SIP


Some comments on Stephen's comments on the draft:
=20

-Moving onto the proposed solution draft itself, in addition for
providing for the identification of the "application" and the
"communication service", I have a few additional thoughts.

The draft seems to have the implication that all services (communication
services in other terminology) that are to be identified have to be
standardised within the IETF.  This is clearly an restriction that will
not be acceptable to the industry as a whole.  My guess is that if such
a statement is mandated by the IETF, it will just be words on paper
which would likely be ignored in practice by other standardisation
organisations.

[AA] This I think is an incorrect interpretation of the draft. My
understanding is that the only services according to the draft that
require some standardisation in IETF are those that by their very nature
cannot interoperate unless there is some common additional understanding
between the endpoints outside of what is currently specified in SIP. The
use of SIP option tags with Require header is proposed by the draft for
this purpose as such services create interoperability issues with SIP.

<Steve>
My comment applies even for these services </Steve>


<AA>
Well then we are in big trouble because what you are proposing is that
its OK for bodies outside of IETF to create non interoperable extensions
for the behavour of SIP entities. This will result in the fragmentation
of SIP into different flavors that don't interoperate. This is the
balkanization of SIP that Paul talks about.even <AA>=20

----------------------------------------------------

The draft also only recognises the case where all explicitly knowledge
of the service is required.  I think it would also be good to allow the
case where knowledge of the service would be helpful, but not explicitly
required.  This could be a case where the service to the end-user would
not be degraded if this was not supported.  This functionality seems to
be missing.

[AA] Service URIs are defined in the draft for these services where the
support by the other endpoint is not required for the session to be
sucessfully established.

The draft seems to indicate that that application servers in the network
are involved only at the request of the originating terminal.  This
seems a strange restriction.  For some network support that supports the
originating user, such as the one provided in the example "I would like
this call recorded for later retrieval", such mechanisms may make sense,
however as a general approach it seems a bit limited.  For example, in
3GPP networks there is the need to invoke application logic in the
originating side and terminating side, not only at the request of the
originating terminal, but as a consequence of the requested service.  If
there is an SIP message contains indication of the requested service, a
proxy can do this (or loose route to an AS that contains the necessary
service logic).=20

[AA] The draft supports both possibilities using the Route header for
originating network servers that need to be involved in the route and
based on the terminating networks routing policy for the Service URI in
the Request URI.

<Steve>
I don't see the advantage of having different solutions for the
originating network and the terminating network.  I don't think that the
originating logic is invoked only at the request of the originating
terminal.
</Steve>

<AA>
Unless IETF is willing to standardise all services and mandate that all
SIP entities understand these (which will never happen) then we cannot
assume that all networks and all endpoints understand and use the same
identifiers. It is possible through configuration for an endpoint to
obtain the identifiers used by its own service provider but there is no
way to ensure that other networks use or understand the same identifiers
for the equivalent or similar service that is compatible at the session
layer. Therefore we need to support the possibility to have different
identifiers in different networks in order to not violate the TS 23.228
requirement:

12.	The usage of communication service identifiers shall not
adversely affect interoperability between IMS networks and
interoperability with external SIP networks and CS networks. The
behaviour of a network receiving the IMS requests without an IMS
communication service identifier is a matter of operator policy. Usage
of communication service identifiers shall not decrease the level of
interoperability with networks and UEs that are unaware of the
communication service identifier.

If the originating service logic isn't invoked based on an identifier
placed in the request by the originating terminal on what basis do you
think it is invoked? How does the use of Accept-Contact hank is an incorrect interpretation of the draft. My
understanding is that the only services according to the draft that
require some standardisation in IETF are those that by their very nature
cannot interoperate unless there is some common additional understanding
between the endpoints outside of what is currently specified in SIP. The
use of SIP option tags with Require header is proposed by the draft for
this purpose as such services create interoperability issues with SIP.

<Steve>
My comment applies even for these services </Steve>


<AA>
Well then we are in big trouble because what you are proposing is that
its OK for bodies outside of IETF to create non interoperable extensions
for the behavour of SIP entities. This will result in the fragmentation
of SIP into different flavors that don't interoperate. This is the
balkanization of SIP that Paul talks about.even <AA>=20

----------------------------------------------------

The draft also only recognises the case where all explicitly knowledge
of the service is required.  I think it would also be good to allow the
case where knowledge of the service would be helpful, but not explicitly
required.  This could be a case where the service to the end-user would
not be degraded if this was not supported.  This functionality seems to
be missing.

[AA] Service URIs are defined in the draft for these services where the
support by the other endpoint is not required for the session to be
sucessfully established.

The draft seems to indicate that that application servers in the network
are involved only at the request of the originating terminal.  This
seems a strange restriction.  For some network support that supports the
originating user, such as the one provided in the example "I would like
this call recorded for later retrieval", such mechanisms may make sense,
however as a general approach it seems a bit limited.  For example, in
3GPP networks there is the need to invoke application logic in the
originating side and terminating side, not only at the request of the
originating terminal, but as a consequence of the requested service.  If
there is an SIP message contains indication of the requested service, a
proxy can do this (or loose route to an AS that contains the necessary
service logic).=20

[AA] The draft supports both possibilities using the Route header for
originating network servers that need to be involved in the route and
based on the terminating networks routing policy for the Service URI in
the Request URI.

<Steve>
I don't see the advantage of having different solutions for the
originating network and the terminating network.  I don't think that the
originating logic is invoked only at the request of the originating
terminal.
</Steve>

<AA>
Unless IETF is willing to standardise all services and mandate that all
SIP entities understand these (which will never happen) then we cannot
assume that all networks and all endpoints understand and use the same
identifiers. It is possible through configuration for an endpoint to
obtain the identifiers used by its own service provider but there is no
way to ensure that other networks use or understand the same identifiers
for the equivalent or similar service that is compatible at the session
layer. Therefore we need to support the possibility to have different
identifiers in different networks in order to not violate the TS 23.228
requirement:

12.	The usage of communication service identifiers shall not
adversely affect interoperability between IMS networks and
interoperability with external SIP networks and CS networks. The
behaviour of a network receiving the IMS requests without an IMS
communication service identifier is a matter of operator policy. Usage
of communication service identifiers shall not decrease the level of
interoperability with networks and UEs that are unaware of the
communication service identifier.

If the originating service logic isn't invoked based on an identifier
placed in the request by the originating terminal on what basis do you
think it is invoked? How does the use of Accept-Contact have advantages
in this case?
<AA>=20

--------------------------------------

 I note here that the proposed service URN does not contain the address
of a specific server containing the application logic.  An AS can look
at that, and loose route to an application server that does contain such
logic (If it is separated from the proxy), however I fail to see the
difference for that proxy if it looks at a feature tag, or it looks at
the service URN for this.  In both cases, it has to parse the message
for the data, and modify the routing information. =20

[AA]Route headers are the SIP mechanism for requesting intermediaries to
be includede on the route of the request. There is nothing in RFC
3840/RFC 3841 that addresses using feature tags in Accept-Contact header
for any routing decisions prior to determining which of the registered
contacts the request should be forked to. Accept-Contact is not a Route
header it is merely a filter that is applied to the set of valid
contacts. Using a parameter in the filter as a service identifier has a
potential significant impact on the use of Accept-Contact for its
original purpose.=20

<Steve>
I don't believe there is anthing that states a Proxy CAN NOT route to
applications or application logic based upon elements in the SIP
message, including what is in the Accept-Contact header (some people
have even proposed SDP ...).
</Steve>

<AA>
As I had stated in an earlier email on this subject a proxy can do a lot
of things (including routing based on your favorite number appearing in
a call-Id) but that doesn't mean they are a good idea or a good design
decision. If you want to ensure you don't break things then you need to
be careful to use SIP headers with the semantics intended for them
otherwise problems can result if extra semantics are put upon existing
headers.
<AA>=20

Further, the route header could be removed in the originating network;
and it is still required to identify the service in the terminating
network.  As proxies may remove the route, it can't be used to transport
the requested service. =20

[AA] Proxies arbitarily removing Route headers is a BAD thing akin to a
bull in a china shop - almost certainly going to break something. RFC
3261 summarises what proxies should be doing with Route headers:
16.12 Summary of Proxy Route Processing

   In the absence of local policy to the contrary, the processing a
   proxy performs on a request containing a Route header field can be
   summarized in the following steps.

      1.  The proxy will inspect the Request-URI.  If it indicates a
          resource owned by this proxy, the proxy will replace it with
          the results of running a location service.  Otherwise, the
          proxy will not change the Request-URI.

      2.  The proxy will inspect the URI in the topmost Route header
          field value.  If it indicates this proxy, the proxy removes it
          from the Route header field (this route node has been
          reached).

      3.  The proxy will forward the request to the resource indicated
          by the URI in the topmost Route header field value or in the
          Request-URI if no Route header field is present.  The proxy
          determines the address, port and transport to use when
          forwarding the request by applying the procedures in [4] to
          that URI.

   If no strict-routing elements are encountered on the path of the
   request, the Request-URI will always indicate the target of the
   request.

<Steve>
The first sentance is interesting "In absence of local policy to the
contrary", indicating that there can be different policies.  I
understand that some proxies do remove this.  While I am on this - I
have lead to believe that R-URI parameters may also be removed by
intermediate proxies - is this correct?
</Steve>

<AA>
My understanding is that local policy may determine that additional
nodes need to be routed to in addition to the Route contained in the
request. However deleting existing Route headers other than the Route
header containing the address of the proxy as part of ve advantages
in this case?
<AA>=20

--------------------------------------

 I note here that the proposed service URN does not contain the address
of a specific server containing the application logic.  An AS can look
at that, and loose route to an application server that does contain such
logic (If it is separated from the proxy), however I fail to see the
difference for that proxy if it looks at a feature tag, or it looks at
the service URN for this.  In both cases, it has to parse the message
for the data, and modify the routing information. =20

[AA]Route headers are the SIP mechanism for requesting intermediaries to
be includede on the route of the request. There is nothing in RFC
3840/RFC 3841 that addresses using feature tags in Accept-Contact header
for any routing decisions prior to determining which of the registered
contacts the request should be forked to. Accept-Contact is not a Route
header it is merely a filter that is applied to the set of valid
contacts. Using a parameter in the filter as a service identifier has a
potential significant impact on the use of Accept-Contact for its
original purpose.=20

<Steve>
I don't believe there is anthing that states a Proxy CAN NOT route to
applications or application logic based upon elements in the SIP
message, including what is in the Accept-Contact header (some people
have even proposed SDP ...).
</Steve>

<AA>
As I had stated in an earlier email on this subject a proxy can do a lot
of things (including routing based on your favorite number appearing in
a call-Id) but that doesn't mean they are a good idea or a good design
decision. If you want to ensure you don't break things then you need to
be careful to use SIP headers with the semantics intended for them
otherwise problems can result if extra semantics are put upon existing
headers.
<AA>=20

Further, the route header could be removed in the originating network;
and it is still required to identify the service in the terminating
network.  As proxies may remove the route, it can't be used to transport
the requested service. =20

[AA] Proxies arbitarily removing Route headers is a BAD thing akin to a
bull in a china shop - almost certainly going to break something. RFC
3261 summarises what proxies should be doing with Route headers:
16.12 Summary of Proxy Route Processing

   In the absence of local policy to the contrary, the processing a
   proxy performs on a request containing a Route header field can be
   summarized in the following steps.

      1.  The proxy will inspect the Request-URI.  If it indicates a
          resource owned by this proxy, the proxy will replace it with
          the results of running a location service.  Otherwise, the
          proxy will not change the Request-URI.

      2.  The proxy will inspect the URI in the topmost Route header
          field value.  If it indicates this proxy, the proxy removes it
          from the Route header field (this route node has been
          reached).

      3.  The proxy will forward the request to the resource indicated
          by the URI in the topmost Route header field value or in the
          Request-URI if no Route header field is present.  The proxy
          determines the address, port and transport to use when
          forwarding the request by applying the procedures in [4] to
          that URI.

   If no strict-routing elements are encountered on the path of the
   request, the Request-URI will always indicate the target of the
   request.

<Steve>
The first sentance is interesting "In absence of local policy to the
contrary", indicating that there can be different policies.  I
understand that some proxies do remove this.  While I am on this - I
have lead to believe that R-URI parameters may also be removed by
intermediate proxies - is this correct?
</Steve>

<AA>
My understanding is that local policy may determine that additional
nodes need to be routed to in addition to the Route contained in the
request. However deleting existing Route headers other than the Route
header containing the address of the proxy as part of the loose routing
procedures is liable to break things and is not part of RFC 3261 routing
(loose routing).
<AA>=20


[AA] Generally a UA wouldn't know the appropriate service identifiers
used in the terminating network or whether a Route header is appropriate
to be used in the terminating network. The draft defines the Service URI
for the purpose of identifying services in the terminating network and
service URIs can be obtained using Presence.

<Steve>
I don't get the above paragraph.  It suggest that the "appropriate
service identifier" used in the terminating network is not known by the
originating UE, but the Service URI is.  I thought that the service URI
is an encoding of the "appropriate service identifier".
</Steve>

<AA>=20
Yes but in the general case the UA doesn't need to know what the service
identifier is it obtains through Presence a Service URI that contains
the Service identifier as a URI parameter. The great thing about using
Presence to obtain a Service URI is that the originating UA doesn't even
need to know anything about service identifiers at all for this to work.
It just uses Presence to find the URI to use for the service which can
be indicated to the user via an indication such as a graphic icon (e,g a
phone for telephony, smiley face for IM, etc) in the presence document.
<AA>=20

-----------------------------------

>From this perspective it is not clear to me the advantage of using a
service URN in a route header compared to the use of a feature tag.

[AA] I hope the above responses clarify that somewhat. Also the Service
URN and Service URIs allow different service identifiers to be used in
the originating and terminating network for session compatiable services
instead of forcing everyone to deploy and understand the same service
and service identifier.

<Steve>
Nope, don't get it, but are information that is carried end-to-end ....
And as far as I see, could be used by intermediate proxies to impact
local routing policy ... Or ?
</Steve>

<AA>=20
I don't understand this response either. The key point is that UAs in
corporate networks and in the internet are not going to deploy 3GPP
flavor services and use the same 3GPP service identifiers or even
possibly any service identifiers at all. The Accept-Contact based
proposal assumes that all endpoints understand the service identifiers
to be used for each service. This concept is not generally applicable
and create interoperability problems.
<AA>=20

The draft seems to mandate GRUU - is this really so, is that really
necessary.  It seems strange to place such a dependency, it also starts
to move the GRUU as being the primary means of which to contact a user.

[AA] My understanding is that GRUU in the draft is used to address how a
request addressed to Service URI is forked only to those contacts that
support the service identified by the that Service URI. A contact would
seem to be fine so long as it had GRUU properties - no need to obtain a
GRUU if you already have one.

This implies that the originating user decides which terminal I should
receive my requested service on, even if I have two devices for that.
As such, I don't believe that there should be a dependency on GRUU.

[AA] If you have obtained a Service URI (probably through presence) for
that user then you can address the request to the Service URI which
represents all the terminals that support that service. However GRUU and
the Service Instance URIs created from GRUUs provide another mechanism
through which a user can obtain a URI to reach that user for the
required service.

Parts of the draft seems to rely on the use of presence.  While presence
could be a supportive mechanism, I doubt that we can mandate that the
number you want to dial has be available in presence as I assume not all
entities are available in presence (diallable network services, users in
the PSTN).  This would be strange to mandated.

[AA] PSTN gateways are presumably in the same domain as the terminating
proxy and therefore these PSTN gateways can be configured with the
appropriate service identifiers for telthe loose routing
procedures is liable to break things and is not part of RFC 3261 routing
(loose routing).
<AA>=20


[AA] Generally a UA wouldn't know the appropriate service identifiers
used in the terminating network or whether a Route header is appropriate
to be used in the terminating network. The draft defines the Service URI
for the purpose of identifying services in the terminating network and
service URIs can be obtained using Presence.

<Steve>
I don't get the above paragraph.  It suggest that the "appropriate
service identifier" used in the terminating network is not known by the
originating UE, but the Service URI is.  I thought that the service URI
is an encoding of the "appropriate service identifier".
</Steve>

<AA>=20
Yes but in the general case the UA doesn't need to know what the service
identifier is it obtains through Presence a Service URI that contains
the Service identifier as a URI parameter. The great thing about using
Presence to obtain a Service URI is that the originating UA doesn't even
need to know anything about service identifiers at all for this to work.
It just uses Presence to find the URI to use for the service which can
be indicated to the user via an indication such as a graphic icon (e,g a
phone for telephony, smiley face for IM, etc) in the presence document.
<AA>=20

-----------------------------------

>From this perspective it is not clear to me the advantage of using a
service URN in a route header compared to the use of a feature tag.

[AA] I hope the above responses clarify that somewhat. Also the Service
URN and Service URIs allow different service identifiers to be used in
the originating and terminating network for session compatiable services
instead of forcing everyone to deploy and understand the same service
and service identifier.

<Steve>
Nope, don't get it, but are information that is carried end-to-end ....
And as far as I see, could be used by intermediate proxies to impact
local routing policy ... Or ?
</Steve>

<AA>=20
I don't understand this response either. The key point is that UAs in
corporate networks and in the internet are not going to deploy 3GPP
flavor services and use the same 3GPP service identifiers or even
possibly any service identifiers at all. The Accept-Contact based
proposal assumes that all endpoints understand the service identifiers
to be used for each service. This concept is not generally applicable
and create interoperability problems.
<AA>=20

The draft seems to mandate GRUU - is this really so, is that really
necessary.  It seems strange to place such a dependency, it also starts
to move the GRUU as being the primary means of which to contact a user.

[AA] My understanding is that GRUU in the draft is used to address how a
request addressed to Service URI is forked only to those contacts that
support the service identified by the that Service URI. A contact would
seem to be fine so long as it had GRUU properties - no need to obtain a
GRUU if you already have one.

This implies that the originating user decides which terminal I should
receive my requested service on, even if I have two devices for that.
As such, I don't believe that there should be a dependency on GRUU.

[AA] If you have obtained a Service URI (probably through presence) for
that user then you can address the request to the Service URI which
represents all the terminals that support that service. However GRUU and
the Service Instance URIs created from GRUUs provide another mechanism
through which a user can obtain a URI to reach that user for the
required service.

Parts of the draft seems to rely on the use of presence.  While presence
could be a supportive mechanism, I doubt that we can mandate that the
number you want to dial has be available in presence as I assume not all
entities are available in presence (diallable network services, users in
the PSTN).  This would be strange to mandated.

[AA] PSTN gateways are presumably in the same domain as the terminating
proxy and therefore these PSTN gateways can be configured with the
appropriate service identifiers for telephony services and can include
this in either a Route header to route to the right application server
or can construct a valid Service URI. The real problem is how do users
in other networks know what the service identifier that's is used in the
terminating network is. The use of Service URIs and Presence provides a
solution to that problem.

<Steve>
Again, presence could be a useful mechanism, but I think it would be a
problem for the draft to rely it.  A user making a call to the PSTN may
not find that user in the presence archicture (and similar to services
like the 800 or 900 services).  (I should note, that you shouldn't make
any assumptions about the location of the PSTN gateway ... But I don't
think that changes anything). =20
</Steve>

<AA>=20
How does a PSTN gateway in another domain from the terminating network
know the service identifier for telephony used in the terminating
network with the Accept-Contact header based proposal?=20
<AA>=20

Moving on from the draft I have seen the questioned raised about "why
can't we expand the power of SDP".  Such an approach would mandate that
SIP methods use SDP, I don't think that is appropriate.=20

[AA] I agree with you here

I hear the appeals that the mechanisms should not be abused, and that we
should not have 1000s of separate "communication services", and I fully
agree with this. I see a smaller number of communication services, not a
larger number; where those communication services are used to access
other end-user-services.

*flame away*

Regards,

Steve

---



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential
information, privileged material (including material protected by the
solicitor-client or other applicable privileges), or constitute
non-public information. Any use of this information by anyone other than
the intended recipient is prohibited. If you have received this
transmission in error, please immediately reply to the sender and delete
this information from your system. Use, dissemination, distribution, or
reproduction of this transmission by unintended recipients is not
authorized and may be unlawful.



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential
information, privileged material (including material protected by the
solicitor-client or other applicable privileges), or constitute
non-public information. Any use of this information by anyone other than
the intended recipient is prohibited. If you have received this
transmission in error, please immediately reply to the sender and delete
this information from your system. Use, dissemination, distribution, or
reproduction of this transmission by unintended recipients is not
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP





ephony services and can include
this in either a Route header to route to the right application server
or can construct a valid Service URI. The real problem is how do users
in other networks know what the service identifier that's is used in the
terminating network is. The use of Service URIs and Presence provides a
solution to that problem.

<Steve>
Again, presence could be a useful mechanism, but I think it would be a
problem for the draft to rely it.  A user making a call to the PSTN may
not find that user in the presence archicture (and similar to services
like the 800 or 900 services).  (I should note, that you shouldn't make
any assumptions about the location of the PSTN gateway ... But I don't
think that changes anything). =20
</Steve>

<AA>=20
How does a PSTN gateway in another domain from the terminating network
know the service identifier for telephony used in the terminating
network with the Accept-Contact header based proposal?=20
<AA>=20

Moving on from the draft I have seen the questioned raised about "why
can't we expand the power of SDP".  Such an approach would mandate that
SIP methods use SDP, I don't think that is appropriate.=20

[AA] I agree with you here

I hear the appeals that the mechanisms should not be abused, and that we
should not have 1000s of separate "communication services", and I fully
agree with this. I see a smaller number of communication services, not a
larger number; where those communication services are used to access
other end-user-services.

*flame away*

Regards,

Steve

---



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential
information, privileged material (including material protected by the
solicitor-client or other applicable privileges), or constitute
non-public information. Any use of this information by anyone other than
the intended recipient is prohibited. If you have received this
transmission in error, please immediately reply to the sender and delete
this information from your system. Use, dissemination, distribution, or
reproduction of this transmission by unintended recipients is not
authorized and may be unlawful.



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential
information, privileged material (including material protected by the
solicitor-client or other applicable privileges), or constitute
non-public information. Any use of this information by anyone other than
the intended recipient is prohibited. If you have received this
transmission in error, please immediately reply to the sender and delete
this information from your system. Use, dissemination, distribution, or
reproduction of this transmission by unintended recipients is not
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP





From sipping-bounces@ietf.org Thu Feb 22 05:07:16 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKAqW-0006Sb-3c; Thu, 22 Feb 2007 05:06:40 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKAqT-0006PX-PK
	for sipping@ietf.org; Thu, 22 Feb 2007 05:06:37 -0500
Received: from sea02-mxf01.citel.com ([205.234.66.116])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKAot-0000j9-So
	for sipping@ietf.org; Thu, 22 Feb 2007 05:05:00 -0500
Received: from [10.8.50.21] (helo=sea02-mxc01.citel.com)
	by sea02-mxf01.citel.com with smtp (Exim 4.43)
	id 1HKAor-0006bx-6T; Thu, 22 Feb 2007 02:04:57 -0800
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification Model 
Date: Thu, 22 Feb 2007 02:04:54 -0800
Message-ID: <76431CABEC5EED489807DB8AEBCCA0BCBA12EB@sea02-mxc01.citel.com>
In-Reply-To: <989763567E51AB48BE015DF7BB960CAD028F09A4@esealmw111>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification Model 
thread-index: AcdV5i0Ma5Tu/sHZEdueygAWy4mm/wAV9tzAAAZvQkAABCBYgA==
References: <61968779B8AC4C4BAB421D4C12F008C009152060@XCH47YKF.rim.net>
	<989763567E51AB48BE015DF7BB960CAD028F09A4@esealmw111>
From: "Steve Langstaff" <steve.langstaff@citel.com>
To: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>,
	<sipping@ietf.org>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: dd7e0c3fd18d19cffdd4de99a114001d
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

If my business card has a FAX number on it, you can be pretty sure that
you can FAX me on it.

If all you have is a generic contact address, you don't know whether you
can FAX me until you try, and possibly fail, to contact me.

So it could be argued, from the "business card" point of view, that a
single AoR is not "preferable", just "different".

> -----Original Message-----
> From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]=20
> Sent: 22 February 2007 08:22
> To: sipping@ietf.org
> Subject: RE: [Sipping] Service Identification Model=20
>=20
> Hi,
>=20
> Isn't relying on a different AoR record for each service a=20
> rather backwards approach?  Today, there have been upto three=20
> numbers on the phones representing different services=20
> (Telephony; fax and Data) - which means business cards which=20
> means different numbers for the different ways I want to=20
> contact a person.  In myunderstanding, that would apply here=20
> as a different AoR record for Telephony; PoC and Messaging. =20
> I think it is much more preferable to have one AoR; and then=20
> to be able to indicate the requested service.
>=20
> Steve.
>=20
> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com]
> Sent: Thursday, February 22, 2007 6:49 AM
> To: Eric Burger; sipping@ietf.org
> Cc: Andrew Allen
> Subject: RE: [Sipping] Service Identification Model=20
>=20
>=20
> Hi Eric,
>=20
> It seems your prefer the approach of using URIs to identify services
> which is part of Jonathan's draft as is the use of Require for those
> services which simply do not interoperate unless the UAS=20
> understands the
> service.
>=20
> Where I am concerned is that you seem to be suggesting that the UA
> separately register an AoR for each service it supports using=20
> REGISTER.
> Certainly for cellular that wouldn't be acceptable as 10 or more
> services could easily be supported by a single device. Registrations
> create a significant load in the IMS network because its not=20
> just a SIP
> REGISTER but also requires a lot of interactions with other network
> elements such as the HSS and Notifications for reg-events etc. Also
> having an order of magnitude increase in multiple=20
> registrations is going
> to be a big issue for the avalanche restart problem for example when a
> cell site goes down and comes back into service.
>=20
> I think Jonathan's approach gives us the possibility for the registrar
> to create the Service URIs based on the feature parameters included in
> the Contact of the REGISTER.=20
>=20
> I think the argument over whether the model is one UA with multiple
> services or multiple UAs each representing a separate service is
> somewhat academic. As I understand a UA can register multiple AoRs and
> can have other AoRs configured or otherwise provided by the service
> provider.=20
>=20
> Andrew
>=20
> -----Original Message-----
> From: Eric Burger [mailto:eburger@bea.com]
> Sent: Wednesday, February 21, 2007 1:29 PM
> To: sipping@ietf.org
> Subject: [Sipping] Service Identification Model=20
>=20
> Folks have been pestering me for a while to look at the Service
> Identification issue.  I finally have gotten around to it.
>=20
>=20
> I guess I have a fundamental problem with the principal=20
> assertion of the
> service-identification draft.  Namely, the assertion that there is a
> single SIP Layer, separate and distinct from two services.  Here is
> Figure 2 from the draft:
>=20
>                            +---------------------------------+
>                            |                                 |
>                            | +-------------+ +-------------+ |
>                            | |      UI     | |      UI     | |
>                            | +-------------+ +-------------+ |
>                            | +-------------+ +-------------+ |
>                            | |             | |             | |
>                            | |  Service 1  | |  Service 2  | |
>                            | |             | |             | |
>                            | +-------------+ +-------------+ |
>                            | +-----------------------------+ |
>                            | |                             | |
>                            | |             SIP             | |
>                            | |            Layer            | |
>                            | |                             | |
>                            | +-----------------------------+ |
>                            |                                 |
>                            +---------------------------------+
>=20
>                                      Physical Device
>=20
> I would offer this drawing confuses a common SIP Stack with a common
> User Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use
> case:
>=20
> > ... In addition users may have
> > multiple devices only one of which has the application that
> understands
> > the game. Other devices that the user has registered against his AoR
> may
> > handle audio for telephony and other voice applications but=20
> if Alice=20
> > invites Bob to play her favourite game it is not very interesting if
> Bob
> > accepts the gaming session on a device that can only accept=20
> the audio=20
> > stream. Also even when the terminal that has the game application=20
> > receives the INVITE the correct application in that=20
> terminal needs to
> be
> > identified to accept the session.
>=20
> I would offer that this is the clinching argument that the request is,
> in fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a
> User Agent.
>=20
> Same argument for PoC versus a regular voice call: One is=20
> looking for a
> PoC User Agent, NOT the PoC service at a given User Agent.
>=20
> Thus, the picture really looks like the following:
>=20
>                            +---------------------------------+
>                            |                                 |
>                            | +-------------+ +-------------+ |
>                            | |    UI       | |    UI       | |
>                            | +-------------+ +-------------+ |
>                            | +-------------+ +-------------+ |
>                            | |             | |             | |
>                            | |    UAS 1    | |    UAS 2    | |
>                            | |             | |             | |
>                            | +-------------+ +-------------+ |
>                            | +-----------------------------+ |
>                            | |                             | |
>                            | |             SIP             | |
>                            | |            Stack            | |
>                            | |                             | |
>                            | +-----------------------------+ |
>                            |                                 |
>                            +---------------------------------+
>=20
>                                      Physical Device
>=20
>=20
> Another argument for different User Agents is the whole=20
> "route based on
> SDP"
> argument.  As the author of RFC 3458, I think I was first to make the
> assertion that knowing the media type did NOT mean you know the
> application to use.  I don't want your phone or proxy or=20
> B2BUA to guess
> that because I have Chess Media, I want the Chess=20
> Application; I want to
> talk with your Chess Application.
>=20
> Now, is the Chess Application a Preference, Requirement, or the thing
> that I am addressing?  The answer drives one towards using=20
> caller prefs,
> the Require header, or the Request-URI.
>=20
> I would offer that it is not a preference: I really want the chess
> application; I probably do not want to talk to you.
>=20
> That leaves us with the Require header or a Request-URI.  My leanings
> are not surprisingly towards the RFC 4240 approach.  Frankly,=20
> if you are
> building an IMS (as an example) that can handle 10,000,000=20
> subscribers,
> you should be able to handle 100,000,000 registrations.  That order of
> magnitude is way easier than the 10,000 to 10,000,000 jump. =20
> I don't buy
> the "registering each service is too much of a burden on the operator"
> argument,
> because it is not that much of a burden, and the network actually does
> the right thing.  Better yet, all of the corner cases work.  Best of
> all, the IETF doesn't have to build an IANA Registry and process a
> thousand PS documents to register urn:service:myco:widget.
>=20
> I suppose I could live with Require:, only if you are way too lazy to
> register each application instance.  It does semantically=20
> work, because
> if I want to do PoC and you do not do PoC, I'll get a 4xx response, to
> which I could try a regular call.
>=20
> ______________________________________________________________
> _________
> Notice:  This email message, together with any attachments,=20
> may contain
> information  of  BEA Systems,  Inc.,  its subsidiaries  and =20
> affiliated
> entities,  that may be confidential,  proprietary, =20
> copyrighted  and/or
> legally privileged, and is intended solely for the use of the=20
> individual
> or entity named in this message. If you are not the intended=20
> recipient,
> and have received this message in error, please immediately=20
> return this
> by email and then delete it.
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use
> sip-implementors@cs.columbia.edu for questions on current sip Use
> sip@ietf.org for new developments of core SIP
>=20
>=20
>=20
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential
> information, privileged material (including material protected by the
> solicitor-client or other applicable privileges), or constitute
> non-public information. Any use of this information by anyone=20
> other than
> the intended recipient is prohibited. If you have received this
> transmission in error, please immediately reply to the sender=20
> and delete
> this information from your system. Use, dissemination,=20
> distribution, or
> reproduction of this transmission by unintended recipients is not
> authorized and may be unlawful.
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use
> sip-implementors@cs.columbia.edu for questions on current sip Use
> sip@ietf.org for new developments of core SIP
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 22 05:54:20 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKBZh-00038R-4Q; Thu, 22 Feb 2007 05:53:21 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKBZg-00038I-CS
	for sipping@ietf.org; Thu, 22 Feb 2007 05:53:20 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKBZY-0007W4-6u
	for sipping@ietf.org; Thu, 22 Feb 2007 05:53:20 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	B597820664
	for <sipping@ietf.org>; Thu, 22 Feb 2007 11:53:11 +0100 (CET)
X-AuditID: c1b4fb3c-aefc6bb0000007de-2d-45dd76177b18 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	97961204CA
	for <sipping@ietf.org>; Thu, 22 Feb 2007 11:53:11 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 11:53:11 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification Model 
Date: Thu, 22 Feb 2007 11:53:10 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD028F0B91@esealmw111>
In-Reply-To: <76431CABEC5EED489807DB8AEBCCA0BCBA12EB@sea02-mxc01.citel.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification Model 
Thread-Index: AcdV5i0Ma5Tu/sHZEdueygAWy4mm/wAV9tzAAAZvQkAABCBYgAABwpng
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 22 Feb 2007 10:53:11.0544 (UTC)
	FILETIME=[A5260780:01C7566F]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8068004c042dabd7f1301bcc80e039df
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi Steve,

I conceed your point with a FAX on a buisiness card as it is often
another (shared) device, however perhaps it may not have been nice if
that practice was carried over to SMS; MMS etc ...=20

BR,

..//Steve

-----Original Message-----
From: Steve Langstaff [mailto:steve.langstaff@citel.com]=20
Sent: Thursday, February 22, 2007 11:05 AM
To: Stephen Terrill (MI/EEM); sipping@ietf.org
Subject: RE: [Sipping] Service Identification Model=20

If my business card has a FAX number on it, you can be pretty sure that
you can FAX me on it.

If all you have is a generic contact address, you don't know whether you
can FAX me until you try, and possibly fail, to contact me.

So it could be argued, from the "business card" point of view, that a
single AoR is not "preferable", just "different".

> -----Original Message-----
> From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
> Sent: 22 February 2007 08:22
> To: sipping@ietf.org
> Subject: RE: [Sipping] Service Identification Model
>=20
> Hi,
>=20
> Isn't relying on a different AoR record for each service a rather=20
> backwards approach?  Today, there have been upto three numbers on the=20
> phones representing different services (Telephony; fax and Data) -=20
> which means business cards which means different numbers for the=20
> different ways I want to contact a person.  In myunderstanding, that=20
> would apply here as a different AoR record for Telephony; PoC and=20
> Messaging.
> I think it is much more preferable to have one AoR; and then to be=20
> able to indicate the requested service.
>=20
> Steve.
>=20
> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com]
> Sent: Thursday, February 22, 2007 6:49 AM
> To: Eric Burger; sipping@ietf.org
> Cc: Andrew Allen
> Subject: RE: [Sipping] Service Identification Model
>=20
>=20
> Hi Eric,
>=20
> It seems your prefer the approach of using URIs to identify services=20
> which is part of Jonathan's draft as is the use of Require for those=20
> services which simply do not interoperate unless the UAS understands=20
> the service.
>=20
> Where I am concerned is that you seem to be suggesting that the UA=20
> separately register an AoR for each service it supports using=20
> REGISTER.
> Certainly for cellular that wouldn't be acceptable as 10 or more=20
> services could easily be supported by a single device. Registrations=20
> create a significant load in the IMS network because its not just a=20
> SIP REGISTER but also requires a lot of interactions with other=20
> network elements such as the HSS and Notifications for reg-events etc.

> Also having an order of magnitude increase in multiple registrations=20
> is going to be a big issue for the avalanche restart problem for=20
> example when a cell site goes down and comes back into service.
>=20
> I think Jonathan's approach gives us the possibility for the registrar

> to create the Service URIs based on the feature parameters included in

> the Contact of the REGISTER.
>=20
> I think the argument over whether the model is one UA with multiple=20
> services or multiple UAs each representing a separate service is=20
> somewhat academic. As I understand a UA can register multiple AoRs and

> can have other AoRs configured or otherwise provided by the service=20
> provider.
>=20
> Andrew
>=20
> -----Original Message-----
> From: Eric Burger [mailto:eburger@bea.com]
> Sent: Wednesday, February 21, 2007 1:29 PM
> To: sipping@ietf.org
> Subject: [Sipping] Service Identification Model
>=20
> Folks have been pestering me for a while to look at the Service=20
> Identification issue.  I finally have gotten around to it.
>=20
>=20
> I guess I have a fundamental problem with the principal assertion of=20
> the service-identification draft.  Namely, the assertion that there is

> a single SIP Layer, separate and distinct from two services.  Here is=20
> Figure 2 from the draft:
>=20
>                            +---------------------------------+
>                            |                                 |
>                            | +-------------+ +-------------+ |
>                            | |      UI     | |      UI     | |
>                            | +-------------+ +-------------+ |
>                            | +-------------+ +-------------+ |
>                            | |             | |             | |
>                            | |  Service 1  | |  Service 2  | |
>                            | |             | |             | |
>                            | +-------------+ +-------------+ |
>                            | +-----------------------------+ |
>                            | |                             | |
>                            | |             SIP             | |
>                            | |            Layer            | |
>                            | |                             | |
>                            | +-----------------------------+ |
>                            |                                 |
>                            +---------------------------------+
>=20
>                                      Physical Device
>=20
> I would offer this drawing confuses a common SIP Stack with a common=20
> User Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use
> case:
>=20
> > ... In addition users may have
> > multiple devices only one of which has the application that
> understands
> > the game. Other devices that the user has registered against his AoR
> may
> > handle audio for telephony and other voice applications but
> if Alice
> > invites Bob to play her favourite game it is not very interesting if
> Bob
> > accepts the gaming session on a device that can only accept
> the audio
> > stream. Also even when the terminal that has the game application=20
> > receives the INVITE the correct application in that
> terminal needs to
> be
> > identified to accept the session.
>=20
> I would offer that this is the clinching argument that the request is,

> in fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a=20
> User Agent.
>=20
> Same argument for PoC versus a regular voice call: One is looking for=20
> a PoC User Agent, NOT the PoC service at a given User Agent.
>=20
> Thus, the picture really looks like the following:
>=20
>                            +---------------------------------+
>                            |                                 |
>                            | +-------------+ +-------------+ |
>                            | |    UI       | |    UI       | |
>                            | +-------------+ +-------------+ |
>                            | +-------------+ +-------------+ |
>                            | |             | |             | |
>                            | |    UAS 1    | |    UAS 2    | |
>                            | |             | |             | |
>                            | +-------------+ +-------------+ |
>                            | +-----------------------------+ |
>                            | |                             | |
>                            | |             SIP             | |
>                            | |            Stack            | |
>                            | |                             | |
>                            | +-----------------------------+ |
>                            |                                 |
>                            +---------------------------------+
>=20
>                                      Physical Device
>=20
>=20
> Another argument for different User Agents is the whole "route based=20
> on SDP"
> argument.  As the author of RFC 3458, I think I was first to make the=20
> assertion that knowing the media type did NOT mean you know the=20
> application to use.  I don't want your phone or proxy or B2BUA to=20
> guess that because I have Chess Media, I want the Chess Application; I

> want to talk with your Chess Application.
>=20
> Now, is the Chess Application a Preference, Requirement, or the thing=20
> that I am addressing?  The answer drives one towards using caller=20
> prefs, the Require header, or the Request-URI.
>=20
> I would offer that it is not a preference: I really want the chess=20
> application; I probably do not want to talk to you.
>=20
> That leaves us with the Require header or a Request-URI.  My leanings=20
> are not surprisingly towards the RFC 4240 approach.  Frankly, if you=20
> are building an IMS (as an example) that can handle 10,000,000=20
> subscribers, you should be able to handle 100,000,000 registrations. =20
> That order of magnitude is way easier than the 10,000 to 10,000,000=20
> jump.
> I don't buy
> the "registering each service is too much of a burden on the operator"
> argument,
> because it is not that much of a burden, and the network actually does

> the right thing.  Better yet, all of the corner cases work.  Best of=20
> all, the IETF doesn't have to build an IANA Registry and process a=20
> thousand PS documents to register urn:service:myco:widget.
>=20
> I suppose I could live with Require:, only if you are way too lazy to=20
> register each application instance.  It does semantically work,=20
> because if I want to do PoC and you do not do PoC, I'll get a 4xx=20
> response, to which I could try a regular call.
>=20
> ______________________________________________________________
> _________
> Notice:  This email message, together with any attachments, may=20
> contain information  of  BEA Systems,  Inc.,  its subsidiaries  and=20
> affiliated entities,  that may be confidential,  proprietary,=20
> copyrighted  and/or legally privileged, and is intended solely for the

> use of the individual or entity named in this message. If you are not=20
> the intended recipient, and have received this message in error,=20
> please immediately return this by email and then delete it.
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20
>=20
>=20
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential

> information, privileged material (including material protected by the=20
> solicitor-client or other applicable privileges), or constitute=20
> non-public information. Any use of this information by anyone other=20
> than the intended recipient is prohibited. If you have received this=20
> transmission in error, please immediately reply to the sender and=20
> delete this information from your system. Use, dissemination,=20
> distribution, or reproduction of this transmission by unintended=20
> recipients is not authorized and may be unlawful.
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 22 08:23:09 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKDt2-0008GT-5N; Thu, 22 Feb 2007 08:21:28 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKDt1-0008GE-61
	for sipping@ietf.org; Thu, 22 Feb 2007 08:21:27 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKDsu-0003N9-OR
	for sipping@ietf.org; Thu, 22 Feb 2007 08:21:27 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	EBE63206CA; Thu, 22 Feb 2007 14:21:13 +0100 (CET)
X-AuditID: c1b4fb3c-b0fcabb0000007de-c3-45dd98c9a6ad 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	DD0DF20417; Thu, 22 Feb 2007 14:21:13 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 14:21:13 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP - Routing decissions
	based on Require/P-R
Date: Thu, 22 Feb 2007 14:21:08 +0100
Message-ID: <7374777208BDC7449D5620EF942325670236219E@esealmw113.eemea.ericsson.se>
In-Reply-To: <6E0BD8DB-DF70-4CBE-AA8F-DB69CECF014E@cs.columbia.edu>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP - Routing decissions
	based on Require/P-R
Thread-Index: AcdULy07S0PScH0GRdSsZffhcF9RPACVQdIg
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Henning Schulzrinne" <hgs@cs.columbia.edu>
X-OriginalArrivalTime: 22 Feb 2007 13:21:13.0459 (UTC)
	FILETIME=[532EB430:01C75684]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 69a74e02bbee44ab4f8eafdbcedd94a1
Cc: sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi Henning,=20

>RFC 3840, Section 10.15 ("Extensions") would be the mechanism to route
according to capabilities.

I am not sure how that answers my question. Could you please clarify?

Regards,

Christer



> >
> > Hi Jonathan,
> >
> >> If you've got a situration where a SIP call *cannot be processed*
> > without
> >> going through application servers, then this is not=20
> something that is=20
> >> signaled with Accept-Contact. Its signaled with Require or=20
> >> Proxy-Require as appropriate.
> >
> > Just for my clarification, where is it said that Require and=20
> > Proxy-Require are used, or at least more appropriate, for routing?
> >
> > As far as I know, they are only used to require capabilities by the=20
> > nodes that receive a SIP request (as you said yourself=20
> earlier) - but=20
> > not to make routing decissions based upon.
> >
> > Regards,
> >
> > Christer
> >
>=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 22 08:34:29 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKE5N-00071w-OI; Thu, 22 Feb 2007 08:34:13 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKE5M-00071M-N3
	for sipping@ietf.org; Thu, 22 Feb 2007 08:34:12 -0500
Received: from brinza.cc.columbia.edu ([128.59.29.8])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKE5K-0005Qa-D1
	for sipping@ietf.org; Thu, 22 Feb 2007 08:34:12 -0500
Received: from [128.59.23.102] (macmini1.cs.columbia.edu [128.59.23.102])
	(user=hgs10 mech=PLAIN bits=0)
	by brinza.cc.columbia.edu (8.13.7/8.13.6) with ESMTP id l1MDXrG8003556
	(version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NOT);
	Thu, 22 Feb 2007 08:33:57 -0500 (EST)
In-Reply-To: <7374777208BDC7449D5620EF942325670236219E@esealmw113.eemea.ericsson.se>
References: <7374777208BDC7449D5620EF942325670236219E@esealmw113.eemea.ericsson.se>
Mime-Version: 1.0 (Apple Message framework v752.3)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <2C7E0232-0EF2-45E7-8822-8ACED11E671D@cs.columbia.edu>
Content-Transfer-Encoding: 7bit
From: Henning Schulzrinne <hgs@cs.columbia.edu>
Subject: Re: [Sipping] Service Identification in SIP - Routing decissions
	based on Require/P-R
Date: Thu, 22 Feb 2007 08:33:51 -0500
To: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
X-Mailer: Apple Mail (2.752.3)
X-No-Spam-Score: Local
X-Scanned-By: MIMEDefang 2.48 on 128.59.29.8
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8b431ad66d60be2d47c7bfeb879db82c
Cc: sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Proxies can choose next hops based on RFC 3840, which is another way  
of saying "routing".

On Feb 22, 2007, at 8:21 AM, Christer Holmberg ((JO/LMF)) wrote:

>
> Hi Henning,
>
>> RFC 3840, Section 10.15 ("Extensions") would be the mechanism to  
>> route
> according to capabilities.
>
> I am not sure how that answers my question. Could you please clarify?
>
> Regards,
>
> Christer
>
>
>
>>>
>>> Hi Jonathan,
>>>
>>>> If you've got a situration where a SIP call *cannot be processed*
>>> without
>>>> going through application servers, then this is not
>> something that is
>>>> signaled with Accept-Contact. Its signaled with Require or
>>>> Proxy-Require as appropriate.
>>>
>>> Just for my clarification, where is it said that Require and
>>> Proxy-Require are used, or at least more appropriate, for routing?
>>>
>>> As far as I know, they are only used to require capabilities by the
>>> nodes that receive a SIP request (as you said yourself
>> earlier) - but
>>> not to make routing decissions based upon.
>>>
>>> Regards,
>>>
>>> Christer
>>>
>>
>>
>
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 22 10:06:49 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKFVx-00036E-8F; Thu, 22 Feb 2007 10:05:45 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKFVw-000369-4k
	for sipping@ietf.org; Thu, 22 Feb 2007 10:05:44 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKFVp-0005Hv-78
	for sipping@ietf.org; Thu, 22 Feb 2007 10:05:44 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	93EA52101E; Thu, 22 Feb 2007 16:05:30 +0100 (CET)
X-AuditID: c1b4fb3e-b0ed5bb0000007e1-3e-45ddb13a789d 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	83150200E8; Thu, 22 Feb 2007 16:05:30 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 16:05:30 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP - Routing decissionsbased
	on Require/P-R
Date: Thu, 22 Feb 2007 16:05:27 +0100
Message-ID: <7374777208BDC7449D5620EF942325670236275D@esealmw113.eemea.ericsson.se>
In-Reply-To: <200702201901.l1KJ1Q6G022448@dragon.ariadne.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP - Routing
	decissionsbased on Require/P-R
Thread-Index: AcdVIZeT9n2OOnN8Syy9FshFBVU1UwBbxNTg
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: <Dale.Worley@comcast.net>,
	<sipping@ietf.org>
X-OriginalArrivalTime: 22 Feb 2007 15:05:30.0442 (UTC)
	FILETIME=[E4A28AA0:01C75692]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: e1e48a527f609d1be2bc8d8a70eb76cb
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,

I am sure a proxy can make routing decissions based on whatever.

My question was simply where it is said that Require/Proxy-Require is
more appropriate.

Regards,

Christer=20

> -----Original Message-----
> From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net]=20
> Sent: 20. helmikuuta 2007 21:01
> To: sipping@ietf.org
> Subject: Re: [Sipping] Service Identification in SIP -=20
> Routing decissionsbased on Require/P-R
>=20
>    From: "Christer Holmberg \(JO/LMF\)"=20
> <christer.holmberg@ericsson.com>
>=20
>    Just for my clarification, where is it said that Require and
>    Proxy-Require are used, or at least more appropriate, for routing?
>=20
>    As far as I know, they are only used to require capabilities by the
>    nodes that receive a SIP request (as you said yourself=20
> earlier) - but
>    not to make routing decissions based upon.
>=20
> Surely a proxy can make routing decisions based upon Require=20
> if it desires to.
>=20
> Dale
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 22 10:10:55 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKFas-0004Im-9o; Thu, 22 Feb 2007 10:10:50 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKFap-0004Ie-E0
	for sipping@ietf.org; Thu, 22 Feb 2007 10:10:47 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKFan-0006Fr-VB
	for sipping@ietf.org; Thu, 22 Feb 2007 10:10:47 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	770FD204F7; Thu, 22 Feb 2007 16:10:45 +0100 (CET)
X-AuditID: c1b4fb3c-b0fcabb0000007de-bc-45ddb275fb7a 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	6380C20400; Thu, 22 Feb 2007 16:10:45 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 16:10:44 +0100
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP - Routing decissions
	based on Require/P-R
Date: Thu, 22 Feb 2007 16:10:42 +0100
Message-ID: <7374777208BDC7449D5620EF942325670236279E@esealmw113.eemea.ericsson.se>
In-Reply-To: <2C7E0232-0EF2-45E7-8822-8ACED11E671D@cs.columbia.edu>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP - Routing decissions
	based on Require/P-R
Thread-Index: AcdWhifnYyCF5WoISbiRcxTn8UWMcAADVQaw
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Henning Schulzrinne" <hgs@cs.columbia.edu>
X-OriginalArrivalTime: 22 Feb 2007 15:10:44.0959 (UTC)
	FILETIME=[A01A06F0:01C75693]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: b4a0a5f5992e2a4954405484e7717d8c
Cc: sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,=20

>Proxies can choose next hops based on RFC 3840, which is another way of
saying "routing".

Yes. But, where in RFC 3840 is it mentioned that the routing is based on
the Require/Proxy-Require header?

Regards,

Christer


>=20
> On Feb 22, 2007, at 8:21 AM, Christer Holmberg ((JO/LMF)) wrote:
>=20
> >
> > Hi Henning,
> >
> >> RFC 3840, Section 10.15 ("Extensions") would be the mechanism to=20
> >> route
> > according to capabilities.
> >
> > I am not sure how that answers my question. Could you=20
> please clarify?
> >
> > Regards,
> >
> > Christer
> >
> >
> >
> >>>
> >>> Hi Jonathan,
> >>>
> >>>> If you've got a situration where a SIP call *cannot be processed*
> >>> without
> >>>> going through application servers, then this is not
> >> something that is
> >>>> signaled with Accept-Contact. Its signaled with Require or=20
> >>>> Proxy-Require as appropriate.
> >>>
> >>> Just for my clarification, where is it said that Require and=20
> >>> Proxy-Require are used, or at least more appropriate, for routing?
> >>>
> >>> As far as I know, they are only used to require=20
> capabilities by the=20
> >>> nodes that receive a SIP request (as you said yourself
> >> earlier) - but
> >>> not to make routing decissions based upon.
> >>>
> >>> Regards,
> >>>
> >>> Christer
> >>>
> >>
> >>
> >
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP Use=20
> > sip-implementors@cs.columbia.edu for questions on current sip Use=20
> > sip@ietf.org for new developments of core SIP
>=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 22 10:20:14 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKFjp-0003Et-E6; Thu, 22 Feb 2007 10:20:05 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKFjo-0003Ee-F2
	for sipping@ietf.org; Thu, 22 Feb 2007 10:20:04 -0500
Received: from alnrmhc12.comcast.net ([206.18.177.52])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKFjn-0007k1-8N
	for sipping@ietf.org; Thu, 22 Feb 2007 10:20:04 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (alnrmhc12) with ESMTP
	id <20070222152002b12009dau8e>; Thu, 22 Feb 2007 15:20:02 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1MFK16g022386
	for <sipping@ietf.org>; Thu, 22 Feb 2007 10:20:01 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1MFK1F9022382;
	Thu, 22 Feb 2007 10:20:01 -0500
Date: Thu, 22 Feb 2007 10:20:01 -0500
Message-Id: <200702221520.l1MFK1F9022382@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <7374777208BDC7449D5620EF942325670236275D@esealmw113.eemea.ericsson.se>
	(christer.holmberg@ericsson.com)
Subject: Re: [Sipping] Service Identification in SIP - Routing decissionsbased
	on Require/P-R
References: <7374777208BDC7449D5620EF942325670236275D@esealmw113.eemea.ericsson.se>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 7bac9cb154eb5790ae3b2913587a40de
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>

   I am sure a proxy can make routing decissions based on whatever.

   My question was simply where it is said that Require/Proxy-Require is
   more appropriate.

More appropriate than what?

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 22 10:36:02 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKFyO-0006Pf-LG; Thu, 22 Feb 2007 10:35:08 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKFyM-0006Of-SJ
	for sipping@ietf.org; Thu, 22 Feb 2007 10:35:06 -0500
Received: from alnrmhc13.comcast.net ([206.18.177.53])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKFyL-0003Ty-H4
	for sipping@ietf.org; Thu, 22 Feb 2007 10:35:06 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (alnrmhc13) with ESMTP
	id <20070222153458b1300e7p9ke>; Thu, 22 Feb 2007 15:34:58 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1MFYw6g023477
	for <sipping@ietf.org>; Thu, 22 Feb 2007 10:34:58 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1MFYwbG023473;
	Thu, 22 Feb 2007 10:34:58 -0500
Date: Thu, 22 Feb 2007 10:34:58 -0500
Message-Id: <200702221534.l1MFYwbG023473@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <2C7E0232-0EF2-45E7-8822-8ACED11E671D@cs.columbia.edu>
	(hgs@cs.columbia.edu)
Subject: Re: [Sipping] Service Identification in SIP - Routing decissions
	based on Require/P-R
References: <7374777208BDC7449D5620EF942325670236219E@esealmw113.eemea.ericsson.se>
	<2C7E0232-0EF2-45E7-8822-8ACED11E671D@cs.columbia.edu>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: d6b246023072368de71562c0ab503126
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   From: Henning Schulzrinne <hgs@cs.columbia.edu>

   Proxies can choose next hops based on RFC 3840, which is another way  
   of saying "routing".

RFC 3261 allows a proxy unlimited latitude in choosing targets to
which to forward a request, if the proxy is authoritative for the
request-URI:

"[T]he element is responsible for the domain in the Request-URI, and
the element MAY use whatever mechanism it desires to determine where
to send the request."

Within that framework, it seems to me quite reasonable for a service
provider's proxy to examine Require or Proxy-Require headers to
determine which application server would be best for handling a call,
and routing the call to it.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 22 10:53:46 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKGG7-0006JB-2w; Thu, 22 Feb 2007 10:53:27 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKGG5-0006GT-Ns
	for sipping@ietf.org; Thu, 22 Feb 2007 10:53:25 -0500
Received: from sccrmhc12.comcast.net ([204.127.200.82])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKGG4-0006R7-Ht
	for sipping@ietf.org; Thu, 22 Feb 2007 10:53:25 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (sccrmhc12) with ESMTP
	id <2007022215531801200cgfj7e>; Thu, 22 Feb 2007 15:53:18 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1MFrE6g024841
	for <sipping@ietf.org>; Thu, 22 Feb 2007 10:53:14 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1MFrEkU024837;
	Thu, 22 Feb 2007 10:53:14 -0500
Date: Thu, 22 Feb 2007 10:53:14 -0500
Message-Id: <200702221553.l1MFrEkU024837@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <6FC4416DDE56C44DA0AEE67BC7CA437112DE3701@zrc2hxm2.corp.nortel.com>
	(bstucker@nortel.com)
Subject: Re: [Sipping] Service Identification in SIP
References: <6FC4416DDE56C44DA0AEE67BC7CA437112DE3701@zrc2hxm2.corp.nortel.com>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 93238566e09e6e262849b4f805833007
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   From: "Brian Stucker" <bstucker@nortel.com>

   Since we not only must agree upon a technical solution, but find it
   desirable to convince 3GPP and others to use the solution
   recommended by the IETF; getting agreement on producing a stable,
   referencable document within the IETF within the 3GPP R7 timeline
   is critical. Otherwise, R7 will go out the door with what it has
   today for service identification, and we'll wind up with an even
   bigger mess on our hands.

That is a problem.

But on the other hand, OSI and SS7 didn't have any IETF input, and
that hasn't delayed the growth of the Internet.

And it seems the cellular industry devised some sort of PoC-using-SIP
without worrying about upward-compatibility with pre-existing SIP.
They can reasonably complain that the IETF wouldn't or couldn't get
PoC standardized in time for their business needs, but in the face of
the fact that they designed a solution that violates the fundamental
principles of good protocol design, they can't complain that the IETF
isn't interested in it.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 22 11:14:22 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKGYz-0003ZF-CT; Thu, 22 Feb 2007 11:12:57 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKGYx-0003Yw-IR
	for sipping@ietf.org; Thu, 22 Feb 2007 11:12:55 -0500
Received: from sj-iport-1-in.cisco.com ([171.71.176.70]
	helo=sj-iport-1.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HKGYw-0001XG-4t
	for sipping@ietf.org; Thu, 22 Feb 2007 11:12:55 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by sj-iport-1.cisco.com with ESMTP; 22 Feb 2007 08:12:53 -0800
X-IronPort-AV: i="4.14,207,1170662400"; 
	d="scan'208"; a="764056524:sNHT48333196"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1MGCrFf015699; 
	Thu, 22 Feb 2007 11:12:53 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1MGCmOI012277; 
	Thu, 22 Feb 2007 11:12:52 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 11:12:40 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 11:12:40 -0500
Message-ID: <45DDC0F7.2040607@cisco.com>
Date: Thu, 22 Feb 2007 11:12:39 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Christer Holmberg (JO/LMF)" <christer.holmberg@ericsson.com>
Subject: Re: [Sipping] Service Identification in SIP - Routing decissions
	based on Require/P-R
References: <7374777208BDC7449D5620EF942325670236279E@esealmw113.eemea.ericsson.se>
In-Reply-To: <7374777208BDC7449D5620EF942325670236279E@esealmw113.eemea.ericsson.se>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 22 Feb 2007 16:12:40.0032 (UTC)
	FILETIME=[46752E00:01C7569C]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=3078; t=1172160773;
	x=1173024773; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP=20-=2
	0Routing=20decissions=0A=20based=20on=20Require/P-R |Sender:=20
	|To:=20=22Christer=20Holmberg=20(JO/LMF)=22=20<christer.holmberg@ericsson
	.com>; bh=krWPUoQlaUco/1T5X+gGIAMTR+CgbjcfxqnlRW5myhs=;
	b=ASu9WC5BkQo4k65NxdVePI7RxBSx6by0LAJCzlLw8471a3jWXUE6BImDgufRfk1n9HawgN++
	WPZBysMK8ivDH8qws2D1ZqfTRv+tLwBJqy2Bh1pKtXJfxEtPD/BH5eht;
Authentication-Results: rtp-dkim-1; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 10ba05e7e8a9aa6adb025f426bef3a30
Cc: sipping <sipping@ietf.org>, Henning Schulzrinne <hgs@cs.columbia.edu>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I think two different things have been confused in this discussion.

1) AFAIK, *any* proxy is free to look at the Require header, and
    use that to choose a next hop proxy.

    (A proxy that is responsible for the domain of the R-URI *could*
    use such logic to choose a translation for the R-URI, though I
    think that would be an unusual thing to do.)

2) RFC 3840 defines the "extensions" feature tag that allows a
    UAS to declare a capability for an extension when registering.
    If this is done, it in effect says: "a request containing
    a Require header with this extension will be accepted if sent
    to this contact address."

    Requests addressed to an AOR can express a preference for a
    UAS that has indicated a capability for an extension. Or it can
    reject contacts that haven't indicated that capability. And of
    course the request could also contain a Require for the same
    capability. This can be used to avoid delivering a request to
    a UAS that doesn't support a required extension.

	Paul

Christer Holmberg (JO/LMF) wrote:
> Hi, 
> 
>> Proxies can choose next hops based on RFC 3840, which is another way of
> saying "routing".
> 
> Yes. But, where in RFC 3840 is it mentioned that the routing is based on
> the Require/Proxy-Require header?
> 
> Regards,
> 
> Christer
> 
> 
>> On Feb 22, 2007, at 8:21 AM, Christer Holmberg ((JO/LMF)) wrote:
>>
>>> Hi Henning,
>>>
>>>> RFC 3840, Section 10.15 ("Extensions") would be the mechanism to 
>>>> route
>>> according to capabilities.
>>>
>>> I am not sure how that answers my question. Could you 
>> please clarify?
>>> Regards,
>>>
>>> Christer
>>>
>>>
>>>
>>>>> Hi Jonathan,
>>>>>
>>>>>> If you've got a situration where a SIP call *cannot be processed*
>>>>> without
>>>>>> going through application servers, then this is not
>>>> something that is
>>>>>> signaled with Accept-Contact. Its signaled with Require or 
>>>>>> Proxy-Require as appropriate.
>>>>> Just for my clarification, where is it said that Require and 
>>>>> Proxy-Require are used, or at least more appropriate, for routing?
>>>>>
>>>>> As far as I know, they are only used to require 
>> capabilities by the 
>>>>> nodes that receive a SIP request (as you said yourself
>>>> earlier) - but
>>>>> not to make routing decissions based upon.
>>>>>
>>>>> Regards,
>>>>>
>>>>> Christer
>>>>>
>>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP Use 
>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>> sip@ietf.org for new developments of core SIP
>>
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 22 17:46:05 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKMf5-0003ze-KV; Thu, 22 Feb 2007 17:43:39 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKMf4-0003zY-A5
	for sipping@ietf.org; Thu, 22 Feb 2007 17:43:38 -0500
Received: from sj-iport-3-in.cisco.com ([171.71.176.72]
	helo=sj-iport-3.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HKMf2-0001aG-7o
	for sipping@ietf.org; Thu, 22 Feb 2007 17:43:38 -0500
Received: from sj-dkim-1.cisco.com ([171.71.179.21])
	by sj-iport-3.cisco.com with ESMTP; 22 Feb 2007 14:43:37 -0800
X-IronPort-AV: i="4.14,207,1170662400"; 
	d="scan'208"; a="466478863:sNHT83545044"
Received: from sj-core-2.cisco.com (sj-core-2.cisco.com [171.71.177.254])
	by sj-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1MMhZrS004285; 
	Thu, 22 Feb 2007 14:43:35 -0800
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by sj-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1MMhSGo015290;
	Thu, 22 Feb 2007 14:43:35 -0800 (PST)
Received: from xmb-rtp-20b.amer.cisco.com ([64.102.31.53]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 17:43:28 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Thu, 22 Feb 2007 17:43:27 -0500
Message-ID: <072C5B76F7CEAB488172C6F64B30B5E302A2F204@xmb-rtp-20b.amer.cisco.com>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C009152056@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdNmH89O4T46iGoRjicZ3hYknVpawB8E6oQAH++leAABzLk8AAYIixwAIPObeAAhKg1EAAq3LHw
References: <61968779B8AC4C4BAB421D4C12F008C009152056@XCH47YKF.rim.net>
From: "Michael Hammer \(mhammer\)" <mhammer@cisco.com>
To: "Andrew Allen" <aallen@rim.com>,
	"Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>,
	<sipping@ietf.org>
X-OriginalArrivalTime: 22 Feb 2007 22:43:28.0327 (UTC)
	FILETIME=[DEBB0D70:01C756D2]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=22118; t=1172184215;
	x=1173048215; c=relaxed/simple; s=sjdkim1004;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=mhammer@cisco.com;
	z=From:=20=22Michael=20Hammer=20\(mhammer\)=22=20<mhammer@cisco.com>
	|Subject:=20RE=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20; bh=Y66az4FqviF2xfnIqr+z7KLrd66xgARv4DEGkLsI/g0=;
	b=MSZq7MUcpLnpMRVuaup6i6lL/EAMD++hoUSrAdCREPH/pCFGLhUqN2zi+/sk9KMNZeJhS2vu
	dlnWIgE/Ja1CiRAGFJC48CSy4UC5MUUXCpPqjqtkVIT8dxLJu5VAKAW7boiCm/Eba4Lxlu3X8D
	jBxOYM3tEI83s5MrRVW7NaFTo=;
Authentication-Results: sj-dkim-1; header.From=mhammer@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim1004 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 0d6b5666eba887052ef5e87a9de0d3b8
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I would hope that if they are hell-bent on doing this, that they at
least also define a service identifier that means "IETF =3D service
intentionally left unspecified", but not in so many characters.  Then
make sure that compatibility is specified when such an identifier is
encountered, such as when an INVITE were to come from a non-3GPP system
with no service identity provided.

Mike


> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com]=20
> Sent: Wednesday, February 21, 2007 10:05 PM
> To: Stephen Terrill (MI/EEM); sipping@ietf.org
> Cc: Andrew Allen
> Subject: RE: [Sipping] Service Identification in SIP
>=20
>=20
> Hi Steve
>=20
> In my view this is one of the interoperability issues with=20
> your proposal. I think the assumption that all networks will=20
> deploy exactly the same services with the same service=20
> identifiers is unrealistic and also undesirable. Are=20
> corporate PBX vendors going to implement the same services as=20
> defined by 3GPP or OMA? The possibility to have the=20
> flexibility to have different service identifiers for session=20
> equivalent services is in my view essential to any IETF=20
> standard and to do otherwise is to create a situation where=20
> basic services like telephony or messaging cannot=20
> interoperate between users in different networks and users in=20
> the internet without the use of inter working gateway. SIP=20
> and SIMPLE were setup specifically to create protocols that=20
> would enable interoperability with IP networks without the=20
> need for gateways.=20
>=20
> If 3GPP network operators choose to deploy the same services=20
> and use common identifiers that is fine but there are whole=20
> other worlds out there as well and service identifiers should=20
> not stand in the way of interoperability between all these=20
> different SIP users. If the IETF solution supports the=20
> ability for different networks to use different identifiers=20
> that doesn't violate 3GPP requirements and doesn't prevent=20
> 3GPP using the same identifiers in different networks. I=20
> don't see how that is against TS 23.228.
>=20
> Andrew
>=20
>=20
> -----Original Message-----
> From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
> Sent: Monday, February 19, 2007 6:00 AM
> To: Andrew Allen; sipping@ietf.org
> Subject: RE: [Sipping] Service Identification in SIP
>=20
> Hi Andrew,
>=20
> I see that in your assumptions, the originating and=20
> terminating networks use different service identifiers, and I=20
> presume that this is for the same service.  In my=20
> understanding, it is the same service identifier that is used=20
> end-to-end (if not it is interworked).  That is why I do not=20
> see the reliance on Presence being necessary.  In my opionion=20
> this totally within the requirements of TS 23.228 (I think=20
> that different service identifiers in different networks for=20
> the same service is against TS 23.228).
>=20
> This comes to the isssue of the PSTN gateway - I presume that=20
> you want to go to the PSTN gateway if the service is=20
> telephony, and within that, voice.  If we know that, then the=20
> we go to any PSTN gateway, and we can't presume a particular=20
> network for a PSTN gateway.=20
>=20
> Best Regards,
>=20
> Steve.
>=20
> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com]
> Sent: Saturday, February 17, 2007 12:51 AM
> To: Stephen Terrill (MI/EEM); sipping@ietf.org
> Cc: Andrew Allen
> Subject: RE: [Sipping] Service Identification in SIP
>=20
>=20
>=20
> Hi Steve,
>=20
> My responses inline
>=20
> Andrew=20
>=20
> -----Original Message-----
> From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
> Sent: Friday, February 16, 2007 3:51 AM
> To: sipping@ietf.org
> Subject: RE: [Sipping] Service Identification in SIP
>=20
> Hello Andrew,
>=20
> Thanks for your comments.  I have a couple of brief responses below.
>=20
> Best Regards,
>=20
> ..//Steve=20
>=20
> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com]
> Sent: Friday, February 16, 2007 7:15 AM
> To: Stephen Terrill (MI/EEM); sipping@ietf.org
> Cc: Andrew Allen
> Subject: RE: [Sipping] Service Identification in SIP
>=20
>=20
> Some comments on Stephen's comments on the draft:
> =20
>=20
> -Moving onto the proposed solution draft itself, in addition for
> providing for the identification of the "application" and the
> "communication service", I have a few additional thoughts.
>=20
> The draft seems to have the implication that all services=20
> (communication
> services in other terminology) that are to be identified have to be
> standardised within the IETF.  This is clearly an restriction=20
> that will
> not be acceptable to the industry as a whole.  My guess is=20
> that if such
> a statement is mandated by the IETF, it will just be words on paper
> which would likely be ignored in practice by other standardisation
> organisations.
>=20
> [AA] This I think is an incorrect interpretation of the draft. My
> understanding is that the only services according to the draft that
> require some standardisation in IETF are those that by their=20
> very nature
> cannot interoperate unless there is some common additional=20
> understanding
> between the endpoints outside of what is currently specified=20
> in SIP. The
> use of SIP option tags with Require header is proposed by the=20
> draft for
> this purpose as such services create interoperability issues with SIP.
>=20
> <Steve>
> My comment applies even for these services </Steve>
>=20
>=20
> <AA>
> Well then we are in big trouble because what you are proposing is that
> its OK for bodies outside of IETF to create non interoperable=20
> extensions
> for the behavour of SIP entities. This will result in the=20
> fragmentation
> of SIP into different flavors that don't interoperate. This is the
> balkanization of SIP that Paul talks about.even <AA>=20
>=20
> ----------------------------------------------------
>=20
> The draft also only recognises the case where all explicitly knowledge
> of the service is required.  I think it would also be good to=20
> allow the
> case where knowledge of the service would be helpful, but not=20
> explicitly
> required.  This could be a case where the service to the=20
> end-user would
> not be degraded if this was not supported.  This=20
> functionality seems to
> be missing.
>=20
> [AA] Service URIs are defined in the draft for these services=20
> where the
> support by the other endpoint is not required for the session to be
> sucessfully established.
>=20
> The draft seems to indicate that that application servers in=20
> the network
> are involved only at the request of the originating terminal.  This
> seems a strange restriction.  For some network support that=20
> supports the
> originating user, such as the one provided in the example "I=20
> would like
> this call recorded for later retrieval", such mechanisms may=20
> make sense,
> however as a general approach it seems a bit limited.  For example, in
> 3GPP networks there is the need to invoke application logic in the
> originating side and terminating side, not only at the request of the
> originating terminal, but as a consequence of the requested=20
> service.  If
> there is an SIP message contains indication of the requested=20
> service, a
> proxy can do this (or loose route to an AS that contains the necessary
> service logic).=20
>=20
> [AA] The draft supports both possibilities using the Route header for
> originating network servers that need to be involved in the route and
> based on the terminating networks routing policy for the=20
> Service URI in
> the Request URI.
>=20
> <Steve>
> I don't see the advantage of having different solutions for the
> originating network and the terminating network.  I don't=20
> think that the
> originating logic is invoked only at the request of the originating
> terminal.
> </Steve>
>=20
> <AA>
> Unless IETF is willing to standardise all services and=20
> mandate that all
> SIP entities understand these (which will never happen) then we cannot
> assume that all networks and all endpoints understand and use the same
> identifiers. It is possible through configuration for an endpoint to
> obtain the identifiers used by its own service provider but=20
> there is no
> way to ensure that other networks use or understand the same=20
> identifiers
> for the equivalent or similar service that is compatible at=20
> the session
> layer. Therefore we need to support the possibility to have different
> identifiers in different networks in order to not violate the=20
> TS 23.228
> requirement:
>=20
> 12.	The usage of communication service identifiers shall not
> adversely affect interoperability between IMS networks and
> interoperability with external SIP networks and CS networks. The
> behaviour of a network receiving the IMS requests without an IMS
> communication service identifier is a matter of operator policy. Usage
> of communication service identifiers shall not decrease the level of
> interoperability with networks and UEs that are unaware of the
> communication service identifier.
>=20
> If the originating service logic isn't invoked based on an identifier
> placed in the request by the originating terminal on what basis do you
> think it is invoked? How does the use of Accept-Contact have=20
> advantages
> in this case?
> <AA>=20
>=20
> --------------------------------------
>=20
>  I note here that the proposed service URN does not contain=20
> the address
> of a specific server containing the application logic.  An AS can look
> at that, and loose route to an application server that does=20
> contain such
> logic (If it is separated from the proxy), however I fail to see the
> difference for that proxy if it looks at a feature tag, or it looks at
> the service URN for this.  In both cases, it has to parse the message
> for the data, and modify the routing information. =20
>=20
> [AA]Route headers are the SIP mechanism for requesting=20
> intermediaries to
> be includede on the route of the request. There is nothing in RFC
> 3840/RFC 3841 that addresses using feature tags in=20
> Accept-Contact header
> for any routing decisions prior to determining which of the registered
> contacts the request should be forked to. Accept-Contact is=20
> not a Route
> header it is merely a filter that is applied to the set of valid
> contacts. Using a parameter in the filter as a service=20
> identifier has a
> potential significant impact on the use of Accept-Contact for its
> original purpose.=20
>=20
> <Steve>
> I don't believe there is anthing that states a Proxy CAN NOT route to
> applications or application logic based upon elements in the SIP
> message, including what is in the Accept-Contact header (some people
> have even proposed SDP ...).
> </Steve>
>=20
> <AA>
> As I had stated in an earlier email on this subject a proxy=20
> can do a lot
> of things (including routing based on your favorite number=20
> appearing in
> a call-Id) but that doesn't mean they are a good idea or a good design
> decision. If you want to ensure you don't break things then=20
> you need to
> be careful to use SIP headers with the semantics intended for them
> otherwise problems can result if extra semantics are put upon existing
> headers.
> <AA>=20
>=20
> Further, the route header could be removed in the originating network;
> and it is still required to identify the service in the terminating
> network.  As proxies may remove the route, it can't be used=20
> to transport
> the requested service. =20
>=20
> [AA] Proxies arbitarily removing Route headers is a BAD thing=20
> akin to a
> bull in a china shop - almost certainly going to break something. RFC
> 3261 summarises what proxies should be doing with Route headers:
> 16.12 Summary of Proxy Route Processing
>=20
>    In the absence of local policy to the contrary, the processing a
>    proxy performs on a request containing a Route header field can be
>    summarized in the following steps.
>=20
>       1.  The proxy will inspect the Request-URI.  If it indicates a
>           resource owned by this proxy, the proxy will replace it with
>           the results of running a location service.  Otherwise, the
>           proxy will not change the Request-URI.
>=20
>       2.  The proxy will inspect the URI in the topmost Route header
>           field value.  If it indicates this proxy, the proxy=20
> removes it
>           from the Route header field (this route node has been
>           reached).
>=20
>       3.  The proxy will forward the request to the resource indicated
>           by the URI in the topmost Route header field value or in the
>           Request-URI if no Route header field is present.  The proxy
>           determines the address, port and transport to use when
>           forwarding the request by applying the procedures in [4] to
>           that URI.
>=20
>    If no strict-routing elements are encountered on the path of the
>    request, the Request-URI will always indicate the target of the
>    request.
>=20
> <Steve>
> The first sentance is interesting "In absence of local policy to the
> contrary", indicating that there can be different policies.  I
> understand that some proxies do remove this.  While I am on this - I
> have lead to believe that R-URI parameters may also be removed by
> intermediate proxies - is this correct?
> </Steve>
>=20
> <AA>
> My understanding is that local policy may determine that additional
> nodes need to be routed to in addition to the Route contained in the
> request. However deleting existing Route headers other than the Route
> header containing the address of the proxy as part of the=20
> loose routing
> procedures is liable to break things and is not part of RFC=20
> 3261 routing
> (loose routing).
> <AA>=20
>=20
>=20
> [AA] Generally a UA wouldn't know the appropriate service identifiers
> used in the terminating network or whether a Route header is=20
> appropriate
> to be used in the terminating network. The draft defines the=20
> Service URI
> for the purpose of identifying services in the terminating network and
> service URIs can be obtained using Presence.
>=20
> <Steve>
> I don't get the above paragraph.  It suggest that the "appropriate
> service identifier" used in the terminating network is not=20
> known by the
> originating UE, but the Service URI is.  I thought that the=20
> service URI
> is an encoding of the "appropriate service identifier".
> </Steve>
>=20
> <AA>=20
> Yes but in the general case the UA doesn't need to know what=20
> the service
> identifier is it obtains through Presence a Service URI that contains
> the Service identifier as a URI parameter. The great thing about using
> Presence to obtain a Service URI is that the originating UA=20
> doesn't even
> need to know anything about service identifiers at all for=20
> this to work.
> It just uses Presence to find the URI to use for the service which can
> be indicated to the user via an indication such as a graphic=20
> icon (e,g a
> phone for telephony, smiley face for IM, etc) in the presence=20
> document.
> <AA>=20
>=20
> -----------------------------------
>=20
> >From this perspective it is not clear to me the advantage of using a
> service URN in a route header compared to the use of a feature tag.
>=20
> [AA] I hope the above responses clarify that somewhat. Also=20
> the Service
> URN and Service URIs allow different service identifiers to be used in
> the originating and terminating network for session=20
> compatiable services
> instead of forcing everyone to deploy and understand the same service
> and service identifier.
>=20
> <Steve>
> Nope, don't get it, but are information that is carried=20
> end-to-end ....
> And as far as I see, could be used by intermediate proxies to impact
> local routing policy ... Or ?
> </Steve>
>=20
> <AA>=20
> I don't understand this response either. The key point is that UAs in
> corporate networks and in the internet are not going to deploy 3GPP
> flavor services and use the same 3GPP service identifiers or even
> possibly any service identifiers at all. The Accept-Contact based
> proposal assumes that all endpoints understand the service identifiers
> to be used for each service. This concept is not generally applicable
> and create interoperability problems.
> <AA>=20
>=20
> The draft seems to mandate GRUU - is this really so, is that really
> necessary.  It seems strange to place such a dependency, it=20
> also starts
> to move the GRUU as being the primary means of which to=20
> contact a user.
>=20
> [AA] My understanding is that GRUU in the draft is used to=20
> address how a
> request addressed to Service URI is forked only to those contacts that
> support the service identified by the that Service URI. A=20
> contact would
> seem to be fine so long as it had GRUU properties - no need=20
> to obtain a
> GRUU if you already have one.
>=20
> This implies that the originating user decides which terminal I should
> receive my requested service on, even if I have two devices for that.
> As such, I don't believe that there should be a dependency on GRUU.
>=20
> [AA] If you have obtained a Service URI (probably through=20
> presence) for
> that user then you can address the request to the Service URI which
> represents all the terminals that support that service.=20
> However GRUU and
> the Service Instance URIs created from GRUUs provide another mechanism
> through which a user can obtain a URI to reach that user for the
> required service.
>=20
> Parts of the draft seems to rely on the use of presence. =20
> While presence
> could be a supportive mechanism, I doubt that we can mandate that the
> number you want to dial has be available in presence as I=20
> assume not all
> entities are available in presence (diallable network=20
> services, users in
> the PSTN).  This would be strange to mandated.
>=20
> [AA] PSTN gateways are presumably in the same domain as the=20
> terminating
> proxy and therefore these PSTN gateways can be configured with the
> appropriate service identifiers for telephony services and can include
> this in either a Route header to route to the right application server
> or can construct a valid Service URI. The real problem is how do users
> in other networks know what the service identifier that's is=20
> used in the
> terminating network is. The use of Service URIs and Presence=20
> provides a
> solution to that problem.
>=20
> <Steve>
> Again, presence could be a useful mechanism, but I think it would be a
> problem for the draft to rely it.  A user making a call to=20
> the PSTN may
> not find that user in the presence archicture (and similar to services
> like the 800 or 900 services).  (I should note, that you=20
> shouldn't make
> any assumptions about the location of the PSTN gateway ... But I don't
> think that changes anything). =20
> </Steve>
>=20
> <AA>=20
> How does a PSTN gateway in another domain from the terminating network
> know the service identifier for telephony used in the terminating
> network with the Accept-Contact header based proposal?=20
> <AA>=20
>=20
> Moving on from the draft I have seen the questioned raised about "why
> can't we expand the power of SDP".  Such an approach would=20
> mandate that
> SIP methods use SDP, I don't think that is appropriate.=20
>=20
> [AA] I agree with you here
>=20
> I hear the appeals that the mechanisms should not be abused,=20
> and that we
> should not have 1000s of separate "communication services",=20
> and I fully
> agree with this. I see a smaller number of communication=20
> services, not a
> larger number; where those communication services are used to access
> other end-user-services.
>=20
> *flame away*
>=20
> Regards,
>=20
> Steve
>=20
> ---
>=20
>=20
>=20
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential
> information, privileged material (including material protected by the
> solicitor-client or other applicable privileges), or constitute
> non-public information. Any use of this information by anyone=20
> other than
> the intended recipient is prohibited. If you have received this
> transmission in error, please immediately reply to the sender=20
> and delete
> this information from your system. Use, dissemination,=20
> distribution, or
> reproduction of this transmission by unintended recipients is not
> authorized and may be unlawful.
>=20
>=20
>=20
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain=20
> confidential information, privileged material (including=20
> material protected by the solicitor-client or other=20
> applicable privileges), or constitute non-public information.=20
> Any use of this information by anyone other than the intended=20
> recipient is prohibited. If you have received this=20
> transmission in error, please immediately reply to the sender=20
> and delete this information from your system. Use,=20
> dissemination, distribution, or reproduction of this=20
> transmission by unintended recipients is not authorized and=20
> may be unlawful.
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 22 18:29:38 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKNN7-0006Nj-8J; Thu, 22 Feb 2007 18:29:09 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKNN6-0006Ls-7g
	for sipping@ietf.org; Thu, 22 Feb 2007 18:29:08 -0500
Received: from sj-iport-6.cisco.com ([171.71.176.117])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKNN5-00070c-5J
	for sipping@ietf.org; Thu, 22 Feb 2007 18:29:08 -0500
Received: from sj-dkim-7.cisco.com ([171.68.10.88])
	by sj-iport-6.cisco.com with ESMTP; 22 Feb 2007 15:29:01 -0800
X-IronPort-AV: i="4.14,207,1170662400"; 
	d="scan'208"; a="115391631:sNHT80604045"
Received: from sj-core-4.cisco.com (sj-core-4.cisco.com [171.68.223.138])
	by sj-dkim-7.cisco.com (8.12.11/8.12.11) with ESMTP id l1MNT1nU001518; 
	Thu, 22 Feb 2007 15:29:01 -0800
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by sj-core-4.cisco.com (8.12.10/8.12.6) with ESMTP id l1MNT1nF023679;
	Thu, 22 Feb 2007 15:29:01 -0800 (PST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 18:29:00 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 18:29:00 -0500
Message-ID: <45DE273B.8070306@cisco.com>
Date: Thu, 22 Feb 2007 18:28:59 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Stephen Terrill (MI/EEM)" <stephen.terrill@ericsson.com>
Subject: Re: [Sipping] Service Identification in SIP
References: <989763567E51AB48BE015DF7BB960CAD028F09A5@esealmw111>
In-Reply-To: <989763567E51AB48BE015DF7BB960CAD028F09A5@esealmw111>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 22 Feb 2007 23:29:00.0108 (UTC)
	FILETIME=[3AFFC0C0:01C756D9]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=22655; t=1172186941;
	x=1173050941; c=relaxed/simple; s=sjdkim7002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20; bh=fqd1qRtSFbvgWd8wIS6gV0CNqVWpeN0gx4qpcYXPa/Y=;
	b=DyRK8XSrdo9i0ym0YoXbNs7kYkY0hnBvOqC5wk1XO3kEJquBiwsI4kEXXlRpg9dBSlMlPbMp
	jvSzPVq4XrYJXZ0zIvrOB6Yb+PGhtQiYgU9fFcsOKgY9qg+CU0DYsp7l;
Authentication-Results: sj-dkim-7; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim7002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 69ad46fbd0e474a7c7f3312d9e320336
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Stephen Terrill (MI/EEM) wrote:
> Hi Andrew,
> 
> I have the opposite view.  I think that if the originating network and
> the terminating network are going to interoperate, then they will have
> the same understanding of the service, hence in my view the same
> "identifier".  To have otherwise is problematic and complicated (the
> operators will have to have a higher admininstrative effort in order to
> understand how they should interconnect).
> 
> The PBX case is more complicated than can be responded to in a single
> sentence - as there are many ways in which the PBXs are connected
> together.  In one case - the PBXs are just treated as a terminal with
> many users; and the service that the operator provides is really the
> same basic service as is provided to other users (with maybe more
> subscriber specific logic).  In other cases, the operator just provides
> routing and connecting support, and here the service provided by the
> operator is different; but I would assume that the peer PBXs had better
> have a common understanding of the service (which they would if it was
> the same vendor), or again there will be inter-operatability issues.

What will happen here if the pbx calls an address on a 3gpp network that 
is using some set of service definitions? Or visa versa.

Will the PBX, that specifies no service but that includes audio and 
video, be able to establish an audio/video call? Or will it somehow have 
to know the service id that the IMS system expects for audio/video calls?

	Paul

> Regarding the URI parameter proposal, don't we have proxies that remove
> additional R-URIs - it seems a natural thing to remove it also in e.g.
> fowarding services etc ...
> 
> Regards,
> 
> Steve.
> 
> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com] 
> Sent: Thursday, February 22, 2007 4:05 AM
> To: Stephen Terrill (MI/EEM); sipping@ietf.org
> Cc: Andrew Allen
> Subject: RE: [Sipping] Service Identification in SIP
> 
> 
> Hi Steve
> 
> In my view this is one of the interoperability issues with your
> proposal. I think the assumption that all networks will deploy exactly
> the same services with the same service identifiers is unrealistic and
> also undesirable. Are corporate PBX vendors going to implement the same
> services as defined by 3GPP or OMA? The possibility to have the
> flexibility to have different service identifiers for session equivalent
> services is in my view essential to any IETF standard and to do
> otherwise is to create a situation where basic services like telephony
> or messaging cannot interoperate between users in different networks and
> users in the internet without the use of inter working gateway. SIP and
> SIMPLE were setup specifically to create protocols that would enable
> interoperability with IP networks without the need for gateways. 
> 
> If 3GPP network operators choose to deploy the same services and use
> common identifiers that is fine but there are whole other worlds out
> there as well and service identifiers should not stand in the way of
> interoperability between all these different SIP users. If the IETF
> solution supports the ability for different networks to use different
> identifiers that doesn't violate 3GPP requirements and doesn't prevent
> 3GPP using the same identifiers in different networks. I don't see how
> that is against TS 23.228.
> 
> Andrew
> 
> 
> -----Original Message-----
> From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
> Sent: Monday, February 19, 2007 6:00 AM
> To: Andrew Allen; sipping@ietf.org
> Subject: RE: [Sipping] Service Identification in SIP
> 
> Hi Andrew,
> 
> I see that in your assumptions, the originating and terminating networks
> use different service identifiers, and I presume that this is for the
> same service.  In my understanding, it is the same service identifier
> that is used end-to-end (if not it is interworked).  That is why I do
> not see the reliance on Presence being necessary.  In my opionion this
> totally within the requirements of TS 23.228 (I think that different
> service identifiers in different networks for the same service is
> against TS 23.228).
> 
> This comes to the isssue of the PSTN gateway - I presume that you want
> to go to the PSTN gateway if the service is telephony, and within that,
> voice.  If we know that, then the we go to any PSTN gateway, and we
> can't presume a particular network for a PSTN gateway. 
> 
> Best Regards,
> 
> Steve.
> 
> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com]
> Sent: Saturday, February 17, 2007 12:51 AM
> To: Stephen Terrill (MI/EEM); sipping@ietf.org
> Cc: Andrew Allen
> Subject: RE: [Sipping] Service Identification in SIP
> 
> 
> 
> Hi Steve,
> 
> My responses inline
> 
> Andrew 
> 
> -----Original Message-----
> From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
> Sent: Friday, February 16, 2007 3:51 AM
> To: sipping@ietf.org
> Subject: RE: [Sipping] Service Identification in SIP
> 
> Hello Andrew,
> 
> Thanks for your comments.  I have a couple of brief responses below.
> 
> Best Regards,
> 
> ..//Steve 
> 
> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com]
> Sent: Friday, February 16, 2007 7:15 AM
> To: Stephen Terrill (MI/EEM); sipping@ietf.org
> Cc: Andrew Allen
> Subject: RE: [Sipping] Service Identification in SIP
> 
> 
> Some comments on Stephen's comments on the draft:
>  
> 
> -Moving onto the proposed solution draft itself, in addition for
> providing for the identification of the "application" and the
> "communication service", I have a few additional thoughts.
> 
> The draft seems to have the implication that all services (communication
> services in other terminology) that are to be identified have to be
> standardised within the IETF.  This is clearly an restriction that will
> not be acceptable to the industry as a whole.  My guess is that if such
> a statement is mandated by the IETF, it will just be words on paper
> which would likely be ignored in practice by other standardisation
> organisations.
> 
> [AA] This I think is an incorrect interpretation of the draft. My
> understanding is that the only services according to the draft that
> require some standardisation in IETF are those that by their very nature
> cannot interoperate unless there is some common additional understanding
> between the endpoints outside of what is currently specified in SIP. The
> use of SIP option tags with Require header is proposed by the draft for
> this purpose as such services create interoperability issues with SIP.
> 
> <Steve>
> My comment applies even for these services </Steve>
> 
> 
> <AA>
> Well then we are in big trouble because what you are proposing is that
> its OK for bodies outside of IETF to create non interoperable extensions
> for the behavour of SIP entities. This will result in the fragmentation
> of SIP into different flavors that don't interoperate. This is the
> balkanization of SIP that Paul talks about.even <AA> 
> 
> ----------------------------------------------------
> 
> The draft also only recognises the case where all explicitly knowledge
> of the service is required.  I think it would also be good to allow the
> case where knowledge of the service would be helpful, but not explicitly
> required.  This could be a case where the service to the end-user would
> not be degraded if this was not supported.  This functionality seems to
> be missing.
> 
> [AA] Service URIs are defined in the draft for these services where the
> support by the other endpoint is not required for the session to be
> sucessfully established.
> 
> The draft seems to indicate that that application servers in the network
> are involved only at the request of the originating terminal.  This
> seems a strange restriction.  For some network support that supports the
> originating user, such as the one provided in the example "I would like
> this call recorded for later retrieval", such mechanisms may make sense,
> however as a general approach it seems a bit limited.  For example, in
> 3GPP networks there is the need to invoke application logic in the
> originating side and terminating side, not only at the request of the
> originating terminal, but as a consequence of the requested service.  If
> there is an SIP message contains indication of the requested service, a
> proxy can do this (or loose route to an AS that contains the necessary
> service logic). 
> 
> [AA] The draft supports both possibilities using the Route header for
> originating network servers that need to be involved in the route and
> based on the terminating networks routing policy for the Service URI in
> the Request URI.
> 
> <Steve>
> I don't see the advantage of having different solutions for the
> originating network and the terminating network.  I don't think that the
> originating logic is invoked only at the request of the originating
> terminal.
> </Steve>
> 
> <AA>
> Unless IETF is willing to standardise all services and mandate that all
> SIP entities understand these (which will never happen) then we cannot
> assume that all networks and all endpoints understand and use the same
> identifiers. It is possible through configuration for an endpoint to
> obtain the identifiers used by its own service provider but there is no
> way to ensure that other networks use or understand the same identifiers
> for the equivalent or similar service that is compatible at the session
> layer. Therefore we need to support the possibility to have different
> identifiers in different networks in order to not violate the TS 23.228
> requirement:
> 
> 12.	The usage of communication service identifiers shall not
> adversely affect interoperability between IMS networks and
> interoperability with external SIP networks and CS networks. The
> behaviour of a network receiving the IMS requests without an IMS
> communication service identifier is a matter of operator policy. Usage
> of communication service identifiers shall not decrease the level of
> interoperability with networks and UEs that are unaware of the
> communication service identifier.
> 
> If the originating service logic isn't invoked based on an identifier
> placed in the request by the originating terminal on what basis do you
> think it is invoked? How does the use of Accept-Contact have advantages
> in this case?
> <AA> 
> 
> --------------------------------------
> 
>  I note here that the proposed service URN does not contain the address
> of a specific server containing the application logic.  An AS can look
> at that, and loose route to an application server that does contain such
> logic (If it is separated from the proxy), however I fail to see the
> difference for that proxy if it looks at a feature tag, or it looks at
> the service URN for this.  In both cases, it has to parse the message
> for the data, and modify the routing information.  
> 
> [AA]Route headers are the SIP mechanism for requesting intermediaries to
> be includede on the route of the request. There is nothing in RFC
> 3840/RFC 3841 that addresses using feature tags in Accept-Contact header
> for any routing decisions prior to determining which of the registered
> contacts the request should be forked to. Accept-Contact is not a Route
> header it is merely a filter that is applied to the set of valid
> contacts. Using a parameter in the filter as a service identifier has a
> potential significant impact on the use of Accept-Contact for its
> original purpose. 
> 
> <Steve>
> I don't believe there is anthing that states a Proxy CAN NOT route to
> applications or application logic based upon elements in the SIP
> message, including what is in the Accept-Contact header (some people
> have even proposed SDP ...).
> </Steve>
> 
> <AA>
> As I had stated in an earlier email on this subject a proxy can do a lot
> of things (including routing based on your favorite number appearing in
> a call-Id) but that doesn't mean they are a good idea or a good design
> decision. If you want to ensure you don't break things then you need to
> be careful to use SIP headers with the semantics intended for them
> otherwise problems can result if extra semantics are put upon existing
> headers.
> <AA> 
> 
> Further, the route header could be removed in the originating network;
> and it is still required to identify the service in the terminating
> network.  As proxies may remove the route, it can't be used to transport
> the requested service.  
> 
> [AA] Proxies arbitarily removing Route headers is a BAD thing akin to a
> bull in a china shop - almost certainly going to break something. RFC
> 3261 summarises what proxies should be doing with Route headers:
> 16.12 Summary of Proxy Route Processing
> 
>    In the absence of local policy to the contrary, the processing a
>    proxy performs on a request containing a Route header field can be
>    summarized in the following steps.
> 
>       1.  The proxy will inspect the Request-URI.  If it indicates a
>           resource owned by this proxy, the proxy will replace it with
>           the results of running a location service.  Otherwise, the
>           proxy will not change the Request-URI.
> 
>       2.  The proxy will inspect the URI in the topmost Route header
>           field value.  If it indicates this proxy, the proxy removes it
>           from the Route header field (this route node has been
>           reached).
> 
>       3.  The proxy will forward the request to the resource indicated
>           by the URI in the topmost Route header field value or in the
>           Request-URI if no Route header field is present.  The proxy
>           determines the address, port and transport to use when
>           forwarding the request by applying the procedures in [4] to
>           that URI.
> 
>    If no strict-routing elements are encountered on the path of the
>    request, the Request-URI will always indicate the target of the
>    request.
> 
> <Steve>
> The first sentance is interesting "In absence of local policy to the
> contrary", indicating that there can be different policies.  I
> understand that some proxies do remove this.  While I am on this - I
> have lead to believe that R-URI parameters may also be removed by
> intermediate proxies - is this correct?
> </Steve>
> 
> <AA>
> My understanding is that local policy may determine that additional
> nodes need to be routed to in addition to the Route contained in the
> request. However deleting existing Route headers other than the Route
> header containing the address of the proxy as part of the loose routing
> procedures is liable to break things and is not part of RFC 3261 routing
> (loose routing).
> <AA> 
> 
> 
> [AA] Generally a UA wouldn't know the appropriate service identifiers
> used in the terminating network or whether a Route header is appropriate
> to be used in the terminating network. The draft defines the Service URI
> for the purpose of identifying services in the terminating network and
> service URIs can be obtained using Presence.
> 
> <Steve>
> I don't get the above paragraph.  It suggest that the "appropriate
> service identifier" used in the terminating network is not known by the
> originating UE, but the Service URI is.  I thought that the service URI
> is an encoding of the "appropriate service identifier".
> </Steve>
> 
> <AA> 
> Yes but in the general case the UA doesn't need to know what the service
> identifier is it obtains through Presence a Service URI that contains
> the Service identifier as a URI parameter. The great thing about using
> Presence to obtain a Service URI is that the originating UA doesn't even
> need to know anything about service identifiers at all for this to work.
> It just uses Presence to find the URI to use for the service which can
> be indicated to the user via an indication such as a graphic icon (e,g a
> phone for telephony, smiley face for IM, etc) in the presence document.
> <AA> 
> 
> -----------------------------------
> 
>>From this perspective it is not clear to me the advantage of using a
> service URN in a route header compared to the use of a feature tag.
> 
> [AA] I hope the above responses clarify that somewhat. Also the Service
> URN and Service URIs allow different service identifiers to be used in
> the originating and terminating network for session compatiable services
> instead of forcing everyone to deploy and understand the same service
> and service identifier.
> 
> <Steve>
> Nope, don't get it, but are information that is carried end-to-end ....
> And as far as I see, could be used by intermediate proxies to impact
> local routing policy ... Or ?
> </Steve>
> 
> <AA> 
> I don't understand this response either. The key point is that UAs in
> corporate networks and in the internet are not going to deploy 3GPP
> flavor services and use the same 3GPP service identifiers or even
> possibly any service identifiers at all. The Accept-Contact based
> proposal assumes that all endpoints understand the service identifiers
> to be used for each service. This concept is not generally applicable
> and create interoperability problems.
> <AA> 
> 
> The draft seems to mandate GRUU - is this really so, is that really
> necessary.  It seems strange to place such a dependency, it also starts
> to move the GRUU as being the primary means of which to contact a user.
> 
> [AA] My understanding is that GRUU in the draft is used to address how a
> request addressed to Service URI is forked only to those contacts that
> support the service identified by the that Service URI. A contact would
> seem to be fine so long as it had GRUU properties - no need to obtain a
> GRUU if you already have one.
> 
> This implies that the originating user decides which terminal I should
> receive my requested service on, even if I have two devices for that.
> As such, I don't believe that there should be a dependency on GRUU.
> 
> [AA] If you have obtained a Service URI (probably through presence) for
> that user then you can address the request to the Service URI which
> represents all the terminals that support that service. However GRUU and
> the Service Instance URIs created from GRUUs provide another mechanism
> through which a user can obtain a URI to reach that user for the
> required service.
> 
> Parts of the draft seems to rely on the use of presence.  While presence
> could be a supportive mechanism, I doubt that we can mandate that the
> number you want to dial has be available in presence as I assume not all
> entities are available in presence (diallable network services, users in
> the PSTN).  This would be strange to mandated.
> 
> [AA] PSTN gateways are presumably in the same domain as the terminating
> proxy and therefore these PSTN gateways can be configured with the
> appropriate service identifiers for telephony services and can include
> this in either a Route header to route to the right application server
> or can construct a valid Service URI. The real problem is how do users
> in other networks know what the service identifier that's is used in the
> terminating network is. The use of Service URIs and Presence provides a
> solution to that problem.
> 
> <Steve>
> Again, presence could be a useful mechanism, but I think it would be a
> problem for the draft to rely it.  A user making a call to the PSTN may
> not find that user in the presence archicture (and similar to services
> like the 800 or 900 services).  (I should note, that you shouldn't make
> any assumptions about the location of the PSTN gateway ... But I don't
> think that changes anything).  
> </Steve>
> 
> <AA> 
> How does a PSTN gateway in another domain from the terminating network
> know the service identifier for telephony used in the terminating
> network with the Accept-Contact header based proposal? 
> <AA> 
> 
> Moving on from the draft I have seen the questioned raised about "why
> can't we expand the power of SDP".  Such an approach would mandate that
> SIP methods use SDP, I don't think that is appropriate. 
> 
> [AA] I agree with you here
> 
> I hear the appeals that the mechanisms should not be abused, and that we
> should not have 1000s of separate "communication services", and I fully
> agree with this. I see a smaller number of communication services, not a
> larger number; where those communication services are used to access
> other end-user-services.
> 
> *flame away*
> 
> Regards,
> 
> Steve
> 
> ---
> 
> 
> 
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential
> information, privileged material (including material protected by the
> solicitor-client or other applicable privileges), or constitute
> non-public information. Any use of this information by anyone other than
> the intended recipient is prohibited. If you have received this
> transmission in error, please immediately reply to the sender and delete
> this information from your system. Use, dissemination, distribution, or
> reproduction of this transmission by unintended recipients is not
> authorized and may be unlawful.
> 
> 
> 
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential
> information, privileged material (including material protected by the
> solicitor-client or other applicable privileges), or constitute
> non-public information. Any use of this information by anyone other than
> the intended recipient is prohibited. If you have received this
> transmission in error, please immediately reply to the sender and delete
> this information from your system. Use, dissemination, distribution, or
> reproduction of this transmission by unintended recipients is not
> authorized and may be unlawful.
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Thu Feb 22 19:05:23 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKNv3-0005A7-KJ; Thu, 22 Feb 2007 19:04:13 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKNv2-0005A2-5x
	for sipping@ietf.org; Thu, 22 Feb 2007 19:04:12 -0500
Received: from hme1.july.broomfield1.level3.net ([209.245.18.8]
	helo=f4bb49-05.idc1.level3.com)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKNv0-0004fR-Sz
	for sipping@ietf.org; Thu, 22 Feb 2007 19:04:12 -0500
Received: from montag.eng.level3.com (montag.eng.l3.com [10.1.68.57])
	by f4bb49-05.idc1.level3.com (8.12.10/8.12.10) with ESMTP id
	l1N045Ya018934 for <sipping@ietf.org>; Fri, 23 Feb 2007 00:04:06 GMT
Subject: [Sipping] Session Establishment Efficiency Rate
From: Daryl Malas <daryl@level3.net>
To: sipping@ietf.org
Content-Type: text/plain
Date: Thu, 22 Feb 2007 17:13:54 -0700
Message-Id: <1172189634.17141.59.camel@montag.eng.level3.com>
Mime-Version: 1.0
X-Mailer: Evolution 2.6.0 (2.6.0-1) 
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.1 (/)
X-Scan-Signature: de4f315c9369b71d7dd5909b42224370
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

A couple more people have expressed to me that this adaptation of the
ITU-T E.411 Network Efficiency Rate metric is very valuable.  The
snapshot of the metric from the draft is:

"This metric is similar to Session Establishment Rate, but it is
intended to exclude the effects of customer behavior and user agent
behavior.  Essentially, the attempts resulting with a final cause code
of 480, 486, and 600 would be added to the SER numerator."

Does anyone else want to see this metric in the draft?  I personally
think it is an interesting metric.

Thanks...

--Daryl



_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP

From sipping-bounces@ietf.org Thu Feb 22 19:05:23 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKNw1-0005Zz-1m; Thu, 22 Feb 2007 19:05:13 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKNvz-0005XV-LR
	for sipping@ietf.org; Thu, 22 Feb 2007 19:05:11 -0500
Received: from sj-iport-3-in.cisco.com ([171.71.176.72]
	helo=sj-iport-3.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HKNvy-00051u-1C
	for sipping@ietf.org; Thu, 22 Feb 2007 19:05:11 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by sj-iport-3.cisco.com with ESMTP; 22 Feb 2007 16:05:09 -0800
X-IronPort-AV: i="4.14,207,1170662400"; 
	d="scan'208"; a="466495882:sNHT87354106"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1N057St020505; 
	Thu, 22 Feb 2007 19:05:07 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1N04wOA005790; 
	Thu, 22 Feb 2007 19:05:02 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 19:04:58 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 19:04:58 -0500
Message-ID: <45DE2FA8.3050501@cisco.com>
Date: Thu, 22 Feb 2007 19:04:56 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Andrew Allen <aallen@rim.com>
Subject: Re: [Sipping] Service Identification Model
References: <61968779B8AC4C4BAB421D4C12F008C009152060@XCH47YKF.rim.net>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C009152060@XCH47YKF.rim.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 23 Feb 2007 00:04:58.0088 (UTC)
	FILETIME=[41415E80:01C756DE]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=9752; t=1172189107;
	x=1173053107; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20Model
	|Sender:=20 |To:=20Andrew=20Allen=20<aallen@rim.com>;
	bh=HN6OegeCHNZ958lqQbpWUw/7TgEK60Dj+DLECSzzpL0=;
	b=PCGjQD6ObYas7/wfJFAKMhBuUd6sxuaK/vYZW/+fbURn5tceWbqq+f0/e5+sr0z8htCpO3SZ
	m6AWU4PyY/HA53bdtHGact5ICKFQGjxw/3gDUV0WGgzb2NF/22LeGz6E;
Authentication-Results: rtp-dkim-2; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 88b11fc64c1bfdb4425294ef5374ca07
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Andrew Allen wrote:
> Hi Eric,
> 
> It seems your prefer the approach of using URIs to identify services
> which is part of Jonathan's draft as is the use of Require for those
> services which simply do not interoperate unless the UAS understands the
> service.
> 
> Where I am concerned is that you seem to be suggesting that the UA
> separately register an AoR for each service it supports using REGISTER.
> Certainly for cellular that wouldn't be acceptable as 10 or more
> services could easily be supported by a single device.

Andrew, can you give me an example of 10 "services" on the device that 
need to be distinguished from one another in this way - that aren't just 
optional behavior of a particular service? (E.g. I see no reason why 
"audio" and "audio/video" need to be separate services.)

I guess as things stand, PoC must be a separate service. And apparently 
Chess. What else?

	Paul

> Registrations
> create a significant load in the IMS network because its not just a SIP
> REGISTER but also requires a lot of interactions with other network
> elements such as the HSS and Notifications for reg-events etc. Also
> having an order of magnitude increase in multiple registrations is going
> to be a big issue for the avalanche restart problem for example when a
> cell site goes down and comes back into service.
> 
> I think Jonathan's approach gives us the possibility for the registrar
> to create the Service URIs based on the feature parameters included in
> the Contact of the REGISTER. 
> 
> I think the argument over whether the model is one UA with multiple
> services or multiple UAs each representing a separate service is
> somewhat academic. As I understand a UA can register multiple AoRs and
> can have other AoRs configured or otherwise provided by the service
> provider. 
> 
> Andrew
> 
> -----Original Message-----
> From: Eric Burger [mailto:eburger@bea.com] 
> Sent: Wednesday, February 21, 2007 1:29 PM
> To: sipping@ietf.org
> Subject: [Sipping] Service Identification Model 
> 
> Folks have been pestering me for a while to look at the Service
> Identification issue.  I finally have gotten around to it.
> 
> 
> I guess I have a fundamental problem with the principal assertion of the
> service-identification draft.  Namely, the assertion that there is a
> single
> SIP Layer, separate and distinct from two services.  Here is Figure 2
> from
> the draft:
> 
>                            +---------------------------------+
>                            |                                 |
>                            | +-------------+ +-------------+ |
>                            | |      UI     | |      UI     | |
>                            | +-------------+ +-------------+ |
>                            | +-------------+ +-------------+ |
>                            | |             | |             | |
>                            | |  Service 1  | |  Service 2  | |
>                            | |             | |             | |
>                            | +-------------+ +-------------+ |
>                            | +-----------------------------+ |
>                            | |                             | |
>                            | |             SIP             | |
>                            | |            Layer            | |
>                            | |                             | |
>                            | +-----------------------------+ |
>                            |                                 |
>                            +---------------------------------+
> 
>                                      Physical Device
> 
> I would offer this drawing confuses a common SIP Stack with a common
> User
> Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use case:
> 
>> ... In addition users may have
>> multiple devices only one of which has the application that
> understands
>> the game. Other devices that the user has registered against his AoR
> may
>> handle audio for telephony and other voice applications but if Alice
>> invites Bob to play her favourite game it is not very interesting if
> Bob
>> accepts the gaming session on a device that can only accept the audio
>> stream. Also even when the terminal that has the game application
>> receives the INVITE the correct application in that terminal needs to
> be
>> identified to accept the session.
> 
> I would offer that this is the clinching argument that the request is,
> in
> fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a User
> Agent.
> 
> Same argument for PoC versus a regular voice call: One is looking for a
> PoC
> User Agent, NOT the PoC service at a given User Agent.
> 
> Thus, the picture really looks like the following:
> 
>                            +---------------------------------+
>                            |                                 |
>                            | +-------------+ +-------------+ |
>                            | |    UI       | |    UI       | |
>                            | +-------------+ +-------------+ |
>                            | +-------------+ +-------------+ |
>                            | |             | |             | |
>                            | |    UAS 1    | |    UAS 2    | |
>                            | |             | |             | |
>                            | +-------------+ +-------------+ |
>                            | +-----------------------------+ |
>                            | |                             | |
>                            | |             SIP             | |
>                            | |            Stack            | |
>                            | |                             | |
>                            | +-----------------------------+ |
>                            |                                 |
>                            +---------------------------------+
> 
>                                      Physical Device
> 
> 
> Another argument for different User Agents is the whole "route based on
> SDP"
> argument.  As the author of RFC 3458, I think I was first to make the
> assertion that knowing the media type did NOT mean you know the
> application
> to use.  I don't want your phone or proxy or B2BUA to guess that because
> I
> have Chess Media, I want the Chess Application; I want to talk with your
> Chess Application.
> 
> Now, is the Chess Application a Preference, Requirement, or the thing
> that I
> am addressing?  The answer drives one towards using caller prefs, the
> Require header, or the Request-URI.
> 
> I would offer that it is not a preference: I really want the chess
> application; I probably do not want to talk to you.
> 
> That leaves us with the Require header or a Request-URI.  My leanings
> are
> not surprisingly towards the RFC 4240 approach.  Frankly, if you are
> building an IMS (as an example) that can handle 10,000,000 subscribers,
> you
> should be able to handle 100,000,000 registrations.  That order of
> magnitude
> is way easier than the 10,000 to 10,000,000 jump.  I don't buy the
> "registering each service is too much of a burden on the operator"
> argument,
> because it is not that much of a burden, and the network actually does
> the
> right thing.  Better yet, all of the corner cases work.  Best of all,
> the
> IETF doesn't have to build an IANA Registry and process a thousand PS
> documents to register urn:service:myco:widget.
> 
> I suppose I could live with Require:, only if you are way too lazy to
> register each application instance.  It does semantically work, because
> if I
> want to do PoC and you do not do PoC, I'll get a 4xx response, to which
> I
> could try a regular call.
> 
> _______________________________________________________________________
> Notice:  This email message, together with any attachments, may contain
> information  of  BEA Systems,  Inc.,  its subsidiaries  and  affiliated
> entities,  that may be confidential,  proprietary,  copyrighted  and/or
> legally privileged, and is intended solely for the use of the individual
> or entity named in this message. If you are not the intended recipient,
> and have received this message in error, please immediately return this
> by email and then delete it.
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 
> 
> 
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential information, privileged material (including material protected by the solicitor-client or other applicable privileges), or constitute non-public information. Any use of this information by anyone other than the intended recipient is prohibited. If you have received this transmission in error, please immediately reply to the sender and delete this information from your system. Use, dissemination, distribution, or reproduction of this transmission by unintended recipients is not authorized and may be unlawful.
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP





From sipping-bounces@ietf.org Thu Feb 22 19:05:23 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKNv3-0005A7-KJ; Thu, 22 Feb 2007 19:04:13 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKNv2-0005A2-5x
	for sipping@ietf.org; Thu, 22 Feb 2007 19:04:12 -0500
Received: from hme1.july.broomfield1.level3.net ([209.245.18.8]
	helo=f4bb49-05.idc1.level3.com)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKNv0-0004fR-Sz
	for sipping@ietf.org; Thu, 22 Feb 2007 19:04:12 -0500
Received: from montag.eng.level3.com (montag.eng.l3.com [10.1.68.57])
	by f4bb49-05.idc1.level3.com (8.12.10/8.12.10) with ESMTP id
	l1N045Ya018934 for <sipping@ietf.org>; Fri, 23 Feb 2007 00:04:06 GMT
Subject: [Sipping] Session Establishment Efficiency Rate
From: Daryl Malas <daryl@level3.net>
To: sipping@ietf.org
Content-Type: text/plain
Date: Thu, 22 Feb 2007 17:13:54 -0700
Message-Id: <1172189634.17141.59.camel@montag.eng.level3.com>
Mime-Version: 1.0
X-Mailer: Evolution 2.6.0 (2.6.0-1) 
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.1 (/)
X-Scan-Signature: de4f315c9369b71d7dd5909b42224370
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

A couple more people have expressed to me that this adaptation of the
ITU-T E.411 Network Efficiency Rate metric is very valuable.  The
snapshot of the metric from the draft is:

"This metric is similar to Session Establishment Rate, but it is
intended to exclude the effects of customer behavior and user agent
behavior.  Essentially, the attempts resulting with a final cause code
of 480, 486, and 600 would be added to the SER numerator."

Does anyone else want to see this metric in the draft?  I personally
think it is an interesting metric.

Thanks...

--Daryl



_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP

From sipping-bounces@ietf.org Thu Feb 22 19:05:23 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKNw1-0005Zz-1m; Thu, 22 Feb 2007 19:05:13 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKNvz-0005XV-LR
	for sipping@ietf.org; Thu, 22 Feb 2007 19:05:11 -0500
Received: from sj-iport-3-in.cisco.com ([171.71.176.72]
	helo=sj-iport-3.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HKNvy-00051u-1C
	for sipping@ietf.org; Thu, 22 Feb 2007 19:05:11 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by sj-iport-3.cisco.com with ESMTP; 22 Feb 2007 16:05:09 -0800
X-IronPort-AV: i="4.14,207,1170662400"; 
	d="scan'208"; a="466495882:sNHT87354106"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1N057St020505; 
	Thu, 22 Feb 2007 19:05:07 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1N04wOA005790; 
	Thu, 22 Feb 2007 19:05:02 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 19:04:58 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 22 Feb 2007 19:04:58 -0500
Message-ID: <45DE2FA8.3050501@cisco.com>
Date: Thu, 22 Feb 2007 19:04:56 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Andrew Allen <aallen@rim.com>
Subject: Re: [Sipping] Service Identification Model
References: <61968779B8AC4C4BAB421D4C12F008C009152060@XCH47YKF.rim.net>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C009152060@XCH47YKF.rim.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 23 Feb 2007 00:04:58.0088 (UTC)
	FILETIME=[41415E80:01C756DE]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=9752; t=1172189107;
	x=1173053107; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20Model
	|Sender:=20 |To:=20Andrew=20Allen=20<aallen@rim.com>;
	bh=HN6OegeCHNZ958lqQbpWUw/7TgEK60Dj+DLECSzzpL0=;
	b=PCGjQD6ObYas7/wfJFAKMhBuUd6sxuaK/vYZW/+fbURn5tceWbqq+f0/e5+sr0z8htCpO3SZ
	m6AWU4PyY/HA53bdtHGact5ICKFQGjxw/3gDUV0WGgzb2NF/22LeGz6E;
Authentication-Results: rtp-dkim-2; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 88b11fc64c1bfdb4425294ef5374ca07
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Andrew Allen wrote:
> Hi Eric,
> 
> It seems your prefer the approach of using URIs to identify services
> which is part of Jonathan's draft as is the use of Require for those
> services which simply do not interoperate unless the UAS understands the
> service.
> 
> Where I am concerned is that you seem to be suggesting that the UA
> separately register an AoR for each service it supports using REGISTER.
> Certainly for cellular that wouldn't be acceptable as 10 or more
> services could easily be supported by a single device.

Andrew, can you give me an example of 10 "services" on the device that 
need to be distinguished from one another in this way - that aren't just 
optional behavior of a particular service? (E.g. I see no reason why 
"audio" and "audio/video" need to be separate services.)

I guess as things stand, PoC must be a separate service. And apparently 
Chess. What else?

	Paul

> Registrations
> create a significant load in the IMS network because its not just a SIP
> REGISTER but also requires a lot of interactions with other network
> elements such as the HSS and Notifications for reg-events etc. Also
> having an order of magnitude increase in multiple registrations is going
> to be a big issue for the avalanche restart problem for example when a
> cell site goes down and comes back into service.
> 
> I think Jonathan's approach gives us the possibility for the registrar
> to create the Service URIs based on the feature parameters included in
> the Contact of the REGISTER. 
> 
> I think the argument over whether the model is one UA with multiple
> services or multiple UAs each representing a separate service is
> somewhat academic. As I understand a UA can register multiple AoRs and
> can have other AoRs configured or otherwise provided by the service
> provider. 
> 
> Andrew
> 
> -----Original Message-----
> From: Eric Burger [mailto:eburger@bea.com] 
> Sent: Wednesday, February 21, 2007 1:29 PM
> To: sipping@ietf.org
> Subject: [Sipping] Service Identification Model 
> 
> Folks have been pestering me for a while to look at the Service
> Identification issue.  I finally have gotten around to it.
> 
> 
> I guess I have a fundamental problem with the principal assertion of the
> service-identification draft.  Namely, the assertion that there is a
> single
> SIP Layer, separate and distinct from two services.  Here is Figure 2
> from
> the draft:
> 
>                            +---------------------------------+
>                            |                                 |
>                            | +-------------+ +-------------+ |
>                            | |      UI     | |      UI     | |
>                            | +-------------+ +-------------+ |
>                            | +-------------+ +-------------+ |
>                            | |             | |             | |
>                            | |  Service 1  | |  Service 2  | |
>                            | |             | |             | |
>                            | +-------------+ +-------------+ |
>                            | +-----------------------------+ |
>                            | |                             | |
>                            | |             SIP             | |
>                            | |            Layer            | |
>                            | |                             | |
>                            | +-----------------------------+ |
>                            |                                 |
>                            +---------------------------------+
> 
>                                      Physical Device
> 
> I would offer this drawing confuses a common SIP Stack with a common
> User
> Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use case:
> 
>> ... In addition users may have
>> multiple devices only one of which has the application that
> understands
>> the game. Other devices that the user has registered against his AoR
> may
>> handle audio for telephony and other voice applications but if Alice
>> invites Bob to play her favourite game it is not very interesting if
> Bob
>> accepts the gaming session on a device that can only accept the audio
>> stream. Also even when the terminal that has the game application
>> receives the INVITE the correct application in that terminal needs to
> be
>> identified to accept the session.
> 
> I would offer that this is the clinching argument that the request is,
> in
> fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a User
> Agent.
> 
> Same argument for PoC versus a regular voice call: One is looking for a
> PoC
> User Agent, NOT the PoC service at a given User Agent.
> 
> Thus, the picture really looks like the following:
> 
>                            +---------------------------------+
>                            |                                 |
>                            | +-------------+ +-------------+ |
>                            | |    UI       | |    UI       | |
>                            | +-------------+ +-------------+ |
>                            | +-------------+ +-------------+ |
>                            | |             | |             | |
>                            | |    UAS 1    | |    UAS 2    | |
>                            | |             | |             | |
>                            | +-------------+ +-------------+ |
>                            | +-----------------------------+ |
>                            | |                             | |
>                            | |             SIP             | |
>                            | |            Stack            | |
>                            | |                             | |
>                            | +-----------------------------+ |
>                            |                                 |
>                            +---------------------------------+
> 
>                                      Physical Device
> 
> 
> Another argument for different User Agents is the whole "route based on
> SDP"
> argument.  As the author of RFC 3458, I think I was first to make the
> assertion that knowing the media type did NOT mean you know the
> application
> to use.  I don't want your phone or proxy or B2BUA to guess that because
> I
> have Chess Media, I want the Chess Application; I want to talk with your
> Chess Application.
> 
> Now, is the Chess Application a Preference, Requirement, or the thing
> that I
> am addressing?  The answer drives one towards using caller prefs, the
> Require header, or the Request-URI.
> 
> I would offer that it is not a preference: I really want the chess
> application; I probably do not want to talk to you.
> 
> That leaves us with the Require header or a Request-URI.  My leanings
> are
> not surprisingly towards the RFC 4240 approach.  Frankly, if you are
> building an IMS (as an example) that can handle 10,000,000 subscribers,
> you
> should be able to handle 100,000,000 registrations.  That order of
> magnitude
> is way easier than the 10,000 to 10,000,000 jump.  I don't buy the
> "registering each service is too much of a burden on the operator"
> argument,
> because it is not that much of a burden, and the network actually does
> the
> right thing.  Better yet, all of the corner cases work.  Best of all,
> the
> IETF doesn't have to build an IANA Registry and process a thousand PS
> documents to register urn:service:myco:widget.
> 
> I suppose I could live with Require:, only if you are way too lazy to
> register each application instance.  It does semantically work, because
> if I
> want to do PoC and you do not do PoC, I'll get a 4xx response, to which
> I
> could try a regular call.
> 
> _______________________________________________________________________
> Notice:  This email message, together with any attachments, may contain
> information  of  BEA Systems,  Inc.,  its subsidiaries  and  affiliated
> entities,  that may be confidential,  proprietary,  copyrighted  and/or
> legally privileged, and is intended solely for the use of the individual
> or entity named in this message. If you are not the intended recipient,
> and have received this message in error, please immediately return this
> by email and then delete it.
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 
> 
> 
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential information, privileged material (including material protected by the solicitor-client or other applicable privileges), or constitute non-public information. Any use of this information by anyone other than the intended recipient is prohibited. If you have received this transmission in error, please immediately reply to the sender and delete this information from your system. Use, dissemination, distribution, or reproduction of this transmission by unintended recipients is not authorized and may be unlawful.
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP





From sipping-bounces@ietf.org Thu Feb 22 20:16:56 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKP2U-0004sN-1B; Thu, 22 Feb 2007 20:15:58 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKP2S-0004sG-Hq
	for sipping@ietf.org; Thu, 22 Feb 2007 20:15:56 -0500
Received: from smtp.mitel.com ([216.191.234.102])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKP2Q-0000LH-7C
	for sipping@ietf.org; Thu, 22 Feb 2007 20:15:56 -0500
Received: from localhost (smtp.mitel.com [127.0.0.1])
	by smtp.mitel.com (Postfix) with ESMTP id ED38B200D7;
	Thu, 22 Feb 2007 20:15:53 -0500 (EST)
Received: from smtp.mitel.com ([127.0.0.1])
	by localhost (smtp.mitel.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id 24599-02; Thu, 22 Feb 2007 20:15:53 -0500 (EST)
Received: from kanmta01.mitel.com (kanmta01 [134.199.37.58])
	by smtp.mitel.com (Postfix) with ESMTP id 519FE200BF;
	Thu, 22 Feb 2007 20:15:53 -0500 (EST)
In-Reply-To: <1172189634.17141.59.camel@montag.eng.level3.com>
To: Daryl Malas <daryl@level3.net>
Subject: Re: [Sipping] Session Establishment Efficiency Rate
MIME-Version: 1.0
X-Mailer: Lotus Notes Release 6.5.5 November 30, 2005
Message-ID: <OF5F1C97AC.89F9CB4E-ON8525728B.0006B1EA-8525728B.0006F1A0@mitel.com>
From: peter_blatherwick@mitel.com
Date: Thu, 22 Feb 2007 20:15:50 -0500
X-MIMETrack: Serialize by Router on kanmta01/Mitel(Release 5.0.12 |February 13,
	2003) at 02/22/2007 08:15:50 PM,
	Serialize complete at 02/22/2007 08:15:50 PM
X-Virus-Scanned: by amavisd-new (virusonly) at mitel.com
X-Spam-Score: 0.3 (/)
X-Scan-Signature: 6e922792024732fb1bb6f346e63517e4
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1610285067=="
Errors-To: sipping-bounces@ietf.org

This is a multipart message in MIME format.
--===============1610285067==
Content-Type: multipart/alternative;
	boundary="=_alternative 0006F19F8525728B_="

This is a multipart message in MIME format.
--=_alternative 0006F19F8525728B_=
Content-Type: text/plain; charset="US-ASCII"

Yes, it seems a good measure of the efficiency of the system mechanics (if 
I understand it right). 

-- Peter Blatherwick





Daryl Malas <daryl@level3.net>
22.02.07 19:13
 
        To:     sipping@ietf.org
        cc: 
        Subject:        [Sipping] Session Establishment Efficiency Rate


A couple more people have expressed to me that this adaptation of the
ITU-T E.411 Network Efficiency Rate metric is very valuable.  The
snapshot of the metric from the draft is:

"This metric is similar to Session Establishment Rate, but it is
intended to exclude the effects of customer behavior and user agent
behavior.  Essentially, the attempts resulting with a final cause code
of 480, 486, and 600 would be added to the SER numerator."

Does anyone else want to see this metric in the draft?  I personally
think it is an interesting metric.

Thanks...

--Daryl



_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP


--=_alternative 0006F19F8525728B_=
Content-Type: text/html; charset="US-ASCII"


<br><font size=2 face="sans-serif">Yes, it seems a good measure of the
efficiency of the system mechanics (if I understand it right). </font>
<br>
<br><font size=2 face="sans-serif">-- Peter Blatherwick</font>
<br>
<br>
<br>
<br>
<table width=100%>
<tr valign=top>
<td>
<td><font size=1 face="sans-serif"><b>Daryl Malas &lt;daryl@level3.net&gt;</b></font>
<p><font size=1 face="sans-serif">22.02.07 19:13</font>
<td><font size=1 face="Arial">&nbsp; &nbsp; &nbsp; &nbsp; </font>
<br><font size=1 face="sans-serif">&nbsp; &nbsp; &nbsp; &nbsp; To:
&nbsp; &nbsp; &nbsp; &nbsp;sipping@ietf.org</font>
<br><font size=1 face="sans-serif">&nbsp; &nbsp; &nbsp; &nbsp; cc:
&nbsp; &nbsp; &nbsp; &nbsp;</font>
<br><font size=1 face="sans-serif">&nbsp; &nbsp; &nbsp; &nbsp; Subject:
&nbsp; &nbsp; &nbsp; &nbsp;[Sipping] Session Establishment Efficiency
Rate</font></table>
<br>
<br>
<br><tt><font size=2>A couple more people have expressed to me that this
adaptation of the<br>
ITU-T E.411 Network Efficiency Rate metric is very valuable. &nbsp;The<br>
snapshot of the metric from the draft is:<br>
<br>
&quot;This metric is similar to Session Establishment Rate, but it is<br>
intended to exclude the effects of customer behavior and user agent<br>
behavior. &nbsp;Essentially, the attempts resulting with a final cause
code<br>
of 480, 486, and 600 would be added to the SER numerator.&quot;<br>
<br>
Does anyone else want to see this metric in the draft? &nbsp;I personally<br>
think it is an interesting metric.<br>
<br>
Thanks...<br>
<br>
--Daryl<br>
<br>
<br>
<br>
_______________________________________________<br>
Sipping mailing list &nbsp;https://www1.ietf.org/mailman/listinfo/sipping<br>
This list is for NEW development of the application of SIP<br>
Use sip-implementors@cs.columbia.edu for questions on current sip<br>
Use sip@ietf.org for new developments of core SIP<br>
</font></tt>
<br>
--=_alternative 0006F19F8525728B_=--


--===============1610285067==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1610285067==--




From sipping-bounces@ietf.org Thu Feb 22 21:24:59 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKQ5z-00017O-H3; Thu, 22 Feb 2007 21:23:39 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKQ5y-00017E-2d
	for sipping@ietf.org; Thu, 22 Feb 2007 21:23:38 -0500
Received: from c243-111.rim.net ([216.9.243.111] helo=MHS01YKF.rim.net)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKQ5w-00050a-GH
	for sipping@ietf.org; Thu, 22 Feb 2007 21:23:38 -0500
Importance: normal
Priority: normal
Received: from ngw01ykf.rim.net ([10.102.108.30]) by MHS01YKF.rim.net with
	Microsoft SMTPSVC(6.0.3790.211); Thu, 22 Feb 2007 21:23:25 -0500
Received: from XCH21YKF.rim.net ([10.102.100.36]) by ngw01ykf.rim.net (SAVSMTP
	3.1.6.45) with SMTP id M2007022221232531699 ;
	Thu, 22 Feb 2007 21:23:25 -0500
Received: from XCH47YKF.rim.net ([10.64.31.217]) by XCH21YKF.rim.net with
	Microsoft SMTPSVC(5.0.2195.6713); Thu, 22 Feb 2007 21:22:53 -0500
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.181
Content-Class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification Model
Date: Thu, 22 Feb 2007 21:22:52 -0500
Message-ID: <61968779B8AC4C4BAB421D4C12F008C009463E2E@XCH47YKF.rim.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification Model
thread-index: AcdW3k4vii1EzpdTQE+w6H+3A+JX3wACqd0Q
From: "Andrew Allen" <aallen@rim.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>
X-OriginalArrivalTime: 23 Feb 2007 02:22:53.0894 (UTC)
	FILETIME=[86052660:01C756F1]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: fcb459c204557d9509ce9c1b55d771f1
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Paul

I think there are more than those 2. I listed several potential services
in my earlier email which you responded to with some points that I
haven't had a chance to respond back to you on that yet (it was a rather
long list). However I still think those are all valid from the 3GPP
point of view of services. As well as the issues of identifying the
service related to the request for routing to Application Servers, and
routing to the correct UAS and identifying the correct application at
the UAS there are also the operator related requirements for
classification of requests for charging, QoS Authorization and
subscription based authorization purposes and for identifying,
accounting and understanding the type of traffic in their network. The
following such requirements exist in the 3GPP stage 2 TS 23.228:


2.   Based on operator policy the S-CSCF or an AS shall be able to
validate an IMS communication service identifier in a SIP request. This
includes e.g. to check the syntactical correctness of a service
identifier, and policing the usage of a communication service
identifier.

8.   Based on operator policy S-CSCF and AS shall consider the IMS
communication service identifier for online and offline charging, e.g.
put appropriate data into call detailed records.

9.   The communication service identifier shall be capable of being an
input into the policy control and charging rules.

10. It shall be possible to use the IMS communication service identifier
as a means to authorise whether a subscriber is allowed to initiate or
receive request for a communication service.


So even when there are other various means to discriminate between one
type of request and another in order to choose the appropriate
Application Server or to choose the appropriate application the
mechanism to identify the request are service specific and therefore
cannot meet the requirements of a common mechanism that meets
requirements 2,8,9 and 10 above.

So to list at least 10 hypothetical services that could be implemented
using SIP that may be present on a mobile handset:

1. Basic Telephony (Conversational QoS Voice only with telephony feature
set) =20
2. Multimedia Telephony (Conversational QoS Voice plus video and other
media streams with telephony feature set)
3. VoIP (Background QoS with no telephony features)
4. SMS
5. MMS
6. Instant Messaging
7. Push to Talk (PoC)
8. Network Game 1 (e.g Chess)
9. Network Game 2 (e.g Martial arts game)
10. Network Game 3 (e.g Dungeons & Dragons)
11. Push Service 1 (e.g. Traffic information service)
12. Push Service 2 (e.g. Weather information service)
13. Push Service 3 (e.g  News information service)
14. Push Service 4 (e.g  Stock information service)
15. Mobile TV service

So there's 15 which may all be different applications on a handset and
all those are existing services or those already envisioned.

Andrew

-----Original Message-----
From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
Sent: Thursday, February 22, 2007 7:05 PM
To: Andrew Allen
Cc: Eric Burger; sipping@ietf.org
Subject: Re: [Sipping] Service Identification Model



Andrew Allen wrote:
> Hi Eric,
>=20
> It seems your prefer the approach of using URIs to identify services
> which is part of Jonathan's draft as is the use of Require for those
> services which simply do not interoperate unless the UAS understands
the
> service.
>=20
> Where I am concerned is that you seem to be suggesting that the UA
> separately register an AoR for each service it supports using
REGISTER.
> Certainly for cellular that wouldn't be acceptable as 10 or more
> services could easily be supported by a single device.

Andrew, can you give me an example of 10 "services" on the device that=20
need to be distinguished from one another in this way - that aren't just

optional behavior of a particular service? (E.g. I see no reason why=20
"audio" and "audio/video" need to be separate services.)

I guess as things stand, PoC must be a separate service. And apparently=20
Chess. What else?

	Paul

> Registrations
> create a significant load in the IMS network because its not just a
SIP
> REGISTER but also requires a lot of interactions with other network
> elements such as the HSS and Notifications for reg-events etc. Also
> having an order of magnitude increase in multiple registrations is
going
> to be a big issue for the avalanche restart problem for example when a
> cell site goes down and comes back into service.
>=20
> I think Jonathan's approach gives us the possibility for the registrar
> to create the Service URIs based on the feature parameters included in
> the Contact of the REGISTER.=20
>=20
> I think the argument over whether the model is one UA with multiple
> services or multiple UAs each representing a separate service is
> somewhat academic. As I understand a UA can register multiple AoRs and
> can have other AoRs configured or otherwise provided by the service
> provider.=20
>=20
> Andrew
>=20
> -----Original Message-----
> From: Eric Burger [mailto:eburger@bea.com]=20
> Sent: Wednesday, February 21, 2007 1:29 PM
> To: sipping@ietf.org
> Subject: [Sipping] Service Identification Model=20
>=20
> Folks have been pestering me for a while to look at the Service
> Identification issue.  I finally have gotten around to it.
>=20
>=20
> I guess I have a fundamental problem with the principal assertion of
the
> service-identification draft.  Namely, the assertion that there is a
> single
> SIP Layer, separate and distinct from two services.  Here is Figure 2
> from
> the draft:
>=20
>                            +---------------------------------+
>                            |                                 |
>                            | +-------------+ +-------------+ |
>                            | |      UI     | |      UI     | |
>                            | +-------------+ +-------------+ |
>                            | +-------------+ +-------------+ |
>                            | |             | |             | |
>                            | |  Service 1  | |  Service 2  | |
>                            | |             | |             | |
>                            | +-------------+ +-------------+ |
>                            | +-----------------------------+ |
>                            | |                             | |
>                            | |             SIP             | |
>                            | |            Layer            | |
>                            | |                             | |
>                            | +-----------------------------+ |
>                            |                                 |
>                            +---------------------------------+
>=20
>                                      Physical Device
>=20
> I would offer this drawing confuses a common SIP Stack with a common
> User
> Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use case:
>=20
>> ... In addition users may have
>> multiple devices only one of which has the application that
> understands
>> the game. Other devices that the user has registered against his AoR
> may
>> handle audio for telephony and other voice applications but if Alice
>> invites Bob to play her favourite game it is not very interesting if
> Bob
>> accepts the gaming session on a device that can only accept the audio
>> stream. Also even when the terminal that has the game application
>> receives the INVITE the correct application in that terminal needs to
> be
>> identified to accept the session.
>=20
> I would offer that this is the clinching argument that the request is,
> in
> fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a
User
> Agent.
>=20
> Same argument for PoC versus a regular voice call: One is looking for
a
> PoC
> User Agent, NOT the PoC service at a given User Agent.
>=20
> Thus, the picture really looks like the following:
>=20
>                            +---------------------------------+
>                            |                                 |
>                            | +-------------+ +-------------+ |
>                            | |    UI       | |    UI       | |
>                            | +-------------+ +-------------+ |
>                            | +-------------+ +-------------+ |
>                            | |             | |             | |
>                            | |    UAS 1    | |    UAS 2    | |
>                            | |             | |             | |
>                            | +-------------+ +-------------+ |
>                            | +-----------------------------+ |
>                            | |                             | |
>                            | |             SIP             | |
>                            | |            Stack            | |
>                            | |                             | |
>                            | +-----------------------------+ |
>                            |                                 |
>                            +---------------------------------+
>=20
>                                      Physical Device
>=20
>=20
> Another argument for different User Agents is the whole "route based
on
> SDP"
> argument.  As the author of RFC 3458, I think I was first to make the
> assertion that knowing the media type did NOT mean you know the
> application
> to use.  I don't want your phone or proxy or B2BUA to guess that
because
> I
> have Chess Media, I want the Chess Application; I want to talk with
your
> Chess Application.
>=20
> Now, is the Chess Application a Preference, Requirement, or the thing
> that I
> am addressing?  The answer drives one towards using caller prefs, the
> Require header, or the Request-URI.
>=20
> I would offer that it is not a preference: I really want the chess
> application; I probably do not want to talk to you.
>=20
> That leaves us with the Require header or a Request-URI.  My leanings
> are
> not surprisingly towards the RFC 4240 approach.  Frankly, if you are
> building an IMS (as an example) that can handle 10,000,000
subscribers,
> you
> should be able to handle 100,000,000 registrations.  That order of
> magnitude
> is way easier than the 10,000 to 10,000,000 jump.  I don't buy the
> "registering each service is too much of a burden on the operator"
> argument,
> because it is not that much of a burden, and the network actually does
> the
> right thing.  Better yet, all of the corner cases work.  Best of all,
> the
> IETF doesn't have to build an IANA Registry and process a thousand PS
> documents to register urn:service:myco:widget.
>=20
> I suppose I could live with Require:, only if you are way too lazy to
> register each application instance.  It does semantically work,
because
> if I
> want to do PoC and you do not do PoC, I'll get a 4xx response, to
which
> I
> could try a regular call.
>=20
>
_______________________________________________________________________
> Notice:  This email message, together with any attachments, may
contain
> information  of  BEA Systems,  Inc.,  its subsidiaries  and
affiliated
> entities,  that may be confidential,  proprietary,  copyrighted
and/or
> legally privileged, and is intended solely for the use of the
individual
> or entity named in this message. If you are not the intended
recipient,
> and have received this message in error, please immediately return
this
> by email and then delete it.
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
>=20
>=20
>=20
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential
information, privileged material (including material protected by the
solicitor-client or other applicable privileges), or constitute
non-public information. Any use of this information by anyone other than
the intended recipient is prohibited. If you have received this
transmission in error, please immediately reply to the sender and delete
this information from your system. Use, dissemination, distribution, or
reproduction of this transmission by unintended recipients is not
authorized and may be unlawful.
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
>=20



---------------------------------------------------------------------
This transmission (including any attachments) may contain confidential =
information, privileged material (including material protected by the =
solicitor-client or other applicable privileges), or constitute =
non-public information. Any use of this information by anyone other than =
the intended recipient is prohibited. If you have received this =
transmission in error, please immediately reply to the sender and delete =
this information from your system. Use, dissemination, distribution, or =
reproduction of this transmission by unintended recipients is not =
authorized and may be unlawful.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 02:46:30 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKV7L-0003ML-QC; Fri, 23 Feb 2007 02:45:23 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKV7K-0003L7-C7
	for sipping@ietf.org; Fri, 23 Feb 2007 02:45:22 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKV7D-0006bj-T8
	for sipping@ietf.org; Fri, 23 Feb 2007 02:45:22 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	E6EA22125A; Fri, 23 Feb 2007 08:44:47 +0100 (CET)
X-AuditID: c1b4fb3e-ae6d0bb0000007e1-33-45de9b6f986a 
Received: from esealmw129.eemea.ericsson.se (unknown [153.88.254.124])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	D937B2125D; Fri, 23 Feb 2007 08:44:47 +0100 (CET)
Received: from esealmw106.eemea.ericsson.se ([153.88.200.69]) by
	esealmw129.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 08:44:47 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Session Establishment Efficiency Rate
Date: Fri, 23 Feb 2007 08:44:48 +0100
Message-ID: <97F2AD7AA4EAD948827639B06CF2745F90A139@esealmw106.eemea.ericsson.se>
In-Reply-To: <1172189634.17141.59.camel@montag.eng.level3.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Session Establishment Efficiency Rate
Thread-Index: AcdW3nyxyftn4PNaRiWqQxE5N0yMdAAP7aAQ
From: "Marian Delkinov \(LN/EAB\)" <marian.delkinov@ericsson.com>
To: "Daryl Malas" <daryl@level3.net>
X-OriginalArrivalTime: 23 Feb 2007 07:44:47.0590 (UTC)
	FILETIME=[7DE14860:01C7571E]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8b30eb7682a596edff707698f4a80f7d
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

=20
Hi Daryl,

(Sorry for not being able to reply earlier on your previous mail. )

Yes, I confirm again, I also find it very valuable to have this metric.
Thanks for considering it.

Best regards!
Mario.
-----Original Message-----
From: Daryl Malas [mailto:daryl@level3.net]=20
Sent: Friday, 23 February, 2007 01:14
To: sipping@ietf.org
Subject: [Sipping] Session Establishment Efficiency Rate

A couple more people have expressed to me that this adaptation of the
ITU-T E.411 Network Efficiency Rate metric is very valuable.  The
snapshot of the metric from the draft is:

"This metric is similar to Session Establishment Rate, but it is
intended to exclude the effects of customer behavior and user agent
behavior.  Essentially, the attempts resulting with a final cause code
of 480, 486, and 600 would be added to the SER numerator."

Does anyone else want to see this metric in the draft?  I personally
think it is an interesting metric.

Thanks...

--Daryl



_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 04:50:00 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKX2s-0001Fr-Qu; Fri, 23 Feb 2007 04:48:54 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKX2r-0001Dm-Bg
	for sipping@ietf.org; Fri, 23 Feb 2007 04:48:53 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKX2h-0002yh-As
	for sipping@ietf.org; Fri, 23 Feb 2007 04:48:53 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	AB1E7211FE; Fri, 23 Feb 2007 10:48:40 +0100 (CET)
X-AuditID: c1b4fb3e-ae6d0bb0000007e1-5a-45deb878fc9e 
Received: from esealmw129.eemea.ericsson.se (unknown [153.88.254.124])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	8F44A205B2; Fri, 23 Feb 2007 10:48:40 +0100 (CET)
Received: from esealmw106.eemea.ericsson.se ([153.88.200.69]) by
	esealmw129.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 10:48:40 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] RE: draft-malas-performance-metrics-06
Date: Fri, 23 Feb 2007 10:48:40 +0100
Message-ID: <97F2AD7AA4EAD948827639B06CF2745F90A36D@esealmw106.eemea.ericsson.se>
In-Reply-To: <9A5D0F8924D0AA4BBEA4B7E0B0D450E2013AA56F@idc1exc0006.corp.global.level3.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] RE: draft-malas-performance-metrics-06
Thread-Index: AcdJwqq2KVrk+P49QbivLfsE/PiKpAF/uY4gAdo4D9A=
From: "Marian Delkinov \(LN/EAB\)" <marian.delkinov@ericsson.com>
To: "Malas, Daryl" <Daryl.Malas@Level3.com>
X-OriginalArrivalTime: 23 Feb 2007 09:48:40.0362 (UTC)
	FILETIME=[CC2848A0:01C7572F]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 1.1 (+)
X-Scan-Signature: bc6181926481d86059e678c9f7cb8b34
Cc: sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi,

(Of course - no problem to copy to the mailing list. I don't mind and
would be happy to see more opinions)

Regarding including diagrams, did you receive more support about it?

Regarding the metrics:

1) Chapter 5.6. Additional Metrics for Consideration
Session Establishment Efficiency Rate: Please, see my comments in a
separate mail with subject "[Sipping] Session Establishment Efficiency
Rate".

2) Session Completion Rate (SCR)

My proposal was:

	 # of Successfully Completed Sessions
SCR% =3D ------------------------------------------ x 100
	 # of Successfully Established Sessions

Your comment:

[DM] In concept I understand what you are trying to get at with this
change; however, this will require defining what an "Established
Session" means.  This will become very tricky.  Instead, I think you can
use the SCR as it is defined and then provide greater granularity by
calculating it for calls above and below a time certain duration.  What
do you think?

[MD] I thought that as "Established Session" we can consider one that
got a 200 OK response as result of an INVITE. Perhaps the definition in
chapter 2 of the draft can be used:

Session Establishment - Session establishment occurs when a 200 OK=20
   response from the UAS has been received, in response to a=20
   corresponding UAC's INVITE setup request, indicating the session=20
   setup request was successful.=20

It's another question how we can get both counters from above in order
to calculate the metric. Some vendors provide # of BYEs and # of OKs,
but in those cases the counters include the repeated BYEs and OKs, which
makes the metric wrong. Some other of course have counters for the
sessions in fact. But I believe that if this gets into an RFC, most of
the vendors will include the session counters as needed.=20

Best regards!
Mario.


=20

-----Original Message-----
From: Malas, Daryl [mailto:Daryl.Malas@Level3.com]=20
Sent: Wednesday, 14 February, 2007 00:02
To: Marian Delkinov (LN/EAB)
Cc: sipping
Subject: [Sipping] RE: draft-malas-performance-metrics-06

Marian...Comments inline...Also, I copied the Sipping mailing list,
because it is helpful to allow others within the Sipping community to
provide insight, thoughts into your comments as well.  I hope you do not
mind.

Thanks...--Daryl

> -----Original Message-----
> From: Marian Delkinov (LN/EAB) [mailto:marian.delkinov@ericsson.com]
> Sent: Tuesday, February 06, 2007 12:45 AM
> To: Malas, Daryl
> Subject: draft-malas-performance-metrics-06
>=20
> Hi,
>=20
> I'm working on implementing some IMS performance metrics into our SLA=20
> solutions. After performing a lot of searching through the different=20
> standardization bodies, the most useful document I found was your
draft.
>=20
>=20
> Would like to comment on some of the metrics and perhaps on the
generic
> approach in the document.
>=20
> Generic approach:
>=20
> It would be good to give some background information about what are
the
> major reasons for SIP calls being lost. A picture similar to the one=20
> used in ITU E411 recommendation for the interpretation of the network=20
> performance data would really help. Please, see example in attachment.
>=20
[DM] I could do this if the general mailing list agrees with the
request.  I would like to get their opinion as well.  In either case, I
will consider it for future revisions of the document.  I agree that
diagrams, in general, usually help provide clarity around concepts.
=20
> Metrics:
>=20
> Since dealing with SLAs, I would like to put the things more into a=20
> Customer perspective. Therefore, I'm more focused on how the service=20
> looks for the customer. First of all what is the service accessibility

> and secondly what is the service performance. Perhaps you can help me=20
> find out the most suitable metrics that you discuss in your document.
>=20
> 1) I try to define the SIP based telephony service effectiveness:
>=20
> The probability that an end user can access the IMS (SIP based)=20
> Telephony Service.
>=20
> In the PSTN/ISDN terms this is defined by ITU in Recommendation 411 as

> Network Effectiveness Rate:
>=20
>             Seizures Delivered to the far end terminal NER% =3D=20
> ---------------------------------------------------- x 100
> 		       Total Number of Seizures
>=20
> Note that in this case the end user behaviour is excluded, so seizures

> that reach answer signal, or user busy, or ring no answer are counted
as
> successful.
>=20
> When I look in your draft document, I don't clearly see similar
metric.
[DM]  If you look in the "Additional Metrics for Consideration" section
of the draft, I think you will find a similar metric to what you are
describing. Again, if you would like to see the metric included in the
document, please state that on the mailing list.  Thanks...

> However if used some of your definitions, it could perhaps be defined
as
> Session Effectiveness Rate (or any other suitable definition, e.g.
> Telephony Service Availability) as follows:
>=20
>          Total # of Session Requests - # of Failed Session Completions

> SEFR% =3D =
--------------------------------------------------------------
x
> 100
> 			Total # of Session Requests
>=20
> Or, if in the above the "# of Failed Session Completion" includes the=20
> end user behaviour, then perhaps a better definition would be:
>=20
>          Total # of Session Requests - # of ISA SEFR% =3D=20
> --------------------------------------------------------------
x
> 100
> 			Total # of Session Requests
>=20
> So, I would be really grateful if you share your opinion about that.
>=20
> 2) I also try to define the SIP based telephony call completion ratio.
> According to your document the Session Completion Rate (SCR) is the=20
> parameter that matches that. However I think of a more useful=20
> definition, that would really present the ability of the network to=20
> maintain successfully initiated session in the time. So, I would like
to
> hear what you think about the following definition proposal:
>=20
> Session Completion Rate is the probability that a successfully=20
> established session is maintained for a predetermined time until it is

> released intentionally by any of the UAs.
>=20
> According to your document it is defined as:
>          # of Successfully Completed Sessions SCR% =3D=20
> ---------------------------------------
>           Total # of Session Requests
>=20
> But if we follow my proposal it would be:
>=20
>          # of Successfully Completed Sessions SCR% =3D=20
> ------------------------------------------ x 100
>          # of Successfully Established Sessions
>=20
>=20
>=20
[DM] In concept I understand what you are trying to get at with this
change; however, this will require defining what an "Established
Session" means.  This will become very tricky.  Instead, I think you can
use the SCR as it is defined and then provide greater granularity by
calculating it for calls above and below a time certain duration.  What
do you think?

This is actually something we can perhaps easily measure, however it
> would be good if we can show this parameter in dependence with the=20
> Session Duration and that I haven't taken into account. In many cases,

> when jitter and packet loss increase, the sessions are terminated=20
> unintentionally. And the probability this to happen raises in time.
For
> example, for short sessions the SCR may look good, but for long
sessions
> it deteriorates.
>=20
> So, I wonder how to link this with the session duration time in order
to
> have a single parameter that really represents the network
performance.
> Will be happy to exchange ideas.
>=20
> Best regards!
> Mario.
>=20
> Marian Delkinov
> Senior Consultant
>=20
> Network Management and Service Assurance
>=20
> ERICSSON AB
> Lindholmspiren 11
> SE-417 56 Gothenburg
> Sweden
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 06:23:53 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKYVP-0004AU-Tj; Fri, 23 Feb 2007 06:22:27 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKYVO-0004AO-St
	for sipping@ietf.org; Fri, 23 Feb 2007 06:22:26 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKYVI-0002E7-9b
	for sipping@ietf.org; Fri, 23 Feb 2007 06:22:26 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	A11FC2089A; Fri, 23 Feb 2007 12:22:15 +0100 (CET)
X-AuditID: c1b4fb3c-af7c7bb0000007de-f5-45dece67dbb0 
Received: from esealmw129.eemea.ericsson.se (unknown [153.88.254.124])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	7BB05207EC; Fri, 23 Feb 2007 12:22:15 +0100 (CET)
Received: from esealmw106.eemea.ericsson.se ([153.88.200.69]) by
	esealmw129.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 12:22:15 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Session Establishment Efficiency Rate
Date: Fri, 23 Feb 2007 12:22:14 +0100
Message-ID: <97F2AD7AA4EAD948827639B06CF2745F90A48A@esealmw106.eemea.ericsson.se>
In-Reply-To: <1172189634.17141.59.camel@montag.eng.level3.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Session Establishment Efficiency Rate
Thread-Index: AcdW3nyxyftn4PNaRiWqQxE5N0yMdAAUZX5Q
From: "Marian Delkinov \(LN/EAB\)" <marian.delkinov@ericsson.com>
To: "Daryl Malas" <daryl@level3.net>
X-OriginalArrivalTime: 23 Feb 2007 11:22:15.0288 (UTC)
	FILETIME=[DEE9FF80:01C7573C]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 538aad3a3c4f01d8b6a6477ca4248793
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi,

If the Network Efficiency Rate is designed to express the ability of the
network to deliver calls (in SIP terms one should perhaps say "establish
sessions"), I think that it might be reasonable to include few more
cases (cause codes) in the numerator: e.g. all redirection 3xx codes,
all 4xx except 408 codes and all 6xx codes. (If you agree that those are
the ones that exclude network failures, congestion and any other network
malfunction).

Actually it sounds fair to me if it's the opposite of the Ineffective
Session Attempts (ISA):

	   Total # of Invite Requests - # of ISA
SEER% =3D -------------------------------------------------------------- =
x
100
	   Total # of Invite Requests

What do you think?

Best regards!
Mario.

-----Original Message-----
From: Daryl Malas [mailto:daryl@level3.net]=20
Sent: Friday, 23 February, 2007 01:14
To: sipping@ietf.org
Subject: [Sipping] Session Establishment Efficiency Rate

A couple more people have expressed to me that this adaptation of the
ITU-T E.411 Network Efficiency Rate metric is very valuable.  The
snapshot of the metric from the draft is:

"This metric is similar to Session Establishment Rate, but it is
intended to exclude the effects of customer behavior and user agent
behavior.  Essentially, the attempts resulting with a final cause code
of 480, 486, and 600 would be added to the SER numerator."

Does anyone else want to see this metric in the draft?  I personally
think it is an interesting metric.

Thanks...

--Daryl



_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 06:34:03 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKYg8-0007kz-SA; Fri, 23 Feb 2007 06:33:32 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKYg7-0007jJ-Aw
	for sipping@ietf.org; Fri, 23 Feb 2007 06:33:31 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKYdJ-0003xH-6Q
	for sipping@ietf.org; Fri, 23 Feb 2007 06:30:51 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	651C821335; Fri, 23 Feb 2007 12:29:40 +0100 (CET)
X-AuditID: c1b4fb3e-afed3bb0000007e1-8f-45ded02433ee 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	5864321331; Fri, 23 Feb 2007 12:29:40 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 12:29:39 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP - Routing
	decissionsbasedon Require/P-R
Date: Fri, 23 Feb 2007 12:29:32 +0100
Message-ID: <7374777208BDC7449D5620EF94232567023932C9@esealmw113.eemea.ericsson.se>
In-Reply-To: <200702221520.l1MFK1F9022382@dragon.ariadne.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP - Routing
	decissionsbasedon Require/P-R
Thread-Index: AcdWlSDeAXXTFc20RtCf+FZxAPqbHgApQ5Mg
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: <Dale.Worley@comcast.net>,
	<sipping@ietf.org>
X-OriginalArrivalTime: 23 Feb 2007 11:29:39.0954 (UTC)
	FILETIME=[E7F4A520:01C7573D]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 08170828343bcf1325e4a0fb4584481c
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,=20

>>>I am sure a proxy can make routing decissions based on whatever.
>=20
>>My question was simply where it is said that=20
>>Require/Proxy-Require is more appropriate.
>=20
>More appropriate than what?

Accept-Contact, for example.

Regards,

Christer

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 06:58:42 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKZ46-0004Hd-RA; Fri, 23 Feb 2007 06:58:18 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKZ45-0004HU-3h
	for sipping@ietf.org; Fri, 23 Feb 2007 06:58:17 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKYcU-0003jy-RI
	for sipping@ietf.org; Fri, 23 Feb 2007 06:30:15 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	4B2D120975; Fri, 23 Feb 2007 12:29:44 +0100 (CET)
X-AuditID: c1b4fb3c-affc8bb0000007de-d8-45ded02889be 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	3E96020956; Fri, 23 Feb 2007 12:29:44 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 12:29:44 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP - Routing decissionsbased
	on Require/P-R
Date: Fri, 23 Feb 2007 12:29:36 +0100
Message-ID: <7374777208BDC7449D5620EF94232567023932CA@esealmw113.eemea.ericsson.se>
In-Reply-To: <200702221534.l1MFYwbG023473@dragon.ariadne.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP - Routing
	decissionsbased on Require/P-R
Thread-Index: AcdWl2pLogHS6wXOQsyIkDK1d4xKOAAoxoiw
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: <Dale.Worley@comcast.net>,
	<sipping@ietf.org>
X-OriginalArrivalTime: 23 Feb 2007 11:29:44.0130 (UTC)
	FILETIME=[EA71DA20:01C7573D]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 538aad3a3c4f01d8b6a6477ca4248793
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,=20

> -----Original Message-----
> From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net]=20
> Sent: 22. helmikuuta 2007 17:35
> To: sipping@ietf.org
> Subject: Re: [Sipping] Service Identification in SIP -=20
> Routing decissionsbased on Require/P-R
>=20
>    From: Henning Schulzrinne <hgs@cs.columbia.edu>
>=20
>    Proxies can choose next hops based on RFC 3840, which is=20
> another way =20
>    of saying "routing".
>=20
> RFC 3261 allows a proxy unlimited latitude in choosing=20
> targets to which to forward a request, if the proxy is=20
> authoritative for the
> request-URI:
>=20
> "[T]he element is responsible for the domain in the=20
> Request-URI, and the element MAY use whatever mechanism it=20
> desires to determine where to send the request."
>=20
> Within that framework, it seems to me quite reasonable for a=20
> service provider's proxy to examine Require or Proxy-Require=20
> headers to determine which application server would be best=20
> for handling a call, and routing the call to it.

Why can't Accept-Contact be used for the same purpose?

My issue is the statement that Require/Proxy-Require is the most
appropriate way of doing the routing decisson, and I am wondering where
that is specified.

Regards,

Christer





>=20
> Dale
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 07:22:56 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKZRU-0004CM-7C; Fri, 23 Feb 2007 07:22:28 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKZRR-0004CE-Sq
	for sipping@ietf.org; Fri, 23 Feb 2007 07:22:25 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKZRL-00010Y-E4
	for sipping@ietf.org; Fri, 23 Feb 2007 07:22:25 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	C20D420408; Fri, 23 Feb 2007 13:22:12 +0100 (CET)
X-AuditID: c1b4fb3e-ae6d0bb0000007e1-a6-45dedc74c0dd 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	A67EB20158; Fri, 23 Feb 2007 13:22:12 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 13:22:12 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Fri, 23 Feb 2007 13:22:10 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD02925808@esealmw111>
In-Reply-To: <45DE273B.8070306@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdW2T3SvudeJOdESAWwP6JhERq5LAAa5PiA
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>
X-OriginalArrivalTime: 23 Feb 2007 12:22:12.0508 (UTC)
	FILETIME=[3F0635C0:01C75745]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 4da339c42fe5be09fa120bb0fcc4a575
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi Paul,

The specification would be open on this (i.e. I cannot see the
specifications saying that this is rejected) as such, the situation when
a PBX (or any terminal) doesn't use a service identifer would be a
matter for operator policy (and maybe internconnect policy).  I would
assume the operator would make a guess at what the requested
communication service was ... E.g. SDP=3Dvoice (without something
indicating PoC) would be a stronge hint to assume this is multimedia
telephony.

BR,

..//Steve

=20

-----Original Message-----
From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
Sent: Friday, February 23, 2007 12:29 AM
To: Stephen Terrill (MI/EEM)
Cc: sipping@ietf.org
Subject: Re: [Sipping] Service Identification in SIP



Stephen Terrill (MI/EEM) wrote:
> Hi Andrew,
>=20
> I have the opposite view.  I think that if the originating network and

> the terminating network are going to interoperate, then they will have

> the same understanding of the service, hence in my view the same=20
> "identifier".  To have otherwise is problematic and complicated (the=20
> operators will have to have a higher admininstrative effort in order=20
> to understand how they should interconnect).
>=20
> The PBX case is more complicated than can be responded to in a single=20
> sentence - as there are many ways in which the PBXs are connected=20
> together.  In one case - the PBXs are just treated as a terminal with=20
> many users; and the service that the operator provides is really the=20
> same basic service as is provided to other users (with maybe more=20
> subscriber specific logic).  In other cases, the operator just=20
> provides routing and connecting support, and here the service provided

> by the operator is different; but I would assume that the peer PBXs=20
> had better have a common understanding of the service (which they=20
> would if it was the same vendor), or again there will be
inter-operatability issues.

What will happen here if the pbx calls an address on a 3gpp network that
is using some set of service definitions? Or visa versa.

Will the PBX, that specifies no service but that includes audio and
video, be able to establish an audio/video call? Or will it somehow have
to know the service id that the IMS system expects for audio/video
calls?

	Paul

> Regarding the URI parameter proposal, don't we have proxies that=20
> remove additional R-URIs - it seems a natural thing to remove it also
in e.g.
> fowarding services etc ...
>=20
> Regards,
>=20
> Steve.
>=20
> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com]
> Sent: Thursday, February 22, 2007 4:05 AM
> To: Stephen Terrill (MI/EEM); sipping@ietf.org
> Cc: Andrew Allen
> Subject: RE: [Sipping] Service Identification in SIP
>=20
>=20
> Hi Steve
>=20
> In my view this is one of the interoperability issues with your=20
> proposal. I think the assumption that all networks will deploy exactly

> the same services with the same service identifiers is unrealistic and

> also undesirable. Are corporate PBX vendors going to implement the=20
> same services as defined by 3GPP or OMA? The possibility to have the=20
> flexibility to have different service identifiers for session=20
> equivalent services is in my view essential to any IETF standard and=20
> to do otherwise is to create a situation where basic services like=20
> telephony or messaging cannot interoperate between users in different=20
> networks and users in the internet without the use of inter working=20
> gateway. SIP and SIMPLE were setup specifically to create protocols=20
> that would enable interoperability with IP networks without the need
for gateways.
>=20
> If 3GPP network operators choose to deploy the same services and use=20
> common identifiers that is fine but there are whole other worlds out=20
> there as well and service identifiers should not stand in the way of=20
> interoperability between all these different SIP users. If the IETF=20
> solution supports the ability for different networks to use different=20
> identifiers that doesn't violate 3GPP requirements and doesn't prevent

> 3GPP using the same identifiers in different networks. I don't see how

> that is against TS 23.228.
>=20
> Andrew
>=20
>=20
> -----Original Message-----
> From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
> Sent: Monday, February 19, 2007 6:00 AM
> To: Andrew Allen; sipping@ietf.org
> Subject: RE: [Sipping] Service Identification in SIP
>=20
> Hi Andrew,
>=20
> I see that in your assumptions, the originating and terminating=20
> networks use different service identifiers, and I presume that this is

> for the same service.  In my understanding, it is the same service=20
> identifier that is used end-to-end (if not it is interworked).  That=20
> is why I do not see the reliance on Presence being necessary.  In my=20
> opionion this totally within the requirements of TS 23.228 (I think=20
> that different service identifiers in different networks for the same=20
> service is against TS 23.228).
>=20
> This comes to the isssue of the PSTN gateway - I presume that you want

> to go to the PSTN gateway if the service is telephony, and within=20
> that, voice.  If we know that, then the we go to any PSTN gateway, and

> we can't presume a particular network for a PSTN gateway.
>=20
> Best Regards,
>=20
> Steve.
>=20
> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com]
> Sent: Saturday, February 17, 2007 12:51 AM
> To: Stephen Terrill (MI/EEM); sipping@ietf.org
> Cc: Andrew Allen
> Subject: RE: [Sipping] Service Identification in SIP
>=20
>=20
>=20
> Hi Steve,
>=20
> My responses inline
>=20
> Andrew
>=20
> -----Original Message-----
> From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
> Sent: Friday, February 16, 2007 3:51 AM
> To: sipping@ietf.org
> Subject: RE: [Sipping] Service Identification in SIP
>=20
> Hello Andrew,
>=20
> Thanks for your comments.  I have a couple of brief responses below.
>=20
> Best Regards,
>=20
> ..//Steve
>=20
> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com]
> Sent: Friday, February 16, 2007 7:15 AM
> To: Stephen Terrill (MI/EEM); sipping@ietf.org
> Cc: Andrew Allen
> Subject: RE: [Sipping] Service Identification in SIP
>=20
>=20
> Some comments on Stephen's comments on the draft:
> =20
>=20
> -Moving onto the proposed solution draft itself, in addition for=20
> providing for the identification of the "application" and the=20
> "communication service", I have a few additional thoughts.
>=20
> The draft seems to have the implication that all services=20
> (communication services in other terminology) that are to be=20
> identified have to be standardised within the IETF.  This is clearly=20
> an restriction that will not be acceptable to the industry as a whole.

> My guess is that if such a statement is mandated by the IETF, it will=20
> just be words on paper which would likely be ignored in practice by=20
> other standardisation organisations.
>=20
> [AA] This I think is an incorrect interpretation of the draft. My=20
> understanding is that the only services according to the draft that=20
> require some standardisation in IETF are those that by their very=20
> nature cannot interoperate unless there is some common additional=20
> understanding between the endpoints outside of what is currently=20
> specified in SIP. The use of SIP option tags with Require header is=20
> proposed by the draft for this purpose as such services create
interoperability issues with SIP.
>=20
> <Steve>
> My comment applies even for these services </Steve>
>=20
>=20
> <AA>
> Well then we are in big trouble because what you are proposing is that

> its OK for bodies outside of IETF to create non interoperable=20
> extensions for the behavour of SIP entities. This will result in the=20
> fragmentation of SIP into different flavors that don't interoperate.=20
> This is the balkanization of SIP that Paul talks about.even <AA>
>=20
> ----------------------------------------------------
>=20
> The draft also only recognises the case where all explicitly knowledge

> of the service is required.  I think it would also be good to allow=20
> the case where knowledge of the service would be helpful, but not=20
> explicitly required.  This could be a case where the service to the=20
> end-user would not be degraded if this was not supported.  This=20
> functionality seems to be missing.
>=20
> [AA] Service URIs are defined in the draft for these services where=20
> the support by the other endpoint is not required for the session to=20
> be sucessfully established.
>=20
> The draft seems to indicate that that application servers in the=20
> network are involved only at the request of the originating terminal.

> This seems a strange restriction.  For some network support that=20
> supports the originating user, such as the one provided in the example

> "I would like this call recorded for later retrieval", such mechanisms

> may make sense, however as a general approach it seems a bit limited.

> For example, in 3GPP networks there is the need to invoke application=20
> logic in the originating side and terminating side, not only at the=20
> request of the originating terminal, but as a consequence of the=20
> requested service.  If there is an SIP message contains indication of=20
> the requested service, a proxy can do this (or loose route to an AS=20
> that contains the necessary service logic).
>=20
> [AA] The draft supports both possibilities using the Route header for=20
> originating network servers that need to be involved in the route and=20
> based on the terminating networks routing policy for the Service URI=20
> in the Request URI.
>=20
> <Steve>
> I don't see the advantage of having different solutions for the=20
> originating network and the terminating network.  I don't think that=20
> the originating logic is invoked only at the request of the=20
> originating terminal.
> </Steve>
>=20
> <AA>
> Unless IETF is willing to standardise all services and mandate that=20
> all SIP entities understand these (which will never happen) then we=20
> cannot assume that all networks and all endpoints understand and use=20
> the same identifiers. It is possible through configuration for an=20
> endpoint to obtain the identifiers used by its own service provider=20
> but there is no way to ensure that other networks use or understand=20
> the same identifiers for the equivalent or similar service that is=20
> compatible at the session layer. Therefore we need to support the=20
> possibility to have different identifiers in different networks in=20
> order to not violate the TS 23.228
> requirement:
>=20
> 12.	The usage of communication service identifiers shall not
> adversely affect interoperability between IMS networks and=20
> interoperability with external SIP networks and CS networks. The=20
> behaviour of a network receiving the IMS requests without an IMS=20
> communication service identifier is a matter of operator policy. Usage

> of communication service identifiers shall not decrease the level of=20
> interoperability with networks and UEs that are unaware of the=20
> communication service identifier.
>=20
> If the originating service logic isn't invoked based on an identifier=20
> placed in the request by the originating terminal on what basis do you

> think it is invoked? How does the use of Accept-Contact have=20
> advantages in this case?
> <AA>
>=20
> --------------------------------------
>=20
>  I note here that the proposed service URN does not contain the=20
> address of a specific server containing the application logic.  An AS=20
> can look at that, and loose route to an application server that does=20
> contain such logic (If it is separated from the proxy), however I fail

> to see the difference for that proxy if it looks at a feature tag, or=20
> it looks at the service URN for this.  In both cases, it has to parse=20
> the message for the data, and modify the routing information.
>=20
> [AA]Route headers are the SIP mechanism for requesting intermediaries=20
> to be includede on the route of the request. There is nothing in RFC=20
> 3840/RFC 3841 that addresses using feature tags in Accept-Contact=20
> header for any routing decisions prior to determining which of the=20
> registered contacts the request should be forked to. Accept-Contact is

> not a Route header it is merely a filter that is applied to the set of

> valid contacts. Using a parameter in the filter as a service=20
> identifier has a potential significant impact on the use of=20
> Accept-Contact for its original purpose.
>=20
> <Steve>
> I don't believe there is anthing that states a Proxy CAN NOT route to=20
> applications or application logic based upon elements in the SIP=20
> message, including what is in the Accept-Contact header (some people=20
> have even proposed SDP ...).
> </Steve>
>=20
> <AA>
> As I had stated in an earlier email on this subject a proxy can do a=20
> lot of things (including routing based on your favorite number=20
> appearing in a call-Id) but that doesn't mean they are a good idea or=20
> a good design decision. If you want to ensure you don't break things=20
> then you need to be careful to use SIP headers with the semantics=20
> intended for them otherwise problems can result if extra semantics are

> put upon existing headers.
> <AA>
>=20
> Further, the route header could be removed in the originating network;

> and it is still required to identify the service in the terminating=20
> network.  As proxies may remove the route, it can't be used to=20
> transport the requested service.
>=20
> [AA] Proxies arbitarily removing Route headers is a BAD thing akin to=20
> a bull in a china shop - almost certainly going to break something.=20
> RFC
> 3261 summarises what proxies should be doing with Route headers:
> 16.12 Summary of Proxy Route Processing
>=20
>    In the absence of local policy to the contrary, the processing a
>    proxy performs on a request containing a Route header field can be
>    summarized in the following steps.
>=20
>       1.  The proxy will inspect the Request-URI.  If it indicates a
>           resource owned by this proxy, the proxy will replace it with
>           the results of running a location service.  Otherwise, the
>           proxy will not change the Request-URI.
>=20
>       2.  The proxy will inspect the URI in the topmost Route header
>           field value.  If it indicates this proxy, the proxy removes
it
>           from the Route header field (this route node has been
>           reached).
>=20
>       3.  The proxy will forward the request to the resource indicated
>           by the URI in the topmost Route header field value or in the
>           Request-URI if no Route header field is present.  The proxy
>           determines the address, port and transport to use when
>           forwarding the request by applying the procedures in [4] to
>           that URI.
>=20
>    If no strict-routing elements are encountered on the path of the
>    request, the Request-URI will always indicate the target of the
>    request.
>=20
> <Steve>
> The first sentance is interesting "In absence of local policy to the=20
> contrary", indicating that there can be different policies.  I=20
> understand that some proxies do remove this.  While I am on this - I=20
> have lead to believe that R-URI parameters may also be removed by=20
> intermediate proxies - is this correct?
> </Steve>
>=20
> <AA>
> My understanding is that local policy may determine that additional=20
> nodes need to be routed to in addition to the Route contained in the=20
> request. However deleting existing Route headers other than the Route=20
> header containing the address of the proxy as part of the loose=20
> routing procedures is liable to break things and is not part of RFC=20
> 3261 routing (loose routing).
> <AA>
>=20
>=20
> [AA] Generally a UA wouldn't know the appropriate service identifiers=20
> used in the terminating network or whether a Route header is=20
> appropriate to be used in the terminating network. The draft defines=20
> the Service URI for the purpose of identifying services in the=20
> terminating network and service URIs can be obtained using Presence.
>=20
> <Steve>
> I don't get the above paragraph.  It suggest that the "appropriate=20
> service identifier" used in the terminating network is not known by=20
> the originating UE, but the Service URI is.  I thought that the=20
> service URI is an encoding of the "appropriate service identifier".
> </Steve>
>=20
> <AA>
> Yes but in the general case the UA doesn't need to know what the=20
> service identifier is it obtains through Presence a Service URI that=20
> contains the Service identifier as a URI parameter. The great thing=20
> about using Presence to obtain a Service URI is that the originating=20
> UA doesn't even need to know anything about service identifiers at all
for this to work.
> It just uses Presence to find the URI to use for the service which can

> be indicated to the user via an indication such as a graphic icon (e,g

> a phone for telephony, smiley face for IM, etc) in the presence
document.
> <AA>
>=20
> -----------------------------------
>=20
>>From this perspective it is not clear to me the advantage of using a
> service URN in a route header compared to the use of a feature tag.
>=20
> [AA] I hope the above responses clarify that somewhat. Also the=20
> Service URN and Service URIs allow different service identifiers to be

> used in the originating and terminating network for session=20
> compatiable services instead of forcing everyone to deploy and=20
> understand the same service and service identifier.
>=20
> <Steve>
> Nope, don't get it, but are information that is carried end-to-end
....
> And as far as I see, could be used by intermediate proxies to impact=20
> local routing policy ... Or ?
> </Steve>
>=20
> <AA>
> I don't understand this response either. The key point is that UAs in=20
> corporate networks and in the internet are not going to deploy 3GPP=20
> flavor services and use the same 3GPP service identifiers or even=20
> possibly any service identifiers at all. The Accept-Contact based=20
> proposal assumes that all endpoints understand the service identifiers

> to be used for each service. This concept is not generally applicable=20
> and create interoperability problems.
> <AA>
>=20
> The draft seems to mandate GRUU - is this really so, is that really=20
> necessary.  It seems strange to place such a dependency, it also=20
> starts to move the GRUU as being the primary means of which to contact
a user.
>=20
> [AA] My understanding is that GRUU in the draft is used to address how

> a request addressed to Service URI is forked only to those contacts=20
> that support the service identified by the that Service URI. A contact

> would seem to be fine so long as it had GRUU properties - no need to=20
> obtain a GRUU if you already have one.
>=20
> This implies that the originating user decides which terminal I should

> receive my requested service on, even if I have two devices for that.
> As such, I don't believe that there should be a dependency on GRUU.
>=20
> [AA] If you have obtained a Service URI (probably through presence)=20
> for that user then you can address the request to the Service URI=20
> which represents all the terminals that support that service. However=20
> GRUU and the Service Instance URIs created from GRUUs provide another=20
> mechanism through which a user can obtain a URI to reach that user for

> the required service.
>=20
> Parts of the draft seems to rely on the use of presence.  While=20
> presence could be a supportive mechanism, I doubt that we can mandate=20
> that the number you want to dial has be available in presence as I=20
> assume not all entities are available in presence (diallable network=20
> services, users in the PSTN).  This would be strange to mandated.
>=20
> [AA] PSTN gateways are presumably in the same domain as the=20
> terminating proxy and therefore these PSTN gateways can be configured=20
> with the appropriate service identifiers for telephony services and=20
> can include this in either a Route header to route to the right=20
> application server or can construct a valid Service URI. The real=20
> problem is how do users in other networks know what the service=20
> identifier that's is used in the terminating network is. The use of=20
> Service URIs and Presence provides a solution to that problem.
>=20
> <Steve>
> Again, presence could be a useful mechanism, but I think it would be a

> problem for the draft to rely it.  A user making a call to the PSTN=20
> may not find that user in the presence archicture (and similar to=20
> services like the 800 or 900 services).  (I should note, that you=20
> shouldn't make any assumptions about the location of the PSTN gateway=20
> ... But I don't think that changes anything).
> </Steve>
>=20
> <AA>
> How does a PSTN gateway in another domain from the terminating network

> know the service identifier for telephony used in the terminating=20
> network with the Accept-Contact header based proposal?
> <AA>
>=20
> Moving on from the draft I have seen the questioned raised about "why=20
> can't we expand the power of SDP".  Such an approach would mandate=20
> that SIP methods use SDP, I don't think that is appropriate.
>=20
> [AA] I agree with you here
>=20
> I hear the appeals that the mechanisms should not be abused, and that=20
> we should not have 1000s of separate "communication services", and I=20
> fully agree with this. I see a smaller number of communication=20
> services, not a larger number; where those communication services are=20
> used to access other end-user-services.
>=20
> *flame away*
>=20
> Regards,
>=20
> Steve
>=20
> ---
>=20
>=20
>=20
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential

> information, privileged material (including material protected by the=20
> solicitor-client or other applicable privileges), or constitute=20
> non-public information. Any use of this information by anyone other=20
> than the intended recipient is prohibited. If you have received this=20
> transmission in error, please immediately reply to the sender and=20
> delete this information from your system. Use, dissemination,=20
> distribution, or reproduction of this transmission by unintended=20
> recipients is not authorized and may be unlawful.
>=20
>=20
>=20
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential

> information, privileged material (including material protected by the=20
> solicitor-client or other applicable privileges), or constitute=20
> non-public information. Any use of this information by anyone other=20
> than the intended recipient is prohibited. If you have received this=20
> transmission in error, please immediately reply to the sender and=20
> delete this information from your system. Use, dissemination,=20
> distribution, or reproduction of this transmission by unintended=20
> recipients is not authorized and may be unlawful.
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 08:09:03 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKa9H-0001Dr-Sd; Fri, 23 Feb 2007 08:07:43 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKa9F-0001Di-Sc
	for sipping@ietf.org; Fri, 23 Feb 2007 08:07:41 -0500
Received: from rtp-iport-2.cisco.com ([64.102.122.149])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKa9E-0005PC-HZ
	for sipping@ietf.org; Fri, 23 Feb 2007 08:07:41 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by rtp-iport-2.cisco.com with ESMTP; 23 Feb 2007 08:07:40 -0500
X-IronPort-AV: i="4.14,211,1170651600"; 
	d="scan'208"; a="114386092:sNHT190439314"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1ND7cCh009307; 
	Fri, 23 Feb 2007 08:07:38 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1ND7UVV002538; 
	Fri, 23 Feb 2007 08:07:37 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 08:07:36 -0500
Received: from [10.86.243.112] ([10.86.243.112]) by xfe-rtp-201.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 08:07:35 -0500
Message-ID: <45DEE717.1070609@cisco.com>
Date: Fri, 23 Feb 2007 08:07:35 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Christer Holmberg (JO/LMF)" <christer.holmberg@ericsson.com>
Subject: Re: [Sipping] Service Identification in SIP - Routing decissionsbased
	on Require/P-R
References: <7374777208BDC7449D5620EF94232567023932CA@esealmw113.eemea.ericsson.se>
In-Reply-To: <7374777208BDC7449D5620EF94232567023932CA@esealmw113.eemea.ericsson.se>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 23 Feb 2007 13:07:35.0733 (UTC)
	FILETIME=[96315E50:01C7574B]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=3317; t=1172236058;
	x=1173100058; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP=20-=2
	0Routing=20decissionsbased=0A=20on=20Require/P-R |Sender:=20
	|To:=20=22Christer=20Holmberg=20(JO/LMF)=22=20<christer.holmberg@ericsson
	.com>; bh=vL7QwB9q7WE8ERPWwIDq54UPirkUQoawIj8gkCB5jqY=;
	b=pRW0yyGuTosYY7UIWCHeLBb911awQvqoThhGrVt59ctro/9uG8StTQiW9MfcuAcZpgfq5Fsx
	30UziAIqRqPd+PMGNMuvSlWW2NAvtkakx4lsNQVTcOK4viqpOUvzrULO;
Authentication-Results: rtp-dkim-2; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 31247fb3be228bb596db9127becad0bc
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

The difference here is that the Require header is definitive that the 
caller doesn't want to reach a UAS unless that UAS supports the stated 
options. Callerprefs is not so definitive. There are some uses of 
callerprefs that can be considered fairly definitive, but in general 
that is not the case.

Jonathan proposed extending 3841 so that the use of Require generates an 
implicit callerpref. And that in turn would influence the selection of 
the registered contacts to try. I can't see a plausible way to infer any 
implicit callerpref from Accept-Contact headers.

In any case, that is all about selecting how to translate the R-URI. It 
is not about selecting intermediate proxies. As Dale has stated, you are 
free to use most anything to select an intermediate proxy. If you want 
to use Require, or Accept-Contact to do that you may do so. But while 
you are free to do so, it is less clear that it is useful or wise to do 
so. If the intent is to select a proxy that is consistent with the 
capabilities of the UAS that will eventually be accepted (as seems to be 
the case with PoC), then it only makes sense if the thing you are 
testing is definitive for the capabilities of the UAS. So again, Require 
may make sense for this, while Accept-Contact is not.

	Paul

Christer Holmberg (JO/LMF) wrote:
> Hi, 
> 
>> -----Original Message-----
>> From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net] 
>> Sent: 22. helmikuuta 2007 17:35
>> To: sipping@ietf.org
>> Subject: Re: [Sipping] Service Identification in SIP - 
>> Routing decissionsbased on Require/P-R
>>
>>    From: Henning Schulzrinne <hgs@cs.columbia.edu>
>>
>>    Proxies can choose next hops based on RFC 3840, which is 
>> another way  
>>    of saying "routing".
>>
>> RFC 3261 allows a proxy unlimited latitude in choosing 
>> targets to which to forward a request, if the proxy is 
>> authoritative for the
>> request-URI:
>>
>> "[T]he element is responsible for the domain in the 
>> Request-URI, and the element MAY use whatever mechanism it 
>> desires to determine where to send the request."
>>
>> Within that framework, it seems to me quite reasonable for a 
>> service provider's proxy to examine Require or Proxy-Require 
>> headers to determine which application server would be best 
>> for handling a call, and routing the call to it.
> 
> Why can't Accept-Contact be used for the same purpose?
> 
> My issue is the statement that Require/Proxy-Require is the most
> appropriate way of doing the routing decisson, and I am wondering where
> that is specified.
> 
> Regards,
> 
> Christer
> 
> 
> 
> 
> 
>> Dale
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP 
>> Use sip-implementors@cs.columbia.edu for questions on current 
>> sip Use sip@ietf.org for new developments of core SIP
>>
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 08:21:42 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKaMi-0004vy-26; Fri, 23 Feb 2007 08:21:36 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKaMg-0004vO-Qb
	for sipping@ietf.org; Fri, 23 Feb 2007 08:21:34 -0500
Received: from usaga01-in.huawei.com ([206.16.17.211])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKaMW-0001AK-6Y
	for sipping@ietf.org; Fri, 23 Feb 2007 08:21:34 -0500
Received: from huawei.com (usaga01-in [172.18.4.6])
	by usaga01-in.huawei.com (iPlanet Messaging Server 5.2 HotFix 1.25
	(built Mar
	3 2004)) with ESMTP id <0JDX00FA63RN3H@usaga01-in.huawei.com> for
	sipping@ietf.org; Fri, 23 Feb 2007 05:21:23 -0800 (PST)
Received: from s73602 (cpe-72-190-0-23.tx.res.rr.com [72.190.0.23])
	by usaga01-in.huawei.com
	(iPlanet Messaging Server 5.2 HotFix 1.25 (built Mar  3 2004))
	with ESMTPA id <0JDX002YI3RM3U@usaga01-in.huawei.com> for
	sipping@ietf.org; Fri, 23 Feb 2007 05:21:23 -0800 (PST)
Date: Fri, 23 Feb 2007 07:21:07 -0600
From: Spencer Dawkins <spencer@mcsr-labs.org>
Subject: Re: [Sipping] Service Identification in SIP - Routing
	decissionsbasedon Require/P-R
To: sipping@ietf.org
Message-id: <00ef01c7574d$7a1567f0$6501a8c0@china.huawei.com>
MIME-version: 1.0
X-MIMEOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
X-Mailer: Microsoft Outlook Express 6.00.2900.2869
Content-type: text/plain; format=flowed; charset=iso-8859-1;
	reply-type=original
Content-transfer-encoding: 7BIT
X-Priority: 3
X-MSMail-priority: Normal
References: <7374777208BDC7449D5620EF94232567023932CA@esealmw113.eemea.ericsson.se>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: e8a67952aa972b528dd04570d58ad8fe
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi, Christer,

I may be misunderstanding the point, but I thought the topic under 
discussion was not general proxy routing, but proxy routing for calls that 
could not be handled correctly without special knowledge - for example, 
PTT/POC.

In this case, Require/Proxy-Require seems appropriate, because the proxy 
really is "required" to do something, or the INVITE will either fail or 
succeed resulting in some strange and unusable state.

Thanks,

Spencer

From: "Christer Holmberg (JO/LMF)" <christer.holmberg@ericsson.com>
To: <Dale.Worley@comcast.net>; <sipping@ietf.org>
Sent: Friday, February 23, 2007 5:29 AM
Subject: RE: [Sipping] Service Identification in SIP - Routing 
decissionsbasedon Require/P-R



Hi,

> -----Original Message-----
> From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net]
> Sent: 22. helmikuuta 2007 17:35
> To: sipping@ietf.org
> Subject: Re: [Sipping] Service Identification in SIP -
> Routing decissionsbased on Require/P-R
>
>    From: Henning Schulzrinne <hgs@cs.columbia.edu>
>
>    Proxies can choose next hops based on RFC 3840, which is
> another way
>    of saying "routing".
>
> RFC 3261 allows a proxy unlimited latitude in choosing
> targets to which to forward a request, if the proxy is
> authoritative for the
> request-URI:
>
> "[T]he element is responsible for the domain in the
> Request-URI, and the element MAY use whatever mechanism it
> desires to determine where to send the request."
>
> Within that framework, it seems to me quite reasonable for a
> service provider's proxy to examine Require or Proxy-Require
> headers to determine which application server would be best
> for handling a call, and routing the call to it.

Why can't Accept-Contact be used for the same purpose?

My issue is the statement that Require/Proxy-Require is the most
appropriate way of doing the routing decisson, and I am wondering where
that is specified.

Regards,

Christer 



_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 08:23:54 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKaOT-0005yD-I7; Fri, 23 Feb 2007 08:23:25 -0500
Received: from [10.90.34.44] (helo=chiedprmail1.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKaOS-0005x6-84
	for sipping@ietf.org; Fri, 23 Feb 2007 08:23:24 -0500
Received: from usaga01-in.huawei.com ([206.16.17.211])
	by chiedprmail1.ietf.org with esmtp (Exim 4.43) id 1HKaOM-000203-AT
	for sipping@ietf.org; Fri, 23 Feb 2007 08:23:24 -0500
Received: from huawei.com (usaga01-in [172.18.4.6])
	by usaga01-in.huawei.com (iPlanet Messaging Server 5.2 HotFix 1.25
	(built Mar
	3 2004)) with ESMTP id <0JDX00FOM3UP3H@usaga01-in.huawei.com> for
	sipping@ietf.org; Fri, 23 Feb 2007 05:23:13 -0800 (PST)
Received: from s73602 (cpe-72-190-0-23.tx.res.rr.com [72.190.0.23])
	by usaga01-in.huawei.com
	(iPlanet Messaging Server 5.2 HotFix 1.25 (built Mar  3 2004))
	with ESMTPA id <0JDX0028J3UN3U@usaga01-in.huawei.com> for
	sipping@ietf.org; Fri, 23 Feb 2007 05:23:13 -0800 (PST)
Date: Fri, 23 Feb 2007 07:22:56 -0600
From: Spencer Dawkins <spencer@mcsr-labs.org>
Subject: Re: [Sipping] Service Identification in SIP - Routing
	decissionsbasedon Require/P-R
To: sipping@ietf.org
Message-id: <00f501c7574d$bb5ad920$6501a8c0@china.huawei.com>
MIME-version: 1.0
X-MIMEOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
X-Mailer: Microsoft Outlook Express 6.00.2900.2869
Content-type: text/plain; format=flowed; charset=iso-8859-1;
	reply-type=response
Content-transfer-encoding: 7BIT
X-Priority: 3
X-MSMail-priority: Normal
References: <7374777208BDC7449D5620EF94232567023932CA@esealmw113.eemea.ericsson.se>
	<45DEE717.1070609@cisco.com>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: e5ba305d0e64821bf3d8bc5d3bb07228
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi, Paul,

What I was trying to say, but clearer than my attempt. Thanks.

Spencer


> The difference here is that the Require header is definitive that the 
> caller doesn't want to reach a UAS unless that UAS supports the stated 
> options. Callerprefs is not so definitive. There are some uses of 
> callerprefs that can be considered fairly definitive, but in general 
> that is not the case.
> 
> Jonathan proposed extending 3841 so that the use of Require generates an 
> implicit callerpref. And that in turn would influence the selection of 
> the registered contacts to try. I can't see a plausible way to infer any 
> implicit callerpref from Accept-Contact headers.
> 
> In any case, that is all about selecting how to translate the R-URI. It 
> is not about selecting intermediate proxies. As Dale has stated, you are 
> free to use most anything to select an intermediate proxy. If you want 
> to use Require, or Accept-Contact to do that you may do so. But while 
> you are free to do so, it is less clear that it is useful or wise to do 
> so. If the intent is to select a proxy that is consistent with the 
> capabilities of the UAS that will eventually be accepted (as seems to be 
> the case with PoC), then it only makes sense if the thing you are 
> testing is definitive for the capabilities of the UAS. So again, Require 
> may make sense for this, while Accept-Contact is not.
> 
> Paul


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 08:49:22 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKan9-0007JI-Dc; Fri, 23 Feb 2007 08:48:55 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKan7-0007JB-Lu
	for sipping@ietf.org; Fri, 23 Feb 2007 08:48:53 -0500
Received: from mailgate.siemenscomms.co.uk ([195.171.110.225]
	helo=bemg01.siemenscomms.co.uk)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKan4-0001IS-Rm
	for sipping@ietf.org; Fri, 23 Feb 2007 08:48:53 -0500
Received: from ntht207e.uksgcs.siemenscomms.co.uk ([137.223.247.82])
	by siemenscomms.co.uk (PMDF V6.0-24 #40642)
	with ESMTP id <0JDX004C0513LC@siemenscomms.co.uk> for sipping@ietf.org;
	Fri, 23 Feb 2007 13:48:39 +0000 (GMT)
Received: by ntht207e.uksgcs.siemenscomms.co.uk with Internet Mail Service
	(5.5.2657.72)	id <FF5YS4DN>; Fri, 23 Feb 2007 13:48:39 +0000
Content-return: allowed
Date: Fri, 23 Feb 2007 13:48:29 +0000
From: "Elwell, John" <john.elwell@siemens.com>
Subject: RE: [Sipping] Service Identification Model
To: Andrew Allen <aallen@rim.com>, Paul Kyzivat <pkyzivat@cisco.com>
Message-id: <50B1CBA96870A34799A506B2313F26670B1EB6D8@ntht201e.siemenscomms.co.uk>
MIME-version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
Content-type: text/plain
Content-transfer-encoding: 7BIT
X-Spam-Score: 0.0 (/)
X-Scan-Signature: f460acdc4aacf7fc5e6f9bd32f8fd8c6
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Andrew,

There is nothing said here about how the S-CSCF or AS prevents abuse of the
service identifier. Surely an astute user will just include whatever service
identifier allows the request to be authorised and charged at the lowest
rate. Provided the network is not adding value that is necessary to make the
desired service work, the caller can lie in the SIP request.

John 

> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com] 
> Sent: 23 February 2007 02:23
> To: Paul Kyzivat
> Cc: sipping@ietf.org
> Subject: RE: [Sipping] Service Identification Model
> 
> 
> Paul
> 
> I think there are more than those 2. I listed several 
> potential services
> in my earlier email which you responded to with some points that I
> haven't had a chance to respond back to you on that yet (it 
> was a rather
> long list). However I still think those are all valid from the 3GPP
> point of view of services. As well as the issues of identifying the
> service related to the request for routing to Application Servers, and
> routing to the correct UAS and identifying the correct application at
> the UAS there are also the operator related requirements for
> classification of requests for charging, QoS Authorization and
> subscription based authorization purposes and for identifying,
> accounting and understanding the type of traffic in their network. The
> following such requirements exist in the 3GPP stage 2 TS 23.228:
> 
> 
> 2.   Based on operator policy the S-CSCF or an AS shall be able to
> validate an IMS communication service identifier in a SIP 
> request. This
> includes e.g. to check the syntactical correctness of a service
> identifier, and policing the usage of a communication service
> identifier.
> 
> 8.   Based on operator policy S-CSCF and AS shall consider the IMS
> communication service identifier for online and offline charging, e.g.
> put appropriate data into call detailed records.
> 
> 9.   The communication service identifier shall be capable of being an
> input into the policy control and charging rules.
> 
> 10. It shall be possible to use the IMS communication service 
> identifier
> as a means to authorise whether a subscriber is allowed to initiate or
> receive request for a communication service.
> 
> 
> So even when there are other various means to discriminate between one
> type of request and another in order to choose the appropriate
> Application Server or to choose the appropriate application the
> mechanism to identify the request are service specific and therefore
> cannot meet the requirements of a common mechanism that meets
> requirements 2,8,9 and 10 above.
> 
> So to list at least 10 hypothetical services that could be implemented
> using SIP that may be present on a mobile handset:
> 
> 1. Basic Telephony (Conversational QoS Voice only with 
> telephony feature
> set)  
> 2. Multimedia Telephony (Conversational QoS Voice plus video and other
> media streams with telephony feature set)
> 3. VoIP (Background QoS with no telephony features)
> 4. SMS
> 5. MMS
> 6. Instant Messaging
> 7. Push to Talk (PoC)
> 8. Network Game 1 (e.g Chess)
> 9. Network Game 2 (e.g Martial arts game)
> 10. Network Game 3 (e.g Dungeons & Dragons)
> 11. Push Service 1 (e.g. Traffic information service)
> 12. Push Service 2 (e.g. Weather information service)
> 13. Push Service 3 (e.g  News information service)
> 14. Push Service 4 (e.g  Stock information service)
> 15. Mobile TV service
> 
> So there's 15 which may all be different applications on a handset and
> all those are existing services or those already envisioned.
> 
> Andrew
> 
> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
> Sent: Thursday, February 22, 2007 7:05 PM
> To: Andrew Allen
> Cc: Eric Burger; sipping@ietf.org
> Subject: Re: [Sipping] Service Identification Model
> 
> 
> 
> Andrew Allen wrote:
> > Hi Eric,
> > 
> > It seems your prefer the approach of using URIs to identify services
> > which is part of Jonathan's draft as is the use of Require for those
> > services which simply do not interoperate unless the UAS understands
> the
> > service.
> > 
> > Where I am concerned is that you seem to be suggesting that the UA
> > separately register an AoR for each service it supports using
> REGISTER.
> > Certainly for cellular that wouldn't be acceptable as 10 or more
> > services could easily be supported by a single device.
> 
> Andrew, can you give me an example of 10 "services" on the 
> device that 
> need to be distinguished from one another in this way - that 
> aren't just
> 
> optional behavior of a particular service? (E.g. I see no reason why 
> "audio" and "audio/video" need to be separate services.)
> 
> I guess as things stand, PoC must be a separate service. And 
> apparently 
> Chess. What else?
> 
> 	Paul
> 
> > Registrations
> > create a significant load in the IMS network because its not just a
> SIP
> > REGISTER but also requires a lot of interactions with other network
> > elements such as the HSS and Notifications for reg-events etc. Also
> > having an order of magnitude increase in multiple registrations is
> going
> > to be a big issue for the avalanche restart problem for 
> example when a
> > cell site goes down and comes back into service.
> > 
> > I think Jonathan's approach gives us the possibility for 
> the registrar
> > to create the Service URIs based on the feature parameters 
> included in
> > the Contact of the REGISTER. 
> > 
> > I think the argument over whether the model is one UA with multiple
> > services or multiple UAs each representing a separate service is
> > somewhat academic. As I understand a UA can register 
> multiple AoRs and
> > can have other AoRs configured or otherwise provided by the service
> > provider. 
> > 
> > Andrew
> > 
> > -----Original Message-----
> > From: Eric Burger [mailto:eburger@bea.com] 
> > Sent: Wednesday, February 21, 2007 1:29 PM
> > To: sipping@ietf.org
> > Subject: [Sipping] Service Identification Model 
> > 
> > Folks have been pestering me for a while to look at the Service
> > Identification issue.  I finally have gotten around to it.
> > 
> > 
> > I guess I have a fundamental problem with the principal assertion of
> the
> > service-identification draft.  Namely, the assertion that there is a
> > single
> > SIP Layer, separate and distinct from two services.  Here 
> is Figure 2
> > from
> > the draft:
> > 
> >                            +---------------------------------+
> >                            |                                 |
> >                            | +-------------+ +-------------+ |
> >                            | |      UI     | |      UI     | |
> >                            | +-------------+ +-------------+ |
> >                            | +-------------+ +-------------+ |
> >                            | |             | |             | |
> >                            | |  Service 1  | |  Service 2  | |
> >                            | |             | |             | |
> >                            | +-------------+ +-------------+ |
> >                            | +-----------------------------+ |
> >                            | |                             | |
> >                            | |             SIP             | |
> >                            | |            Layer            | |
> >                            | |                             | |
> >                            | +-----------------------------+ |
> >                            |                                 |
> >                            +---------------------------------+
> > 
> >                                      Physical Device
> > 
> > I would offer this drawing confuses a common SIP Stack with a common
> > User
> > Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use case:
> > 
> >> ... In addition users may have
> >> multiple devices only one of which has the application that
> > understands
> >> the game. Other devices that the user has registered 
> against his AoR
> > may
> >> handle audio for telephony and other voice applications 
> but if Alice
> >> invites Bob to play her favourite game it is not very 
> interesting if
> > Bob
> >> accepts the gaming session on a device that can only 
> accept the audio
> >> stream. Also even when the terminal that has the game application
> >> receives the INVITE the correct application in that 
> terminal needs to
> > be
> >> identified to accept the session.
> > 
> > I would offer that this is the clinching argument that the 
> request is,
> > in
> > fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a
> User
> > Agent.
> > 
> > Same argument for PoC versus a regular voice call: One is 
> looking for
> a
> > PoC
> > User Agent, NOT the PoC service at a given User Agent.
> > 
> > Thus, the picture really looks like the following:
> > 
> >                            +---------------------------------+
> >                            |                                 |
> >                            | +-------------+ +-------------+ |
> >                            | |    UI       | |    UI       | |
> >                            | +-------------+ +-------------+ |
> >                            | +-------------+ +-------------+ |
> >                            | |             | |             | |
> >                            | |    UAS 1    | |    UAS 2    | |
> >                            | |             | |             | |
> >                            | +-------------+ +-------------+ |
> >                            | +-----------------------------+ |
> >                            | |                             | |
> >                            | |             SIP             | |
> >                            | |            Stack            | |
> >                            | |                             | |
> >                            | +-----------------------------+ |
> >                            |                                 |
> >                            +---------------------------------+
> > 
> >                                      Physical Device
> > 
> > 
> > Another argument for different User Agents is the whole "route based
> on
> > SDP"
> > argument.  As the author of RFC 3458, I think I was first 
> to make the
> > assertion that knowing the media type did NOT mean you know the
> > application
> > to use.  I don't want your phone or proxy or B2BUA to guess that
> because
> > I
> > have Chess Media, I want the Chess Application; I want to talk with
> your
> > Chess Application.
> > 
> > Now, is the Chess Application a Preference, Requirement, or 
> the thing
> > that I
> > am addressing?  The answer drives one towards using caller 
> prefs, the
> > Require header, or the Request-URI.
> > 
> > I would offer that it is not a preference: I really want the chess
> > application; I probably do not want to talk to you.
> > 
> > That leaves us with the Require header or a Request-URI.  
> My leanings
> > are
> > not surprisingly towards the RFC 4240 approach.  Frankly, if you are
> > building an IMS (as an example) that can handle 10,000,000
> subscribers,
> > you
> > should be able to handle 100,000,000 registrations.  That order of
> > magnitude
> > is way easier than the 10,000 to 10,000,000 jump.  I don't buy the
> > "registering each service is too much of a burden on the operator"
> > argument,
> > because it is not that much of a burden, and the network 
> actually does
> > the
> > right thing.  Better yet, all of the corner cases work.  
> Best of all,
> > the
> > IETF doesn't have to build an IANA Registry and process a 
> thousand PS
> > documents to register urn:service:myco:widget.
> > 
> > I suppose I could live with Require:, only if you are way 
> too lazy to
> > register each application instance.  It does semantically work,
> because
> > if I
> > want to do PoC and you do not do PoC, I'll get a 4xx response, to
> which
> > I
> > could try a regular call.
> > 
> >
> ______________________________________________________________
> _________
> > Notice:  This email message, together with any attachments, may
> contain
> > information  of  BEA Systems,  Inc.,  its subsidiaries  and
> affiliated
> > entities,  that may be confidential,  proprietary,  copyrighted
> and/or
> > legally privileged, and is intended solely for the use of the
> individual
> > or entity named in this message. If you are not the intended
> recipient,
> > and have received this message in error, please immediately return
> this
> > by email and then delete it.
> > 
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP
> > Use sip-implementors@cs.columbia.edu for questions on current sip
> > Use sip@ietf.org for new developments of core SIP
> > 
> > 
> > 
> > 
> ---------------------------------------------------------------------
> > This transmission (including any attachments) may contain 
> confidential
> information, privileged material (including material protected by the
> solicitor-client or other applicable privileges), or constitute
> non-public information. Any use of this information by anyone 
> other than
> the intended recipient is prohibited. If you have received this
> transmission in error, please immediately reply to the sender 
> and delete
> this information from your system. Use, dissemination, 
> distribution, or
> reproduction of this transmission by unintended recipients is not
> authorized and may be unlawful.
> > 
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP
> > Use sip-implementors@cs.columbia.edu for questions on current sip
> > Use sip@ietf.org for new developments of core SIP
> > 
> 
> 
> 
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain 
> confidential information, privileged material (including 
> material protected by the solicitor-client or other 
> applicable privileges), or constitute non-public information. 
> Any use of this information by anyone other than the intended 
> recipient is prohibited. If you have received this 
> transmission in error, please immediately reply to the sender 
> and delete this information from your system. Use, 
> dissemination, distribution, or reproduction of this 
> transmission by unintended recipients is not authorized and 
> may be unlawful.
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 09:08:40 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKb5c-0002hL-Ee; Fri, 23 Feb 2007 09:08:00 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKb5a-0002hB-Hg
	for sipping@ietf.org; Fri, 23 Feb 2007 09:07:58 -0500
Received: from exprod6og53.obsmtp.com ([64.18.1.187])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1HKb5X-000585-JB
	for sipping@ietf.org; Fri, 23 Feb 2007 09:07:58 -0500
Received: from source ([192.150.20.142]) by exprod6ob53.postini.com
	([64.18.5.12]) with SMTP; Fri, 23 Feb 2007 06:06:50 PST
Received: from inner-relay-1.corp.adobe.com ([153.32.1.51])
	by outbound-smtp-2.corp.adobe.com (8.12.10/8.12.10) with ESMTP id
	l1NE6PSd017479; Fri, 23 Feb 2007 06:06:26 -0800 (PST)
Received: from fe1.corp.adobe.com (fe1.corp.adobe.com [10.8.192.70])
	by inner-relay-1.corp.adobe.com (8.12.10/8.12.10) with ESMTP id
	l1NE60PP017762; Fri, 23 Feb 2007 06:06:01 -0800 (PST)
Received: from namail5.corp.adobe.com ([10.8.192.88]) by fe1.corp.adobe.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 06:05:58 -0800
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification Model
Date: Fri, 23 Feb 2007 06:05:38 -0800
Message-ID: <24CCCC428EFEA2469BF046DB3C7A8D224E3E39@namail5.corp.adobe.com>
In-Reply-To: <50B1CBA96870A34799A506B2313F26670B1EB6D8@ntht201e.siemenscomms.co.uk>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification Model
Thread-Index: AcdXUZ1am+IkIo8vRSmiAaKVTlUhsgAARXMg
References: <50B1CBA96870A34799A506B2313F26670B1EB6D8@ntht201e.siemenscomms.co.uk>
From: "Henry Sinnreich" <hsinnrei@adobe.com>
To: "Elwell, John" <john.elwell@siemens.com>, "Andrew Allen" <aallen@rim.com>, 
	"Paul Kyzivat" <pkyzivat@cisco.com>
X-OriginalArrivalTime: 23 Feb 2007 14:05:58.0908 (UTC)
	FILETIME=[BE3F73C0:01C75753]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 85e99493ec37f9acef29c7843dbf2e68
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

> Surely an astute user will just include whatever service identifier=20
> allows the request to be authorised and charged at the lowest rate.

Wow! Giving the evil user a choice for what he/she wants to pay :-)

> the caller can lie in the SIP request

So shopping is equaled here to lying?=20
Is this Alice in Wonderland?

Thanks, Henry

-----Original Message-----
From: Elwell, John [mailto:john.elwell@siemens.com]=20
Sent: Friday, February 23, 2007 7:48 AM
To: Andrew Allen; Paul Kyzivat
Cc: sipping@ietf.org
Subject: RE: [Sipping] Service Identification Model

Andrew,

There is nothing said here about how the S-CSCF or AS prevents abuse of
the
service identifier. Surely an astute user will just include whatever
service
identifier allows the request to be authorised and charged at the lowest
rate. Provided the network is not adding value that is necessary to make
the
desired service work, the caller can lie in the SIP request.

John=20

> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com]=20
> Sent: 23 February 2007 02:23
> To: Paul Kyzivat
> Cc: sipping@ietf.org
> Subject: RE: [Sipping] Service Identification Model
>=20
>=20
> Paul
>=20
> I think there are more than those 2. I listed several=20
> potential services
> in my earlier email which you responded to with some points that I
> haven't had a chance to respond back to you on that yet (it=20
> was a rather
> long list). However I still think those are all valid from the 3GPP
> point of view of services. As well as the issues of identifying the
> service related to the request for routing to Application Servers, and
> routing to the correct UAS and identifying the correct application at
> the UAS there are also the operator related requirements for
> classification of requests for charging, QoS Authorization and
> subscription based authorization purposes and for identifying,
> accounting and understanding the type of traffic in their network. The
> following such requirements exist in the 3GPP stage 2 TS 23.228:
>=20
>=20
> 2.   Based on operator policy the S-CSCF or an AS shall be able to
> validate an IMS communication service identifier in a SIP=20
> request. This
> includes e.g. to check the syntactical correctness of a service
> identifier, and policing the usage of a communication service
> identifier.
>=20
> 8.   Based on operator policy S-CSCF and AS shall consider the IMS
> communication service identifier for online and offline charging, e.g.
> put appropriate data into call detailed records.
>=20
> 9.   The communication service identifier shall be capable of being an
> input into the policy control and charging rules.
>=20
> 10. It shall be possible to use the IMS communication service=20
> identifier
> as a means to authorise whether a subscriber is allowed to initiate or
> receive request for a communication service.
>=20
>=20
> So even when there are other various means to discriminate between one
> type of request and another in order to choose the appropriate
> Application Server or to choose the appropriate application the
> mechanism to identify the request are service specific and therefore
> cannot meet the requirements of a common mechanism that meets
> requirements 2,8,9 and 10 above.
>=20
> So to list at least 10 hypothetical services that could be implemented
> using SIP that may be present on a mobile handset:
>=20
> 1. Basic Telephony (Conversational QoS Voice only with=20
> telephony feature
> set) =20
> 2. Multimedia Telephony (Conversational QoS Voice plus video and other
> media streams with telephony feature set)
> 3. VoIP (Background QoS with no telephony features)
> 4. SMS
> 5. MMS
> 6. Instant Messaging
> 7. Push to Talk (PoC)
> 8. Network Game 1 (e.g Chess)
> 9. Network Game 2 (e.g Martial arts game)
> 10. Network Game 3 (e.g Dungeons & Dragons)
> 11. Push Service 1 (e.g. Traffic information service)
> 12. Push Service 2 (e.g. Weather information service)
> 13. Push Service 3 (e.g  News information service)
> 14. Push Service 4 (e.g  Stock information service)
> 15. Mobile TV service
>=20
> So there's 15 which may all be different applications on a handset and
> all those are existing services or those already envisioned.
>=20
> Andrew
>=20
> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
> Sent: Thursday, February 22, 2007 7:05 PM
> To: Andrew Allen
> Cc: Eric Burger; sipping@ietf.org
> Subject: Re: [Sipping] Service Identification Model
>=20
>=20
>=20
> Andrew Allen wrote:
> > Hi Eric,
> >=20
> > It seems your prefer the approach of using URIs to identify services
> > which is part of Jonathan's draft as is the use of Require for those
> > services which simply do not interoperate unless the UAS understands
> the
> > service.
> >=20
> > Where I am concerned is that you seem to be suggesting that the UA
> > separately register an AoR for each service it supports using
> REGISTER.
> > Certainly for cellular that wouldn't be acceptable as 10 or more
> > services could easily be supported by a single device.
>=20
> Andrew, can you give me an example of 10 "services" on the=20
> device that=20
> need to be distinguished from one another in this way - that=20
> aren't just
>=20
> optional behavior of a particular service? (E.g. I see no reason why=20
> "audio" and "audio/video" need to be separate services.)
>=20
> I guess as things stand, PoC must be a separate service. And=20
> apparently=20
> Chess. What else?
>=20
> 	Paul
>=20
> > Registrations
> > create a significant load in the IMS network because its not just a
> SIP
> > REGISTER but also requires a lot of interactions with other network
> > elements such as the HSS and Notifications for reg-events etc. Also
> > having an order of magnitude increase in multiple registrations is
> going
> > to be a big issue for the avalanche restart problem for=20
> example when a
> > cell site goes down and comes back into service.
> >=20
> > I think Jonathan's approach gives us the possibility for=20
> the registrar
> > to create the Service URIs based on the feature parameters=20
> included in
> > the Contact of the REGISTER.=20
> >=20
> > I think the argument over whether the model is one UA with multiple
> > services or multiple UAs each representing a separate service is
> > somewhat academic. As I understand a UA can register=20
> multiple AoRs and
> > can have other AoRs configured or otherwise provided by the service
> > provider.=20
> >=20
> > Andrew
> >=20
> > -----Original Message-----
> > From: Eric Burger [mailto:eburger@bea.com]=20
> > Sent: Wednesday, February 21, 2007 1:29 PM
> > To: sipping@ietf.org
> > Subject: [Sipping] Service Identification Model=20
> >=20
> > Folks have been pestering me for a while to look at the Service
> > Identification issue.  I finally have gotten around to it.
> >=20
> >=20
> > I guess I have a fundamental problem with the principal assertion of
> the
> > service-identification draft.  Namely, the assertion that there is a
> > single
> > SIP Layer, separate and distinct from two services.  Here=20
> is Figure 2
> > from
> > the draft:
> >=20
> >                            +---------------------------------+
> >                            |                                 |
> >                            | +-------------+ +-------------+ |
> >                            | |      UI     | |      UI     | |
> >                            | +-------------+ +-------------+ |
> >                            | +-------------+ +-------------+ |
> >                            | |             | |             | |
> >                            | |  Service 1  | |  Service 2  | |
> >                            | |             | |             | |
> >                            | +-------------+ +-------------+ |
> >                            | +-----------------------------+ |
> >                            | |                             | |
> >                            | |             SIP             | |
> >                            | |            Layer            | |
> >                            | |                             | |
> >                            | +-----------------------------+ |
> >                            |                                 |
> >                            +---------------------------------+
> >=20
> >                                      Physical Device
> >=20
> > I would offer this drawing confuses a common SIP Stack with a common
> > User
> > Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use case:
> >=20
> >> ... In addition users may have
> >> multiple devices only one of which has the application that
> > understands
> >> the game. Other devices that the user has registered=20
> against his AoR
> > may
> >> handle audio for telephony and other voice applications=20
> but if Alice
> >> invites Bob to play her favourite game it is not very=20
> interesting if
> > Bob
> >> accepts the gaming session on a device that can only=20
> accept the audio
> >> stream. Also even when the terminal that has the game application
> >> receives the INVITE the correct application in that=20
> terminal needs to
> > be
> >> identified to accept the session.
> >=20
> > I would offer that this is the clinching argument that the=20
> request is,
> > in
> > fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a
> User
> > Agent.
> >=20
> > Same argument for PoC versus a regular voice call: One is=20
> looking for
> a
> > PoC
> > User Agent, NOT the PoC service at a given User Agent.
> >=20
> > Thus, the picture really looks like the following:
> >=20
> >                            +---------------------------------+
> >                            |                                 |
> >                            | +-------------+ +-------------+ |
> >                            | |    UI       | |    UI       | |
> >                            | +-------------+ +-------------+ |
> >                            | +-------------+ +-------------+ |
> >                            | |             | |             | |
> >                            | |    UAS 1    | |    UAS 2    | |
> >                            | |             | |             | |
> >                            | +-------------+ +-------------+ |
> >                            | +-----------------------------+ |
> >                            | |                             | |
> >                            | |             SIP             | |
> >                            | |            Stack            | |
> >                            | |                             | |
> >                            | +-----------------------------+ |
> >                            |                                 |
> >                            +---------------------------------+
> >=20
> >                                      Physical Device
> >=20
> >=20
> > Another argument for different User Agents is the whole "route based
> on
> > SDP"
> > argument.  As the author of RFC 3458, I think I was first=20
> to make the
> > assertion that knowing the media type did NOT mean you know the
> > application
> > to use.  I don't want your phone or proxy or B2BUA to guess that
> because
> > I
> > have Chess Media, I want the Chess Application; I want to talk with
> your
> > Chess Application.
> >=20
> > Now, is the Chess Application a Preference, Requirement, or=20
> the thing
> > that I
> > am addressing?  The answer drives one towards using caller=20
> prefs, the
> > Require header, or the Request-URI.
> >=20
> > I would offer that it is not a preference: I really want the chess
> > application; I probably do not want to talk to you.
> >=20
> > That leaves us with the Require header or a Request-URI. =20
> My leanings
> > are
> > not surprisingly towards the RFC 4240 approach.  Frankly, if you are
> > building an IMS (as an example) that can handle 10,000,000
> subscribers,
> > you
> > should be able to handle 100,000,000 registrations.  That order of
> > magnitude
> > is way easier than the 10,000 to 10,000,000 jump.  I don't buy the
> > "registering each service is too much of a burden on the operator"
> > argument,
> > because it is not that much of a burden, and the network=20
> actually does
> > the
> > right thing.  Better yet, all of the corner cases work. =20
> Best of all,
> > the
> > IETF doesn't have to build an IANA Registry and process a=20
> thousand PS
> > documents to register urn:service:myco:widget.
> >=20
> > I suppose I could live with Require:, only if you are way=20
> too lazy to
> > register each application instance.  It does semantically work,
> because
> > if I
> > want to do PoC and you do not do PoC, I'll get a 4xx response, to
> which
> > I
> > could try a regular call.
> >=20
> >
> ______________________________________________________________
> _________
> > Notice:  This email message, together with any attachments, may
> contain
> > information  of  BEA Systems,  Inc.,  its subsidiaries  and
> affiliated
> > entities,  that may be confidential,  proprietary,  copyrighted
> and/or
> > legally privileged, and is intended solely for the use of the
> individual
> > or entity named in this message. If you are not the intended
> recipient,
> > and have received this message in error, please immediately return
> this
> > by email and then delete it.
> >=20
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP
> > Use sip-implementors@cs.columbia.edu for questions on current sip
> > Use sip@ietf.org for new developments of core SIP
> >=20
> >=20
> >=20
> >=20
> ---------------------------------------------------------------------
> > This transmission (including any attachments) may contain=20
> confidential
> information, privileged material (including material protected by the
> solicitor-client or other applicable privileges), or constitute
> non-public information. Any use of this information by anyone=20
> other than
> the intended recipient is prohibited. If you have received this
> transmission in error, please immediately reply to the sender=20
> and delete
> this information from your system. Use, dissemination,=20
> distribution, or
> reproduction of this transmission by unintended recipients is not
> authorized and may be unlawful.
> >=20
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP
> > Use sip-implementors@cs.columbia.edu for questions on current sip
> > Use sip@ietf.org for new developments of core SIP
> >=20
>=20
>=20
>=20
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain=20
> confidential information, privileged material (including=20
> material protected by the solicitor-client or other=20
> applicable privileges), or constitute non-public information.=20
> Any use of this information by anyone other than the intended=20
> recipient is prohibited. If you have received this=20
> transmission in error, please immediately reply to the sender=20
> and delete this information from your system. Use,=20
> dissemination, distribution, or reproduction of this=20
> transmission by unintended recipients is not authorized and=20
> may be unlawful.
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 09:20:37 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKbGp-0002g9-Nu; Fri, 23 Feb 2007 09:19:35 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKbGo-0002fY-FB
	for sipping@ietf.org; Fri, 23 Feb 2007 09:19:34 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKbGV-0007Y8-Bo
	for sipping@ietf.org; Fri, 23 Feb 2007 09:19:34 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	D5A7520934
	for <sipping@ietf.org>; Fri, 23 Feb 2007 15:19:12 +0100 (CET)
X-AuditID: c1b4fb3c-b07c9bb0000007de-3d-45def7e08814 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	B9C8620900
	for <sipping@ietf.org>; Fri, 23 Feb 2007 15:19:12 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 15:19:04 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification Model
Date: Fri, 23 Feb 2007 15:19:03 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD029258E6@esealmw111>
In-Reply-To: <24CCCC428EFEA2469BF046DB3C7A8D224E3E39@namail5.corp.adobe.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification Model
Thread-Index: AcdXUZ1am+IkIo8vRSmiAaKVTlUhsgAARXMgAACwBHA=
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 23 Feb 2007 14:19:04.0658 (UTC)
	FILETIME=[92976320:01C75755]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 563af5038a5e1dade28c8affc0fff375
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I agree - you will be charged for what you have requested.=20

-----Original Message-----
From: Henry Sinnreich [mailto:hsinnrei@adobe.com]=20
Sent: Friday, February 23, 2007 3:06 PM
To: Elwell, John; Andrew Allen; Paul Kyzivat
Cc: sipping@ietf.org
Subject: RE: [Sipping] Service Identification Model

> Surely an astute user will just include whatever service identifier=20
> allows the request to be authorised and charged at the lowest rate.

Wow! Giving the evil user a choice for what he/she wants to pay :-)

> the caller can lie in the SIP request

So shopping is equaled here to lying?=20
Is this Alice in Wonderland?

Thanks, Henry

-----Original Message-----
From: Elwell, John [mailto:john.elwell@siemens.com]
Sent: Friday, February 23, 2007 7:48 AM
To: Andrew Allen; Paul Kyzivat
Cc: sipping@ietf.org
Subject: RE: [Sipping] Service Identification Model

Andrew,

There is nothing said here about how the S-CSCF or AS prevents abuse of
the service identifier. Surely an astute user will just include whatever
service identifier allows the request to be authorised and charged at
the lowest rate. Provided the network is not adding value that is
necessary to make the desired service work, the caller can lie in the
SIP request.

John=20

> -----Original Message-----
> From: Andrew Allen [mailto:aallen@rim.com]
> Sent: 23 February 2007 02:23
> To: Paul Kyzivat
> Cc: sipping@ietf.org
> Subject: RE: [Sipping] Service Identification Model
>=20
>=20
> Paul
>=20
> I think there are more than those 2. I listed several potential=20
> services in my earlier email which you responded to with some points=20
> that I haven't had a chance to respond back to you on that yet (it was

> a rather long list). However I still think those are all valid from=20
> the 3GPP point of view of services. As well as the issues of=20
> identifying the service related to the request for routing to=20
> Application Servers, and routing to the correct UAS and identifying=20
> the correct application at the UAS there are also the operator related

> requirements for classification of requests for charging, QoS=20
> Authorization and subscription based authorization purposes and for=20
> identifying, accounting and understanding the type of traffic in their

> network. The following such requirements exist in the 3GPP stage 2 TS=20
> 23.228:
>=20
>=20
> 2.   Based on operator policy the S-CSCF or an AS shall be able to
> validate an IMS communication service identifier in a SIP request.=20
> This includes e.g. to check the syntactical correctness of a service=20
> identifier, and policing the usage of a communication service=20
> identifier.
>=20
> 8.   Based on operator policy S-CSCF and AS shall consider the IMS
> communication service identifier for online and offline charging, e.g.
> put appropriate data into call detailed records.
>=20
> 9.   The communication service identifier shall be capable of being an
> input into the policy control and charging rules.
>=20
> 10. It shall be possible to use the IMS communication service=20
> identifier as a means to authorise whether a subscriber is allowed to=20
> initiate or receive request for a communication service.
>=20
>=20
> So even when there are other various means to discriminate between one

> type of request and another in order to choose the appropriate=20
> Application Server or to choose the appropriate application the=20
> mechanism to identify the request are service specific and therefore=20
> cannot meet the requirements of a common mechanism that meets=20
> requirements 2,8,9 and 10 above.
>=20
> So to list at least 10 hypothetical services that could be implemented

> using SIP that may be present on a mobile handset:
>=20
> 1. Basic Telephony (Conversational QoS Voice only with telephony=20
> feature
> set)
> 2. Multimedia Telephony (Conversational QoS Voice plus video and other

> media streams with telephony feature set) 3. VoIP (Background QoS with

> no telephony features) 4. SMS 5. MMS 6. Instant Messaging 7. Push to=20
> Talk (PoC) 8. Network Game 1 (e.g Chess) 9. Network Game 2 (e.g=20
> Martial arts game) 10. Network Game 3 (e.g Dungeons & Dragons) 11.=20
> Push Service 1 (e.g. Traffic information service) 12. Push Service 2=20
> (e.g. Weather information service) 13. Push Service 3 (e.g  News=20
> information service) 14. Push Service 4 (e.g  Stock information=20
> service) 15. Mobile TV service
>=20
> So there's 15 which may all be different applications on a handset and

> all those are existing services or those already envisioned.
>=20
> Andrew
>=20
> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
> Sent: Thursday, February 22, 2007 7:05 PM
> To: Andrew Allen
> Cc: Eric Burger; sipping@ietf.org
> Subject: Re: [Sipping] Service Identification Model
>=20
>=20
>=20
> Andrew Allen wrote:
> > Hi Eric,
> >=20
> > It seems your prefer the approach of using URIs to identify services

> > which is part of Jonathan's draft as is the use of Require for those

> > services which simply do not interoperate unless the UAS understands
> the
> > service.
> >=20
> > Where I am concerned is that you seem to be suggesting that the UA=20
> > separately register an AoR for each service it supports using
> REGISTER.
> > Certainly for cellular that wouldn't be acceptable as 10 or more=20
> > services could easily be supported by a single device.
>=20
> Andrew, can you give me an example of 10 "services" on the device that

> need to be distinguished from one another in this way - that aren't=20
> just
>=20
> optional behavior of a particular service? (E.g. I see no reason why=20
> "audio" and "audio/video" need to be separate services.)
>=20
> I guess as things stand, PoC must be a separate service. And=20
> apparently Chess. What else?
>=20
> 	Paul
>=20
> > Registrations
> > create a significant load in the IMS network because its not just a
> SIP
> > REGISTER but also requires a lot of interactions with other network=20
> > elements such as the HSS and Notifications for reg-events etc. Also=20
> > having an order of magnitude increase in multiple registrations is
> going
> > to be a big issue for the avalanche restart problem for
> example when a
> > cell site goes down and comes back into service.
> >=20
> > I think Jonathan's approach gives us the possibility for
> the registrar
> > to create the Service URIs based on the feature parameters
> included in
> > the Contact of the REGISTER.=20
> >=20
> > I think the argument over whether the model is one UA with multiple=20
> > services or multiple UAs each representing a separate service is=20
> > somewhat academic. As I understand a UA can register
> multiple AoRs and
> > can have other AoRs configured or otherwise provided by the service=20
> > provider.
> >=20
> > Andrew
> >=20
> > -----Original Message-----
> > From: Eric Burger [mailto:eburger@bea.com]
> > Sent: Wednesday, February 21, 2007 1:29 PM
> > To: sipping@ietf.org
> > Subject: [Sipping] Service Identification Model
> >=20
> > Folks have been pestering me for a while to look at the Service=20
> > Identification issue.  I finally have gotten around to it.
> >=20
> >=20
> > I guess I have a fundamental problem with the principal assertion of
> the
> > service-identification draft.  Namely, the assertion that there is a

> > single SIP Layer, separate and distinct from two services.  Here
> is Figure 2
> > from
> > the draft:
> >=20
> >                            +---------------------------------+
> >                            |                                 |
> >                            | +-------------+ +-------------+ |
> >                            | |      UI     | |      UI     | |
> >                            | +-------------+ +-------------+ |
> >                            | +-------------+ +-------------+ |
> >                            | |             | |             | |
> >                            | |  Service 1  | |  Service 2  | |
> >                            | |             | |             | |
> >                            | +-------------+ +-------------+ |
> >                            | +-----------------------------+ |
> >                            | |                             | |
> >                            | |             SIP             | |
> >                            | |            Layer            | |
> >                            | |                             | |
> >                            | +-----------------------------+ |
> >                            |                                 |
> >                            +---------------------------------+
> >=20
> >                                      Physical Device
> >=20
> > I would offer this drawing confuses a common SIP Stack with a common

> > User Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use=20
> > case:
> >=20
> >> ... In addition users may have
> >> multiple devices only one of which has the application that
> > understands
> >> the game. Other devices that the user has registered
> against his AoR
> > may
> >> handle audio for telephony and other voice applications
> but if Alice
> >> invites Bob to play her favourite game it is not very
> interesting if
> > Bob
> >> accepts the gaming session on a device that can only
> accept the audio
> >> stream. Also even when the terminal that has the game application=20
> >> receives the INVITE the correct application in that
> terminal needs to
> > be
> >> identified to accept the session.
> >=20
> > I would offer that this is the clinching argument that the
> request is,
> > in
> > fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a
> User
> > Agent.
> >=20
> > Same argument for PoC versus a regular voice call: One is
> looking for
> a
> > PoC
> > User Agent, NOT the PoC service at a given User Agent.
> >=20
> > Thus, the picture really looks like the following:
> >=20
> >                            +---------------------------------+
> >                            |                                 |
> >                            | +-------------+ +-------------+ |
> >                            | |    UI       | |    UI       | |
> >                            | +-------------+ +-------------+ |
> >                            | +-------------+ +-------------+ |
> >                            | |             | |             | |
> >                            | |    UAS 1    | |    UAS 2    | |
> >                            | |             | |             | |
> >                            | +-------------+ +-------------+ |
> >                            | +-----------------------------+ |
> >                            | |                             | |
> >                            | |             SIP             | |
> >                            | |            Stack            | |
> >                            | |                             | |
> >                            | +-----------------------------+ |
> >                            |                                 |
> >                            +---------------------------------+
> >=20
> >                                      Physical Device
> >=20
> >=20
> > Another argument for different User Agents is the whole "route based
> on
> > SDP"
> > argument.  As the author of RFC 3458, I think I was first
> to make the
> > assertion that knowing the media type did NOT mean you know the=20
> > application to use.  I don't want your phone or proxy or B2BUA to=20
> > guess that
> because
> > I
> > have Chess Media, I want the Chess Application; I want to talk with
> your
> > Chess Application.
> >=20
> > Now, is the Chess Application a Preference, Requirement, or
> the thing
> > that I
> > am addressing?  The answer drives one towards using caller
> prefs, the
> > Require header, or the Request-URI.
> >=20
> > I would offer that it is not a preference: I really want the chess=20
> > application; I probably do not want to talk to you.
> >=20
> > That leaves us with the Require header or a Request-URI. =20
> My leanings
> > are
> > not surprisingly towards the RFC 4240 approach.  Frankly, if you are

> > building an IMS (as an example) that can handle 10,000,000
> subscribers,
> > you
> > should be able to handle 100,000,000 registrations.  That order of=20
> > magnitude is way easier than the 10,000 to 10,000,000 jump.  I don't

> > buy the "registering each service is too much of a burden on the=20
> > operator"
> > argument,
> > because it is not that much of a burden, and the network
> actually does
> > the
> > right thing.  Better yet, all of the corner cases work. =20
> Best of all,
> > the
> > IETF doesn't have to build an IANA Registry and process a
> thousand PS
> > documents to register urn:service:myco:widget.
> >=20
> > I suppose I could live with Require:, only if you are way
> too lazy to
> > register each application instance.  It does semantically work,
> because
> > if I
> > want to do PoC and you do not do PoC, I'll get a 4xx response, to
> which
> > I
> > could try a regular call.
> >=20
> >
> ______________________________________________________________
> _________
> > Notice:  This email message, together with any attachments, may
> contain
> > information  of  BEA Systems,  Inc.,  its subsidiaries  and
> affiliated
> > entities,  that may be confidential,  proprietary,  copyrighted
> and/or
> > legally privileged, and is intended solely for the use of the
> individual
> > or entity named in this message. If you are not the intended
> recipient,
> > and have received this message in error, please immediately return
> this
> > by email and then delete it.
> >=20
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP Use=20
> > sip-implementors@cs.columbia.edu for questions on current sip Use=20
> > sip@ietf.org for new developments of core SIP
> >=20
> >=20
> >=20
> >=20
> ---------------------------------------------------------------------
> > This transmission (including any attachments) may contain
> confidential
> information, privileged material (including material protected by the=20
> solicitor-client or other applicable privileges), or constitute=20
> non-public information. Any use of this information by anyone other=20
> than the intended recipient is prohibited. If you have received this=20
> transmission in error, please immediately reply to the sender and=20
> delete this information from your system. Use, dissemination,=20
> distribution, or reproduction of this transmission by unintended=20
> recipients is not authorized and may be unlawful.
> >=20
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP Use=20
> > sip-implementors@cs.columbia.edu for questions on current sip Use=20
> > sip@ietf.org for new developments of core SIP
> >=20
>=20
>=20
>=20
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential

> information, privileged material (including material protected by the=20
> solicitor-client or other applicable privileges), or constitute=20
> non-public information.
> Any use of this information by anyone other than the intended=20
> recipient is prohibited. If you have received this transmission in=20
> error, please immediately reply to the sender and delete this=20
> information from your system. Use, dissemination, distribution, or=20
> reproduction of this transmission by unintended recipients is not=20
> authorized and may be unlawful.
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 09:35:58 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKbVs-0000gu-98; Fri, 23 Feb 2007 09:35:08 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKbVq-0000gn-Tu
	for sipping@ietf.org; Fri, 23 Feb 2007 09:35:06 -0500
Received: from dsl001-129-069.dfw1.dsl.speakeasy.net ([72.1.129.69]
	helo=vicuna.estacado.net) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HKbVn-0002mc-U3
	for sipping@ietf.org; Fri, 23 Feb 2007 09:35:06 -0500
Received: from [10.0.0.188] (adsl-70-235-190-149.dsl.rcsntx.sbcglobal.net
	[70.235.190.149]) (authenticated bits=0)
	by vicuna.estacado.net (8.13.8/8.13.8) with ESMTP id l1NEYq2H095554
	(version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO);
	Fri, 23 Feb 2007 08:34:57 -0600 (CST)
	(envelope-from bcampen@estacado.net)
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C009463E2E@XCH47YKF.rim.net>
References: <61968779B8AC4C4BAB421D4C12F008C009463E2E@XCH47YKF.rim.net>
Mime-Version: 1.0 (Apple Message framework v752.2)
Message-Id: <FC318FB3-A497-4BED-BBC6-3B5B630FC589@estacado.net>
From: Byron Campen <bcampen@estacado.net>
Subject: Re: [Sipping] Service Identification Model
Date: Fri, 23 Feb 2007 08:34:40 -0600
To: Andrew Allen <aallen@rim.com>
X-Mailer: Apple Mail (2.752.2)
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 58b614506802734014829a093beb6879
Cc: Paul Kyzivat <pkyzivat@cisco.com>, sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1517427085=="
Errors-To: sipping-bounces@ietf.org


--===============1517427085==
Content-Type: multipart/signed; micalg=sha1; boundary=Apple-Mail-6-660558670;
	protocol="application/pkcs7-signature"


--Apple-Mail-6-660558670
Content-Transfer-Encoding: 7bit
Content-Type: text/plain;
	charset=US-ASCII;
	delsp=yes;
	format=flowed

>
> 1. Basic Telephony (Conversational QoS Voice only with telephony  
> feature
> set)
> 2. Multimedia Telephony (Conversational QoS Voice plus video and other
> media streams with telephony feature set)
> 3. VoIP (Background QoS with no telephony features)
> 4. SMS
> 5. MMS
> 6. Instant Messaging
> 7. Push to Talk (PoC)
> 8. Network Game 1 (e.g Chess)
> 9. Network Game 2 (e.g Martial arts game)
> 10. Network Game 3 (e.g Dungeons & Dragons)
> 11. Push Service 1 (e.g. Traffic information service)
> 12. Push Service 2 (e.g. Weather information service)
> 13. Push Service 3 (e.g  News information service)
> 14. Push Service 4 (e.g  Stock information service)
> 15. Mobile TV service
>
> So there's 15 which may all be different applications on a handset and
> all those are existing services or those already envisioned.
>

	I only see 10 here that would need to ever send a register request.  
(11-15 are strictly client-server) 4-6 will have distinct Content- 
Type headers and payloads. You would need only 1 UA for these. So  
we're down to 8. 8-10 will, once again, be distinct when you look at  
the payload and Content-Type. So we're down to 6. There is no reason  
at all that the UA that handles voice is incapable of handling IM,  
and you can tell the difference by looking at the request type, so  
we're down to 5. As for the voice stuff, you can distinguish 2 from  
anything else through the payload. So, we could quite easily roll 1-2  
together. (4) 3 can be distinguished by watching for any in-dialog  
call-control stuff. (Having a service identifier that essentially  
says "I promise not to use any fancy telephony features on this call.  
No, really, I promise!" is never going to work) So we're down to 3.

	SIP is more than flexible enough to handle all of this stuff in a  
single UA. The UAs are perfectly capable of figuring out what  
"service" the other wants to use; why are your middleboxes unable to  
do the same? Adding a parallel method to make things easy on the  
middleboxes is just asking for trouble, because you're conveying the  
same information twice. It is ripe for abuse. Of course, the easy way  
to do all this is to just let UAs send what they want, and charge per  
KB, but hey, what do I know about simplicity, flexibility,  
scalability, ease-of-deployment, and customer satisfaction? I'm just  
a IETFer.

Best regards,
Byron Campen


> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
> Sent: Thursday, February 22, 2007 7:05 PM
> To: Andrew Allen
> Cc: Eric Burger; sipping@ietf.org
> Subject: Re: [Sipping] Service Identification Model
>
>
>
> Andrew Allen wrote:
>> Hi Eric,
>>
>> It seems your prefer the approach of using URIs to identify services
>> which is part of Jonathan's draft as is the use of Require for those
>> services which simply do not interoperate unless the UAS understands
> the
>> service.
>>
>> Where I am concerned is that you seem to be suggesting that the UA
>> separately register an AoR for each service it supports using
> REGISTER.
>> Certainly for cellular that wouldn't be acceptable as 10 or more
>> services could easily be supported by a single device.
>
> Andrew, can you give me an example of 10 "services" on the device that
> need to be distinguished from one another in this way - that aren't  
> just
>
> optional behavior of a particular service? (E.g. I see no reason why
> "audio" and "audio/video" need to be separate services.)
>
> I guess as things stand, PoC must be a separate service. And  
> apparently
> Chess. What else?
>
> 	Paul
>
>> Registrations
>> create a significant load in the IMS network because its not just a
> SIP
>> REGISTER but also requires a lot of interactions with other network
>> elements such as the HSS and Notifications for reg-events etc. Also
>> having an order of magnitude increase in multiple registrations is
> going
>> to be a big issue for the avalanche restart problem for example  
>> when a
>> cell site goes down and comes back into service.
>>
>> I think Jonathan's approach gives us the possibility for the  
>> registrar
>> to create the Service URIs based on the feature parameters  
>> included in
>> the Contact of the REGISTER.
>>
>> I think the argument over whether the model is one UA with multiple
>> services or multiple UAs each representing a separate service is
>> somewhat academic. As I understand a UA can register multiple AoRs  
>> and
>> can have other AoRs configured or otherwise provided by the service
>> provider.
>>
>> Andrew
>>
>> -----Original Message-----
>> From: Eric Burger [mailto:eburger@bea.com]
>> Sent: Wednesday, February 21, 2007 1:29 PM
>> To: sipping@ietf.org
>> Subject: [Sipping] Service Identification Model
>>
>> Folks have been pestering me for a while to look at the Service
>> Identification issue.  I finally have gotten around to it.
>>
>>
>> I guess I have a fundamental problem with the principal assertion of
> the
>> service-identification draft.  Namely, the assertion that there is a
>> single
>> SIP Layer, separate and distinct from two services.  Here is Figure 2
>> from
>> the draft:
>>
>>                            +---------------------------------+
>>                            |                                 |
>>                            | +-------------+ +-------------+ |
>>                            | |      UI     | |      UI     | |
>>                            | +-------------+ +-------------+ |
>>                            | +-------------+ +-------------+ |
>>                            | |             | |             | |
>>                            | |  Service 1  | |  Service 2  | |
>>                            | |             | |             | |
>>                            | +-------------+ +-------------+ |
>>                            | +-----------------------------+ |
>>                            | |                             | |
>>                            | |             SIP             | |
>>                            | |            Layer            | |
>>                            | |                             | |
>>                            | +-----------------------------+ |
>>                            |                                 |
>>                            +---------------------------------+
>>
>>                                      Physical Device
>>
>> I would offer this drawing confuses a common SIP Stack with a common
>> User
>> Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use case:
>>
>>> ... In addition users may have
>>> multiple devices only one of which has the application that
>> understands
>>> the game. Other devices that the user has registered against his AoR
>> may
>>> handle audio for telephony and other voice applications but if Alice
>>> invites Bob to play her favourite game it is not very interesting if
>> Bob
>>> accepts the gaming session on a device that can only accept the  
>>> audio
>>> stream. Also even when the terminal that has the game application
>>> receives the INVITE the correct application in that terminal  
>>> needs to
>> be
>>> identified to accept the session.
>>
>> I would offer that this is the clinching argument that the request  
>> is,
>> in
>> fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a
> User
>> Agent.
>>
>> Same argument for PoC versus a regular voice call: One is looking for
> a
>> PoC
>> User Agent, NOT the PoC service at a given User Agent.
>>
>> Thus, the picture really looks like the following:
>>
>>                            +---------------------------------+
>>                            |                                 |
>>                            | +-------------+ +-------------+ |
>>                            | |    UI       | |    UI       | |
>>                            | +-------------+ +-------------+ |
>>                            | +-------------+ +-------------+ |
>>                            | |             | |             | |
>>                            | |    UAS 1    | |    UAS 2    | |
>>                            | |             | |             | |
>>                            | +-------------+ +-------------+ |
>>                            | +-----------------------------+ |
>>                            | |                             | |
>>                            | |             SIP             | |
>>                            | |            Stack            | |
>>                            | |                             | |
>>                            | +-----------------------------+ |
>>                            |                                 |
>>                            +---------------------------------+
>>
>>                                      Physical Device
>>
>>
>> Another argument for different User Agents is the whole "route based
> on
>> SDP"
>> argument.  As the author of RFC 3458, I think I was first to make the
>> assertion that knowing the media type did NOT mean you know the
>> application
>> to use.  I don't want your phone or proxy or B2BUA to guess that
> because
>> I
>> have Chess Media, I want the Chess Application; I want to talk with
> your
>> Chess Application.
>>
>> Now, is the Chess Application a Preference, Requirement, or the thing
>> that I
>> am addressing?  The answer drives one towards using caller prefs, the
>> Require header, or the Request-URI.
>>
>> I would offer that it is not a preference: I really want the chess
>> application; I probably do not want to talk to you.
>>
>> That leaves us with the Require header or a Request-URI.  My leanings
>> are
>> not surprisingly towards the RFC 4240 approach.  Frankly, if you are
>> building an IMS (as an example) that can handle 10,000,000
> subscribers,
>> you
>> should be able to handle 100,000,000 registrations.  That order of
>> magnitude
>> is way easier than the 10,000 to 10,000,000 jump.  I don't buy the
>> "registering each service is too much of a burden on the operator"
>> argument,
>> because it is not that much of a burden, and the network actually  
>> does
>> the
>> right thing.  Better yet, all of the corner cases work.  Best of all,
>> the
>> IETF doesn't have to build an IANA Registry and process a thousand PS
>> documents to register urn:service:myco:widget.
>>
>> I suppose I could live with Require:, only if you are way too lazy to
>> register each application instance.  It does semantically work,
> because
>> if I
>> want to do PoC and you do not do PoC, I'll get a 4xx response, to
> which
>> I
>> could try a regular call.
>>
>>
> ______________________________________________________________________ 
> _
>> Notice:  This email message, together with any attachments, may
> contain
>> information  of  BEA Systems,  Inc.,  its subsidiaries  and
> affiliated
>> entities,  that may be confidential,  proprietary,  copyrighted
> and/or
>> legally privileged, and is intended solely for the use of the
> individual
>> or entity named in this message. If you are not the intended
> recipient,
>> and have received this message in error, please immediately return
> this
>> by email and then delete it.
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP
>> Use sip-implementors@cs.columbia.edu for questions on current sip
>> Use sip@ietf.org for new developments of core SIP
>>
>>
>>
>> ---------------------------------------------------------------------
>> This transmission (including any attachments) may contain  
>> confidential
> information, privileged material (including material protected by the
> solicitor-client or other applicable privileges), or constitute
> non-public information. Any use of this information by anyone other  
> than
> the intended recipient is prohibited. If you have received this
> transmission in error, please immediately reply to the sender and  
> delete
> this information from your system. Use, dissemination,  
> distribution, or
> reproduction of this transmission by unintended recipients is not
> authorized and may be unlawful.
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP
>> Use sip-implementors@cs.columbia.edu for questions on current sip
>> Use sip@ietf.org for new developments of core SIP
>>
>
>
>
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain  
> confidential information, privileged material (including material  
> protected by the solicitor-client or other applicable privileges),  
> or constitute non-public information. Any use of this information  
> by anyone other than the intended recipient is prohibited. If you  
> have received this transmission in error, please immediately reply  
> to the sender and delete this information from your system. Use,  
> dissemination, distribution, or reproduction of this transmission  
> by unintended recipients is not authorized and may be unlawful.
>
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP


--Apple-Mail-6-660558670
Content-Transfer-Encoding: base64
Content-Type: application/pkcs7-signature;
	name=smime.p7s
Content-Disposition: attachment;
	filename=smime.p7s
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--Apple-Mail-6-660558670--


--===============1517427085==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1517427085==--




From sipping-bounces@ietf.org Fri Feb 23 09:38:56 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKbZP-0002Cd-MK; Fri, 23 Feb 2007 09:38:47 -0500
Received: from [10.90.34.44] (helo=chiedprmail1.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKbZO-0002CU-1D
	for sipping@ietf.org; Fri, 23 Feb 2007 09:38:46 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by chiedprmail1.ietf.org with esmtp (Exim 4.43) id 1HKbZG-0003Dd-CG
	for sipping@ietf.org; Fri, 23 Feb 2007 09:38:45 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	C745820857; Fri, 23 Feb 2007 15:38:07 +0100 (CET)
X-AuditID: c1b4fb3c-af7c7bb0000007de-40-45defc4fc71d 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	ACB6720616; Fri, 23 Feb 2007 15:38:07 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 15:37:58 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP - Routing decissionsbased
	on Require/P-R
Date: Fri, 23 Feb 2007 15:37:50 +0100
Message-ID: <7374777208BDC7449D5620EF94232567023BD791@esealmw113.eemea.ericsson.se>
In-Reply-To: <45DEE717.1070609@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP - Routing
	decissionsbased on Require/P-R
Thread-Index: AcdXS5qYFgouhuF8TYyOm40aPufdNAACx5mw
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>
X-OriginalArrivalTime: 23 Feb 2007 14:37:58.0885 (UTC)
	FILETIME=[36A4B150:01C75758]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 9a2be21919e71dc6faef12b370c4ecf5
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,

>The difference here is that the Require header is definitive=20
>that the caller doesn't want to reach a UAS unless that UAS=20
>supports the stated options. Callerprefs is not so=20
>definitive. There are some uses of callerprefs that can be=20
>considered fairly definitive, but in general that is not the case.
>=20
>Jonathan proposed extending 3841 so that the use of Require=20
>generates an implicit callerpref. And that in turn would=20
>influence the selection of the registered contacts to try. I=20
>can't see a plausible way to infer any implicit callerpref=20
>from Accept-Contact headers.
>=20
>In any case, that is all about selecting how to translate the=20
>R-URI. It is not about selecting intermediate proxies. As=20
>Dale has stated, you are free to use most anything to select=20
>an intermediate proxy. If you want to use Require, or=20
>Accept-Contact to do that you may do so. But while you are=20
>free to do so, it is less clear that it is useful or wise to=20
>do so. If the intent is to select a proxy that is consistent=20
>with the capabilities of the UAS that will eventually be=20
>accepted (as seems to be the case with PoC), then it only=20
>makes sense if the thing you are testing is definitive for=20
>the capabilities of the UAS. So again, Require may make sense=20
>for this, while Accept-Contact is not.

Don't you get the same effect by including explicit require in the
Accept-Contact and adding "pref" to the Require Header?=20

Regards,

Christer




>=20
> 	Paul
>=20
> Christer Holmberg (JO/LMF) wrote:
> > Hi,
> >=20
> >> -----Original Message-----
> >> From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net]
> >> Sent: 22. helmikuuta 2007 17:35
> >> To: sipping@ietf.org
> >> Subject: Re: [Sipping] Service Identification in SIP - Routing=20
> >> decissionsbased on Require/P-R
> >>
> >>    From: Henning Schulzrinne <hgs@cs.columbia.edu>
> >>
> >>    Proxies can choose next hops based on RFC 3840, which=20
> is another=20
> >> way
> >>    of saying "routing".
> >>
> >> RFC 3261 allows a proxy unlimited latitude in choosing targets to=20
> >> which to forward a request, if the proxy is authoritative for the
> >> request-URI:
> >>
> >> "[T]he element is responsible for the domain in the=20
> Request-URI, and=20
> >> the element MAY use whatever mechanism it desires to=20
> determine where=20
> >> to send the request."
> >>
> >> Within that framework, it seems to me quite reasonable for=20
> a service=20
> >> provider's proxy to examine Require or Proxy-Require headers to=20
> >> determine which application server would be best for=20
> handling a call,=20
> >> and routing the call to it.
> >=20
> > Why can't Accept-Contact be used for the same purpose?
> >=20
> > My issue is the statement that Require/Proxy-Require is the most=20
> > appropriate way of doing the routing decisson, and I am wondering=20
> > where that is specified.
> >=20
> > Regards,
> >=20
> > Christer
> >=20
> >=20
> >=20
> >=20
> >=20
> >> Dale
> >>
> >> _______________________________________________
> >> Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> >> This list is for NEW development of the application of SIP Use=20
> >> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> >> sip@ietf.org for new developments of core SIP
> >>
> >=20
> > _______________________________________________
> > Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> > This list is for NEW development of the application of SIP Use=20
> > sip-implementors@cs.columbia.edu for questions on current sip Use=20
> > sip@ietf.org for new developments of core SIP
> >=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 09:46:20 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKbgP-0001Ha-G0; Fri, 23 Feb 2007 09:46:01 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKbgO-0001HL-1d
	for sipping@ietf.org; Fri, 23 Feb 2007 09:46:00 -0500
Received: from rtp-iport-2.cisco.com ([64.102.122.149])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKbgL-0006CP-1x
	for sipping@ietf.org; Fri, 23 Feb 2007 09:46:00 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by rtp-iport-2.cisco.com with ESMTP; 23 Feb 2007 09:45:57 -0500
X-IronPort-AV: i="4.14,211,1170651600"; 
	d="scan'208"; a="114394300:sNHT54816964"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1NEjukn015946; 
	Fri, 23 Feb 2007 09:45:56 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1NEjtVd025726; 
	Fri, 23 Feb 2007 09:45:56 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 09:45:32 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 09:45:32 -0500
Message-ID: <45DEFE0C.603@cisco.com>
Date: Fri, 23 Feb 2007 09:45:32 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Paul Kyzivat <pkyzivat@cisco.com>
Subject: Re: [Sipping] Service Identification in SIP - Routing decissionsbased
	on Require/P-R
References: <7374777208BDC7449D5620EF94232567023932CA@esealmw113.eemea.ericsson.se>
	<45DEE717.1070609@cisco.com>
In-Reply-To: <45DEE717.1070609@cisco.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 23 Feb 2007 14:45:32.0579 (UTC)
	FILETIME=[4510E730:01C75759]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=6398; t=1172241956;
	x=1173105956; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP=20-=2
	0Routing=20decissionsbased=0A=20on=20Require/P-R |Sender:=20
	|To:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>;
	bh=rkm1CXKKyNtiqRQBl0SpS+mcRLcNTnwKIEe0psBTefI=;
	b=Tc/TXbmBUOCh/OO1b9VelFqLwceGfXL5Z6XmV00EyANgfyKzGHIDDYHgxHr3HoIbc5vGosVm
	yO+JtalDhq8lVKr8qKZIBIvpUjqZUgMwVKNU2EqrlrFrlWVeHU6OUMR6;
Authentication-Results: rtp-dkim-1; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 932cba6e0228cc603da43d861a7e09d8
Cc: sipping@ietf.org,
	"Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Some further thoughts on this subject:

Typically, some of the capabilities used in a call are *negotiated* 
between the caller and the callee. As a result, those that will be used 
are not known for certain at the time the request is sent. And the 
results of that negotiation may differ depending on which of various 
candidate targets is selected.

Apparently one key intended use of the service id is to bring in one or 
more intermediate nodes that are specific to the capabilities that will 
be used in the call. These intermediate nodes must be selected *before* 
the UAS is selected and the negotiation is done. For that to work, one 
of two things must happen:

- the proxy that selects those intermediate nodes
   must be prescient about what capabilities will be negotiated

- or the selection of UAS and the negotiation of capabilities to be
   used must be somehow constrained to be consistent with the
   capabilities supported by the selected intermediate nodes.

If we agree that prescience is infeasible, I guess we must assume that 
the selection and negotiation will be constrained. We of course have a 
mechanism for that: it is the Require header. There are also other 
mechanisms that provide constraints of some sort. For instance the SDP 
offer constrains what is permitted in the answer. (It does *not* however 
constrain subsequent offers and answers in the same session.)

If the request is already sufficiently constraining for the purposes of 
the intermediate nodes then nothing more need be done. But if the 
request isn't sufficiently constraining, then the intermediate node 
ought to add a constraint to the request, to ensure its requirements are 
satisfied. E.g. it could add a Require header to the request. Or, if it 
is a B2BUA, it might adjust the offer in the request to be suitably 
restrictive.

How does all this relate to the discussion?

If the service id is indicated by an Accept-Contact, then it is not in 
general definitive for what capabilities will eventually be used in a 
call. (This is true because honoring of callerprefs by proxies is 
*optional*, and because in general callerprefs involves the 
*prioritzation* of contacts rather than being a strict acceptance test.

If the Accept-Contact is used to select an intermediate node that makes 
some assumptions about the nature of the resulting call, then it needs 
to somehow ensure that the UAS ultimately selected will be consistent 
with those assumptions. It will need to use something other than 
Accept-Contact to do that. So at a minimum A-C is not a complete 
solution to this problem.

	Paul

Paul Kyzivat wrote:
> The difference here is that the Require header is definitive that the 
> caller doesn't want to reach a UAS unless that UAS supports the stated 
> options. Callerprefs is not so definitive. There are some uses of 
> callerprefs that can be considered fairly definitive, but in general 
> that is not the case.
> 
> Jonathan proposed extending 3841 so that the use of Require generates an 
> implicit callerpref. And that in turn would influence the selection of 
> the registered contacts to try. I can't see a plausible way to infer any 
> implicit callerpref from Accept-Contact headers.
> 
> In any case, that is all about selecting how to translate the R-URI. It 
> is not about selecting intermediate proxies. As Dale has stated, you are 
> free to use most anything to select an intermediate proxy. If you want 
> to use Require, or Accept-Contact to do that you may do so. But while 
> you are free to do so, it is less clear that it is useful or wise to do 
> so. If the intent is to select a proxy that is consistent with the 
> capabilities of the UAS that will eventually be accepted (as seems to be 
> the case with PoC), then it only makes sense if the thing you are 
> testing is definitive for the capabilities of the UAS. So again, Require 
> may make sense for this, while Accept-Contact is not.
> 
>     Paul
> 
> Christer Holmberg (JO/LMF) wrote:
>> Hi,
>>> -----Original Message-----
>>> From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net] Sent: 
>>> 22. helmikuuta 2007 17:35
>>> To: sipping@ietf.org
>>> Subject: Re: [Sipping] Service Identification in SIP - Routing 
>>> decissionsbased on Require/P-R
>>>
>>>    From: Henning Schulzrinne <hgs@cs.columbia.edu>
>>>
>>>    Proxies can choose next hops based on RFC 3840, which is another 
>>> way     of saying "routing".
>>>
>>> RFC 3261 allows a proxy unlimited latitude in choosing targets to 
>>> which to forward a request, if the proxy is authoritative for the
>>> request-URI:
>>>
>>> "[T]he element is responsible for the domain in the Request-URI, and 
>>> the element MAY use whatever mechanism it desires to determine where 
>>> to send the request."
>>>
>>> Within that framework, it seems to me quite reasonable for a service 
>>> provider's proxy to examine Require or Proxy-Require headers to 
>>> determine which application server would be best for handling a call, 
>>> and routing the call to it.
>>
>> Why can't Accept-Contact be used for the same purpose?
>>
>> My issue is the statement that Require/Proxy-Require is the most
>> appropriate way of doing the routing decisson, and I am wondering where
>> that is specified.
>>
>> Regards,
>>
>> Christer
>>
>>
>>
>>
>>
>>> Dale
>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP Use 
>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>> sip@ietf.org for new developments of core SIP
>>>
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP
>> Use sip-implementors@cs.columbia.edu for questions on current sip
>> Use sip@ietf.org for new developments of core SIP
>>
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 10:14:10 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKc6h-0000z0-GX; Fri, 23 Feb 2007 10:13:11 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKc6f-0000xt-WF
	for sipping@ietf.org; Fri, 23 Feb 2007 10:13:10 -0500
Received: from zrtps0kp.nortel.com ([47.140.192.56])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKc6d-0003uE-Lz
	for sipping@ietf.org; Fri, 23 Feb 2007 10:13:09 -0500
Received: from zcarhxm1.corp.nortel.com (zcarhxm1.corp.nortel.com
	[47.129.230.97])
	by zrtps0kp.nortel.com (Switch-2.2.6/Switch-2.2.0) with ESMTP id
	l1NFD2s08303; Fri, 23 Feb 2007 10:13:02 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Session Establishment Efficiency Rate
Date: Fri, 23 Feb 2007 10:13:00 -0500
Message-ID: <1444BF0ED21A0E44B1E6C9F8F9C5BFB606EB00E7@zcarhxm1.corp.nortel.com>
In-Reply-To: <1172189634.17141.59.camel@montag.eng.level3.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Session Establishment Efficiency Rate
Thread-Index: AcdW3n089iyN/mXHQmikt18LcZ6r+gAfRSfw
From: "Tadeusz Drwiega" <drwiega@nortel.com>
To: "Daryl Malas" <daryl@level3.net>, <sipping@ietf.org>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a7d6aff76b15f3f56fcb94490e1052e4
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

There should be a metric corresponding to NER for SIP. You may call it
session establishment efficiency ratio. SER does not help to assess how
the network performance, because it also depends on user behavior. SER
may be relatively low, say 0.8, and the network may perform very well.
On the other hand SEER should be close to 1.

Regards,
Tadeusz


> -----Original Message-----
> From: Daryl Malas [mailto:daryl@level3.net]=20
> Sent: Thursday, February 22, 2007 7:14 PM
> To: sipping@ietf.org
> Subject: [Sipping] Session Establishment Efficiency Rate
>=20
>=20
> A couple more people have expressed to me that this=20
> adaptation of the ITU-T E.411 Network Efficiency Rate metric=20
> is very valuable.  The snapshot of the metric from the draft is:
>=20
> "This metric is similar to Session Establishment Rate, but it=20
> is intended to exclude the effects of customer behavior and=20
> user agent behavior.  Essentially, the attempts resulting=20
> with a final cause code of 480, 486, and 600 would be added=20
> to the SER numerator."
>=20
> Does anyone else want to see this metric in the draft?  I=20
> personally think it is an interesting metric.
>=20
> Thanks...
>=20
> --Daryl
>=20
>=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP

From sipping-bounces@ietf.org Fri Feb 23 10:14:10 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKc65-0000gm-Lz; Fri, 23 Feb 2007 10:12:33 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKc64-0000gh-In
	for sipping@ietf.org; Fri, 23 Feb 2007 10:12:32 -0500
Received: from rtp-iport-1.cisco.com ([64.102.122.148])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKc62-0003oK-S1
	for sipping@ietf.org; Fri, 23 Feb 2007 10:12:32 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by rtp-iport-1.cisco.com with ESMTP; 23 Feb 2007 10:12:29 -0500
X-IronPort-AV: i="4.14,211,1170651600"; 
	d="scan'208"; a="53520965:sNHT74156430"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1NFCSri025559; 
	Fri, 23 Feb 2007 10:12:28 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1NFC6Vt002276; 
	Fri, 23 Feb 2007 10:12:24 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 10:12:18 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 10:12:17 -0500
Message-ID: <45DF0451.7020209@cisco.com>
Date: Fri, 23 Feb 2007 10:12:17 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Byron Campen <bcampen@estacado.net>
Subject: Re: [Sipping] Service Identification Model
References: <61968779B8AC4C4BAB421D4C12F008C009463E2E@XCH47YKF.rim.net>
	<FC318FB3-A497-4BED-BBC6-3B5B630FC589@estacado.net>
In-Reply-To: <FC318FB3-A497-4BED-BBC6-3B5B630FC589@estacado.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 23 Feb 2007 15:12:17.0814 (UTC)
	FILETIME=[01DC5360:01C7575D]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=16116; t=1172243548;
	x=1173107548; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20Model
	|Sender:=20 |To:=20Byron=20Campen=20<bcampen@estacado.net>;
	bh=B015KvhT+TgjJqrBZIjXeHnFVomu3kx3uW9CV8Q9qh0=;
	b=tQJBCBnaHopnGr4OXoTR16mCv+0JY7bUyItUBwVmbsvCUycY5/KZwfNm1XHNnkvOwyijXVV1
	zN5SLD5/iR9t3xHigUUb6CwpZSLJQll7pJBImlXOFWvnOTwqGO9X0TTD;
Authentication-Results: rtp-dkim-2; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: b058151374d77ee76edaac850f7449fb
Cc: Andrew Allen <aallen@rim.com>, sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Byron Campen wrote:
>>
>> 1. Basic Telephony (Conversational QoS Voice only with telephony feature
>> set)
>> 2. Multimedia Telephony (Conversational QoS Voice plus video and other
>> media streams with telephony feature set)

Why are 1 & 2 different services? 1 is a proper subset of 2.

>> 3. VoIP (Background QoS with no telephony features)

And why is 3 different from 1 and 2? Are you saying that to be able to 
use a feature I must decide *before* I make the call that I might want to?

Even if that is the case, that sounds like a policy issue between a user 
and his own provider, rather than an e2e negotiation. (My features are 
provided by my UA and/or my provider.)

>> 4. SMS
>> 5. MMS

Why are 4 and 5 different services? They are both datagrams. The 
difference in media types is irrelevant. If its delivered via MESSAGE, 
then it is already distinguishable from calls. This should be able to be 
part of the same service as IM.

>> 6. Instant Messaging

Is just another media type. Like voice and video, these can be 
negotiated for the session. Why balkanize as a different service?

>> 7. Push to Talk (PoC)

We've beaten this one to death already. Current implementation forces 
need for essentially an overlay network to support it. But its far from 
an exemplar of how to do it.

>> 8. Network Game 1 (e.g Chess)
>> 9. Network Game 2 (e.g Martial arts game)
>> 10. Network Game 3 (e.g Dungeons & Dragons)

I tend to agree with Eric that service URIs are good for these if there 
is no better way. But in practice I think these could easily be 
distinguished in SDP using o/a.

>> 11. Push Service 1 (e.g. Traffic information service)
>> 12. Push Service 2 (e.g. Weather information service)
>> 13. Push Service 3 (e.g  News information service)
>> 14. Push Service 4 (e.g  Stock information service)

There may be *something* here. I need to know more about the 
characteristics of these. If they consist of human intelligible media, 
then I think they could adequately be dealt with via callerprefs. (This 
may be a bit like the Actor feature tag.) Specifically, callerprefs are 
fine if its ok that the call end up elsewhere if the callerprefs 
selection doesn't happen or there isn't any target with that capability.

If the information isn't intelligible media, then it ought to be 
distinguishable as a special media type in the offer.

>> 15. Mobile TV service

AFAIK, the station doesn't call me to deliver me tv service. I would be 
initiating the call. And presumably it will be to a target service URI 
or URN. So I see no need for a service id here.

>> So there's 15 which may all be different applications on a handset and
>> all those are existing services or those already envisioned.

So I only see:

- 1 for basic telephony, multimedia telephony, SMS, MMS, IM

- *maybe* something for games

- *maybe* something for push services

	Paul

>     I only see 10 here that would need to ever send a register request. 
> (11-15 are strictly client-server) 4-6 will have distinct Content-Type 
> headers and payloads. You would need only 1 UA for these. So we're down 
> to 8. 8-10 will, once again, be distinct when you look at the payload 
> and Content-Type. So we're down to 6. There is no reason at all that the 
> UA that handles voice is incapable of handling IM, and you can tell the 
> difference by looking at the request type, so we're down to 5. As for 
> the voice stuff, you can distinguish 2 from anything else through the 
> payload. So, we could quite easily roll 1-2 together. (4) 3 can be 
> distinguished by watching for any in-dialog call-control stuff. (Having 
> a service identifier that essentially says "I promise not to use any 
> fancy telephony features on this call. No, really, I promise!" is never 
> going to work) So we're down to 3.
> 
>     SIP is more than flexible enough to handle all of this stuff in a 
> single UA. The UAs are perfectly capable of figuring out what "service" 
> the other wants to use; why are your middleboxes unable to do the same? 
> Adding a parallel method to make things easy on the middleboxes is just 
> asking for trouble, because you're conveying the same information twice. 
> It is ripe for abuse. Of course, the easy way to do all this is to just 
> let UAs send what they want, and charge per KB, but hey, what do I know 
> about simplicity, flexibility, scalability, ease-of-deployment, and 
> customer satisfaction? I'm just a IETFer.
> 
> Best regards,
> Byron Campen
> 
> 
>> -----Original Message-----
>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
>> Sent: Thursday, February 22, 2007 7:05 PM
>> To: Andrew Allen
>> Cc: Eric Burger; sipping@ietf.org
>> Subject: Re: [Sipping] Service Identification Model
>>
>>
>>
>> Andrew Allen wrote:
>>> Hi Eric,
>>>
>>> It seems your prefer the approach of using URIs to identify services
>>> which is part of Jonathan's draft as is the use of Require for those
>>> services which simply do not interoperate unless the UAS understands
>> the
>>> service.
>>>
>>> Where I am concerned is that you seem to be suggesting that the UA
>>> separately register an AoR for each service it supports using
>> REGISTER.
>>> Certainly for cellular that wouldn't be acceptable as 10 or more
>>> services could easily be supported by a single device.
>>
>> Andrew, can you give me an example of 10 "services" on the device that
>> need to be distinguished from one another in this way - that aren't just
>>
>> optional behavior of a particular service? (E.g. I see no reason why
>> "audio" and "audio/video" need to be separate services.)
>>
>> I guess as things stand, PoC must be a separate service. And apparently
>> Chess. What else?
>>
>>     Paul
>>
>>> Registrations
>>> create a significant load in the IMS network because its not just a
>> SIP
>>> REGISTER but also requires a lot of interactions with other network
>>> elements such as the HSS and Notifications for reg-events etc. Also
>>> having an order of magnitude increase in multiple registrations is
>> going
>>> to be a big issue for the avalanche restart problem for example when a
>>> cell site goes down and comes back into service.
>>>
>>> I think Jonathan's approach gives us the possibility for the registrar
>>> to create the Service URIs based on the feature parameters included in
>>> the Contact of the REGISTER.
>>>
>>> I think the argument over whether the model is one UA with multiple
>>> services or multiple UAs each representing a separate service is
>>> somewhat academic. As I understand a UA can register multiple AoRs and
>>> can have other AoRs configured or otherwise provided by the service
>>> provider.
>>>
>>> Andrew
>>>
>>> -----Original Message-----
>>> From: Eric Burger [mailto:eburger@bea.com]
>>> Sent: Wednesday, February 21, 2007 1:29 PM
>>> To: sipping@ietf.org
>>> Subject: [Sipping] Service Identification Model
>>>
>>> Folks have been pestering me for a while to look at the Service
>>> Identification issue.  I finally have gotten around to it.
>>>
>>>
>>> I guess I have a fundamental problem with the principal assertion of
>> the
>>> service-identification draft.  Namely, the assertion that there is a
>>> single
>>> SIP Layer, separate and distinct from two services.  Here is Figure 2
>>> from
>>> the draft:
>>>
>>>                            +---------------------------------+
>>>                            |                                 |
>>>                            | +-------------+ +-------------+ |
>>>                            | |      UI     | |      UI     | |
>>>                            | +-------------+ +-------------+ |
>>>                            | +-------------+ +-------------+ |
>>>                            | |             | |             | |
>>>                            | |  Service 1  | |  Service 2  | |
>>>                            | |             | |             | |
>>>                            | +-------------+ +-------------+ |
>>>                            | +-----------------------------+ |
>>>                            | |                             | |
>>>                            | |             SIP             | |
>>>                            | |            Layer            | |
>>>                            | |                             | |
>>>                            | +-----------------------------+ |
>>>                            |                                 |
>>>                            +---------------------------------+
>>>
>>>                                      Physical Device
>>>
>>> I would offer this drawing confuses a common SIP Stack with a common
>>> User
>>> Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use case:
>>>
>>>> ... In addition users may have
>>>> multiple devices only one of which has the application that
>>> understands
>>>> the game. Other devices that the user has registered against his AoR
>>> may
>>>> handle audio for telephony and other voice applications but if Alice
>>>> invites Bob to play her favourite game it is not very interesting if
>>> Bob
>>>> accepts the gaming session on a device that can only accept the audio
>>>> stream. Also even when the terminal that has the game application
>>>> receives the INVITE the correct application in that terminal needs to
>>> be
>>>> identified to accept the session.
>>>
>>> I would offer that this is the clinching argument that the request is,
>>> in
>>> fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a
>> User
>>> Agent.
>>>
>>> Same argument for PoC versus a regular voice call: One is looking for
>> a
>>> PoC
>>> User Agent, NOT the PoC service at a given User Agent.
>>>
>>> Thus, the picture really looks like the following:
>>>
>>>                            +---------------------------------+
>>>                            |                                 |
>>>                            | +-------------+ +-------------+ |
>>>                            | |    UI       | |    UI       | |
>>>                            | +-------------+ +-------------+ |
>>>                            | +-------------+ +-------------+ |
>>>                            | |             | |             | |
>>>                            | |    UAS 1    | |    UAS 2    | |
>>>                            | |             | |             | |
>>>                            | +-------------+ +-------------+ |
>>>                            | +-----------------------------+ |
>>>                            | |                             | |
>>>                            | |             SIP             | |
>>>                            | |            Stack            | |
>>>                            | |                             | |
>>>                            | +-----------------------------+ |
>>>                            |                                 |
>>>                            +---------------------------------+
>>>
>>>                                      Physical Device
>>>
>>>
>>> Another argument for different User Agents is the whole "route based
>> on
>>> SDP"
>>> argument.  As the author of RFC 3458, I think I was first to make the
>>> assertion that knowing the media type did NOT mean you know the
>>> application
>>> to use.  I don't want your phone or proxy or B2BUA to guess that
>> because
>>> I
>>> have Chess Media, I want the Chess Application; I want to talk with
>> your
>>> Chess Application.
>>>
>>> Now, is the Chess Application a Preference, Requirement, or the thing
>>> that I
>>> am addressing?  The answer drives one towards using caller prefs, the
>>> Require header, or the Request-URI.
>>>
>>> I would offer that it is not a preference: I really want the chess
>>> application; I probably do not want to talk to you.
>>>
>>> That leaves us with the Require header or a Request-URI.  My leanings
>>> are
>>> not surprisingly towards the RFC 4240 approach.  Frankly, if you are
>>> building an IMS (as an example) that can handle 10,000,000
>> subscribers,
>>> you
>>> should be able to handle 100,000,000 registrations.  That order of
>>> magnitude
>>> is way easier than the 10,000 to 10,000,000 jump.  I don't buy the
>>> "registering each service is too much of a burden on the operator"
>>> argument,
>>> because it is not that much of a burden, and the network actually does
>>> the
>>> right thing.  Better yet, all of the corner cases work.  Best of all,
>>> the
>>> IETF doesn't have to build an IANA Registry and process a thousand PS
>>> documents to register urn:service:myco:widget.
>>>
>>> I suppose I could live with Require:, only if you are way too lazy to
>>> register each application instance.  It does semantically work,
>> because
>>> if I
>>> want to do PoC and you do not do PoC, I'll get a 4xx response, to
>> which
>>> I
>>> could try a regular call.
>>>
>>>
>> _______________________________________________________________________
>>> Notice:  This email message, together with any attachments, may
>> contain
>>> information  of  BEA Systems,  Inc.,  its subsidiaries  and
>> affiliated
>>> entities,  that may be confidential,  proprietary,  copyrighted
>> and/or
>>> legally privileged, and is intended solely for the use of the
>> individual
>>> or entity named in this message. If you are not the intended
>> recipient,
>>> and have received this message in error, please immediately return
>> this
>>> by email and then delete it.
>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP
>>> Use sip-implementors@cs.columbia.edu for questions on current sip
>>> Use sip@ietf.org for new developments of core SIP
>>>
>>>
>>>
>>> ---------------------------------------------------------------------
>>> This transmission (including any attachments) may contain confidential
>> information, privileged material (including material protected by the
>> solicitor-client or other applicable privileges), or constitute
>> non-public information. Any use of this information by anyone other than
>> the intended recipient is prohibited. If you have received this
>> transmission in error, please immediately reply to the sender and delete
>> this information from your system. Use, dissemination, distribution, or
>> reproduction of this transmission by unintended recipients is not
>> authorized and may be unlawful.
>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP
>>> Use sip-implementors@cs.columbia.edu for questions on current sip
>>> Use sip@ietf.org for new developments of core SIP
>>>
>>
>>
>>
>> ---------------------------------------------------------------------
>> This transmission (including any attachments) may contain confidential 
>> information, privileged material (including material protected by the 
>> solicitor-client or other applicable privileges), or constitute 
>> non-public information. Any use of this information by anyone other 
>> than the intended recipient is prohibited. If you have received this 
>> transmission in error, please immediately reply to the sender and 
>> delete this information from your system. Use, dissemination, 
>> distribution, or reproduction of this transmission by unintended 
>> recipients is not authorized and may be unlawful.
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP
>> Use sip-implementors@cs.columbia.edu for questions on current sip
>> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP





From sipping-bounces@ietf.org Fri Feb 23 10:14:10 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKc6h-0000z0-GX; Fri, 23 Feb 2007 10:13:11 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKc6f-0000xt-WF
	for sipping@ietf.org; Fri, 23 Feb 2007 10:13:10 -0500
Received: from zrtps0kp.nortel.com ([47.140.192.56])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKc6d-0003uE-Lz
	for sipping@ietf.org; Fri, 23 Feb 2007 10:13:09 -0500
Received: from zcarhxm1.corp.nortel.com (zcarhxm1.corp.nortel.com
	[47.129.230.97])
	by zrtps0kp.nortel.com (Switch-2.2.6/Switch-2.2.0) with ESMTP id
	l1NFD2s08303; Fri, 23 Feb 2007 10:13:02 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Session Establishment Efficiency Rate
Date: Fri, 23 Feb 2007 10:13:00 -0500
Message-ID: <1444BF0ED21A0E44B1E6C9F8F9C5BFB606EB00E7@zcarhxm1.corp.nortel.com>
In-Reply-To: <1172189634.17141.59.camel@montag.eng.level3.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Session Establishment Efficiency Rate
Thread-Index: AcdW3n089iyN/mXHQmikt18LcZ6r+gAfRSfw
From: "Tadeusz Drwiega" <drwiega@nortel.com>
To: "Daryl Malas" <daryl@level3.net>, <sipping@ietf.org>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a7d6aff76b15f3f56fcb94490e1052e4
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

There should be a metric corresponding to NER for SIP. You may call it
session establishment efficiency ratio. SER does not help to assess how
the network performance, because it also depends on user behavior. SER
may be relatively low, say 0.8, and the network may perform very well.
On the other hand SEER should be close to 1.

Regards,
Tadeusz


> -----Original Message-----
> From: Daryl Malas [mailto:daryl@level3.net]=20
> Sent: Thursday, February 22, 2007 7:14 PM
> To: sipping@ietf.org
> Subject: [Sipping] Session Establishment Efficiency Rate
>=20
>=20
> A couple more people have expressed to me that this=20
> adaptation of the ITU-T E.411 Network Efficiency Rate metric=20
> is very valuable.  The snapshot of the metric from the draft is:
>=20
> "This metric is similar to Session Establishment Rate, but it=20
> is intended to exclude the effects of customer behavior and=20
> user agent behavior.  Essentially, the attempts resulting=20
> with a final cause code of 480, 486, and 600 would be added=20
> to the SER numerator."
>=20
> Does anyone else want to see this metric in the draft?  I=20
> personally think it is an interesting metric.
>=20
> Thanks...
>=20
> --Daryl
>=20
>=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP

From sipping-bounces@ietf.org Fri Feb 23 10:14:10 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKc65-0000gm-Lz; Fri, 23 Feb 2007 10:12:33 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKc64-0000gh-In
	for sipping@ietf.org; Fri, 23 Feb 2007 10:12:32 -0500
Received: from rtp-iport-1.cisco.com ([64.102.122.148])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKc62-0003oK-S1
	for sipping@ietf.org; Fri, 23 Feb 2007 10:12:32 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by rtp-iport-1.cisco.com with ESMTP; 23 Feb 2007 10:12:29 -0500
X-IronPort-AV: i="4.14,211,1170651600"; 
	d="scan'208"; a="53520965:sNHT74156430"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1NFCSri025559; 
	Fri, 23 Feb 2007 10:12:28 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1NFC6Vt002276; 
	Fri, 23 Feb 2007 10:12:24 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 10:12:18 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 10:12:17 -0500
Message-ID: <45DF0451.7020209@cisco.com>
Date: Fri, 23 Feb 2007 10:12:17 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Byron Campen <bcampen@estacado.net>
Subject: Re: [Sipping] Service Identification Model
References: <61968779B8AC4C4BAB421D4C12F008C009463E2E@XCH47YKF.rim.net>
	<FC318FB3-A497-4BED-BBC6-3B5B630FC589@estacado.net>
In-Reply-To: <FC318FB3-A497-4BED-BBC6-3B5B630FC589@estacado.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 23 Feb 2007 15:12:17.0814 (UTC)
	FILETIME=[01DC5360:01C7575D]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=16116; t=1172243548;
	x=1173107548; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20Model
	|Sender:=20 |To:=20Byron=20Campen=20<bcampen@estacado.net>;
	bh=B015KvhT+TgjJqrBZIjXeHnFVomu3kx3uW9CV8Q9qh0=;
	b=tQJBCBnaHopnGr4OXoTR16mCv+0JY7bUyItUBwVmbsvCUycY5/KZwfNm1XHNnkvOwyijXVV1
	zN5SLD5/iR9t3xHigUUb6CwpZSLJQll7pJBImlXOFWvnOTwqGO9X0TTD;
Authentication-Results: rtp-dkim-2; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: b058151374d77ee76edaac850f7449fb
Cc: Andrew Allen <aallen@rim.com>, sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Byron Campen wrote:
>>
>> 1. Basic Telephony (Conversational QoS Voice only with telephony feature
>> set)
>> 2. Multimedia Telephony (Conversational QoS Voice plus video and other
>> media streams with telephony feature set)

Why are 1 & 2 different services? 1 is a proper subset of 2.

>> 3. VoIP (Background QoS with no telephony features)

And why is 3 different from 1 and 2? Are you saying that to be able to 
use a feature I must decide *before* I make the call that I might want to?

Even if that is the case, that sounds like a policy issue between a user 
and his own provider, rather than an e2e negotiation. (My features are 
provided by my UA and/or my provider.)

>> 4. SMS
>> 5. MMS

Why are 4 and 5 different services? They are both datagrams. The 
difference in media types is irrelevant. If its delivered via MESSAGE, 
then it is already distinguishable from calls. This should be able to be 
part of the same service as IM.

>> 6. Instant Messaging

Is just another media type. Like voice and video, these can be 
negotiated for the session. Why balkanize as a different service?

>> 7. Push to Talk (PoC)

We've beaten this one to death already. Current implementation forces 
need for essentially an overlay network to support it. But its far from 
an exemplar of how to do it.

>> 8. Network Game 1 (e.g Chess)
>> 9. Network Game 2 (e.g Martial arts game)
>> 10. Network Game 3 (e.g Dungeons & Dragons)

I tend to agree with Eric that service URIs are good for these if there 
is no better way. But in practice I think these could easily be 
distinguished in SDP using o/a.

>> 11. Push Service 1 (e.g. Traffic information service)
>> 12. Push Service 2 (e.g. Weather information service)
>> 13. Push Service 3 (e.g  News information service)
>> 14. Push Service 4 (e.g  Stock information service)

There may be *something* here. I need to know more about the 
characteristics of these. If they consist of human intelligible media, 
then I think they could adequately be dealt with via callerprefs. (This 
may be a bit like the Actor feature tag.) Specifically, callerprefs are 
fine if its ok that the call end up elsewhere if the callerprefs 
selection doesn't happen or there isn't any target with that capability.

If the information isn't intelligible media, then it ought to be 
distinguishable as a special media type in the offer.

>> 15. Mobile TV service

AFAIK, the station doesn't call me to deliver me tv service. I would be 
initiating the call. And presumably it will be to a target service URI 
or URN. So I see no need for a service id here.

>> So there's 15 which may all be different applications on a handset and
>> all those are existing services or those already envisioned.

So I only see:

- 1 for basic telephony, multimedia telephony, SMS, MMS, IM

- *maybe* something for games

- *maybe* something for push services

	Paul

>     I only see 10 here that would need to ever send a register request. 
> (11-15 are strictly client-server) 4-6 will have distinct Content-Type 
> headers and payloads. You would need only 1 UA for these. So we're down 
> to 8. 8-10 will, once again, be distinct when you look at the payload 
> and Content-Type. So we're down to 6. There is no reason at all that the 
> UA that handles voice is incapable of handling IM, and you can tell the 
> difference by looking at the request type, so we're down to 5. As for 
> the voice stuff, you can distinguish 2 from anything else through the 
> payload. So, we could quite easily roll 1-2 together. (4) 3 can be 
> distinguished by watching for any in-dialog call-control stuff. (Having 
> a service identifier that essentially says "I promise not to use any 
> fancy telephony features on this call. No, really, I promise!" is never 
> going to work) So we're down to 3.
> 
>     SIP is more than flexible enough to handle all of this stuff in a 
> single UA. The UAs are perfectly capable of figuring out what "service" 
> the other wants to use; why are your middleboxes unable to do the same? 
> Adding a parallel method to make things easy on the middleboxes is just 
> asking for trouble, because you're conveying the same information twice. 
> It is ripe for abuse. Of course, the easy way to do all this is to just 
> let UAs send what they want, and charge per KB, but hey, what do I know 
> about simplicity, flexibility, scalability, ease-of-deployment, and 
> customer satisfaction? I'm just a IETFer.
> 
> Best regards,
> Byron Campen
> 
> 
>> -----Original Message-----
>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
>> Sent: Thursday, February 22, 2007 7:05 PM
>> To: Andrew Allen
>> Cc: Eric Burger; sipping@ietf.org
>> Subject: Re: [Sipping] Service Identification Model
>>
>>
>>
>> Andrew Allen wrote:
>>> Hi Eric,
>>>
>>> It seems your prefer the approach of using URIs to identify services
>>> which is part of Jonathan's draft as is the use of Require for those
>>> services which simply do not interoperate unless the UAS understands
>> the
>>> service.
>>>
>>> Where I am concerned is that you seem to be suggesting that the UA
>>> separately register an AoR for each service it supports using
>> REGISTER.
>>> Certainly for cellular that wouldn't be acceptable as 10 or more
>>> services could easily be supported by a single device.
>>
>> Andrew, can you give me an example of 10 "services" on the device that
>> need to be distinguished from one another in this way - that aren't just
>>
>> optional behavior of a particular service? (E.g. I see no reason why
>> "audio" and "audio/video" need to be separate services.)
>>
>> I guess as things stand, PoC must be a separate service. And apparently
>> Chess. What else?
>>
>>     Paul
>>
>>> Registrations
>>> create a significant load in the IMS network because its not just a
>> SIP
>>> REGISTER but also requires a lot of interactions with other network
>>> elements such as the HSS and Notifications for reg-events etc. Also
>>> having an order of magnitude increase in multiple registrations is
>> going
>>> to be a big issue for the avalanche restart problem for example when a
>>> cell site goes down and comes back into service.
>>>
>>> I think Jonathan's approach gives us the possibility for the registrar
>>> to create the Service URIs based on the feature parameters included in
>>> the Contact of the REGISTER.
>>>
>>> I think the argument over whether the model is one UA with multiple
>>> services or multiple UAs each representing a separate service is
>>> somewhat academic. As I understand a UA can register multiple AoRs and
>>> can have other AoRs configured or otherwise provided by the service
>>> provider.
>>>
>>> Andrew
>>>
>>> -----Original Message-----
>>> From: Eric Burger [mailto:eburger@bea.com]
>>> Sent: Wednesday, February 21, 2007 1:29 PM
>>> To: sipping@ietf.org
>>> Subject: [Sipping] Service Identification Model
>>>
>>> Folks have been pestering me for a while to look at the Service
>>> Identification issue.  I finally have gotten around to it.
>>>
>>>
>>> I guess I have a fundamental problem with the principal assertion of
>> the
>>> service-identification draft.  Namely, the assertion that there is a
>>> single
>>> SIP Layer, separate and distinct from two services.  Here is Figure 2
>>> from
>>> the draft:
>>>
>>>                            +---------------------------------+
>>>                            |                                 |
>>>                            | +-------------+ +-------------+ |
>>>                            | |      UI     | |      UI     | |
>>>                            | +-------------+ +-------------+ |
>>>                            | +-------------+ +-------------+ |
>>>                            | |             | |             | |
>>>                            | |  Service 1  | |  Service 2  | |
>>>                            | |             | |             | |
>>>                            | +-------------+ +-------------+ |
>>>                            | +-----------------------------+ |
>>>                            | |                             | |
>>>                            | |             SIP             | |
>>>                            | |            Layer            | |
>>>                            | |                             | |
>>>                            | +-----------------------------+ |
>>>                            |                                 |
>>>                            +---------------------------------+
>>>
>>>                                      Physical Device
>>>
>>> I would offer this drawing confuses a common SIP Stack with a common
>>> User
>>> Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use case:
>>>
>>>> ... In addition users may have
>>>> multiple devices only one of which has the application that
>>> understands
>>>> the game. Other devices that the user has registered against his AoR
>>> may
>>>> handle audio for telephony and other voice applications but if Alice
>>>> invites Bob to play her favourite game it is not very interesting if
>>> Bob
>>>> accepts the gaming session on a device that can only accept the audio
>>>> stream. Also even when the terminal that has the game application
>>>> receives the INVITE the correct application in that terminal needs to
>>> be
>>>> identified to accept the session.
>>>
>>> I would offer that this is the clinching argument that the request is,
>>> in
>>> fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a
>> User
>>> Agent.
>>>
>>> Same argument for PoC versus a regular voice call: One is looking for
>> a
>>> PoC
>>> User Agent, NOT the PoC service at a given User Agent.
>>>
>>> Thus, the picture really looks like the following:
>>>
>>>                            +---------------------------------+
>>>                            |                                 |
>>>                            | +-------------+ +-------------+ |
>>>                            | |    UI       | |    UI       | |
>>>                            | +-------------+ +-------------+ |
>>>                            | +-------------+ +-------------+ |
>>>                            | |             | |             | |
>>>                            | |    UAS 1    | |    UAS 2    | |
>>>                            | |             | |             | |
>>>                            | +-------------+ +-------------+ |
>>>                            | +-----------------------------+ |
>>>                            | |                             | |
>>>                            | |             SIP             | |
>>>                            | |            Stack            | |
>>>                            | |                             | |
>>>                            | +-----------------------------+ |
>>>                            |                                 |
>>>                            +---------------------------------+
>>>
>>>                                      Physical Device
>>>
>>>
>>> Another argument for different User Agents is the whole "route based
>> on
>>> SDP"
>>> argument.  As the author of RFC 3458, I think I was first to make the
>>> assertion that knowing the media type did NOT mean you know the
>>> application
>>> to use.  I don't want your phone or proxy or B2BUA to guess that
>> because
>>> I
>>> have Chess Media, I want the Chess Application; I want to talk with
>> your
>>> Chess Application.
>>>
>>> Now, is the Chess Application a Preference, Requirement, or the thing
>>> that I
>>> am addressing?  The answer drives one towards using caller prefs, the
>>> Require header, or the Request-URI.
>>>
>>> I would offer that it is not a preference: I really want the chess
>>> application; I probably do not want to talk to you.
>>>
>>> That leaves us with the Require header or a Request-URI.  My leanings
>>> are
>>> not surprisingly towards the RFC 4240 approach.  Frankly, if you are
>>> building an IMS (as an example) that can handle 10,000,000
>> subscribers,
>>> you
>>> should be able to handle 100,000,000 registrations.  That order of
>>> magnitude
>>> is way easier than the 10,000 to 10,000,000 jump.  I don't buy the
>>> "registering each service is too much of a burden on the operator"
>>> argument,
>>> because it is not that much of a burden, and the network actually does
>>> the
>>> right thing.  Better yet, all of the corner cases work.  Best of all,
>>> the
>>> IETF doesn't have to build an IANA Registry and process a thousand PS
>>> documents to register urn:service:myco:widget.
>>>
>>> I suppose I could live with Require:, only if you are way too lazy to
>>> register each application instance.  It does semantically work,
>> because
>>> if I
>>> want to do PoC and you do not do PoC, I'll get a 4xx response, to
>> which
>>> I
>>> could try a regular call.
>>>
>>>
>> _______________________________________________________________________
>>> Notice:  This email message, together with any attachments, may
>> contain
>>> information  of  BEA Systems,  Inc.,  its subsidiaries  and
>> affiliated
>>> entities,  that may be confidential,  proprietary,  copyrighted
>> and/or
>>> legally privileged, and is intended solely for the use of the
>> individual
>>> or entity named in this message. If you are not the intended
>> recipient,
>>> and have received this message in error, please immediately return
>> this
>>> by email and then delete it.
>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP
>>> Use sip-implementors@cs.columbia.edu for questions on current sip
>>> Use sip@ietf.org for new developments of core SIP
>>>
>>>
>>>
>>> ---------------------------------------------------------------------
>>> This transmission (including any attachments) may contain confidential
>> information, privileged material (including material protected by the
>> solicitor-client or other applicable privileges), or constitute
>> non-public information. Any use of this information by anyone other than
>> the intended recipient is prohibited. If you have received this
>> transmission in error, please immediately reply to the sender and delete
>> this information from your system. Use, dissemination, distribution, or
>> reproduction of this transmission by unintended recipients is not
>> authorized and may be unlawful.
>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP
>>> Use sip-implementors@cs.columbia.edu for questions on current sip
>>> Use sip@ietf.org for new developments of core SIP
>>>
>>
>>
>>
>> ---------------------------------------------------------------------
>> This transmission (including any attachments) may contain confidential 
>> information, privileged material (including material protected by the 
>> solicitor-client or other applicable privileges), or constitute 
>> non-public information. Any use of this information by anyone other 
>> than the intended recipient is prohibited. If you have received this 
>> transmission in error, please immediately reply to the sender and 
>> delete this information from your system. Use, dissemination, 
>> distribution, or reproduction of this transmission by unintended 
>> recipients is not authorized and may be unlawful.
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP
>> Use sip-implementors@cs.columbia.edu for questions on current sip
>> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP





From sipping-bounces@ietf.org Fri Feb 23 10:16:57 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKcAL-0006T6-9E; Fri, 23 Feb 2007 10:16:57 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKcAJ-0006RL-S0
	for sipping@ietf.org; Fri, 23 Feb 2007 10:16:55 -0500
Received: from sj-iport-4.cisco.com ([171.68.10.86])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKcAI-0004W9-CB
	for sipping@ietf.org; Fri, 23 Feb 2007 10:16:55 -0500
Received: from sj-dkim-8.cisco.com ([171.68.10.93])
	by sj-iport-4.cisco.com with ESMTP; 23 Feb 2007 07:16:54 -0800
X-IronPort-AV: i="4.14,212,1170662400"; 
	d="scan'208"; a="41941254:sNHT61794783"
Received: from sj-core-3.cisco.com (sj-core-3.cisco.com [171.68.223.137])
	by sj-dkim-8.cisco.com (8.12.11/8.12.11) with ESMTP id l1NFGr3g002533; 
	Fri, 23 Feb 2007 07:16:53 -0800
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by sj-core-3.cisco.com (8.12.10/8.12.6) with ESMTP id l1NFG7iQ019156;
	Fri, 23 Feb 2007 07:16:53 -0800 (PST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 10:16:52 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 10:16:52 -0500
Message-ID: <45DF0564.1070402@cisco.com>
Date: Fri, 23 Feb 2007 10:16:52 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Christer Holmberg (JO/LMF)" <christer.holmberg@ericsson.com>
Subject: Re: [Sipping] Service Identification in SIP - Routing decissionsbased
	on Require/P-R
References: <7374777208BDC7449D5620EF94232567023BD791@esealmw113.eemea.ericsson.se>
In-Reply-To: <7374777208BDC7449D5620EF94232567023BD791@esealmw113.eemea.ericsson.se>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 23 Feb 2007 15:16:52.0454 (UTC)
	FILETIME=[A58F1060:01C7575D]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=4184; t=1172243813;
	x=1173107813; c=relaxed/simple; s=sjdkim8002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP=20-=2
	0Routing=20decissionsbased=0A=20on=20Require/P-R |Sender:=20;
	bh=ZDkfTa6yhtuFbcQAng5wvj13EhHqRm6JxPmqnsEp4o4=;
	b=M7a/XBRwqeEDBZvGOh82R2Z32LZjEH9Ms/tcoJMJiG48xaFLAR8kr2OolC8QUd2ngAShuhBa
	scrIyCNXpO9+06Hfsih0n8jTW6aUKreD5fA31/AXjmatFbxwgpXzX7Fz;
Authentication-Results: sj-dkim-8; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim8002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 36c793b20164cfe75332aa66ddb21196
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Christer Holmberg (JO/LMF) wrote:
> Hi,
> 
>> The difference here is that the Require header is definitive 
>> that the caller doesn't want to reach a UAS unless that UAS 
>> supports the stated options. Callerprefs is not so 
>> definitive. There are some uses of callerprefs that can be 
>> considered fairly definitive, but in general that is not the case.
>>
>> Jonathan proposed extending 3841 so that the use of Require 
>> generates an implicit callerpref. And that in turn would 
>> influence the selection of the registered contacts to try. I 
>> can't see a plausible way to infer any implicit callerpref 
>>from Accept-Contact headers.
>> In any case, that is all about selecting how to translate the 
>> R-URI. It is not about selecting intermediate proxies. As 
>> Dale has stated, you are free to use most anything to select 
>> an intermediate proxy. If you want to use Require, or 
>> Accept-Contact to do that you may do so. But while you are 
>> free to do so, it is less clear that it is useful or wise to 
>> do so. If the intent is to select a proxy that is consistent 
>> with the capabilities of the UAS that will eventually be 
>> accepted (as seems to be the case with PoC), then it only 
>> makes sense if the thing you are testing is definitive for 
>> the capabilities of the UAS. So again, Require may make sense 
>> for this, while Accept-Contact is not.
> 
> Don't you get the same effect by including explicit require in the
> Accept-Contact and adding "pref" to the Require Header? 

Not quite.

Require applies to the UAS, not proxies. Proxy-Require:pref would be 
closer. But Proxy-Require:pref only requires that the proxy *understand* 
callerprefs. It can still ignore those preferences based on its own 
policy. And it requires that *all* the proxies on the path understand 
callerprefs, not just the one that is responsible for the domain of the 
R-URI, where the preferences should be acted upon.

	Paul

> Regards,
> 
> Christer
> 
> 
> 
> 
>> 	Paul
>>
>> Christer Holmberg (JO/LMF) wrote:
>>> Hi,
>>>
>>>> -----Original Message-----
>>>> From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net]
>>>> Sent: 22. helmikuuta 2007 17:35
>>>> To: sipping@ietf.org
>>>> Subject: Re: [Sipping] Service Identification in SIP - Routing 
>>>> decissionsbased on Require/P-R
>>>>
>>>>    From: Henning Schulzrinne <hgs@cs.columbia.edu>
>>>>
>>>>    Proxies can choose next hops based on RFC 3840, which 
>> is another 
>>>> way
>>>>    of saying "routing".
>>>>
>>>> RFC 3261 allows a proxy unlimited latitude in choosing targets to 
>>>> which to forward a request, if the proxy is authoritative for the
>>>> request-URI:
>>>>
>>>> "[T]he element is responsible for the domain in the 
>> Request-URI, and 
>>>> the element MAY use whatever mechanism it desires to 
>> determine where 
>>>> to send the request."
>>>>
>>>> Within that framework, it seems to me quite reasonable for 
>> a service 
>>>> provider's proxy to examine Require or Proxy-Require headers to 
>>>> determine which application server would be best for 
>> handling a call, 
>>>> and routing the call to it.
>>> Why can't Accept-Contact be used for the same purpose?
>>>
>>> My issue is the statement that Require/Proxy-Require is the most 
>>> appropriate way of doing the routing decisson, and I am wondering 
>>> where that is specified.
>>>
>>> Regards,
>>>
>>> Christer
>>>
>>>
>>>
>>>
>>>
>>>> Dale
>>>>
>>>> _______________________________________________
>>>> Sipping mailing list  
>> https://www1.ietf.org/mailman/listinfo/sipping
>>>> This list is for NEW development of the application of SIP Use 
>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>> sip@ietf.org for new developments of core SIP
>>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP Use 
>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>> sip@ietf.org for new developments of core SIP
>>>
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 10:48:12 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKceS-0004yc-Q9; Fri, 23 Feb 2007 10:48:04 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKceR-0004yC-77
	for sipping@ietf.org; Fri, 23 Feb 2007 10:48:03 -0500
Received: from dsl001-129-069.dfw1.dsl.speakeasy.net ([72.1.129.69]
	helo=vicuna.estacado.net) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HKceP-0003bb-O2
	for sipping@ietf.org; Fri, 23 Feb 2007 10:48:03 -0500
Received: from [172.17.2.60] ([172.17.2.60]) (authenticated bits=0)
	by vicuna.estacado.net (8.13.8/8.13.8) with ESMTP id l1NFlvGb003065
	(version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO);
	Fri, 23 Feb 2007 09:47:58 -0600 (CST)
	(envelope-from bcampen@estacado.net)
In-Reply-To: <45DF0451.7020209@cisco.com>
References: <61968779B8AC4C4BAB421D4C12F008C009463E2E@XCH47YKF.rim.net>
	<FC318FB3-A497-4BED-BBC6-3B5B630FC589@estacado.net>
	<45DF0451.7020209@cisco.com>
Mime-Version: 1.0 (Apple Message framework v752.3)
Message-Id: <3B7302A3-1224-4B1B-8A98-96695C0A056C@estacado.net>
From: Byron Campen <bcampen@estacado.net>
Subject: Re: [Sipping] Service Identification Model
Date: Fri, 23 Feb 2007 09:47:51 -0600
To: Paul Kyzivat <pkyzivat@cisco.com>
X-Mailer: Apple Mail (2.752.3)
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 6cca30437e2d04f45110f2ff8dc1b1d5
Cc: Andrew Allen <aallen@rim.com>, sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1062419373=="
Errors-To: sipping-bounces@ietf.org


--===============1062419373==
Content-Type: multipart/signed; micalg=sha1; boundary=Apple-Mail-3-664949061;
	protocol="application/pkcs7-signature"


--Apple-Mail-3-664949061
Content-Transfer-Encoding: 7bit
Content-Type: text/plain;
	charset=US-ASCII;
	delsp=yes;
	format=flowed

> So I only see:
>
> - 1 for basic telephony, multimedia telephony, SMS, MMS, IM
>
> - *maybe* something for games
>
> - *maybe* something for push services
>
> 	Paul

	Seems like push services would be handled by SUBSCRIBE/NOTIFY, with  
appropriate event packages. Nothing fancy there. Certainly doesn't  
warrant a service identifier, Require header, or an AOR (at all).

Best regards,
Byron Campen


--Apple-Mail-3-664949061
Content-Transfer-Encoding: base64
Content-Type: application/pkcs7-signature;
	name=smime.p7s
Content-Disposition: attachment;
	filename=smime.p7s
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--Apple-Mail-3-664949061--


--===============1062419373==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1062419373==--




From sipping-bounces@ietf.org Fri Feb 23 10:48:12 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKcdT-0003k4-QU; Fri, 23 Feb 2007 10:47:03 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKcdS-0003bP-9K
	for sipping@ietf.org; Fri, 23 Feb 2007 10:47:02 -0500
Received: from hme1.july.broomfield1.level3.net ([209.245.18.8]
	helo=f4bb49-05.idc1.level3.com)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKcdQ-0003RW-6A
	for sipping@ietf.org; Fri, 23 Feb 2007 10:47:02 -0500
Received: from montag.eng.level3.com (montag.eng.l3.com [10.1.68.57])
	by f4bb49-05.idc1.level3.com (8.12.10/8.12.10) with ESMTP id
	l1NFkpYa005144; Fri, 23 Feb 2007 15:46:52 GMT
Subject: RE: [Sipping] RE: draft-malas-performance-metrics-06
From: Daryl Malas <daryl@level3.net>
To: "Marian Delkinov (LN/EAB)" <marian.delkinov@ericsson.com>
In-Reply-To: <97F2AD7AA4EAD948827639B06CF2745F90A36D@esealmw106.eemea.ericsson.se>
References: <97F2AD7AA4EAD948827639B06CF2745F90A36D@esealmw106.eemea.ericsson.se>
Content-Type: text/plain
Date: Fri, 23 Feb 2007 08:56:42 -0700
Message-Id: <1172246203.17141.66.camel@montag.eng.level3.com>
Mime-Version: 1.0
X-Mailer: Evolution 2.6.0 (2.6.0-1) 
Content-Transfer-Encoding: 7bit
X-Spam-Score: 1.2 (+)
X-Scan-Signature: 325b777e1a3a618c889460b612a65510
Cc: "Malas, Daryl" <Daryl.Malas@Level3.com>, sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Okay, I just wanted to make sure our definitions of "Established
Session" and "Session Establishment" were the same.  I assumed so, but
wanted to make sure you were not inferring a different definition.

Thanks for clarifying...---Daryl 

On Fri, 2007-02-23 at 10:48 +0100, Marian Delkinov (LN/EAB) wrote:
> Hi,
> 
> (Of course - no problem to copy to the mailing list. I don't mind and
> would be happy to see more opinions)
> 
> Regarding including diagrams, did you receive more support about it?
> 
> Regarding the metrics:
> 
> 1) Chapter 5.6. Additional Metrics for Consideration
> Session Establishment Efficiency Rate: Please, see my comments in a
> separate mail with subject "[Sipping] Session Establishment Efficiency
> Rate".
> 
> 2) Session Completion Rate (SCR)
> 
> My proposal was:
> 
> 	 # of Successfully Completed Sessions
> SCR% = ------------------------------------------ x 100
> 	 # of Successfully Established Sessions
> 
> Your comment:
> 
> [DM] In concept I understand what you are trying to get at with this
> change; however, this will require defining what an "Established
> Session" means.  This will become very tricky.  Instead, I think you can
> use the SCR as it is defined and then provide greater granularity by
> calculating it for calls above and below a time certain duration.  What
> do you think?
> 
> [MD] I thought that as "Established Session" we can consider one that
> got a 200 OK response as result of an INVITE. Perhaps the definition in
> chapter 2 of the draft can be used:
> 
> Session Establishment - Session establishment occurs when a 200 OK 
>    response from the UAS has been received, in response to a 
>    corresponding UAC's INVITE setup request, indicating the session 
>    setup request was successful. 
> 
> It's another question how we can get both counters from above in order
> to calculate the metric. Some vendors provide # of BYEs and # of OKs,
> but in those cases the counters include the repeated BYEs and OKs, which
> makes the metric wrong. Some other of course have counters for the
> sessions in fact. But I believe that if this gets into an RFC, most of
> the vendors will include the session counters as needed. 
> 
> Best regards!
> Mario.
> 
> 
>  
> 
> -----Original Message-----
> From: Malas, Daryl [mailto:Daryl.Malas@Level3.com] 
> Sent: Wednesday, 14 February, 2007 00:02
> To: Marian Delkinov (LN/EAB)
> Cc: sipping
> Subject: [Sipping] RE: draft-malas-performance-metrics-06
> 
> Marian...Comments inline...Also, I copied the Sipping mailing list,
> because it is helpful to allow others within the Sipping community to
> provide insight, thoughts into your comments as well.  I hope you do not
> mind.
> 
> Thanks...--Daryl
> 
> > -----Original Message-----
> > From: Marian Delkinov (LN/EAB) [mailto:marian.delkinov@ericsson.com]
> > Sent: Tuesday, February 06, 2007 12:45 AM
> > To: Malas, Daryl
> > Subject: draft-malas-performance-metrics-06
> > 
> > Hi,
> > 
> > I'm working on implementing some IMS performance metrics into our SLA 
> > solutions. After performing a lot of searching through the different 
> > standardization bodies, the most useful document I found was your
> draft.
> > 
> > 
> > Would like to comment on some of the metrics and perhaps on the
> generic
> > approach in the document.
> > 
> > Generic approach:
> > 
> > It would be good to give some background information about what are
> the
> > major reasons for SIP calls being lost. A picture similar to the one 
> > used in ITU E411 recommendation for the interpretation of the network 
> > performance data would really help. Please, see example in attachment.
> > 
> [DM] I could do this if the general mailing list agrees with the
> request.  I would like to get their opinion as well.  In either case, I
> will consider it for future revisions of the document.  I agree that
> diagrams, in general, usually help provide clarity around concepts.
>  
> > Metrics:
> > 
> > Since dealing with SLAs, I would like to put the things more into a 
> > Customer perspective. Therefore, I'm more focused on how the service 
> > looks for the customer. First of all what is the service accessibility
> 
> > and secondly what is the service performance. Perhaps you can help me 
> > find out the most suitable metrics that you discuss in your document.
> > 
> > 1) I try to define the SIP based telephony service effectiveness:
> > 
> > The probability that an end user can access the IMS (SIP based) 
> > Telephony Service.
> > 
> > In the PSTN/ISDN terms this is defined by ITU in Recommendation 411 as
> 
> > Network Effectiveness Rate:
> > 
> >             Seizures Delivered to the far end terminal NER% = 
> > ---------------------------------------------------- x 100
> > 		       Total Number of Seizures
> > 
> > Note that in this case the end user behaviour is excluded, so seizures
> 
> > that reach answer signal, or user busy, or ring no answer are counted
> as
> > successful.
> > 
> > When I look in your draft document, I don't clearly see similar
> metric.
> [DM]  If you look in the "Additional Metrics for Consideration" section
> of the draft, I think you will find a similar metric to what you are
> describing. Again, if you would like to see the metric included in the
> document, please state that on the mailing list.  Thanks...
> 
> > However if used some of your definitions, it could perhaps be defined
> as
> > Session Effectiveness Rate (or any other suitable definition, e.g.
> > Telephony Service Availability) as follows:
> > 
> >          Total # of Session Requests - # of Failed Session Completions
> 
> > SEFR% = --------------------------------------------------------------
> x
> > 100
> > 			Total # of Session Requests
> > 
> > Or, if in the above the "# of Failed Session Completion" includes the 
> > end user behaviour, then perhaps a better definition would be:
> > 
> >          Total # of Session Requests - # of ISA SEFR% = 
> > --------------------------------------------------------------
> x
> > 100
> > 			Total # of Session Requests
> > 
> > So, I would be really grateful if you share your opinion about that.
> > 
> > 2) I also try to define the SIP based telephony call completion ratio.
> > According to your document the Session Completion Rate (SCR) is the 
> > parameter that matches that. However I think of a more useful 
> > definition, that would really present the ability of the network to 
> > maintain successfully initiated session in the time. So, I would like
> to
> > hear what you think about the following definition proposal:
> > 
> > Session Completion Rate is the probability that a successfully 
> > established session is maintained for a predetermined time until it is
> 
> > released intentionally by any of the UAs.
> > 
> > According to your document it is defined as:
> >          # of Successfully Completed Sessions SCR% = 
> > ---------------------------------------
> >           Total # of Session Requests
> > 
> > But if we follow my proposal it would be:
> > 
> >          # of Successfully Completed Sessions SCR% = 
> > ------------------------------------------ x 100
> >          # of Successfully Established Sessions
> > 
> > 
> > 
> [DM] In concept I understand what you are trying to get at with this
> change; however, this will require defining what an "Established
> Session" means.  This will become very tricky.  Instead, I think you can
> use the SCR as it is defined and then provide greater granularity by
> calculating it for calls above and below a time certain duration.  What
> do you think?
> 
> This is actually something we can perhaps easily measure, however it
> > would be good if we can show this parameter in dependence with the 
> > Session Duration and that I haven't taken into account. In many cases,
> 
> > when jitter and packet loss increase, the sessions are terminated 
> > unintentionally. And the probability this to happen raises in time.
> For
> > example, for short sessions the SCR may look good, but for long
> sessions
> > it deteriorates.
> > 
> > So, I wonder how to link this with the session duration time in order
> to
> > have a single parameter that really represents the network
> performance.
> > Will be happy to exchange ideas.
> > 
> > Best regards!
> > Mario.
> > 
> > Marian Delkinov
> > Senior Consultant
> > 
> > Network Management and Service Assurance
> > 
> > ERICSSON AB
> > Lindholmspiren 11
> > SE-417 56 Gothenburg
> > Sweden
> > 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 10:50:11 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKcgV-0000B8-Ci; Fri, 23 Feb 2007 10:50:11 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKcgQ-00007M-Tu
	for sipping@ietf.org; Fri, 23 Feb 2007 10:50:06 -0500
Received: from hme1.july.broomfield1.level3.net ([209.245.18.8]
	helo=f4bb49-05.idc1.level3.com)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKcgN-0003vl-Jn
	for sipping@ietf.org; Fri, 23 Feb 2007 10:50:06 -0500
Received: from montag.eng.level3.com (montag.eng.l3.com [10.1.68.57])
	by f4bb49-05.idc1.level3.com (8.12.10/8.12.10) with ESMTP id
	l1NFnuYa007215; Fri, 23 Feb 2007 15:49:57 GMT
Subject: RE: [Sipping] Session Establishment Efficiency Rate
From: Daryl Malas <daryl@level3.net>
To: "Marian Delkinov (LN/EAB)" <marian.delkinov@ericsson.com>
In-Reply-To: <97F2AD7AA4EAD948827639B06CF2745F90A48A@esealmw106.eemea.ericsson.se>
References: <97F2AD7AA4EAD948827639B06CF2745F90A48A@esealmw106.eemea.ericsson.se>
Content-Type: text/plain
Date: Fri, 23 Feb 2007 08:59:47 -0700
Message-Id: <1172246388.17141.69.camel@montag.eng.level3.com>
Mime-Version: 1.0
X-Mailer: Evolution 2.6.0 (2.6.0-1) 
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 5a9a1bd6c2d06a21d748b7d0070ddcb8
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I don't think 3xx codes should be included, since they would not be
considered a "Final" release cause code.  --Daryl

On Fri, 2007-02-23 at 12:22 +0100, Marian Delkinov (LN/EAB) wrote:
> Hi,
> 
> If the Network Efficiency Rate is designed to express the ability of the
> network to deliver calls (in SIP terms one should perhaps say "establish
> sessions"), I think that it might be reasonable to include few more
> cases (cause codes) in the numerator: e.g. all redirection 3xx codes,
> all 4xx except 408 codes and all 6xx codes. (If you agree that those are
> the ones that exclude network failures, congestion and any other network
> malfunction).
> 
> Actually it sounds fair to me if it's the opposite of the Ineffective
> Session Attempts (ISA):
> 
> 	   Total # of Invite Requests - # of ISA
> SEER% = -------------------------------------------------------------- x
> 100
> 	   Total # of Invite Requests
> 
> What do you think?
> 
> Best regards!
> Mario.
> 
> -----Original Message-----
> From: Daryl Malas [mailto:daryl@level3.net] 
> Sent: Friday, 23 February, 2007 01:14
> To: sipping@ietf.org
> Subject: [Sipping] Session Establishment Efficiency Rate
> 
> A couple more people have expressed to me that this adaptation of the
> ITU-T E.411 Network Efficiency Rate metric is very valuable.  The
> snapshot of the metric from the draft is:
> 
> "This metric is similar to Session Establishment Rate, but it is
> intended to exclude the effects of customer behavior and user agent
> behavior.  Essentially, the attempts resulting with a final cause code
> of 480, 486, and 600 would be added to the SER numerator."
> 
> Does anyone else want to see this metric in the draft?  I personally
> think it is an interesting metric.
> 
> Thanks...
> 
> --Daryl
> 
> 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use
> sip-implementors@cs.columbia.edu for questions on current sip Use
> sip@ietf.org for new developments of core SIP


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 13:23:57 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKf2o-0007Wk-Or; Fri, 23 Feb 2007 13:21:22 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKf2m-0007VA-BI
	for sipping@ietf.org; Fri, 23 Feb 2007 13:21:20 -0500
Received: from usremg02.bea.com ([66.248.192.22])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKf2l-00041d-It
	for sipping@ietf.org; Fri, 23 Feb 2007 13:21:20 -0500
Received: from usremr01.bea.com (usremr01.bea.com [10.160.29.91])
	by usremg02.bea.com (Switch-3.2.2/Switch-3.2.2) with ESMTP id
	l1NILBO4021296
	(version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=FAIL);
	Fri, 23 Feb 2007 10:21:11 -0800
Received: from repbex01.amer.bea.com (repbex01.bea.com [10.160.26.98])
	by usremr01.bea.com (Switch-3.2.2/Switch-3.2.2) with ESMTP id
	l1NILAGO019663; Fri, 23 Feb 2007 10:21:10 -0800
Received: from 172.24.29.139 ([172.24.29.139]) by repbex01.amer.bea.com
	([10.160.26.98]) with Microsoft Exchange Server HTTP-DAV ; 
	Fri, 23 Feb 2007 18:21:09 +0000
User-Agent: Microsoft-Entourage/11.3.3.061214
Date: Fri, 23 Feb 2007 11:32:03 -0500
Subject: Re: [Sipping] Service Identification Model 
From: Eric Burger <eburger@bea.com>
To: "Stephen Terrill (MI/EEM)" <stephen.terrill@ericsson.com>
Message-ID: <C2048133.22D6%eburger@bea.com>
Thread-Topic: [Sipping] Service Identification Model 
Thread-Index: AcdV5i0Ma5Tu/sHZEdueygAWy4mm/wAV9tzAAAZvQkAABCBYgAABwpngAD41MSE=
In-Reply-To: <989763567E51AB48BE015DF7BB960CAD028F0B91@esealmw111>
Mime-version: 1.0
Content-type: text/plain;
	charset="US-ASCII"
Content-transfer-encoding: 7bit
X-PMX-Version: 4.7.1.128075, Antispam-Engine: 2.5.0.283055,
	Antispam-Data: 2007.2.21.161433
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 7f3fa64b9851a63d7f3174ef64114da7
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

What is the difference between:

sip:eburger+sms@example.com

sip:eburger@example.com   <message-context: pager> (See RFC 3458)

sip:eburger@example.com   <Require: sms>

???


On 2/22/07 2:53 AM, "Stephen Terrill (MI/EEM)"
<stephen.terrill@ericsson.com> wrote:

> Hi Steve,
> 
> I conceed your point with a FAX on a buisiness card as it is often
> another (shared) device, however perhaps it may not have been nice if
> that practice was carried over to SMS; MMS etc ...
> 
> BR,
> 
> ..//Steve
> 
> -----Original Message-----
> From: Steve Langstaff [mailto:steve.langstaff@citel.com]
> Sent: Thursday, February 22, 2007 11:05 AM
> To: Stephen Terrill (MI/EEM); sipping@ietf.org
> Subject: RE: [Sipping] Service Identification Model
> 
> If my business card has a FAX number on it, you can be pretty sure that
> you can FAX me on it.
> 
> If all you have is a generic contact address, you don't know whether you
> can FAX me until you try, and possibly fail, to contact me.
> 
> So it could be argued, from the "business card" point of view, that a
> single AoR is not "preferable", just "different".
> 
>> -----Original Message-----
>> From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
>> Sent: 22 February 2007 08:22
>> To: sipping@ietf.org
>> Subject: RE: [Sipping] Service Identification Model
>> 
>> Hi,
>> 
>> Isn't relying on a different AoR record for each service a rather
>> backwards approach?  Today, there have been upto three numbers on the
>> phones representing different services (Telephony; fax and Data) -
>> which means business cards which means different numbers for the
>> different ways I want to contact a person.  In myunderstanding, that
>> would apply here as a different AoR record for Telephony; PoC and
>> Messaging.
>> I think it is much more preferable to have one AoR; and then to be
>> able to indicate the requested service.
>> 
>> Steve.
>> 
>> -----Original Message-----
>> From: Andrew Allen [mailto:aallen@rim.com]
>> Sent: Thursday, February 22, 2007 6:49 AM
>> To: Eric Burger; sipping@ietf.org
>> Cc: Andrew Allen
>> Subject: RE: [Sipping] Service Identification Model
>> 
>> 
>> Hi Eric,
>> 
>> It seems your prefer the approach of using URIs to identify services
>> which is part of Jonathan's draft as is the use of Require for those
>> services which simply do not interoperate unless the UAS understands
>> the service.
>> 
>> Where I am concerned is that you seem to be suggesting that the UA
>> separately register an AoR for each service it supports using
>> REGISTER.
>> Certainly for cellular that wouldn't be acceptable as 10 or more
>> services could easily be supported by a single device. Registrations
>> create a significant load in the IMS network because its not just a
>> SIP REGISTER but also requires a lot of interactions with other
>> network elements such as the HSS and Notifications for reg-events etc.
> 
>> Also having an order of magnitude increase in multiple registrations
>> is going to be a big issue for the avalanche restart problem for
>> example when a cell site goes down and comes back into service.
>> 
>> I think Jonathan's approach gives us the possibility for the registrar
> 
>> to create the Service URIs based on the feature parameters included in
> 
>> the Contact of the REGISTER.
>> 
>> I think the argument over whether the model is one UA with multiple
>> services or multiple UAs each representing a separate service is
>> somewhat academic. As I understand a UA can register multiple AoRs and
> 
>> can have other AoRs configured or otherwise provided by the service
>> provider.
>> 
>> Andrew
>> 
>> -----Original Message-----
>> From: Eric Burger [mailto:eburger@bea.com]
>> Sent: Wednesday, February 21, 2007 1:29 PM
>> To: sipping@ietf.org
>> Subject: [Sipping] Service Identification Model
>> 
>> Folks have been pestering me for a while to look at the Service
>> Identification issue.  I finally have gotten around to it.
>> 
>> 
>> I guess I have a fundamental problem with the principal assertion of
>> the service-identification draft.  Namely, the assertion that there is
> 
>> a single SIP Layer, separate and distinct from two services.  Here is
>> Figure 2 from the draft:
>> 
>>                            +---------------------------------+
>>                            |                                 |
>>                            | +-------------+ +-------------+ |
>>                            | |      UI     | |      UI     | |
>>                            | +-------------+ +-------------+ |
>>                            | +-------------+ +-------------+ |
>>                            | |             | |             | |
>>                            | |  Service 1  | |  Service 2  | |
>>                            | |             | |             | |
>>                            | +-------------+ +-------------+ |
>>                            | +-----------------------------+ |
>>                            | |                             | |
>>                            | |             SIP             | |
>>                            | |            Layer            | |
>>                            | |                             | |
>>                            | +-----------------------------+ |
>>                            |                                 |
>>                            +---------------------------------+
>> 
>>                                      Physical Device
>> 
>> I would offer this drawing confuses a common SIP Stack with a common
>> User Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use
>> case:
>> 
>>> ... In addition users may have
>>> multiple devices only one of which has the application that
>> understands
>>> the game. Other devices that the user has registered against his AoR
>> may
>>> handle audio for telephony and other voice applications but
>> if Alice
>>> invites Bob to play her favourite game it is not very interesting if
>> Bob
>>> accepts the gaming session on a device that can only accept
>> the audio
>>> stream. Also even when the terminal that has the game application
>>> receives the INVITE the correct application in that
>> terminal needs to
>> be
>>> identified to accept the session.
>> 
>> I would offer that this is the clinching argument that the request is,
> 
>> in fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a
>> User Agent.
>> 
>> Same argument for PoC versus a regular voice call: One is looking for
>> a PoC User Agent, NOT the PoC service at a given User Agent.
>> 
>> Thus, the picture really looks like the following:
>> 
>>                            +---------------------------------+
>>                            |                                 |
>>                            | +-------------+ +-------------+ |
>>                            | |    UI       | |    UI       | |
>>                            | +-------------+ +-------------+ |
>>                            | +-------------+ +-------------+ |
>>                            | |             | |             | |
>>                            | |    UAS 1    | |    UAS 2    | |
>>                            | |             | |             | |
>>                            | +-------------+ +-------------+ |
>>                            | +-----------------------------+ |
>>                            | |                             | |
>>                            | |             SIP             | |
>>                            | |            Stack            | |
>>                            | |                             | |
>>                            | +-----------------------------+ |
>>                            |                                 |
>>                            +---------------------------------+
>> 
>>                                      Physical Device
>> 
>> 
>> Another argument for different User Agents is the whole "route based
>> on SDP"
>> argument.  As the author of RFC 3458, I think I was first to make the
>> assertion that knowing the media type did NOT mean you know the
>> application to use.  I don't want your phone or proxy or B2BUA to
>> guess that because I have Chess Media, I want the Chess Application; I
> 
>> want to talk with your Chess Application.
>> 
>> Now, is the Chess Application a Preference, Requirement, or the thing
>> that I am addressing?  The answer drives one towards using caller
>> prefs, the Require header, or the Request-URI.
>> 
>> I would offer that it is not a preference: I really want the chess
>> application; I probably do not want to talk to you.
>> 
>> That leaves us with the Require header or a Request-URI.  My leanings
>> are not surprisingly towards the RFC 4240 approach.  Frankly, if you
>> are building an IMS (as an example) that can handle 10,000,000
>> subscribers, you should be able to handle 100,000,000 registrations.
>> That order of magnitude is way easier than the 10,000 to 10,000,000
>> jump.
>> I don't buy
>> the "registering each service is too much of a burden on the operator"
>> argument,
>> because it is not that much of a burden, and the network actually does
> 
>> the right thing.  Better yet, all of the corner cases work.  Best of
>> all, the IETF doesn't have to build an IANA Registry and process a
>> thousand PS documents to register urn:service:myco:widget.
>> 
>> I suppose I could live with Require:, only if you are way too lazy to
>> register each application instance.  It does semantically work,
>> because if I want to do PoC and you do not do PoC, I'll get a 4xx
>> response, to which I could try a regular call.
>> 
>> ______________________________________________________________
>> _________
>> Notice:  This email message, together with any attachments, may
>> contain information  of  BEA Systems,  Inc.,  its subsidiaries  and
>> affiliated entities,  that may be confidential,  proprietary,
>> copyrighted  and/or legally privileged, and is intended solely for the
> 
>> use of the individual or entity named in this message. If you are not
>> the intended recipient, and have received this message in error,
>> please immediately return this by email and then delete it.
>> 
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP Use
>> sip-implementors@cs.columbia.edu for questions on current sip Use
>> sip@ietf.org for new developments of core SIP
>> 
>> 
>> 
>> ---------------------------------------------------------------------
>> This transmission (including any attachments) may contain confidential
> 
>> information, privileged material (including material protected by the
>> solicitor-client or other applicable privileges), or constitute
>> non-public information. Any use of this information by anyone other
>> than the intended recipient is prohibited. If you have received this
>> transmission in error, please immediately reply to the sender and
>> delete this information from your system. Use, dissemination,
>> distribution, or reproduction of this transmission by unintended
>> recipients is not authorized and may be unlawful.
>> 
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP Use
>> sip-implementors@cs.columbia.edu for questions on current sip Use
>> sip@ietf.org for new developments of core SIP
>> 
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP Use
>> sip-implementors@cs.columbia.edu for questions on current sip Use
>> sip@ietf.org for new developments of core SIP
>> 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP

_______________________________________________________________________
Notice:  This email message, together with any attachments, may contain
information  of  BEA Systems,  Inc.,  its subsidiaries  and  affiliated
entities,  that may be confidential,  proprietary,  copyrighted  and/or
legally privileged, and is intended solely for the use of the individual
or entity named in this message. If you are not the intended recipient,
and have received this message in error, please immediately return this
by email and then delete it.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 18:41:12 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKjzO-0001Ew-KX; Fri, 23 Feb 2007 18:38:10 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKjzM-0001EY-KI
	for sipping@ietf.org; Fri, 23 Feb 2007 18:38:08 -0500
Received: from jes-fe1.zx.nl ([194.187.76.132])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKjzH-0003UR-AD
	for sipping@ietf.org; Fri, 23 Feb 2007 18:38:08 -0500
Received: from Inbox ([84.241.198.249])
	by jes-fe1.zx.nl (Sun Java System Messaging Server 6.2-7.05 (built Sep
	5 2006)) with ESMTP id <0JDX0036NWAGN610@jes-fe1.zx.nl> for
	sipping@ietf.org; Sat, 24 Feb 2007 00:37:41 +0100 (CET)
Date: Sat, 24 Feb 2007 00:40:08 +0100
From: Jeroen van Bemmel <jbemmel@zonnet.nl>
Subject: RE: [Sipping] Service Identification in SIP - Routing decissionsbased
	on Require/P-R
To: Paul Kyzivat <pkyzivat@cisco.com>, Paul Kyzivat <pkyzivat@cisco.com>
Message-id: <0JDX0036OWAIN610@jes-fe1.zx.nl>
MIME-version: 1.0
Content-type: text/plain; charset=iso-8859-1
Content-transfer-encoding: quoted-printable
Importance: normal
X-Priority: 3
X-Spam-Score: 1.2 (+)
X-Scan-Signature: 5a9a1bd6c2d06a21d748b7d0070ddcb8
Cc: sipping@ietf.org,
	"Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Paul,
The model realized in IMS is perhaps more accurately described as a UAC sel=
ecting a UAS with a specific intent (e.g. PoC), that UAS in turn acting as =
a UAC to select an appropriate UAS (still for PoC, on behalf of the initiat=
ing UAC) and finally that second UAS acting again as UAC to select a target=
 UAS (device) which supports PoC

If you have a device and I have one, we can try to setup a session and nego=
tiate the least common denominator as features. The PoC use case is more ab=
out selecting an appropriate next hop from a heterogeneous set of many (tim=
es 3)

Considering PoC as a capability / feature is confusing, I'll argue that fro=
m a user perspective it is more "I want to PoC with X" as opposed to "I wan=
t to communicate with X and I'll use whatever matching capabilities our dev=
ices support". It is about user intent, which appears difficult to capture =
in a generic INVITE. Perhaps a new method "PoC"?

Regards,
Jeroen

-----Oorspronkelijk bericht -----
Van: "Paul Kyzivat" <pkyzivat@cisco.com>
Aan: "Paul Kyzivat" <pkyzivat@cisco.com>
CC: sipping@ietf.org; "Christer Holmberg (JO/LMF)" <christer.holmberg@erics=
son.com>
Verzonden: 23-2-07 15:45
Onderwerp: Re: [Sipping] Service Identification in SIP - Routing decissions=
based	on Require/P-R

Some further thoughts on this subject:

Typically, some of the capabilities used in a call are *negotiated*=20
between the caller and the callee. As a result, those that will be used=20
are not known for certain at the time the request is sent. And the=20
results of that negotiation may differ depending on which of various=20
candidate targets is selected.

Apparently one key intended use of the service id is to bring in one or=20
more intermediate nodes that are specific to the capabilities that will=20
be used in the call. These intermediate nodes must be selected *before*=20
the UAS is selected and the negotiation is done. For that to work, one=20
of two things must happen:

- the proxy that selects those intermediate nodes
   must be prescient about what capabilities will be negotiated

- or the selection of UAS and the negotiation of capabilities to be
   used must be somehow constrained to be consistent with the
   capabilities supported by the selected intermediate nodes.

If we agree that prescience is infeasible, I guess we must assume that=20
the selection and negotiation will be constrained. We of course have a=20
mechanism for that: it is the Require header. There are also other=20
mechanisms that provide constraints of some sort. For instance the SDP=20
offer constrains what is permitted in the answer. (It does *not* however=20


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Fri Feb 23 21:19:01 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKmUJ-0004Vc-Ca; Fri, 23 Feb 2007 21:18:15 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKmUI-0004VV-9U
	for sipping@ietf.org; Fri, 23 Feb 2007 21:18:14 -0500
Received: from sj-iport-5.cisco.com ([171.68.10.87])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKmUF-0004Mp-MG
	for sipping@ietf.org; Fri, 23 Feb 2007 21:18:14 -0500
Received: from sj-dkim-6.cisco.com ([171.68.10.81])
	by sj-iport-5.cisco.com with ESMTP; 23 Feb 2007 18:18:12 -0800
X-IronPort-AV: i="4.14,213,1170662400"; 
	d="scan'208"; a="392880038:sNHT54493956"
Received: from sj-core-3.cisco.com (sj-core-3.cisco.com [171.68.223.137])
	by sj-dkim-6.cisco.com (8.12.11/8.12.11) with ESMTP id l1O2IBG8023904; 
	Fri, 23 Feb 2007 18:18:11 -0800
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by sj-core-3.cisco.com (8.12.10/8.12.6) with ESMTP id l1O2I8hq006434;
	Fri, 23 Feb 2007 18:18:10 -0800 (PST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 21:18:10 -0500
Received: from [10.86.242.125] ([10.86.242.125]) by xfe-rtp-202.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 23 Feb 2007 21:18:09 -0500
Message-ID: <45DFA060.50600@cisco.com>
Date: Fri, 23 Feb 2007 21:18:08 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Jeroen van Bemmel <jbemmel@zonnet.nl>
Subject: Re: [Sipping] Service Identification in SIP - Routing decissionsbased
	on Require/P-R
References: <0JDX0036OWAIN610@jes-fe1.zx.nl>
In-Reply-To: <0JDX0036OWAIN610@jes-fe1.zx.nl>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 24 Feb 2007 02:18:09.0689 (UTC)
	FILETIME=[0708A890:01C757BA]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=5722; t=1172283491;
	x=1173147491; c=relaxed/simple; s=sjdkim6002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP=20-=2
	0Routing=20decissionsbased=0A=20on=20Require/P-R |Sender:=20;
	bh=UCokJ2gADYpR89m2RXVYcKO9wlDp4EpP92h0Zw7c1Jw=;
	b=BzVVhmBeyXGxq2v6koDfo2a3o8vONl44fIhiOjYO8FZVfLHB75aSyRLNSp21yUFKjF2kwNG9
	ePfpBoIj5+SMpvHn9SkGvOxFMbBTra70owtsqeAUUfSj9oGtgKcxQ/C3;
Authentication-Results: sj-dkim-6; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim6002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 1676547e4f33b5e63227e9c02bd359e3
Cc: sipping@ietf.org,
	"Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Jeroen van Bemmel wrote:
> Paul,
> The model realized in IMS is perhaps more accurately described
> as a UAC selecting a UAS with a specific intent (e.g. PoC),
> that UAS in turn acting as a UAC to select an appropriate UAS
> (still for PoC, on behalf of the initiating UAC) and finally
> that second UAS acting again as UAC to select a target UAS (device)
> which supports PoC

I disagree. The originating UAC inserts an address identifying the 
terminating AOR. So the originating UAC is not selecting the originating 
PoC server.

If one of the proposals in jonathan's draft (inserting a Route header 
with a URN for a PoC server) then I would agree that the originating UAC 
as selecting both the originating PoC server and the target AOR.

I  also don't see that the originating PoC server is doing anything 
about selecting either a terminating PoC server or the final terminating 
UAS. It is just leaving criteria as they were set by the originating UAC.

Its unclear to me whether the originating UAC and/or the terminating PoC 
server is the primary influence over which terminating UAS is selected, 
and the flip side of that is also unclear - whether the originating UAC 
is selecting the terminating PoC server. To determine that, we would 
need to examine some non-trivial use cases.

Consider a case where there are two different UAs both registered to the 
same AOR, as well as another UA that doesn't support PoC. Then some 
other UAC calls this AoR. It could potentially select among the 
registered UAs using callerprefs. (IMS even specifies that the 
terminating S-CSCF shoujld apply callerprefs.) If a terminating PoC 
server is brought in, AFAIK it leaves the callerprefs alone, so that the 
terminating UAS is still selected via criteria provided by the 
originating UAC.

I do think it would be good if things were arranged so that it *did* 
work as you suggest. But I think that will require some different 
mechanisms.

> If you have a device and I have one, we can try to setup a
> session and negotiate the least common denominator as features.
> The PoC use case is more about selecting an appropriate next hop
> from a heterogeneous set of many (times 3)

If that is the purpose of the service id, then it has not been very well 
explained. This call model is very strange - pretty far from anything 
sip was designed for. If this is the goal, its going to take more work 
to figure out if/how to accomplish it.

> Considering PoC as a capability / feature is confusing, I'll argue
> that from a user perspective it is more "I want to PoC with X" as
> opposed to "I want to communicate with X and I'll use whatever
> matching capabilities our devices support". It is about user intent,
> which appears difficult to capture in a generic INVITE. Perhaps a
> new method "PoC"?

I think this could be achieved as follows:

- each UA that wants to receive PoC calls establishes a relationship
   with a PoC server that will be supporting it. (TBD how this is
   done.) The PoC server registers with the AoR on behalf of the UA.
   If there are any capabilities that the UA wants advertised, then
   the PoC server provides them in its registration. PoC would be
   advertised as one of the capabilities.
   (If there are multiple UAs that do PoC, then each will have an
   association with a PoC server and there will be a registration
   for each.)

- incoming calls to the AoR simply consider the registered PoC server
   as one of the registered contacts. Selection among those contacts
   is normal. There could be forking to all. Or callerprefs could be
   used to select a contact that supports PoC, or non-PoC.

- for originating PoC calls, the UA could include a Route header that
   sends the request first to its associated PoC server. And it could
   include callerprefs to select PoC at the termination.

	Paul


> Regards,
> Jeroen
> 
> -----Oorspronkelijk bericht -----
> Van: "Paul Kyzivat" <pkyzivat@cisco.com>
> Aan: "Paul Kyzivat" <pkyzivat@cisco.com>
> CC: sipping@ietf.org; "Christer Holmberg (JO/LMF)" <christer.holmberg@ericsson.com>
> Verzonden: 23-2-07 15:45
> Onderwerp: Re: [Sipping] Service Identification in SIP - Routing decissionsbased	on Require/P-R
> 
> Some further thoughts on this subject:
> 
> Typically, some of the capabilities used in a call are *negotiated* 
> between the caller and the callee. As a result, those that will be used 
> are not known for certain at the time the request is sent. And the 
> results of that negotiation may differ depending on which of various 
> candidate targets is selected.
> 
> Apparently one key intended use of the service id is to bring in one or 
> more intermediate nodes that are specific to the capabilities that will 
> be used in the call. These intermediate nodes must be selected *before* 
> the UAS is selected and the negotiation is done. For that to work, one 
> of two things must happen:
> 
> - the proxy that selects those intermediate nodes
>    must be prescient about what capabilities will be negotiated
> 
> - or the selection of UAS and the negotiation of capabilities to be
>    used must be somehow constrained to be consistent with the
>    capabilities supported by the selected intermediate nodes.
> 
> If we agree that prescience is infeasible, I guess we must assume that 
> the selection and negotiation will be constrained. We of course have a 
> mechanism for that: it is the Require header. There are also other 
> mechanisms that provide constraints of some sort. For instance the SDP 
> offer constrains what is permitted in the answer. (It does *not* however 
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sat Feb 24 08:13:42 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HKweQ-0001qu-AJ; Sat, 24 Feb 2007 08:09:22 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HKweO-0001qo-Ns
	for sipping@ietf.org; Sat, 24 Feb 2007 08:09:20 -0500
Received: from rwcrmhc12.comcast.net ([216.148.227.152])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HKweM-00018F-9T
	for sipping@ietf.org; Sat, 24 Feb 2007 08:09:20 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (rwcrmhc12) with ESMTP
	id <20070224130917m1200j7cn2e>; Sat, 24 Feb 2007 13:09:17 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1OD9G6g028850
	for <sipping@ietf.org>; Sat, 24 Feb 2007 08:09:16 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1OD9GxI028846;
	Sat, 24 Feb 2007 08:09:16 -0500
Date: Sat, 24 Feb 2007 08:09:16 -0500
Message-Id: <200702241309.l1OD9GxI028846@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <7374777208BDC7449D5620EF94232567023932CA@esealmw113.eemea.ericsson.se>
	(christer.holmberg@ericsson.com)
Subject: Re: [Sipping] Service Identification in SIP - Routing decissionsbased
	on Require/P-R
References: <7374777208BDC7449D5620EF94232567023932CA@esealmw113.eemea.ericsson.se>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: c1c65599517f9ac32519d043c37c5336
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>

   > Within that framework, it seems to me quite reasonable for a 
   > service provider's proxy to examine Require or Proxy-Require 
   > headers to determine which application server would be best 
   > for handling a call, and routing the call to it.

   Why can't Accept-Contact be used for the same purpose?

   My issue is the statement that Require/Proxy-Require is the most
   appropriate way of doing the routing decisson, and I am wondering
   where that is specified.

Well, it isn't going to be *specified* anywhere, it can only arise as
an implication based on examining the implementation mechanisms we
have at our disposal.

In regard to this sort of routing, my impression from what others have
said is that Accept-Contact (or "callerprefs") is advisory, that a
proxy is allowed to disregard its statements based on local policy,
whereas Require is mandatory (though it is in principle implemented
only at the destination UA, and ignored by proxies).

So from the point of view of the semantics alone, Require is closer to
what we want.

Of course, Proxy-Require is Not A Good Idea, because it requires that
all intermediate proxies believe that they understand the extension,
and (assuming that the service is implemented in a way that does not
require special service in the truly WAN link), that is not what we
want.

And of course, a proxy at the destination domain is free to examine
the Require header and omit forwarding the request to UAs that it
*knows* do not support the requested extension, as the UAC can not
discover that such a shortcut has been taken.

But the discussion is rather academic -- In an IMS-like architecture,
the proxies in the destination domain, the application servers, and
the UA aren't really true SIP agents in the way that proxies and UAs
are in a more decentralized system.  Rather, all those agents
cooperate to produce the appearance of a SIP UA to any external
observer.  So all those agents have a great deal of latitude in
exactly how they use SIP, as long as the overall effect is
standard-compliant.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sun Feb 25 09:21:33 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLK9m-0001Bq-6D; Sun, 25 Feb 2007 09:15:18 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLK9l-0001Bh-6N
	for sipping@ietf.org; Sun, 25 Feb 2007 09:15:17 -0500
Received: from smtp1.versatel.nl ([62.58.50.88])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLK9j-0004uf-D3
	for sipping@ietf.org; Sun, 25 Feb 2007 09:15:17 -0500
Received: (qmail 10376 invoked by uid 0); 25 Feb 2007 14:16:23 -0000
Received: from ip198-11-212-87.adsl2.versatel.nl (HELO BEMBUSTER)
	([87.212.11.198]) (envelope-sender <jbemmel@zonnet.nl>)
	by smtp1.versatel.nl (qmail-ldap-1.03) with SMTP
	for < >; 25 Feb 2007 14:16:23 -0000
Message-ID: <005c01c758e7$594e45a0$0601a8c0@BEMBUSTER>
From: "Jeroen van Bemmel" <jbemmel@zonnet.nl>
To: "Paul Kyzivat" <pkyzivat@cisco.com>
References: <0JDX0036OWAIN610@jes-fe1.zx.nl> <45DFA060.50600@cisco.com>
Subject: Re: [Sipping] Service Identification in SIP - Routing decissionsbased
	on Require/P-R
Date: Sun, 25 Feb 2007 15:15:05 +0100
MIME-Version: 1.0
Content-Type: text/plain; format=flowed; charset="iso-8859-1";
	reply-type=response
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 6.00.2900.3028
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.3028
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 43317e64100dd4d87214c51822b582d1
Cc: sipping@ietf.org,
	"Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Paul,

My points are that:
a) this is not a proxy selection problem, but a UAS selection problem (and 
mechanisms for proxy selection are lacking, see outbound discussions)
b) this is not a single selection problem, but three distinct cases. Most 
likely a single mechanism will not cover all of these

I understand the PoC issue as follows: when compared to a regular voice 
call, signaling for PoC uses
- the same request method (INVITE)
- the same mime type (SDP)
- the same media (audio)

As a result, there is currently no way to configure the initial filter 
criteria such that PoC signaling goes to one (PoC) UAS, and regular voice 
calls goes to another (voice) UAS. (the separation is due to deployed 
legacy, we can keep argueing this model but this is the starting point for 
IMS systems).

The 3GPP requirement should state "we need a way to distinguish between 
voice calls and PoC calls". Instead, they went one step further and state 
"we need a service identifier".

There is an additional, implicit 3GPP requirement that the chosen solution 
should be consistent with the current IMS architecture. That means that a 
Route header based solution (e.g. service URN) is not acceptable: there is 
already a mechanism for next hop selection (IFC), adding pre-loaded route 
headers would introduce a second mechanism. A similar argument applies to 
the mechanisms you are proposing below (registration with PoC server): this 
would be too different from the current mechanisms / architecture

And then there is a third requirement, apparent from the proposed solution: 
the solution should be generic enough such that it can also be used in other 
cases besides PoC. Possibly also the lengthy process of developing a 
standards track specification for each case is seen as a hurdle.

>From IETF side the requirement is also that the solution should be 
consistent with the current SIP architecture (as defined in the various 
RFCs). This means that if there is an existing, standardized mechanism that 
can solve this problem, that would be preferrable over a new mechanism / 
header (service identifier).

A second argument from IETF side, is that (a) the IETF does not standardize 
services and (b) people view this model of separating things into arbitrary 
services as "evil". Perhaps if we consider PoC as a voice call that requires 
"floor-control", similar to a voice call that would use "session-timer" or 
other features, we can mitigate this issue. There will probably always 
remain people who don't agree here, but assuming our goal is to come to a 
solution, the question is now: can current mechanisms solve the PoC case?

I believe the answer is YES, using a combination of caller prefs and 
Require, as follows:
1 - for originating UAS: IFC are configured to select UAS based on presence 
of "Require: floor-control"
2 - for terminating UAS: same
3 - for end-point selection: use callerprefs with an implicit preference for 
support of floor-control (derived from the Require header)

Alternative for (3) is to use explicit preferences instead. What may be 
confusing some people is the use of the word "preferences" in the name of 
the mechanism (it sounds optional, "nice to have", whereas PoC is 
intentional, "required"). I believe that the combination of callerprefs + 
Require does give the desired semantics and results.

Added benefit (though perhaps not initially/first concern for 3GPP) is that 
this also takes care of the case where a PoC request ends up at a UAS that 
does not support it.

Regards,
Jeroen

Paul Kyzivat wrote:
> Jeroen van Bemmel wrote:
>> Paul,
>> The model realized in IMS is perhaps more accurately described
>> as a UAC selecting a UAS with a specific intent (e.g. PoC),
>> that UAS in turn acting as a UAC to select an appropriate UAS
>> (still for PoC, on behalf of the initiating UAC) and finally
>> that second UAS acting again as UAC to select a target UAS (device)
>> which supports PoC
>
> I disagree. The originating UAC inserts an address identifying the
> terminating AOR. So the originating UAC is not selecting the
> originating PoC server.
>
> If one of the proposals in jonathan's draft (inserting a Route header
> with a URN for a PoC server) then I would agree that the originating
> UAC as selecting both the originating PoC server and the target AOR.
>
> I  also don't see that the originating PoC server is doing anything
> about selecting either a terminating PoC server or the final
> terminating UAS. It is just leaving criteria as they were set by the
> originating UAC.
> Its unclear to me whether the originating UAC and/or the terminating
> PoC server is the primary influence over which terminating UAS is
> selected, and the flip side of that is also unclear - whether the
> originating UAC is selecting the terminating PoC server. To determine
> that, we would need to examine some non-trivial use cases.
>
> Consider a case where there are two different UAs both registered to
> the same AOR, as well as another UA that doesn't support PoC. Then
> some other UAC calls this AoR. It could potentially select among the
> registered UAs using callerprefs. (IMS even specifies that the
> terminating S-CSCF shoujld apply callerprefs.) If a terminating PoC
> server is brought in, AFAIK it leaves the callerprefs alone, so that
> the terminating UAS is still selected via criteria provided by the
> originating UAC.
>
> I do think it would be good if things were arranged so that it *did*
> work as you suggest. But I think that will require some different
> mechanisms.
>
>> If you have a device and I have one, we can try to setup a
>> session and negotiate the least common denominator as features.
>> The PoC use case is more about selecting an appropriate next hop
>> from a heterogeneous set of many (times 3)
>
> If that is the purpose of the service id, then it has not been very
> well explained. This call model is very strange - pretty far from
> anything sip was designed for. If this is the goal, its going to take
> more work to figure out if/how to accomplish it.
>
>> Considering PoC as a capability / feature is confusing, I'll argue
>> that from a user perspective it is more "I want to PoC with X" as
>> opposed to "I want to communicate with X and I'll use whatever
>> matching capabilities our devices support". It is about user intent,
>> which appears difficult to capture in a generic INVITE. Perhaps a
>> new method "PoC"?
>
> I think this could be achieved as follows:
>
> - each UA that wants to receive PoC calls establishes a relationship
>   with a PoC server that will be supporting it. (TBD how this is
>   done.) The PoC server registers with the AoR on behalf of the UA.
>   If there are any capabilities that the UA wants advertised, then
>   the PoC server provides them in its registration. PoC would be
>   advertised as one of the capabilities.
>   (If there are multiple UAs that do PoC, then each will have an
>   association with a PoC server and there will be a registration
>   for each.)
>
> - incoming calls to the AoR simply consider the registered PoC server
>   as one of the registered contacts. Selection among those contacts
>   is normal. There could be forking to all. Or callerprefs could be
>   used to select a contact that supports PoC, or non-PoC.
>
> - for originating PoC calls, the UA could include a Route header that
>   sends the request first to its associated PoC server. And it could
>   include callerprefs to select PoC at the termination.
>
> Paul
>
>
>> Regards,
>> Jeroen
>>
>> -----Oorspronkelijk bericht -----
>> Van: "Paul Kyzivat" <pkyzivat@cisco.com>
>> Aan: "Paul Kyzivat" <pkyzivat@cisco.com>
>> CC: sipping@ietf.org; "Christer Holmberg (JO/LMF)"
>> <christer.holmberg@ericsson.com> Verzonden: 23-2-07 15:45
>> Onderwerp: Re: [Sipping] Service Identification in SIP - Routing
>> decissionsbased on Require/P-R Some further thoughts on this subject:
>>
>> Typically, some of the capabilities used in a call are *negotiated*
>> between the caller and the callee. As a result, those that will be
>> used are not known for certain at the time the request is sent. And the
>> results of that negotiation may differ depending on which of various
>> candidate targets is selected.
>>
>> Apparently one key intended use of the service id is to bring in one
>> or more intermediate nodes that are specific to the capabilities that
>> will be used in the call. These intermediate nodes must be selected
>> *before* the UAS is selected and the negotiation is done. For that to 
>> work,
>> one of two things must happen:
>>
>> - the proxy that selects those intermediate nodes
>>    must be prescient about what capabilities will be negotiated
>>
>> - or the selection of UAS and the negotiation of capabilities to be
>>    used must be somehow constrained to be consistent with the
>>    capabilities supported by the selected intermediate nodes.
>>
>> If we agree that prescience is infeasible, I guess we must assume
>> that the selection and negotiation will be constrained. We of course have
>> a mechanism for that: it is the Require header. There are also other
>> mechanisms that provide constraints of some sort. For instance the
>> SDP offer constrains what is permitted in the answer. (It does *not*
>> however 


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sun Feb 25 09:50:58 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLKhs-0006q9-Ch; Sun, 25 Feb 2007 09:50:32 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLKhq-0006q4-Qk
	for sipping@ietf.org; Sun, 25 Feb 2007 09:50:31 -0500
Received: from rwcrmhc12.comcast.net ([204.127.192.82])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLKho-0003qt-J2
	for sipping@ietf.org; Sun, 25 Feb 2007 09:50:30 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (rwcrmhc12) with ESMTP
	id <20070225145027m1200j9d3ce>; Sun, 25 Feb 2007 14:50:27 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1PEoQC7025073
	for <sipping@ietf.org>; Sun, 25 Feb 2007 09:50:26 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1PEoQ5N025069;
	Sun, 25 Feb 2007 09:50:26 -0500
Date: Sun, 25 Feb 2007 09:50:26 -0500
Message-Id: <200702251450.l1PEoQ5N025069@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <45DFA060.50600@cisco.com> (pkyzivat@cisco.com)
Subject: Re: [Sipping] Service Identification in SIP - Routing decissionsbased
	on Require/P-R
References: <0JDX0036OWAIN610@jes-fe1.zx.nl> <45DFA060.50600@cisco.com>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 1ac7cc0a4cd376402b85bc1961a86ac2
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   From: Paul Kyzivat <pkyzivat@cisco.com>

   > Considering PoC as a capability / feature is confusing, I'll argue
   > [edited]

   I think this could be achieved as follows:

   - each UA that wants to receive PoC calls establishes a relationship
      with a PoC server that will be supporting it. (TBD how this is
      [edited]

OK, this is starting to make sense.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sun Feb 25 09:56:09 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLKn8-0002sJ-5a; Sun, 25 Feb 2007 09:55:58 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLKn7-0002sA-3o
	for sipping@ietf.org; Sun, 25 Feb 2007 09:55:57 -0500
Received: from rwcrmhc15.comcast.net ([216.148.227.155])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLKn3-0004j4-1d
	for sipping@ietf.org; Sun, 25 Feb 2007 09:55:57 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (rwcrmhc15) with ESMTP
	id <20070225145552m1500at2dfe>; Sun, 25 Feb 2007 14:55:52 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1PEtpC7025496
	for <sipping@ietf.org>; Sun, 25 Feb 2007 09:55:51 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1PEtpLO025492;
	Sun, 25 Feb 2007 09:55:51 -0500
Date: Sun, 25 Feb 2007 09:55:51 -0500
Message-Id: <200702251455.l1PEtpLO025492@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <0JDX0036OWAIN610@jes-fe1.zx.nl> (jbemmel@zonnet.nl)
Subject: Re: [Sipping] Service Identification in SIP - Routing decissionsbased
	on Require/P-R
References: <0JDX0036OWAIN610@jes-fe1.zx.nl>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: ffa9dfbbe7cc58b3fa6b8ae3e57b0aa3
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   From: Jeroen van Bemmel <jbemmel@zonnet.nl>

   Considering PoC as a capability / feature is confusing, I'll argue
   that from a user perspective it is more "I want to PoC with X" as
   opposed to "I want to communicate with X and I'll use whatever
   matching capabilities our devices support". It is about user
   intent, which appears difficult to capture in a generic
   INVITE. Perhaps a new method "PoC"?

OK, but *why* do the users think that?  From a naive point of view,
push-to-talk is simply a user interface issue, and I can't see why the
user would be dissatisfied if the system fell back into an ordinary
full-duplex audio call, which, after all, is *better* than a PTT
connection.

I'm sure that there is some reason the cellular industry set up PoC in
a different way than ordinary calls, but I don't see why we have to
carry it forward into SIP.

And if PoC really is a "separate silo" from ordinary telephony, why
can't it just be handled by some systematic modification of AORs?
When the UA is doing a PoC mode, it can append ";poc" to the To and
>From URIs, and the destination proxy will know to route the call to an
appropriate application server.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sun Feb 25 09:57:33 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLKoe-0003eb-S4; Sun, 25 Feb 2007 09:57:32 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLKod-0003ch-6U
	for sipping@ietf.org; Sun, 25 Feb 2007 09:57:31 -0500
Received: from alnrmhc12.comcast.net ([204.127.225.92])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLKoY-0004xf-Uz
	for sipping@ietf.org; Sun, 25 Feb 2007 09:57:31 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (alnrmhc12) with ESMTP
	id <20070225145726b12009cl7je>; Sun, 25 Feb 2007 14:57:26 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1PEvPC7025613
	for <sipping@ietf.org>; Sun, 25 Feb 2007 09:57:25 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1PEvPVJ025609;
	Sun, 25 Feb 2007 09:57:25 -0500
Date: Sun, 25 Feb 2007 09:57:25 -0500
Message-Id: <200702251457.l1PEvPVJ025609@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <45DEE717.1070609@cisco.com> (pkyzivat@cisco.com)
Subject: Re: [Sipping] Service Identification in SIP - Routing decissionsbased
	on Require/P-R
References: <7374777208BDC7449D5620EF94232567023932CA@esealmw113.eemea.ericsson.se>
	<45DEE717.1070609@cisco.com>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 30ac594df0e66ffa5a93eb4c48bcb014
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   From: Paul Kyzivat <pkyzivat@cisco.com>

   I can't see a plausible way to infer any implicit callerpref from
   Accept-Contact headers.

Am I missing something?  I thought that Accept-Contact *was*
"callerpref".

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sun Feb 25 10:51:51 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLLcG-0005KV-VY; Sun, 25 Feb 2007 10:48:48 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLLcF-0005KI-DJ
	for sipping@ietf.org; Sun, 25 Feb 2007 10:48:47 -0500
Received: from smtp1.versatel.nl ([62.58.50.88])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLLcD-0004PL-Ui
	for sipping@ietf.org; Sun, 25 Feb 2007 10:48:47 -0500
Received: (qmail 1956 invoked by uid 0); 25 Feb 2007 15:49:54 -0000
Received: from ip198-11-212-87.adsl2.versatel.nl (HELO BEMBUSTER)
	([87.212.11.198]) (envelope-sender <jbemmel@zonnet.nl>)
	by smtp1.versatel.nl (qmail-ldap-1.03) with SMTP
	for < >; 25 Feb 2007 15:49:54 -0000
Message-ID: <009601c758f4$7cef71c0$0601a8c0@BEMBUSTER>
From: "Jeroen van Bemmel" <jbemmel@zonnet.nl>
To: <sipping@ietf.org>,
	<Dale.Worley@comcast.net>
References: <0JDX0036OWAIN610@jes-fe1.zx.nl>
	<200702251455.l1PEtpLO025492@dragon.ariadne.com>
Subject: Re: [Sipping] Service Identification in SIP - Routing
	decissionsbasedon Require/P-R
Date: Sun, 25 Feb 2007 16:49:09 +0100
MIME-Version: 1.0
Content-Type: text/plain; format=flowed; charset="iso-8859-1";
	reply-type=original
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 6.00.2900.3028
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.3028
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 02ec665d00de228c50c93ed6b5e4fc1a
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Dale,

Dale.Worley@comcast.net wrote:
>   From: Jeroen van Bemmel <jbemmel@zonnet.nl>
>
>   Considering PoC as a capability / feature is confusing, I'll argue
>   that from a user perspective it is more "I want to PoC with X" as
>   opposed to "I want to communicate with X and I'll use whatever
>   matching capabilities our devices support". It is about user
>   intent, which appears difficult to capture in a generic
>   INVITE. Perhaps a new method "PoC"?
>
> OK, but *why* do the users think that?  From a naive point of view,
> push-to-talk is simply a user interface issue, and I can't see why the
> user would be dissatisfied if the system fell back into an ordinary
> full-duplex audio call, which, after all, is *better* than a PTT
> connection.
>

What is "better" when depends on the usage context and the subjective 
definition of "better". I am sure there are cases where users would be 
sattisfied with a regular audio call after they pressed the "PoC" button. 
Likewise, you will also find users who would be unpleasantly surprised. 
Let's work from the assumption that there exist distinct user experiences 
"regular voice" and "push-to-talk", and that a user can choose one or the 
other. Assume that users are not willing to accept "regular voice" when they 
choose "push-to-talk", and vice versa

> I'm sure that there is some reason the cellular industry set up PoC in
> a different way than ordinary calls, but I don't see why we have to
> carry it forward into SIP.
>

I must admit I am not familiar with the details of PoC, but I imagine for 
example that for a regular voice conference you would need audio mixing, 
whereas for a group PoC session you would not. The difference should somehow 
be signaled.

> And if PoC really is a "separate silo" from ordinary telephony, why
> can't it just be handled by some systematic modification of AORs?
> When the UA is doing a PoC mode, it can append ";poc" to the To and
>> From URIs, and the destination proxy will know to route the call to
>> an
> appropriate application server.
>

That is equivalent to saying "why can't it just be handled by systematically 
including a header that says 'PoC'", which is exactly what 3GPP is doing 
with its proposal for a service identifier. Is there really a difference?

Regards,
Jeroen

> Dale
>
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP 


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sun Feb 25 11:23:49 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLM9T-0008Sy-HE; Sun, 25 Feb 2007 11:23:07 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLM9S-0008QN-48
	for sipping@ietf.org; Sun, 25 Feb 2007 11:23:06 -0500
Received: from zrtps0kn.nortel.com ([47.140.192.55])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLM9O-0003Pd-0g
	for sipping@ietf.org; Sun, 25 Feb 2007 11:23:06 -0500
Received: from zcarhxs1.corp.nortel.com (zcarhxs1.corp.nortel.com
	[47.129.230.89])
	by zrtps0kn.nortel.com (Switch-2.2.6/Switch-2.2.0) with ESMTP id
	l1PGMgb23582; Sun, 25 Feb 2007 11:22:42 -0500 (EST)
Received: from [47.130.17.207] ([47.130.17.207] RDNS failed) by
	zcarhxs1.corp.nortel.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Sun, 25 Feb 2007 11:22:27 -0500
Message-ID: <45E1B7A9.6070707@nortel.com>
Date: Sun, 25 Feb 2007 11:22:01 -0500
From: "Tom-PT Taylor" <taylor@nortel.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Paul Kyzivat <pkyzivat@cisco.com>
Subject: Re: [Sipping] Service Identification Model
References: <61968779B8AC4C4BAB421D4C12F008C009463E2E@XCH47YKF.rim.net>
	<FC318FB3-A497-4BED-BBC6-3B5B630FC589@estacado.net>
	<45DF0451.7020209@cisco.com>
In-Reply-To: <45DF0451.7020209@cisco.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 25 Feb 2007 16:22:27.0138 (UTC)
	FILETIME=[23A3B220:01C758F9]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 1df8e4abc9851cb4adb45bd64d8514ae
Cc: Andrew Allen <aallen@rim.com>, sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Paul's analysis is reasonable. Basically, starting with Andrew's list, I 
see service identification as a business issue. To optimize operation in 
the IMS architecture, it would seem reasonable in several cases for the 
originating S-CSCF to identify the tariffed service(s) implied by the 
messaging from the UA and add the appropriate administrative header 
fields so downstream elements don't have to repeat the analysis.

Paul Kyzivat wrote:
> 
> Byron Campen wrote:
>>> 1. Basic Telephony (Conversational QoS Voice only with telephony feature
>>> set)
>>> 2. Multimedia Telephony (Conversational QoS Voice plus video and other
>>> media streams with telephony feature set)
> 
> Why are 1 & 2 different services? 1 is a proper subset of 2.
> 
>>> 3. VoIP (Background QoS with no telephony features)
> 
> And why is 3 different from 1 and 2? Are you saying that to be able to 
> use a feature I must decide *before* I make the call that I might want to?
> 
> Even if that is the case, that sounds like a policy issue between a user 
> and his own provider, rather than an e2e negotiation. (My features are 
> provided by my UA and/or my provider.)
> 
>>> 4. SMS
>>> 5. MMS
> 
> Why are 4 and 5 different services? They are both datagrams. The 
> difference in media types is irrelevant. If its delivered via MESSAGE, 
> then it is already distinguishable from calls. This should be able to be 
> part of the same service as IM.
> 
>>> 6. Instant Messaging
> 
> Is just another media type. Like voice and video, these can be 
> negotiated for the session. Why balkanize as a different service?
> 
>>> 7. Push to Talk (PoC)
> 
> We've beaten this one to death already. Current implementation forces 
> need for essentially an overlay network to support it. But its far from 
> an exemplar of how to do it.
> 
>>> 8. Network Game 1 (e.g Chess)
>>> 9. Network Game 2 (e.g Martial arts game)
>>> 10. Network Game 3 (e.g Dungeons & Dragons)
> 
> I tend to agree with Eric that service URIs are good for these if there 
> is no better way. But in practice I think these could easily be 
> distinguished in SDP using o/a.
> 
>>> 11. Push Service 1 (e.g. Traffic information service)
>>> 12. Push Service 2 (e.g. Weather information service)
>>> 13. Push Service 3 (e.g  News information service)
>>> 14. Push Service 4 (e.g  Stock information service)
> 
> There may be *something* here. I need to know more about the 
> characteristics of these. If they consist of human intelligible media, 
> then I think they could adequately be dealt with via callerprefs. (This 
> may be a bit like the Actor feature tag.) Specifically, callerprefs are 
> fine if its ok that the call end up elsewhere if the callerprefs 
> selection doesn't happen or there isn't any target with that capability.
> 
> If the information isn't intelligible media, then it ought to be 
> distinguishable as a special media type in the offer.
> 
>>> 15. Mobile TV service
> 
> AFAIK, the station doesn't call me to deliver me tv service. I would be 
> initiating the call. And presumably it will be to a target service URI 
> or URN. So I see no need for a service id here.
> 
>>> So there's 15 which may all be different applications on a handset and
>>> all those are existing services or those already envisioned.
> 
> So I only see:
> 
> - 1 for basic telephony, multimedia telephony, SMS, MMS, IM
> 
> - *maybe* something for games
> 
> - *maybe* something for push services
> 
> 	Paul
> 
>>     I only see 10 here that would need to ever send a register request. 
>> (11-15 are strictly client-server) 4-6 will have distinct Content-Type 
>> headers and payloads. You would need only 1 UA for these. So we're down 
>> to 8. 8-10 will, once again, be distinct when you look at the payload 
>> and Content-Type. So we're down to 6. There is no reason at all that the 
>> UA that handles voice is incapable of handling IM, and you can tell the 
>> difference by looking at the request type, so we're down to 5. As for 
>> the voice stuff, you can distinguish 2 from anything else through the 
>> payload. So, we could quite easily roll 1-2 together. (4) 3 can be 
>> distinguished by watching for any in-dialog call-control stuff. (Having 
>> a service identifier that essentially says "I promise not to use any 
>> fancy telephony features on this call. No, really, I promise!" is never 
>> going to work) So we're down to 3.
>>
>>     SIP is more than flexible enough to handle all of this stuff in a 
>> single UA. The UAs are perfectly capable of figuring out what "service" 
>> the other wants to use; why are your middleboxes unable to do the same? 
>> Adding a parallel method to make things easy on the middleboxes is just 
>> asking for trouble, because you're conveying the same information twice. 
>> It is ripe for abuse. Of course, the easy way to do all this is to just 
>> let UAs send what they want, and charge per KB, but hey, what do I know 
>> about simplicity, flexibility, scalability, ease-of-deployment, and 
>> customer satisfaction? I'm just a IETFer.
>>
>> Best regards,
>> Byron Campen
>>
>>
>>> -----Original Message-----
>>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
>>> Sent: Thursday, February 22, 2007 7:05 PM
>>> To: Andrew Allen
>>> Cc: Eric Burger; sipping@ietf.org
>>> Subject: Re: [Sipping] Service Identification Model
>>>
>>>
>>>
>>> Andrew Allen wrote:
>>>> Hi Eric,
>>>>
>>>> It seems your prefer the approach of using URIs to identify services
>>>> which is part of Jonathan's draft as is the use of Require for those
>>>> services which simply do not interoperate unless the UAS understands
>>> the
>>>> service.
>>>>
>>>> Where I am concerned is that you seem to be suggesting that the UA
>>>> separately register an AoR for each service it supports using
>>> REGISTER.
>>>> Certainly for cellular that wouldn't be acceptable as 10 or more
>>>> services could easily be supported by a single device.
>>> Andrew, can you give me an example of 10 "services" on the device that
>>> need to be distinguished from one another in this way - that aren't just
>>>
>>> optional behavior of a particular service? (E.g. I see no reason why
>>> "audio" and "audio/video" need to be separate services.)
>>>
>>> I guess as things stand, PoC must be a separate service. And apparently
>>> Chess. What else?
>>>
>>>     Paul
>>>
>>>> Registrations
>>>> create a significant load in the IMS network because its not just a
>>> SIP
>>>> REGISTER but also requires a lot of interactions with other network
>>>> elements such as the HSS and Notifications for reg-events etc. Also
>>>> having an order of magnitude increase in multiple registrations is
>>> going
>>>> to be a big issue for the avalanche restart problem for example when a
>>>> cell site goes down and comes back into service.
>>>>
>>>> I think Jonathan's approach gives us the possibility for the registrar
>>>> to create the Service URIs based on the feature parameters included in
>>>> the Contact of the REGISTER.
>>>>
>>>> I think the argument over whether the model is one UA with multiple
>>>> services or multiple UAs each representing a separate service is
>>>> somewhat academic. As I understand a UA can register multiple AoRs and
>>>> can have other AoRs configured or otherwise provided by the service
>>>> provider.
>>>>
>>>> Andrew
>>>>
>>>> -----Original Message-----
>>>> From: Eric Burger [mailto:eburger@bea.com]
>>>> Sent: Wednesday, February 21, 2007 1:29 PM
>>>> To: sipping@ietf.org
>>>> Subject: [Sipping] Service Identification Model
>>>>
>>>> Folks have been pestering me for a while to look at the Service
>>>> Identification issue.  I finally have gotten around to it.
>>>>
>>>>
>>>> I guess I have a fundamental problem with the principal assertion of
>>> the
>>>> service-identification draft.  Namely, the assertion that there is a
>>>> single
>>>> SIP Layer, separate and distinct from two services.  Here is Figure 2
>>>> from
>>>> the draft:
>>>>
>>>>                            +---------------------------------+
>>>>                            |                                 |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | |      UI     | |      UI     | |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | |             | |             | |
>>>>                            | |  Service 1  | |  Service 2  | |
>>>>                            | |             | |             | |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | +-----------------------------+ |
>>>>                            | |                             | |
>>>>                            | |             SIP             | |
>>>>                            | |            Layer            | |
>>>>                            | |                             | |
>>>>                            | +-----------------------------+ |
>>>>                            |                                 |
>>>>                            +---------------------------------+
>>>>
>>>>                                      Physical Device
>>>>
>>>> I would offer this drawing confuses a common SIP Stack with a common
>>>> User
>>>> Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use case:
>>>>
>>>>> ... In addition users may have
>>>>> multiple devices only one of which has the application that
>>>> understands
>>>>> the game. Other devices that the user has registered against his AoR
>>>> may
>>>>> handle audio for telephony and other voice applications but if Alice
>>>>> invites Bob to play her favourite game it is not very interesting if
>>>> Bob
>>>>> accepts the gaming session on a device that can only accept the audio
>>>>> stream. Also even when the terminal that has the game application
>>>>> receives the INVITE the correct application in that terminal needs to
>>>> be
>>>>> identified to accept the session.
>>>> I would offer that this is the clinching argument that the request is,
>>>> in
>>>> fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a
>>> User
>>>> Agent.
>>>>
>>>> Same argument for PoC versus a regular voice call: One is looking for
>>> a
>>>> PoC
>>>> User Agent, NOT the PoC service at a given User Agent.
>>>>
>>>> Thus, the picture really looks like the following:
>>>>
>>>>                            +---------------------------------+
>>>>                            |                                 |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | |    UI       | |    UI       | |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | |             | |             | |
>>>>                            | |    UAS 1    | |    UAS 2    | |
>>>>                            | |             | |             | |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | +-----------------------------+ |
>>>>                            | |                             | |
>>>>                            | |             SIP             | |
>>>>                            | |            Stack            | |
>>>>                            | |                             | |
>>>>                            | +-----------------------------+ |
>>>>                            |                                 |
>>>>                            +---------------------------------+
>>>>
>>>>                                      Physical Device
>>>>
>>>>
>>>> Another argument for different User Agents is the whole "route based
>>> on
>>>> SDP"
>>>> argument.  As the author of RFC 3458, I think I was first to make the
>>>> assertion that knowing the media type did NOT mean you know the
>>>> application
>>>> to use.  I don't want your phone or proxy or B2BUA to guess that
>>> because
>>>> I
>>>> have Chess Media, I want the Chess Application; I want to talk with
>>> your
>>>> Chess Application.
>>>>
>>>> Now, is the Chess Application a Preference, Requirement, or the thing
>>>> that I
>>>> am addressing?  The answer drives one towards using caller prefs, the
>>>> Require header, or the Request-URI.
>>>>
>>>> I would offer that it is not a preference: I really want the chess
>>>> application; I probably do not want to talk to you.
>>>>
>>>> That leaves us with the Require header or a Request-URI.  My leanings
>>>> are
>>>> not surprisingly towards the RFC 4240 approach.  Frankly, if you are
>>>> building an IMS (as an example) that can handle 10,000,000
>>> subscribers,
>>>> you
>>>> should be able to handle 100,000,000 registrations.  That order of
>>>> magnitude
>>>> is way easier than the 10,000 to 10,000,000 jump.  I don't buy the
>>>> "registering each service is too much of a burden on the operator"
>>>> argument,
>>>> because it is not that much of a burden, and the network actually does
>>>> the
>>>> right thing.  Better yet, all of the corner cases work.  Best of all,
>>>> the
>>>> IETF doesn't have to build an IANA Registry and process a thousand PS
>>>> documents to register urn:service:myco:widget.
>>>>
>>>> I suppose I could live with Require:, only if you are way too lazy to
>>>> register each application instance.  It does semantically work,
>>> because
>>>> if I
>>>> want to do PoC and you do not do PoC, I'll get a 4xx response, to
>>> which
>>>> I
>>>> could try a regular call.
>>>>
>>>>
>>> _______________________________________________________________________
>>>> Notice:  This email message, together with any attachments, may
>>> contain
>>>> information  of  BEA Systems,  Inc.,  its subsidiaries  and
>>> affiliated
>>>> entities,  that may be confidential,  proprietary,  copyrighted
>>> and/or
>>>> legally privileged, and is intended solely for the use of the
>>> individual
>>>> or entity named in this message. If you are not the intended
>>> recipient,
>>>> and have received this message in error, please immediately return
>>> this
>>>> by email and then delete it.
>>>>
>>>> _______________________________________________
>>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>>> This list is for NEW development of the application of SIP
>>>> Use sip-implementors@cs.columbia.edu for questions on current sip
>>>> Use sip@ietf.org for new developments of core SIP
>>>>
>>>>
>>>>
>>>> ---------------------------------------------------------------------
>>>> This transmission (including any attachments) may contain confidential
>>> information, privileged material (including material protected by the
>>> solicitor-client or other applicable privileges), or constitute
>>> non-public information. Any use of this information by anyone other than
>>> the intended recipient is prohibited. If you have received this
>>> transmission in error, please immediately reply to the sender and delete
>>> this information from your system. Use, dissemination, distribution, or
>>> reproduction of this transmission by unintended recipients is not
>>> authorized and may be unlawful.
>>>> _______________________________________________
>>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>>> This list is for NEW development of the application of SIP
>>>> Use sip-implementors@cs.columbia.edu for questions on current sip
>>>> Use sip@ietf.org for new developments of core SIP
>>>>
>>>
>>>
>>> ---------------------------------------------------------------------
>>> This transmission (including any attachments) may contain confidential 
>>> information, privileged material (including material protected by the 
>>> solicitor-client or other applicable privileges), or constitute 
>>> non-public information. Any use of this information by anyone other 
>>> than the intended recipient is prohibited. If you have received this 
>>> transmission in error, please immediately reply to the sender and 
>>> delete this information from your system. Use, dissemination, 
>>> distribution, or reproduction of this transmission by unintended 
>>> recipients is not authorized and may be unlawful.
>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP
>>> Use sip-implementors@cs.columbia.edu for questions on current sip
>>> Use sip@ietf.org for new developments of core SIP
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Sun Feb 25 12:13:52 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLMul-0008B9-5i; Sun, 25 Feb 2007 12:11:59 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLMui-0007wT-Iv
	for sipping@ietf.org; Sun, 25 Feb 2007 12:11:56 -0500
Received: from dsl001-129-069.dfw1.dsl.speakeasy.net ([72.1.129.69]
	helo=vicuna.estacado.net) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HLMug-0002nY-Dz
	for sipping@ietf.org; Sun, 25 Feb 2007 12:11:55 -0500
Received: from [10.0.0.188] (adsl-70-235-190-149.dsl.rcsntx.sbcglobal.net
	[70.235.190.149]) (authenticated bits=0)
	by vicuna.estacado.net (8.13.8/8.13.8) with ESMTP id l1PHBhu5093191
	(version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO);
	Sun, 25 Feb 2007 11:11:48 -0600 (CST)
	(envelope-from bcampen@estacado.net)
In-Reply-To: <45E1B7A9.6070707@nortel.com>
References: <61968779B8AC4C4BAB421D4C12F008C009463E2E@XCH47YKF.rim.net>
	<FC318FB3-A497-4BED-BBC6-3B5B630FC589@estacado.net>
	<45DF0451.7020209@cisco.com> <45E1B7A9.6070707@nortel.com>
Mime-Version: 1.0 (Apple Message framework v752.2)
Message-Id: <681D70B6-F649-4CAF-8E43-72118FCE8575@estacado.net>
From: Byron Campen <bcampen@estacado.net>
Subject: Re: [Sipping] Service Identification Model
Date: Sun, 25 Feb 2007 11:11:31 -0600
To: Tom-PT Taylor <taylor@nortel.com>
X-Mailer: Apple Mail (2.752.2)
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 34d35111647d654d033d58d318c0d21a
Cc: Andrew Allen <aallen@rim.com>, Paul Kyzivat <pkyzivat@cisco.com>,
	sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============0634622591=="
Errors-To: sipping-bounces@ietf.org


--===============0634622591==
Content-Type: multipart/signed; micalg=sha1; boundary=Apple-Mail-9-842769952;
	protocol="application/pkcs7-signature"


--Apple-Mail-9-842769952
Content-Transfer-Encoding: 7bit
Content-Type: text/plain;
	charset=US-ASCII;
	delsp=yes;
	format=flowed

	This is more palatable than having the UAs provide these  
identifiers. I can still see interop problems down the road where UAs  
could confuse (or deliberately mislead) the S-CSCF by trying to do  
something with which it is not familiar. It seems to me that  
approaching this from the angle of "we have a fixed set of services,  
with individual pricing" is fundamentally and unavoidably at odds  
with the basic structure of SIP. Pricing should be based on the  
amount of work that the provider has to do. This is extremely easy to  
figure out. Bandwidth costs x, location service (ie, call  
establishment) costs y, other functionality (that your servers have  
to do work to accomplish) will cost whatever you feel is appropriate,  
etc. Putting in all the extra infrastructure/logic to simulate the  
charging for old 2G networks, or even the PSTN, seems pointless to  
me. You end up wasting money, time(ie, money), and effort(ie, more  
money). Why not use a charging system that isn't a total pain to  
implement? If you base your charging on the actual work you're doing,  
there is less to exploit in your system, and you can make _more_  
money because you don't have all the overhead from the PSTN billing  
emulation layer! Adopting this strategy does not impede your ability  
to set your pricing to any level you want. It in no way impedes your  
ability to make money.

Best regards,
Byron Campen

> Paul's analysis is reasonable. Basically, starting with Andrew's  
> list, I see service identification as a business issue. To optimize  
> operation in the IMS architecture, it would seem reasonable in  
> several cases for the originating S-CSCF to identify the tariffed  
> service(s) implied by the messaging from the UA and add the  
> appropriate administrative header fields so downstream elements  
> don't have to repeat the analysis.
>


--Apple-Mail-9-842769952
Content-Transfer-Encoding: base64
Content-Type: application/pkcs7-signature;
	name=smime.p7s
Content-Disposition: attachment;
	filename=smime.p7s
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--Apple-Mail-9-842769952--


--===============0634622591==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============0634622591==--




From sipping-bounces@ietf.org Sun Feb 25 21:34:30 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLVdP-00021z-Ux; Sun, 25 Feb 2007 21:30:39 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLVdN-00021u-OT
	for sipping@ietf.org; Sun, 25 Feb 2007 21:30:37 -0500
Received: from rtp-iport-2.cisco.com ([64.102.122.149])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLVdM-0005gm-CL
	for sipping@ietf.org; Sun, 25 Feb 2007 21:30:37 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by rtp-iport-2.cisco.com with ESMTP; 25 Feb 2007 21:30:36 -0500
X-IronPort-AV: i="4.14,217,1170651600"; 
	d="scan'208"; a="114545304:sNHT45088490"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1Q2UZMN002146; 
	Sun, 25 Feb 2007 21:30:35 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1Q2UVVV018186; 
	Sun, 25 Feb 2007 21:30:35 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Sun, 25 Feb 2007 21:30:32 -0500
Received: from [10.86.242.125] ([10.86.242.125]) by xfe-rtp-202.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Sun, 25 Feb 2007 21:30:32 -0500
Message-ID: <45E24647.40002@cisco.com>
Date: Sun, 25 Feb 2007 21:30:31 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Dale.Worley@comcast.net
Subject: Re: [Sipping] Service Identification in SIP - Routing decissionsbased
	on Require/P-R
References: <7374777208BDC7449D5620EF94232567023932CA@esealmw113.eemea.ericsson.se>	<45DEE717.1070609@cisco.com>
	<200702251457.l1PEvPVJ025609@dragon.ariadne.com>
In-Reply-To: <200702251457.l1PEvPVJ025609@dragon.ariadne.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 02:30:32.0506 (UTC)
	FILETIME=[169D09A0:01C7594E]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=776; t=1172457035;
	x=1173321035; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP=20-=2
	0Routing=20decissionsbased=0A=20on=20Require/P-R |Sender:=20
	|To:=20Dale.Worley@comcast.net;
	bh=nXyE8CmZPXYJZssTkTJwpntrp1xifaU56hxdlGczMx0=;
	b=MR+xId3ESNkVcQ5fENbPewgYjNGRFNzWU11M/i4qlKDyQSHrD2qNusB9EfjuWpHBTtpNVj7T
	NLjX+eOnM+K26iN8whWFkoxi/XppvFkRSeZETCrSLLs6qgHWcQf5Uy74;
Authentication-Results: rtp-dkim-1; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 7d33c50f3756db14428398e2bdedd581
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Dale.Worley@comcast.net wrote:
>    From: Paul Kyzivat <pkyzivat@cisco.com>
> 
>    I can't see a plausible way to infer any implicit callerpref from
>    Accept-Contact headers.
> 
> Am I missing something?  I thought that Accept-Contact *was*
> "callerpref".

What *was* I thinking? I clearly didn't say what I meant. :-(

The context was making a routing decision to an application server. What 
I *meant* to say was that I can't see any plausible way to make a 
selection of an application server based on an Accept-Contact header, 
because that isn't definitive for the capabilities of the UAS that will 
be selected. That is in contrast to the Require header, which is 
definitive, so using it to select an application server is plausible.

	Paul

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 03:10:00 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLauN-0007sY-K5; Mon, 26 Feb 2007 03:08:31 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLauK-0007hq-Eb
	for sipping@ietf.org; Mon, 26 Feb 2007 03:08:28 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLauE-0007RF-Ae
	for sipping@ietf.org; Mon, 26 Feb 2007 03:08:28 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	DCC6920563
	for <sipping@ietf.org>; Mon, 26 Feb 2007 09:08:21 +0100 (CET)
X-AuditID: c1b4fb3e-af6d2bb0000007e1-2d-45e29575d650 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	BF2FE2010A
	for <sipping@ietf.org>; Mon, 26 Feb 2007 09:08:21 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 09:08:21 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification Model
Date: Mon, 26 Feb 2007 09:08:16 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD02925D3B@esealmw111>
In-Reply-To: <45DF0451.7020209@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification Model
Thread-Index: AcdXXY85vsVlosCSTTmvMKfm9wFwaACHzb2g
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 26 Feb 2007 08:08:21.0781 (UTC)
	FILETIME=[480AF450:01C7597D]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 4ec3642ae9025e273a4a263d640f3300
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi,

Just a comment on the Messaging here;  While I fully understand the view
that isntant messaging is another media, I also understand that the
evolution of messaging in OMA is that instant messaging is actally
defined as a seperate service to telephony.  SMS->MMS->Instant
messaging.

Cheers,

..//Steve

-----Original Message-----
From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
Sent: Friday, February 23, 2007 4:12 PM
To: Byron Campen
Cc: Andrew Allen; sipping@ietf.org
Subject: Re: [Sipping] Service Identification Model



Byron Campen wrote:
>>
>> 1. Basic Telephony (Conversational QoS Voice only with telephony=20
>> feature
>> set)
>> 2. Multimedia Telephony (Conversational QoS Voice plus video and=20
>> other media streams with telephony feature set)

Why are 1 & 2 different services? 1 is a proper subset of 2.

>> 3. VoIP (Background QoS with no telephony features)

And why is 3 different from 1 and 2? Are you saying that to be able to
use a feature I must decide *before* I make the call that I might want
to?

Even if that is the case, that sounds like a policy issue between a user
and his own provider, rather than an e2e negotiation. (My features are
provided by my UA and/or my provider.)

>> 4. SMS
>> 5. MMS

Why are 4 and 5 different services? They are both datagrams. The
difference in media types is irrelevant. If its delivered via MESSAGE,
then it is already distinguishable from calls. This should be able to be
part of the same service as IM.

>> 6. Instant Messaging

Is just another media type. Like voice and video, these can be
negotiated for the session. Why balkanize as a different service?

>> 7. Push to Talk (PoC)

We've beaten this one to death already. Current implementation forces
need for essentially an overlay network to support it. But its far from
an exemplar of how to do it.

>> 8. Network Game 1 (e.g Chess)
>> 9. Network Game 2 (e.g Martial arts game) 10. Network Game 3 (e.g=20
>> Dungeons & Dragons)

I tend to agree with Eric that service URIs are good for these if there
is no better way. But in practice I think these could easily be
distinguished in SDP using o/a.

>> 11. Push Service 1 (e.g. Traffic information service) 12. Push=20
>> Service 2 (e.g. Weather information service) 13. Push Service 3 (e.g

>> News information service) 14. Push Service 4 (e.g  Stock information=20
>> service)

There may be *something* here. I need to know more about the
characteristics of these. If they consist of human intelligible media,
then I think they could adequately be dealt with via callerprefs. (This
may be a bit like the Actor feature tag.) Specifically, callerprefs are
fine if its ok that the call end up elsewhere if the callerprefs
selection doesn't happen or there isn't any target with that capability.

If the information isn't intelligible media, then it ought to be
distinguishable as a special media type in the offer.

>> 15. Mobile TV service

AFAIK, the station doesn't call me to deliver me tv service. I would be
initiating the call. And presumably it will be to a target service URI
or URN. So I see no need for a service id here.

>> So there's 15 which may all be different applications on a handset=20
>> and all those are existing services or those already envisioned.

So I only see:

- 1 for basic telephony, multimedia telephony, SMS, MMS, IM

- *maybe* something for games

- *maybe* something for push services

	Paul

>     I only see 10 here that would need to ever send a register
request.=20
> (11-15 are strictly client-server) 4-6 will have distinct Content-Type

> headers and payloads. You would need only 1 UA for these. So we're=20
> down to 8. 8-10 will, once again, be distinct when you look at the=20
> payload and Content-Type. So we're down to 6. There is no reason at=20
> all that the UA that handles voice is incapable of handling IM, and=20
> you can tell the difference by looking at the request type, so we're=20
> down to 5. As for the voice stuff, you can distinguish 2 from anything

> else through the payload. So, we could quite easily roll 1-2 together.

> (4) 3 can be distinguished by watching for any in-dialog call-control=20
> stuff. (Having a service identifier that essentially says "I promise=20
> not to use any fancy telephony features on this call. No, really, I=20
> promise!" is never going to work) So we're down to 3.
>=20
>     SIP is more than flexible enough to handle all of this stuff in a=20
> single UA. The UAs are perfectly capable of figuring out what
"service"
> the other wants to use; why are your middleboxes unable to do the
same?=20
> Adding a parallel method to make things easy on the middleboxes is=20
> just asking for trouble, because you're conveying the same information
twice.
> It is ripe for abuse. Of course, the easy way to do all this is to=20
> just let UAs send what they want, and charge per KB, but hey, what do=20
> I know about simplicity, flexibility, scalability, ease-of-deployment,

> and customer satisfaction? I'm just a IETFer.
>=20
> Best regards,
> Byron Campen
>=20
>=20
>> -----Original Message-----
>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
>> Sent: Thursday, February 22, 2007 7:05 PM
>> To: Andrew Allen
>> Cc: Eric Burger; sipping@ietf.org
>> Subject: Re: [Sipping] Service Identification Model
>>
>>
>>
>> Andrew Allen wrote:
>>> Hi Eric,
>>>
>>> It seems your prefer the approach of using URIs to identify services

>>> which is part of Jonathan's draft as is the use of Require for those

>>> services which simply do not interoperate unless the UAS understands
>> the
>>> service.
>>>
>>> Where I am concerned is that you seem to be suggesting that the UA=20
>>> separately register an AoR for each service it supports using
>> REGISTER.
>>> Certainly for cellular that wouldn't be acceptable as 10 or more=20
>>> services could easily be supported by a single device.
>>
>> Andrew, can you give me an example of 10 "services" on the device=20
>> that need to be distinguished from one another in this way - that=20
>> aren't just
>>
>> optional behavior of a particular service? (E.g. I see no reason why=20
>> "audio" and "audio/video" need to be separate services.)
>>
>> I guess as things stand, PoC must be a separate service. And=20
>> apparently Chess. What else?
>>
>>     Paul
>>
>>> Registrations
>>> create a significant load in the IMS network because its not just a
>> SIP
>>> REGISTER but also requires a lot of interactions with other network=20
>>> elements such as the HSS and Notifications for reg-events etc. Also=20
>>> having an order of magnitude increase in multiple registrations is
>> going
>>> to be a big issue for the avalanche restart problem for example when

>>> a cell site goes down and comes back into service.
>>>
>>> I think Jonathan's approach gives us the possibility for the=20
>>> registrar to create the Service URIs based on the feature parameters

>>> included in the Contact of the REGISTER.
>>>
>>> I think the argument over whether the model is one UA with multiple=20
>>> services or multiple UAs each representing a separate service is=20
>>> somewhat academic. As I understand a UA can register multiple AoRs=20
>>> and can have other AoRs configured or otherwise provided by the=20
>>> service provider.
>>>
>>> Andrew
>>>
>>> -----Original Message-----
>>> From: Eric Burger [mailto:eburger@bea.com]
>>> Sent: Wednesday, February 21, 2007 1:29 PM
>>> To: sipping@ietf.org
>>> Subject: [Sipping] Service Identification Model
>>>
>>> Folks have been pestering me for a while to look at the Service=20
>>> Identification issue.  I finally have gotten around to it.
>>>
>>>
>>> I guess I have a fundamental problem with the principal assertion of
>> the
>>> service-identification draft.  Namely, the assertion that there is a

>>> single SIP Layer, separate and distinct from two services.  Here is=20
>>> Figure 2 from the draft:
>>>
>>>                            +---------------------------------+
>>>                            |                                 |
>>>                            | +-------------+ +-------------+ |
>>>                            | |      UI     | |      UI     | |
>>>                            | +-------------+ +-------------+ |
>>>                            | +-------------+ +-------------+ |
>>>                            | |             | |             | |
>>>                            | |  Service 1  | |  Service 2  | |
>>>                            | |             | |             | |
>>>                            | +-------------+ +-------------+ |
>>>                            | +-----------------------------+ |
>>>                            | |                             | |
>>>                            | |             SIP             | |
>>>                            | |            Layer            | |
>>>                            | |                             | |
>>>                            | +-----------------------------+ |
>>>                            |                                 |
>>>                            +---------------------------------+
>>>
>>>                                      Physical Device
>>>
>>> I would offer this drawing confuses a common SIP Stack with a common

>>> User Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use=20
>>> case:
>>>
>>>> ... In addition users may have
>>>> multiple devices only one of which has the application that
>>> understands
>>>> the game. Other devices that the user has registered against his=20
>>>> AoR
>>> may
>>>> handle audio for telephony and other voice applications but if=20
>>>> Alice invites Bob to play her favourite game it is not very=20
>>>> interesting if
>>> Bob
>>>> accepts the gaming session on a device that can only accept the=20
>>>> audio stream. Also even when the terminal that has the game=20
>>>> application receives the INVITE the correct application in that=20
>>>> terminal needs to
>>> be
>>>> identified to accept the session.
>>>
>>> I would offer that this is the clinching argument that the request=20
>>> is, in fact, addressing a SEPARATE USER AGENT, not a special SERVICE

>>> at a
>> User
>>> Agent.
>>>
>>> Same argument for PoC versus a regular voice call: One is looking=20
>>> for
>> a
>>> PoC
>>> User Agent, NOT the PoC service at a given User Agent.
>>>
>>> Thus, the picture really looks like the following:
>>>
>>>                            +---------------------------------+
>>>                            |                                 |
>>>                            | +-------------+ +-------------+ |
>>>                            | |    UI       | |    UI       | |
>>>                            | +-------------+ +-------------+ |
>>>                            | +-------------+ +-------------+ |
>>>                            | |             | |             | |
>>>                            | |    UAS 1    | |    UAS 2    | |
>>>                            | |             | |             | |
>>>                            | +-------------+ +-------------+ |
>>>                            | +-----------------------------+ |
>>>                            | |                             | |
>>>                            | |             SIP             | |
>>>                            | |            Stack            | |
>>>                            | |                             | |
>>>                            | +-----------------------------+ |
>>>                            |                                 |
>>>                            +---------------------------------+
>>>
>>>                                      Physical Device
>>>
>>>
>>> Another argument for different User Agents is the whole "route based
>> on
>>> SDP"
>>> argument.  As the author of RFC 3458, I think I was first to make=20
>>> the assertion that knowing the media type did NOT mean you know the=20
>>> application to use.  I don't want your phone or proxy or B2BUA to=20
>>> guess that
>> because
>>> I
>>> have Chess Media, I want the Chess Application; I want to talk with
>> your
>>> Chess Application.
>>>
>>> Now, is the Chess Application a Preference, Requirement, or the=20
>>> thing that I am addressing?  The answer drives one towards using=20
>>> caller prefs, the Require header, or the Request-URI.
>>>
>>> I would offer that it is not a preference: I really want the chess=20
>>> application; I probably do not want to talk to you.
>>>
>>> That leaves us with the Require header or a Request-URI.  My=20
>>> leanings are not surprisingly towards the RFC 4240 approach. =20
>>> Frankly, if you are building an IMS (as an example) that can handle=20
>>> 10,000,000
>> subscribers,
>>> you
>>> should be able to handle 100,000,000 registrations.  That order of=20
>>> magnitude is way easier than the 10,000 to 10,000,000 jump.  I don't

>>> buy the "registering each service is too much of a burden on the=20
>>> operator"
>>> argument,
>>> because it is not that much of a burden, and the network actually=20
>>> does the right thing.  Better yet, all of the corner cases work. =20
>>> Best of all, the IETF doesn't have to build an IANA Registry and=20
>>> process a thousand PS documents to register urn:service:myco:widget.
>>>
>>> I suppose I could live with Require:, only if you are way too lazy=20
>>> to register each application instance.  It does semantically work,
>> because
>>> if I
>>> want to do PoC and you do not do PoC, I'll get a 4xx response, to
>> which
>>> I
>>> could try a regular call.
>>>
>>>
>> _____________________________________________________________________
>> __
>>> Notice:  This email message, together with any attachments, may
>> contain
>>> information  of  BEA Systems,  Inc.,  its subsidiaries  and
>> affiliated
>>> entities,  that may be confidential,  proprietary,  copyrighted
>> and/or
>>> legally privileged, and is intended solely for the use of the
>> individual
>>> or entity named in this message. If you are not the intended
>> recipient,
>>> and have received this message in error, please immediately return
>> this
>>> by email and then delete it.
>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP Use=20
>>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
>>> sip@ietf.org for new developments of core SIP
>>>
>>>
>>>
>>> --------------------------------------------------------------------
>>> - This transmission (including any attachments) may contain=20
>>> confidential
>> information, privileged material (including material protected by the

>> solicitor-client or other applicable privileges), or constitute=20
>> non-public information. Any use of this information by anyone other=20
>> than the intended recipient is prohibited. If you have received this=20
>> transmission in error, please immediately reply to the sender and=20
>> delete this information from your system. Use, dissemination,=20
>> distribution, or reproduction of this transmission by unintended=20
>> recipients is not authorized and may be unlawful.
>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP Use=20
>>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
>>> sip@ietf.org for new developments of core SIP
>>>
>>
>>
>>
>> ---------------------------------------------------------------------
>> This transmission (including any attachments) may contain=20
>> confidential information, privileged material (including material=20
>> protected by the solicitor-client or other applicable privileges), or

>> constitute non-public information. Any use of this information by=20
>> anyone other than the intended recipient is prohibited. If you have=20
>> received this transmission in error, please immediately reply to the=20
>> sender and delete this information from your system. Use,=20
>> dissemination, distribution, or reproduction of this transmission by=20
>> unintended recipients is not authorized and may be unlawful.
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP Use=20
>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
>> sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 03:41:58 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLbPQ-0005zg-5e; Mon, 26 Feb 2007 03:40:36 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLbPO-0005zW-9I
	for sipping@ietf.org; Mon, 26 Feb 2007 03:40:34 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLbPH-0002NG-Ji
	for sipping@ietf.org; Mon, 26 Feb 2007 03:40:34 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	0AFEF2100D; Mon, 26 Feb 2007 09:40:23 +0100 (CET)
X-AuditID: c1b4fb3e-af6d2bb0000007e1-d4-45e29cf6221a 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	E64B12100C; Mon, 26 Feb 2007 09:40:22 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 09:40:22 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP - Routing decissionsbased
	on Require/P-R
Date: Mon, 26 Feb 2007 09:40:21 +0100
Message-ID: <7374777208BDC7449D5620EF94232567023F8BCA@esealmw113.eemea.ericsson.se>
In-Reply-To: <45DF0564.1070402@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP - Routing
	decissionsbased on Require/P-R
Thread-Index: AcdXXaxaat1xCsj/QCmmaTiv8aFJ7ACIKraw
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>
X-OriginalArrivalTime: 26 Feb 2007 08:40:22.0987 (UTC)
	FILETIME=[C12BB9B0:01C75981]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 789c141a303c09204b537a4078e2a63f
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,=20

> >> The difference here is that the Require header is=20
> definitive that the=20
> >> caller doesn't want to reach a UAS unless that UAS supports the=20
> >> stated options. Callerprefs is not so definitive. There=20
> are some uses=20
> >> of callerprefs that can be considered fairly definitive, but in=20
> >> general that is not the case.
> >>
> >> Jonathan proposed extending 3841 so that the use of Require =20
> >>generates an implicit callerpref. And that in turn would  influence=20
> >>the selection of the registered contacts to try. I  can't see a=20
> >>plausible way to infer any implicit callerpref from Accept-Contact=20
> >>headers.
> >> In any case, that is all about selecting how to translate=20
> the  R-URI.=20
> >>It is not about selecting intermediate proxies. As  Dale=20
> has stated,=20
> >>you are free to use most anything to select  an=20
> intermediate proxy. If=20
> >>you want to use Require, or  Accept-Contact to do that you=20
> may do so.=20
> >>But while you are  free to do so, it is less clear that it=20
> is useful=20
> >>or wise to  do so. If the intent is to select a proxy that is=20
> >>consistent  with the capabilities of the UAS that will=20
> eventually be =20
> >>accepted (as seems to be the case with PoC), then it only =20
> makes sense=20
> >>if the thing you are testing is definitive for  the capabilities of=20
> >>the UAS. So again, Require may make sense  for this, while=20
> >>Accept-Contact is not.
> >=20
>>Don't you get the same effect by including explicit require in the=20
>>Accept-Contact and adding "pref" to the Require Header?
>=20
>Not quite.
>=20
>Require applies to the UAS, not proxies. Proxy-Require:pref=20
>would be closer. But Proxy-Require:pref only requires that=20
>the proxy *understand* callerprefs. It can still ignore those=20
>preferences based on its own policy. And it requires that=20
>*all* the proxies on the path understand callerprefs, not=20
>just the one that is responsible for the domain of the R-URI,=20
>where the preferences should be acted upon.

No matter if you use "Require:service-X" or "Require:pref", the proxy is
not requierd to understand the extension as such.

In both cases the proxy simply performs routing based on the value (in
the "pref" case it also looks at the Accept-Contact).

I do agree that Proxy-Require is not good. It would be good to point
Proxy-Require to specific proxies (as we noted also in the outbound
discussion), but unfortunately that is not possible.

Regards,

Christer








>=20
> 	Paul
>=20
> > Regards,
> >=20
> > Christer
> >=20
> >=20
> >=20
> >=20
> >> 	Paul
> >>
> >> Christer Holmberg (JO/LMF) wrote:
> >>> Hi,
> >>>
> >>>> -----Original Message-----
> >>>> From: Dale.Worley@comcast.net [mailto:Dale.Worley@comcast.net]
> >>>> Sent: 22. helmikuuta 2007 17:35
> >>>> To: sipping@ietf.org
> >>>> Subject: Re: [Sipping] Service Identification in SIP - Routing=20
> >>>> decissionsbased on Require/P-R
> >>>>
> >>>>    From: Henning Schulzrinne <hgs@cs.columbia.edu>
> >>>>
> >>>>    Proxies can choose next hops based on RFC 3840, which
> >> is another
> >>>> way
> >>>>    of saying "routing".
> >>>>
> >>>> RFC 3261 allows a proxy unlimited latitude in choosing=20
> targets to=20
> >>>> which to forward a request, if the proxy is authoritative for the
> >>>> request-URI:
> >>>>
> >>>> "[T]he element is responsible for the domain in the
> >> Request-URI, and
> >>>> the element MAY use whatever mechanism it desires to
> >> determine where
> >>>> to send the request."
> >>>>
> >>>> Within that framework, it seems to me quite reasonable for
> >> a service
> >>>> provider's proxy to examine Require or Proxy-Require headers to=20
> >>>> determine which application server would be best for
> >> handling a call,
> >>>> and routing the call to it.
> >>> Why can't Accept-Contact be used for the same purpose?
> >>>
> >>> My issue is the statement that Require/Proxy-Require is the most=20
> >>> appropriate way of doing the routing decisson, and I am wondering=20
> >>> where that is specified.
> >>>
> >>> Regards,
> >>>
> >>> Christer
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>> Dale
> >>>>
> >>>> _______________________________________________
> >>>> Sipping mailing list
> >> https://www1.ietf.org/mailman/listinfo/sipping
> >>>> This list is for NEW development of the application of SIP Use=20
> >>>> sip-implementors@cs.columbia.edu for questions on=20
> current sip Use=20
> >>>> sip@ietf.org for new developments of core SIP
> >>>>
> >>> _______________________________________________
> >>> Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> >>> This list is for NEW development of the application of SIP Use=20
> >>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> >>> sip@ietf.org for new developments of core SIP
> >>>
> >=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 05:35:03 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLdB8-00062A-2F; Mon, 26 Feb 2007 05:33:58 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLdB7-000624-7L
	for sipping@ietf.org; Mon, 26 Feb 2007 05:33:57 -0500
Received: from wx-out-0506.google.com ([66.249.82.232])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLdB5-0003rS-UZ
	for sipping@ietf.org; Mon, 26 Feb 2007 05:33:57 -0500
Received: by wx-out-0506.google.com with SMTP id h31so1367668wxd
	for <sipping@ietf.org>; Mon, 26 Feb 2007 02:33:55 -0800 (PST)
DKIM-Signature: a=rsa-sha1; c=relaxed/relaxed; d=gmail.com; s=beta;
	h=domainkey-signature:received:received:message-id:date:from:sender:to:subject:mime-version:content-type:content-transfer-encoding:content-disposition:x-google-sender-auth;
	b=r7Dx+5u0TFT5W/xDJeLYyM4s8vOy7a1raDhB7KI/ZZ+NpfnHnWAeClOIO0ic2SJjC6VfVyPwX7IIr/RsNhOAQNbxqV4mlHJjGkZAdAhR28f+KHFsQ7Fos7d6a1SCnlCPNarTr/gDDxXlwQxeOy92BSwB2HLcXMhxjdmiSApIoRA=
DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta;
	h=received:message-id:date:from:sender:to:subject:mime-version:content-type:content-transfer-encoding:content-disposition:x-google-sender-auth;
	b=XnaxrKXEEOIWTgo18y4NjZ+BXyavqSP9qWQDTxoG/gbfVCb6QPZmHzEi3/T+h8sYO8KfMjv4ridqd9A50kAkZcYlX8Xtukqqto0WMbgWuvISZVrqizfdwynaOkWuv/DdAazHyKkLPvqEjXmGa7KjCWorVosIZT5tvdIO9MLwpU4=
Received: by 10.90.52.18 with SMTP id z18mr4455850agz.1172486035596;
	Mon, 26 Feb 2007 02:33:55 -0800 (PST)
Received: by 10.90.53.9 with HTTP; Mon, 26 Feb 2007 02:33:55 -0800 (PST)
Message-ID: <458913680702260233v3731d4c2t6116004d4542996@mail.gmail.com>
Date: Mon, 26 Feb 2007 11:33:55 +0100
From: "Xavier Marjou" <xavier.marjou@orange-ftgroup.com>
To: sipping@ietf.org
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
X-Google-Sender-Auth: 9c543e7812a32c42
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 0fa76816851382eb71b0a882ccdc29ac
Subject: [Sipping] Transparent B2BUA
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

If there is agenda time and interest, I would like to present this
draft in Prague.

Xavier

---------- Forwarded message ----------

Date: Thu, 08 Feb 2007 18:50:02 -0500
From: Internet-Drafts@ietf.org
Subject: I-D ACTION:draft-marjou-sipping-b2bua-00.txt
To: i-d-announce@ietf.org
Content-Type: text/plain; charset="us-ascii"

A New Internet-Draft is available from the on-line Internet-Drafts
directories.


      Title           : Requirements for a Session Initiation
Protocol (SIP) Transparent Back-To-Back User-Agent (B2BUA)
      Author(s)       : X. Marjou
      Filename        : draft-marjou-sipping-b2bua-00.txt
      Pages           : 10
      Date            : 2007-2-8


 A SIP Back-To-Back User Agent (B2BUA) refers to the concatenation of
 a SIP User Agent Client (UAC) and a SIP User Agent Server (UAS).  A
 transparent B2BUA is a particular type of B2BUA that forwards SIP
 messages in a SIP proxy-like way, and that also benefits from some
 features of a User Agent (UA) element.  This document recommends best
 current practices for the implementation of such a transparent
B2BUA.   Developing transparent B2BUAs that meet this set of
requirements will
 greatly increase the likelihood that SIP applications will function
 properly.




A URL for this Internet-Draft is:
http://www.ietf.org/internet-drafts/draft-marjou-sipping-b2bua-00.txt

To remove yourself from the I-D Announcement list, send a message to
i-d-announce-request@ietf.org with the word unsubscribe in the body of
the message.
You can also visit https://www1.ietf.org/mailman/listinfo/I-D-announce
to change your subscription settings.

Internet-Drafts are also available by anonymous FTP. Login with the
username "anonymous" and a password of your e-mail address. After
logging in, type "cd internet-drafts" and then
"get draft-marjou-sipping-b2bua-00.txt".

A list of Internet-Drafts directories can be found in
http://www.ietf.org/shadow.html
or ftp://ftp.ietf.org/ietf/1shadow-sites.txt

Internet-Drafts can also be obtained by e-mail.

Send a message to:
      mailserv@ietf.org.
In the body type:
      "FILE /internet-drafts/draft-marjou-sipping-b2bua-00.txt".

NOTE:   The mail server at ietf.org can return the document in
      MIME-encoded form by using the "mpack" utility.  To use this
      feature, insert the command "ENCODING mime" before the "FILE"
      command.  To decode the response(s), you will need "munpack" or
      a MIME-compliant mail reader.  Different MIME-compliant mail readers
      exhibit different behavior, especially when dealing with
      "multipart" MIME messages (i.e. documents which have been split
      up into multiple messages), so check your local documentation on
      how to manipulate these messages.

Below is the data which will enable a MIME compliant mail reader
implementation to automatically retrieve the ASCII version of the
Internet-Draft.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 09:44:16 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLh13-0001pp-Pd; Mon, 26 Feb 2007 09:39:49 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLh11-0001pV-6U
	for sipping@ietf.org; Mon, 26 Feb 2007 09:39:47 -0500
Received: from rtp-iport-2.cisco.com ([64.102.122.149])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLh0v-0001mf-R5
	for sipping@ietf.org; Mon, 26 Feb 2007 09:39:47 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by rtp-iport-2.cisco.com with ESMTP; 26 Feb 2007 09:39:42 -0500
X-IronPort-AV: i="4.14,221,1170651600"; 
	d="scan'208"; a="114561102:sNHT50226888"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1QEdfCO027818; 
	Mon, 26 Feb 2007 09:39:41 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1QEdEW7001684; 
	Mon, 26 Feb 2007 09:39:41 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 09:39:11 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-201.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 09:39:11 -0500
Message-ID: <45E2F10E.3070906@cisco.com>
Date: Mon, 26 Feb 2007 09:39:10 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: "Stephen Terrill (MI/EEM)" <stephen.terrill@ericsson.com>
Subject: Re: [Sipping] Service Identification in SIP
References: <989763567E51AB48BE015DF7BB960CAD028F09A5@esealmw111>
In-Reply-To: <989763567E51AB48BE015DF7BB960CAD028F09A5@esealmw111>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 14:39:11.0180 (UTC)
	FILETIME=[E0F948C0:01C759B3]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=2995; t=1172500781;
	x=1173364781; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20
	|To:=20=22Stephen=20Terrill=20(MI/EEM)=22=20<stephen.terrill@ericsson.com
	>; bh=+KXO8nwnoALKE+DVGX74HsX9KMbYoZNVkN5qrdtYX6E=;
	b=KbrRBP+EMrT/YNHH4bmlnd1F7ZVosx8AnGJcm/nkjpsgb3YZXV/md0mb6QH/7GoG7s8ZIXMi
	lUWnqAqPMyN+UUSTSxJ+iX2y8dYDnbCPA/1iGoxglLNdH9sajmxvkxpR;
Authentication-Results: rtp-dkim-2; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: f607d15ccc2bc4eaf3ade8ffa8af02a0
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Stephen Terrill (MI/EEM) wrote:

> Hi Andrew,
> 
> I have the opposite view.  I think that if the originating network and
> the terminating network are going to interoperate, then they will have
> the same understanding of the service, hence in my view the same
> "identifier".  To have otherwise is problematic and complicated (the
> operators will have to have a higher admininstrative effort in order to
> understand how they should interconnect).

This is exactly what will lead to total balkanization and siloization of 
IP-based communications systems.

We achieve interoperability by defining standards around how SIP works. 
What you are basically saying above is that IETF standards are not 
sufficient for operators to interwork and interconnect, and instead they 
will need to all get together and discuss what services run on their 
network, define identifiers for them, and agree on how those match up. 
In what way is that not a recipe for complete balkanization and 
interoperability failures? In what way is such a view not anathema to 
the entire purpose of standardization?

If we have done our jobs right, all you need to interwork is to look at 
the SIP message, and based on the standards we have defined, you will 
know exactly what is being requested. If SDP contains audio and video, 
you know that a phone is asking for an audio/video call. If it contains 
MSRP then you know its asking for a text chat session. We go through a 
lot of effort to standardize these things so that we have 
interoperability between vendors and networks. We carefully design 
fallbacks and backwards compatibility schemes so that we get 
least-common-denominator service when heterogeneous devices 
interconnect. I believe it is critical that we not bypass all of that 
machinery by requiring that providers define and agree upon services for 
interworking.

This is why I suggest that, if you have in fact defined a 'service' 
which cannot interoperate at all unless the other side has understood 
it, then you have in fact introduced a non-backwards compatible 
extension to SIP. You signal that with Require. Does this make it hard 
to go and standardize lots of these? ABSOLUTELY! That is what we want. 
The bar should be quite high for defining non-backwards compatible 
extensions, so that we preserve interoperability across heterogeneous 
networks. I hope that there are as few of these as possible, not lots. 
If PoC is already there, then so be it - let us define a tag for it, put 
it in Require, and be done. I hope that is the first and last such 
service tag we need to standardize.

-Jonathan R.


-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 09:46:32 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLh7Y-0004QT-JF; Mon, 26 Feb 2007 09:46:32 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLh7W-0004Oy-N2
	for sipping@ietf.org; Mon, 26 Feb 2007 09:46:30 -0500
Received: from sj-iport-6.cisco.com ([171.71.176.117])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLh7T-0002nY-Hh
	for sipping@ietf.org; Mon, 26 Feb 2007 09:46:30 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by sj-iport-6.cisco.com with ESMTP; 26 Feb 2007 06:46:26 -0800
X-IronPort-AV: i="4.14,221,1170662400"; 
	d="scan'208"; a="116197820:sNHT75658536"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1QEkQ7s006732; 
	Mon, 26 Feb 2007 09:46:26 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1QEkMVX003821; 
	Mon, 26 Feb 2007 09:46:26 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 09:46:22 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-202.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 09:46:21 -0500
Message-ID: <45E2F2BD.9080508@cisco.com>
Date: Mon, 26 Feb 2007 09:46:21 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: "Michael Hammer (mhammer)" <mhammer@cisco.com>
Subject: Re: [Sipping] Service Identification in SIP
References: <61968779B8AC4C4BAB421D4C12F008C009152056@XCH47YKF.rim.net>
	<072C5B76F7CEAB488172C6F64B30B5E302A2F204@xmb-rtp-20b.amer.cisco.com>
In-Reply-To: <072C5B76F7CEAB488172C6F64B30B5E302A2F204@xmb-rtp-20b.amer.cisco.com>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 14:46:21.0541 (UTC)
	FILETIME=[E17D2950:01C759B4]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=23951; t=1172501186;
	x=1173365186; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20
	|To:=20=22Michael=20Hammer=20(mhammer)=22=20<mhammer@cisco.com>;
	bh=3fzqfOyxJGlRbh+Q3eg5mq4dtPCFaT5t8DPVYPcV2/Q=;
	b=cmgiwtwVNk4DHrA2usxGHXNJxS7bId4qpE284tNN+rJ4ljhm6Ide+0jZg18g2ziz9Ce/H1TV
	HXF0OJEyEN6wWSP5EYCg7XSlVUXMDQ4s6wljtYsCj6944G/Bk2+PMeW4;
Authentication-Results: rtp-dkim-1; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 2af167865907217f0b49c659e31a77f7
Cc: Andrew Allen <aallen@rim.com>, sipping@ietf.org,
	"Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

My fear is that the overall experience will be exceptionally poor, even 
if the endpoints interoperate.

As the 3gpp requirements indicate, they want to drive lots of different 
things from this service identifier. For example, QoS treatment, billing 
and charging, invocation of network applications, features, and so on. 
My worry is that, if 3gpp defines a service for, say, "messaging", and 
it makes use of a service tag, and a non-3gpp endpoint calls a 3gpp 
endpoint with MSRP and without such a service tag, even if you get the 
two endpoints to connect, the overall experience for both users will be 
bad because the network QoS treatment will be inadequte, the 3gpp user 
won't be billed correctly, supplementary features won't work (say, 
message logging or chat history), and so on.

The notion that I need to have, for each and every call, some kind of 
"service identifier" in the message, in order to successfully process a 
call (where successfully includes everything that needs to be done for 
correct user experience), is anathema to the very notion of SIP, IP, and 
internetworking. It is nothing short of complete balkanization and 
segmentation of communications systems into non-interoperable islands 
that will substantially and permanantly raise the cost and complexity of 
IP communications.

Despite my dire warnings about the end of civilization as we know it, if 
you really want to go off and build totally isolated communications 
networks, then we have a mechanism for it - its called Require. Use it. 
Sparingly.

-Jonathan R.


Michael Hammer (mhammer) wrote:

> I would hope that if they are hell-bent on doing this, that they at
> least also define a service identifier that means "IETF = service
> intentionally left unspecified", but not in so many characters.  Then
> make sure that compatibility is specified when such an identifier is
> encountered, such as when an INVITE were to come from a non-3GPP system
> with no service identity provided.
> 
> Mike
> 
> 
> 
>>-----Original Message-----
>>From: Andrew Allen [mailto:aallen@rim.com] 
>>Sent: Wednesday, February 21, 2007 10:05 PM
>>To: Stephen Terrill (MI/EEM); sipping@ietf.org
>>Cc: Andrew Allen
>>Subject: RE: [Sipping] Service Identification in SIP
>>
>>
>>Hi Steve
>>
>>In my view this is one of the interoperability issues with 
>>your proposal. I think the assumption that all networks will 
>>deploy exactly the same services with the same service 
>>identifiers is unrealistic and also undesirable. Are 
>>corporate PBX vendors going to implement the same services as 
>>defined by 3GPP or OMA? The possibility to have the 
>>flexibility to have different service identifiers for session 
>>equivalent services is in my view essential to any IETF 
>>standard and to do otherwise is to create a situation where 
>>basic services like telephony or messaging cannot 
>>interoperate between users in different networks and users in 
>>the internet without the use of inter working gateway. SIP 
>>and SIMPLE were setup specifically to create protocols that 
>>would enable interoperability with IP networks without the 
>>need for gateways. 
>>
>>If 3GPP network operators choose to deploy the same services 
>>and use common identifiers that is fine but there are whole 
>>other worlds out there as well and service identifiers should 
>>not stand in the way of interoperability between all these 
>>different SIP users. If the IETF solution supports the 
>>ability for different networks to use different identifiers 
>>that doesn't violate 3GPP requirements and doesn't prevent 
>>3GPP using the same identifiers in different networks. I 
>>don't see how that is against TS 23.228.
>>
>>Andrew
>>
>>
>>-----Original Message-----
>>From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
>>Sent: Monday, February 19, 2007 6:00 AM
>>To: Andrew Allen; sipping@ietf.org
>>Subject: RE: [Sipping] Service Identification in SIP
>>
>>Hi Andrew,
>>
>>I see that in your assumptions, the originating and 
>>terminating networks use different service identifiers, and I 
>>presume that this is for the same service.  In my 
>>understanding, it is the same service identifier that is used 
>>end-to-end (if not it is interworked).  That is why I do not 
>>see the reliance on Presence being necessary.  In my opionion 
>>this totally within the requirements of TS 23.228 (I think 
>>that different service identifiers in different networks for 
>>the same service is against TS 23.228).
>>
>>This comes to the isssue of the PSTN gateway - I presume that 
>>you want to go to the PSTN gateway if the service is 
>>telephony, and within that, voice.  If we know that, then the 
>>we go to any PSTN gateway, and we can't presume a particular 
>>network for a PSTN gateway. 
>>
>>Best Regards,
>>
>>Steve.
>>
>>-----Original Message-----
>>From: Andrew Allen [mailto:aallen@rim.com]
>>Sent: Saturday, February 17, 2007 12:51 AM
>>To: Stephen Terrill (MI/EEM); sipping@ietf.org
>>Cc: Andrew Allen
>>Subject: RE: [Sipping] Service Identification in SIP
>>
>>
>>
>>Hi Steve,
>>
>>My responses inline
>>
>>Andrew 
>>
>>-----Original Message-----
>>From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]
>>Sent: Friday, February 16, 2007 3:51 AM
>>To: sipping@ietf.org
>>Subject: RE: [Sipping] Service Identification in SIP
>>
>>Hello Andrew,
>>
>>Thanks for your comments.  I have a couple of brief responses below.
>>
>>Best Regards,
>>
>>..//Steve 
>>
>>-----Original Message-----
>>From: Andrew Allen [mailto:aallen@rim.com]
>>Sent: Friday, February 16, 2007 7:15 AM
>>To: Stephen Terrill (MI/EEM); sipping@ietf.org
>>Cc: Andrew Allen
>>Subject: RE: [Sipping] Service Identification in SIP
>>
>>
>>Some comments on Stephen's comments on the draft:
>> 
>>
>>-Moving onto the proposed solution draft itself, in addition for
>>providing for the identification of the "application" and the
>>"communication service", I have a few additional thoughts.
>>
>>The draft seems to have the implication that all services 
>>(communication
>>services in other terminology) that are to be identified have to be
>>standardised within the IETF.  This is clearly an restriction 
>>that will
>>not be acceptable to the industry as a whole.  My guess is 
>>that if such
>>a statement is mandated by the IETF, it will just be words on paper
>>which would likely be ignored in practice by other standardisation
>>organisations.
>>
>>[AA] This I think is an incorrect interpretation of the draft. My
>>understanding is that the only services according to the draft that
>>require some standardisation in IETF are those that by their 
>>very nature
>>cannot interoperate unless there is some common additional 
>>understanding
>>between the endpoints outside of what is currently specified 
>>in SIP. The
>>use of SIP option tags with Require header is proposed by the 
>>draft for
>>this purpose as such services create interoperability issues with SIP.
>>
>><Steve>
>>My comment applies even for these services </Steve>
>>
>>
>><AA>
>>Well then we are in big trouble because what you are proposing is that
>>its OK for bodies outside of IETF to create non interoperable 
>>extensions
>>for the behavour of SIP entities. This will result in the 
>>fragmentation
>>of SIP into different flavors that don't interoperate. This is the
>>balkanization of SIP that Paul talks about.even <AA> 
>>
>>----------------------------------------------------
>>
>>The draft also only recognises the case where all explicitly knowledge
>>of the service is required.  I think it would also be good to 
>>allow the
>>case where knowledge of the service would be helpful, but not 
>>explicitly
>>required.  This could be a case where the service to the 
>>end-user would
>>not be degraded if this was not supported.  This 
>>functionality seems to
>>be missing.
>>
>>[AA] Service URIs are defined in the draft for these services 
>>where the
>>support by the other endpoint is not required for the session to be
>>sucessfully established.
>>
>>The draft seems to indicate that that application servers in 
>>the network
>>are involved only at the request of the originating terminal.  This
>>seems a strange restriction.  For some network support that 
>>supports the
>>originating user, such as the one provided in the example "I 
>>would like
>>this call recorded for later retrieval", such mechanisms may 
>>make sense,
>>however as a general approach it seems a bit limited.  For example, in
>>3GPP networks there is the need to invoke application logic in the
>>originating side and terminating side, not only at the request of the
>>originating terminal, but as a consequence of the requested 
>>service.  If
>>there is an SIP message contains indication of the requested 
>>service, a
>>proxy can do this (or loose route to an AS that contains the necessary
>>service logic). 
>>
>>[AA] The draft supports both possibilities using the Route header for
>>originating network servers that need to be involved in the route and
>>based on the terminating networks routing policy for the 
>>Service URI in
>>the Request URI.
>>
>><Steve>
>>I don't see the advantage of having different solutions for the
>>originating network and the terminating network.  I don't 
>>think that the
>>originating logic is invoked only at the request of the originating
>>terminal.
>></Steve>
>>
>><AA>
>>Unless IETF is willing to standardise all services and 
>>mandate that all
>>SIP entities understand these (which will never happen) then we cannot
>>assume that all networks and all endpoints understand and use the same
>>identifiers. It is possible through configuration for an endpoint to
>>obtain the identifiers used by its own service provider but 
>>there is no
>>way to ensure that other networks use or understand the same 
>>identifiers
>>for the equivalent or similar service that is compatible at 
>>the session
>>layer. Therefore we need to support the possibility to have different
>>identifiers in different networks in order to not violate the 
>>TS 23.228
>>requirement:
>>
>>12.	The usage of communication service identifiers shall not
>>adversely affect interoperability between IMS networks and
>>interoperability with external SIP networks and CS networks. The
>>behaviour of a network receiving the IMS requests without an IMS
>>communication service identifier is a matter of operator policy. Usage
>>of communication service identifiers shall not decrease the level of
>>interoperability with networks and UEs that are unaware of the
>>communication service identifier.
>>
>>If the originating service logic isn't invoked based on an identifier
>>placed in the request by the originating terminal on what basis do you
>>think it is invoked? How does the use of Accept-Contact have 
>>advantages
>>in this case?
>><AA> 
>>
>>--------------------------------------
>>
>> I note here that the proposed service URN does not contain 
>>the address
>>of a specific server containing the application logic.  An AS can look
>>at that, and loose route to an application server that does 
>>contain such
>>logic (If it is separated from the proxy), however I fail to see the
>>difference for that proxy if it looks at a feature tag, or it looks at
>>the service URN for this.  In both cases, it has to parse the message
>>for the data, and modify the routing information.  
>>
>>[AA]Route headers are the SIP mechanism for requesting 
>>intermediaries to
>>be includede on the route of the request. There is nothing in RFC
>>3840/RFC 3841 that addresses using feature tags in 
>>Accept-Contact header
>>for any routing decisions prior to determining which of the registered
>>contacts the request should be forked to. Accept-Contact is 
>>not a Route
>>header it is merely a filter that is applied to the set of valid
>>contacts. Using a parameter in the filter as a service 
>>identifier has a
>>potential significant impact on the use of Accept-Contact for its
>>original purpose. 
>>
>><Steve>
>>I don't believe there is anthing that states a Proxy CAN NOT route to
>>applications or application logic based upon elements in the SIP
>>message, including what is in the Accept-Contact header (some people
>>have even proposed SDP ...).
>></Steve>
>>
>><AA>
>>As I had stated in an earlier email on this subject a proxy 
>>can do a lot
>>of things (including routing based on your favorite number 
>>appearing in
>>a call-Id) but that doesn't mean they are a good idea or a good design
>>decision. If you want to ensure you don't break things then 
>>you need to
>>be careful to use SIP headers with the semantics intended for them
>>otherwise problems can result if extra semantics are put upon existing
>>headers.
>><AA> 
>>
>>Further, the route header could be removed in the originating network;
>>and it is still required to identify the service in the terminating
>>network.  As proxies may remove the route, it can't be used 
>>to transport
>>the requested service.  
>>
>>[AA] Proxies arbitarily removing Route headers is a BAD thing 
>>akin to a
>>bull in a china shop - almost certainly going to break something. RFC
>>3261 summarises what proxies should be doing with Route headers:
>>16.12 Summary of Proxy Route Processing
>>
>>   In the absence of local policy to the contrary, the processing a
>>   proxy performs on a request containing a Route header field can be
>>   summarized in the following steps.
>>
>>      1.  The proxy will inspect the Request-URI.  If it indicates a
>>          resource owned by this proxy, the proxy will replace it with
>>          the results of running a location service.  Otherwise, the
>>          proxy will not change the Request-URI.
>>
>>      2.  The proxy will inspect the URI in the topmost Route header
>>          field value.  If it indicates this proxy, the proxy 
>>removes it
>>          from the Route header field (this route node has been
>>          reached).
>>
>>      3.  The proxy will forward the request to the resource indicated
>>          by the URI in the topmost Route header field value or in the
>>          Request-URI if no Route header field is present.  The proxy
>>          determines the address, port and transport to use when
>>          forwarding the request by applying the procedures in [4] to
>>          that URI.
>>
>>   If no strict-routing elements are encountered on the path of the
>>   request, the Request-URI will always indicate the target of the
>>   request.
>>
>><Steve>
>>The first sentance is interesting "In absence of local policy to the
>>contrary", indicating that there can be different policies.  I
>>understand that some proxies do remove this.  While I am on this - I
>>have lead to believe that R-URI parameters may also be removed by
>>intermediate proxies - is this correct?
>></Steve>
>>
>><AA>
>>My understanding is that local policy may determine that additional
>>nodes need to be routed to in addition to the Route contained in the
>>request. However deleting existing Route headers other than the Route
>>header containing the address of the proxy as part of the 
>>loose routing
>>procedures is liable to break things and is not part of RFC 
>>3261 routing
>>(loose routing).
>><AA> 
>>
>>
>>[AA] Generally a UA wouldn't know the appropriate service identifiers
>>used in the terminating network or whether a Route header is 
>>appropriate
>>to be used in the terminating network. The draft defines the 
>>Service URI
>>for the purpose of identifying services in the terminating network and
>>service URIs can be obtained using Presence.
>>
>><Steve>
>>I don't get the above paragraph.  It suggest that the "appropriate
>>service identifier" used in the terminating network is not 
>>known by the
>>originating UE, but the Service URI is.  I thought that the 
>>service URI
>>is an encoding of the "appropriate service identifier".
>></Steve>
>>
>><AA> 
>>Yes but in the general case the UA doesn't need to know what 
>>the service
>>identifier is it obtains through Presence a Service URI that contains
>>the Service identifier as a URI parameter. The great thing about using
>>Presence to obtain a Service URI is that the originating UA 
>>doesn't even
>>need to know anything about service identifiers at all for 
>>this to work.
>>It just uses Presence to find the URI to use for the service which can
>>be indicated to the user via an indication such as a graphic 
>>icon (e,g a
>>phone for telephony, smiley face for IM, etc) in the presence 
>>document.
>><AA> 
>>
>>-----------------------------------
>>
>>>From this perspective it is not clear to me the advantage of using a
>>service URN in a route header compared to the use of a feature tag.
>>
>>[AA] I hope the above responses clarify that somewhat. Also 
>>the Service
>>URN and Service URIs allow different service identifiers to be used in
>>the originating and terminating network for session 
>>compatiable services
>>instead of forcing everyone to deploy and understand the same service
>>and service identifier.
>>
>><Steve>
>>Nope, don't get it, but are information that is carried 
>>end-to-end ....
>>And as far as I see, could be used by intermediate proxies to impact
>>local routing policy ... Or ?
>></Steve>
>>
>><AA> 
>>I don't understand this response either. The key point is that UAs in
>>corporate networks and in the internet are not going to deploy 3GPP
>>flavor services and use the same 3GPP service identifiers or even
>>possibly any service identifiers at all. The Accept-Contact based
>>proposal assumes that all endpoints understand the service identifiers
>>to be used for each service. This concept is not generally applicable
>>and create interoperability problems.
>><AA> 
>>
>>The draft seems to mandate GRUU - is this really so, is that really
>>necessary.  It seems strange to place such a dependency, it 
>>also starts
>>to move the GRUU as being the primary means of which to 
>>contact a user.
>>
>>[AA] My understanding is that GRUU in the draft is used to 
>>address how a
>>request addressed to Service URI is forked only to those contacts that
>>support the service identified by the that Service URI. A 
>>contact would
>>seem to be fine so long as it had GRUU properties - no need 
>>to obtain a
>>GRUU if you already have one.
>>
>>This implies that the originating user decides which terminal I should
>>receive my requested service on, even if I have two devices for that.
>>As such, I don't believe that there should be a dependency on GRUU.
>>
>>[AA] If you have obtained a Service URI (probably through 
>>presence) for
>>that user then you can address the request to the Service URI which
>>represents all the terminals that support that service. 
>>However GRUU and
>>the Service Instance URIs created from GRUUs provide another mechanism
>>through which a user can obtain a URI to reach that user for the
>>required service.
>>
>>Parts of the draft seems to rely on the use of presence.  
>>While presence
>>could be a supportive mechanism, I doubt that we can mandate that the
>>number you want to dial has be available in presence as I 
>>assume not all
>>entities are available in presence (diallable network 
>>services, users in
>>the PSTN).  This would be strange to mandated.
>>
>>[AA] PSTN gateways are presumably in the same domain as the 
>>terminating
>>proxy and therefore these PSTN gateways can be configured with the
>>appropriate service identifiers for telephony services and can include
>>this in either a Route header to route to the right application server
>>or can construct a valid Service URI. The real problem is how do users
>>in other networks know what the service identifier that's is 
>>used in the
>>terminating network is. The use of Service URIs and Presence 
>>provides a
>>solution to that problem.
>>
>><Steve>
>>Again, presence could be a useful mechanism, but I think it would be a
>>problem for the draft to rely it.  A user making a call to 
>>the PSTN may
>>not find that user in the presence archicture (and similar to services
>>like the 800 or 900 services).  (I should note, that you 
>>shouldn't make
>>any assumptions about the location of the PSTN gateway ... But I don't
>>think that changes anything).  
>></Steve>
>>
>><AA> 
>>How does a PSTN gateway in another domain from the terminating network
>>know the service identifier for telephony used in the terminating
>>network with the Accept-Contact header based proposal? 
>><AA> 
>>
>>Moving on from the draft I have seen the questioned raised about "why
>>can't we expand the power of SDP".  Such an approach would 
>>mandate that
>>SIP methods use SDP, I don't think that is appropriate. 
>>
>>[AA] I agree with you here
>>
>>I hear the appeals that the mechanisms should not be abused, 
>>and that we
>>should not have 1000s of separate "communication services", 
>>and I fully
>>agree with this. I see a smaller number of communication 
>>services, not a
>>larger number; where those communication services are used to access
>>other end-user-services.
>>
>>*flame away*
>>
>>Regards,
>>
>>Steve
>>
>>---
>>
>>
>>
>>---------------------------------------------------------------------
>>This transmission (including any attachments) may contain confidential
>>information, privileged material (including material protected by the
>>solicitor-client or other applicable privileges), or constitute
>>non-public information. Any use of this information by anyone 
>>other than
>>the intended recipient is prohibited. If you have received this
>>transmission in error, please immediately reply to the sender 
>>and delete
>>this information from your system. Use, dissemination, 
>>distribution, or
>>reproduction of this transmission by unintended recipients is not
>>authorized and may be unlawful.
>>
>>
>>
>>---------------------------------------------------------------------
>>This transmission (including any attachments) may contain 
>>confidential information, privileged material (including 
>>material protected by the solicitor-client or other 
>>applicable privileges), or constitute non-public information. 
>>Any use of this information by anyone other than the intended 
>>recipient is prohibited. If you have received this 
>>transmission in error, please immediately reply to the sender 
>>and delete this information from your system. Use, 
>>dissemination, distribution, or reproduction of this 
>>transmission by unintended recipients is not authorized and 
>>may be unlawful.
>>
>>_______________________________________________
>>Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>This list is for NEW development of the application of SIP
>>Use sip-implementors@cs.columbia.edu for questions on current sip
>>Use sip@ietf.org for new developments of core SIP
>>
> 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 10:01:10 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLhL0-0008KJ-10; Mon, 26 Feb 2007 10:00:26 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLhKy-0008KE-OI
	for sipping@ietf.org; Mon, 26 Feb 2007 10:00:24 -0500
Received: from rtp-iport-1.cisco.com ([64.102.122.148])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLhKx-0004h0-F5
	for sipping@ietf.org; Mon, 26 Feb 2007 10:00:24 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by rtp-iport-1.cisco.com with ESMTP; 26 Feb 2007 10:00:24 -0500
X-IronPort-AV: i="4.14,221,1170651600"; 
	d="scan'208"; a="53682919:sNHT46872136"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1QF0NLq006733
	for <sipping@ietf.org>; Mon, 26 Feb 2007 10:00:23 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1QF0MVj008555
	for <sipping@ietf.org>; Mon, 26 Feb 2007 10:00:23 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 10:00:18 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-202.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 10:00:18 -0500
Message-ID: <45E2F602.40802@cisco.com>
Date: Mon, 26 Feb 2007 10:00:18 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: IETF Sipping List <sipping@ietf.org>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 15:00:18.0729 (UTC)
	FILETIME=[D47DE990:01C759B6]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=2311; t=1172502023;
	x=1173366023; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Service=20identification=20-=20clarifying=20draft-rosenberg-s
	ipping-service-identification |Sender:=20
	|To:=20IETF=20Sipping=20List=20<sipping@ietf.org>;
	bh=WevU3JDytaZdeDp9vCqvneEXrerTIIqZvDSlRn6zRag=;
	b=G183sRAiokBERd7FqjCT4PpJqPThs92+2ojnkk+YC9sOvLn80u83Cx23b4jvkotrz/cG8My1
	LYeYho2CdlmnlCfpH5OAoF58CYOJ+lNxJwueAvSnpzyTZKXEXGqawBOH;
Authentication-Results: rtp-dkim-2; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 1.1 (+)
X-Scan-Signature: cab78e1e39c4b328567edb48482b6a69
Subject: [Sipping] Service identification - clarifying
	draft-rosenberg-sipping-service-identification
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

The threads on this topic are getting long and intertwined. So, I wanted 
to send a few notes off on this topic.

First and foremost, I must apologize that my draft is far from crisp on 
the points its trying to make, and it gets obscured in what is 
admittedly an extremely contrived example.

The draft is trying to make several key points:

1. The right answer to this whole mess is to infer what the service is 
from what is actually signaled in SIP through standards-based 
extensions. If an INVITE comes with SDP that includes audio and video 
m/c-lines, then you know that this is a multimedia call. It is the usage 
of standard sip techniques, all of which are designed to work in 
heterogeneous environments, that gives you the ability to work between 
different domains, different operators with different endpoints and 
different billing plans.

2. In rare cases, you cannot accomplish #1. Or, you don't want to. Or, 
you've already designed a service which won't work without shared 
knowledge of the 'service' on each side. In these cases, SIP provides a 
mechanism for making sure that the other side supports your special 
context - Require. So, use it. But try not to, since it leads to 
non-interoperability.

3. Another issue, which is quite different IMHO, is that we have cases 
where an originating UA would like to have additional servers inserted 
into the request path for specific functions. FOr example, I, as a 
caller, want to get this recording server into this call. I want to make 
that decision on a call by call basis. My proposal for this is to use a 
Route header for a service URN. This is NOT the same problem as service 
identification since it doesn't have any of the horrible 
interoperability problems we've been discussing; the caller and called 
party don't need to agree on, or know about, this thing. Its merely a 
matter of agreement between me and my provider about what gets included.

HTH,
Jonathan R.
-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 10:06:14 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLhQc-0004YL-8A; Mon, 26 Feb 2007 10:06:14 -0500
Received: from [10.90.34.44] (helo=chiedprmail1.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLhQa-0004Y4-JZ
	for sipping@ietf.org; Mon, 26 Feb 2007 10:06:12 -0500
Received: from rtp-iport-2.cisco.com ([64.102.122.149])
	by chiedprmail1.ietf.org with esmtp (Exim 4.43) id 1HLhQV-0001OS-Is
	for sipping@ietf.org; Mon, 26 Feb 2007 10:06:12 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by rtp-iport-2.cisco.com with ESMTP; 26 Feb 2007 10:06:07 -0500
X-IronPort-AV: i="4.14,221,1170651600"; 
	d="scan'208"; a="114563012:sNHT79471242"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1QF67kq017180; 
	Mon, 26 Feb 2007 10:06:07 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1QF5vOM006507; 
	Mon, 26 Feb 2007 10:06:07 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 10:06:01 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 10:06:01 -0500
Message-ID: <45E2F759.5030302@cisco.com>
Date: Mon, 26 Feb 2007 10:06:01 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Stephen Terrill (MI/EEM)" <stephen.terrill@ericsson.com>
Subject: Re: [Sipping] Service Identification Model
References: <989763567E51AB48BE015DF7BB960CAD02925D3B@esealmw111>
In-Reply-To: <989763567E51AB48BE015DF7BB960CAD02925D3B@esealmw111>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 15:06:01.0541 (UTC)
	FILETIME=[A0D2DF50:01C759B7]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=22076; t=1172502367;
	x=1173366367; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20Model
	|Sender:=20
	|To:=20=22Stephen=20Terrill=20(MI/EEM)=22=20<stephen.terrill@ericsson.com
	>; bh=bti/+6LlmGOe6+TH7acwKVkjQLGNWn9XX3J9weJvOH8=;
	b=O/VXpBOZOMCiGEsKTOJEaMpnWF9Ooq7YrWNsdxUvedADFsnxCCEx/wvTF4o6+US03vwW8w4B
	Kvezv4v2yevuASEzsfxECFz8W83g+C++WvJXznRDxJSLvdz/8yJflJ1H;
Authentication-Results: rtp-dkim-1; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 22f8e36c8d8be0bcbb9bf02fb6ce7335
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Stephen Terrill (MI/EEM) wrote:
> Hi,
> 
> Just a comment on the Messaging here;  While I fully understand the view
> that isntant messaging is another media, I also understand that the
> evolution of messaging in OMA is that instant messaging is actally
> defined as a seperate service to telephony.  SMS->MMS->Instant
> messaging.

There's that "S" word again!  In this case, by "different service" do 
you mean "different UI", or "different billing model", or "different 
implementation model"?

I can understand that user expectations will be for a different UI, 
though I think providers who can support both the traditional UIs and a 
unified UI will be more popular in the long run.

I can also understand "different billing model", although I am afraid 
what that really means is an intent to gouge the user for more $$$ than 
is justified by what the provider actually provides. But I am not 
convinced anything special is needed to support whatever billing model 
is desired.

Its the "different implementation model" that is most distressing. The 
users don't care about the implementation model, so any history in that 
regard is irrelevant.

Regarding the UI, I see the following key differences:
- telephony traditionally uses an "alert first" model, and exchanges
   media only after the recipient as accepted the incoming call.
   This is presumed to be applicable to video as well.

- in contrast, with traditional UIs for IM, presence may be
   used as a hint whether the target will accept the message,
   and the first message is sent without any *guarantee* that it
   will be accepted, or seen immediately. Depending on which IM
   model you adhere to, the message may be rejected if the recipient
   isn't ready to receive, or it may be stored for later delivery.
   With SMS/MMS and page mode SIP messaging, every message is a
   "first" message, but this may be disguised with a session oriented
   UI, or not. On the receiving side, there may be an implicit "auto
   answer" that accepts the message and displays it without alerting.
   Or there may be an "alert" that messages are present, but viewing
   them is more like viewing emails.

- The UI for PoC is in many ways similar to IM.

It is clear that people would be upset if their familiar UI paradigms 
were changed. The question is whether that has anything to do with the 
signaling. Ideally this would be strictly a local issue for a UA to 
implement.

In the case of page mode SIP messaging, using MESSAGE out of dialog, 
this is straightforward. MESSAGE can also be used in-dialog, though the 
occasions when this is appropriate are not so clear. From a UI 
perspective, having a separate UI for page mode IM vs voice is obviously 
straightforward. Some innovation in UI design is required to handle page 
mode IM within a dialog set up for other reasons - i.e. voice. But no 
"service" is needed in either case. A UI that doesn't support 
origination of page mode messages in-dialog could nevertheless support 
them on receipt simply by treating them the same as out-of-dialog messages.

IMO there remains work to do on session mode IM using MSRP, in terms of 
mapping it to expected UIs. Mapping to the model where there separate 
alerting before the first message is delivered is not straightforward. 
It could be done in at least a couple of ways:
- the UI could be set up to "auto answer" INVITEs with MSRP,
   and then present the first message in the traditional way.
   The sender would collect the first message before sending
   the invite, and then send it as soon as an MSRP session is
   established.

- the first message could be sent as "early media", and that
   early media could be displayed in lieu of some other form of
   alerting. Sending a response from the UI could be treated as
   an indication to answer the incoming call.

My point here is that there is work to be done, but I don't see why 
there is need for "service identification" to make it work. In fact, the 
similarity between PoC and IM on one hand, and PoC and voice on the 
other hand, suggests that there is a lot of value in figuring out how to 
unify these rather than partitioning them.

After all, who here hasn't been having an IM session and wanted to 
expand it to a voice session as well? This is natural in SIP if the two 
media are not partitioned into different "services".

	Paul


When received, the display of the message
   serves as its own "alert".

> Cheers,
> 
> ..//Steve
> 
> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
> Sent: Friday, February 23, 2007 4:12 PM
> To: Byron Campen
> Cc: Andrew Allen; sipping@ietf.org
> Subject: Re: [Sipping] Service Identification Model
> 
> 
> 
> Byron Campen wrote:
>>> 1. Basic Telephony (Conversational QoS Voice only with telephony 
>>> feature
>>> set)
>>> 2. Multimedia Telephony (Conversational QoS Voice plus video and 
>>> other media streams with telephony feature set)
> 
> Why are 1 & 2 different services? 1 is a proper subset of 2.
> 
>>> 3. VoIP (Background QoS with no telephony features)
> 
> And why is 3 different from 1 and 2? Are you saying that to be able to
> use a feature I must decide *before* I make the call that I might want
> to?
> 
> Even if that is the case, that sounds like a policy issue between a user
> and his own provider, rather than an e2e negotiation. (My features are
> provided by my UA and/or my provider.)
> 
>>> 4. SMS
>>> 5. MMS
> 
> Why are 4 and 5 different services? They are both datagrams. The
> difference in media types is irrelevant. If its delivered via MESSAGE,
> then it is already distinguishable from calls. This should be able to be
> part of the same service as IM.
> 
>>> 6. Instant Messaging
> 
> Is just another media type. Like voice and video, these can be
> negotiated for the session. Why balkanize as a different service?
> 
>>> 7. Push to Talk (PoC)
> 
> We've beaten this one to death already. Current implementation forces
> need for essentially an overlay network to support it. But its far from
> an exemplar of how to do it.
> 
>>> 8. Network Game 1 (e.g Chess)
>>> 9. Network Game 2 (e.g Martial arts game) 10. Network Game 3 (e.g 
>>> Dungeons & Dragons)
> 
> I tend to agree with Eric that service URIs are good for these if there
> is no better way. But in practice I think these could easily be
> distinguished in SDP using o/a.
> 
>>> 11. Push Service 1 (e.g. Traffic information service) 12. Push 
>>> Service 2 (e.g. Weather information service) 13. Push Service 3 (e.g
> 
>>> News information service) 14. Push Service 4 (e.g  Stock information 
>>> service)
> 
> There may be *something* here. I need to know more about the
> characteristics of these. If they consist of human intelligible media,
> then I think they could adequately be dealt with via callerprefs. (This
> may be a bit like the Actor feature tag.) Specifically, callerprefs are
> fine if its ok that the call end up elsewhere if the callerprefs
> selection doesn't happen or there isn't any target with that capability.
> 
> If the information isn't intelligible media, then it ought to be
> distinguishable as a special media type in the offer.
> 
>>> 15. Mobile TV service
> 
> AFAIK, the station doesn't call me to deliver me tv service. I would be
> initiating the call. And presumably it will be to a target service URI
> or URN. So I see no need for a service id here.
> 
>>> So there's 15 which may all be different applications on a handset 
>>> and all those are existing services or those already envisioned.
> 
> So I only see:
> 
> - 1 for basic telephony, multimedia telephony, SMS, MMS, IM
> 
> - *maybe* something for games
> 
> - *maybe* something for push services
> 
> 	Paul
> 
>>     I only see 10 here that would need to ever send a register
> request. 
>> (11-15 are strictly client-server) 4-6 will have distinct Content-Type
> 
>> headers and payloads. You would need only 1 UA for these. So we're 
>> down to 8. 8-10 will, once again, be distinct when you look at the 
>> payload and Content-Type. So we're down to 6. There is no reason at 
>> all that the UA that handles voice is incapable of handling IM, and 
>> you can tell the difference by looking at the request type, so we're 
>> down to 5. As for the voice stuff, you can distinguish 2 from anything
> 
>> else through the payload. So, we could quite easily roll 1-2 together.
> 
>> (4) 3 can be distinguished by watching for any in-dialog call-control 
>> stuff. (Having a service identifier that essentially says "I promise 
>> not to use any fancy telephony features on this call. No, really, I 
>> promise!" is never going to work) So we're down to 3.
>>
>>     SIP is more than flexible enough to handle all of this stuff in a 
>> single UA. The UAs are perfectly capable of figuring out what
> "service"
>> the other wants to use; why are your middleboxes unable to do the
> same? 
>> Adding a parallel method to make things easy on the middleboxes is 
>> just asking for trouble, because you're conveying the same information
> twice.
>> It is ripe for abuse. Of course, the easy way to do all this is to 
>> just let UAs send what they want, and charge per KB, but hey, what do 
>> I know about simplicity, flexibility, scalability, ease-of-deployment,
> 
>> and customer satisfaction? I'm just a IETFer.
>>
>> Best regards,
>> Byron Campen
>>
>>
>>> -----Original Message-----
>>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
>>> Sent: Thursday, February 22, 2007 7:05 PM
>>> To: Andrew Allen
>>> Cc: Eric Burger; sipping@ietf.org
>>> Subject: Re: [Sipping] Service Identification Model
>>>
>>>
>>>
>>> Andrew Allen wrote:
>>>> Hi Eric,
>>>>
>>>> It seems your prefer the approach of using URIs to identify services
> 
>>>> which is part of Jonathan's draft as is the use of Require for those
> 
>>>> services which simply do not interoperate unless the UAS understands
>>> the
>>>> service.
>>>>
>>>> Where I am concerned is that you seem to be suggesting that the UA 
>>>> separately register an AoR for each service it supports using
>>> REGISTER.
>>>> Certainly for cellular that wouldn't be acceptable as 10 or more 
>>>> services could easily be supported by a single device.
>>> Andrew, can you give me an example of 10 "services" on the device 
>>> that need to be distinguished from one another in this way - that 
>>> aren't just
>>>
>>> optional behavior of a particular service? (E.g. I see no reason why 
>>> "audio" and "audio/video" need to be separate services.)
>>>
>>> I guess as things stand, PoC must be a separate service. And 
>>> apparently Chess. What else?
>>>
>>>     Paul
>>>
>>>> Registrations
>>>> create a significant load in the IMS network because its not just a
>>> SIP
>>>> REGISTER but also requires a lot of interactions with other network 
>>>> elements such as the HSS and Notifications for reg-events etc. Also 
>>>> having an order of magnitude increase in multiple registrations is
>>> going
>>>> to be a big issue for the avalanche restart problem for example when
> 
>>>> a cell site goes down and comes back into service.
>>>>
>>>> I think Jonathan's approach gives us the possibility for the 
>>>> registrar to create the Service URIs based on the feature parameters
> 
>>>> included in the Contact of the REGISTER.
>>>>
>>>> I think the argument over whether the model is one UA with multiple 
>>>> services or multiple UAs each representing a separate service is 
>>>> somewhat academic. As I understand a UA can register multiple AoRs 
>>>> and can have other AoRs configured or otherwise provided by the 
>>>> service provider.
>>>>
>>>> Andrew
>>>>
>>>> -----Original Message-----
>>>> From: Eric Burger [mailto:eburger@bea.com]
>>>> Sent: Wednesday, February 21, 2007 1:29 PM
>>>> To: sipping@ietf.org
>>>> Subject: [Sipping] Service Identification Model
>>>>
>>>> Folks have been pestering me for a while to look at the Service 
>>>> Identification issue.  I finally have gotten around to it.
>>>>
>>>>
>>>> I guess I have a fundamental problem with the principal assertion of
>>> the
>>>> service-identification draft.  Namely, the assertion that there is a
> 
>>>> single SIP Layer, separate and distinct from two services.  Here is 
>>>> Figure 2 from the draft:
>>>>
>>>>                            +---------------------------------+
>>>>                            |                                 |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | |      UI     | |      UI     | |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | |             | |             | |
>>>>                            | |  Service 1  | |  Service 2  | |
>>>>                            | |             | |             | |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | +-----------------------------+ |
>>>>                            | |                             | |
>>>>                            | |             SIP             | |
>>>>                            | |            Layer            | |
>>>>                            | |                             | |
>>>>                            | +-----------------------------+ |
>>>>                            |                                 |
>>>>                            +---------------------------------+
>>>>
>>>>                                      Physical Device
>>>>
>>>> I would offer this drawing confuses a common SIP Stack with a common
> 
>>>> User Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use 
>>>> case:
>>>>
>>>>> ... In addition users may have
>>>>> multiple devices only one of which has the application that
>>>> understands
>>>>> the game. Other devices that the user has registered against his 
>>>>> AoR
>>>> may
>>>>> handle audio for telephony and other voice applications but if 
>>>>> Alice invites Bob to play her favourite game it is not very 
>>>>> interesting if
>>>> Bob
>>>>> accepts the gaming session on a device that can only accept the 
>>>>> audio stream. Also even when the terminal that has the game 
>>>>> application receives the INVITE the correct application in that 
>>>>> terminal needs to
>>>> be
>>>>> identified to accept the session.
>>>> I would offer that this is the clinching argument that the request 
>>>> is, in fact, addressing a SEPARATE USER AGENT, not a special SERVICE
> 
>>>> at a
>>> User
>>>> Agent.
>>>>
>>>> Same argument for PoC versus a regular voice call: One is looking 
>>>> for
>>> a
>>>> PoC
>>>> User Agent, NOT the PoC service at a given User Agent.
>>>>
>>>> Thus, the picture really looks like the following:
>>>>
>>>>                            +---------------------------------+
>>>>                            |                                 |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | |    UI       | |    UI       | |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | |             | |             | |
>>>>                            | |    UAS 1    | |    UAS 2    | |
>>>>                            | |             | |             | |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | +-----------------------------+ |
>>>>                            | |                             | |
>>>>                            | |             SIP             | |
>>>>                            | |            Stack            | |
>>>>                            | |                             | |
>>>>                            | +-----------------------------+ |
>>>>                            |                                 |
>>>>                            +---------------------------------+
>>>>
>>>>                                      Physical Device
>>>>
>>>>
>>>> Another argument for different User Agents is the whole "route based
>>> on
>>>> SDP"
>>>> argument.  As the author of RFC 3458, I think I was first to make 
>>>> the assertion that knowing the media type did NOT mean you know the 
>>>> application to use.  I don't want your phone or proxy or B2BUA to 
>>>> guess that
>>> because
>>>> I
>>>> have Chess Media, I want the Chess Application; I want to talk with
>>> your
>>>> Chess Application.
>>>>
>>>> Now, is the Chess Application a Preference, Requirement, or the 
>>>> thing that I am addressing?  The answer drives one towards using 
>>>> caller prefs, the Require header, or the Request-URI.
>>>>
>>>> I would offer that it is not a preference: I really want the chess 
>>>> application; I probably do not want to talk to you.
>>>>
>>>> That leaves us with the Require header or a Request-URI.  My 
>>>> leanings are not surprisingly towards the RFC 4240 approach.  
>>>> Frankly, if you are building an IMS (as an example) that can handle 
>>>> 10,000,000
>>> subscribers,
>>>> you
>>>> should be able to handle 100,000,000 registrations.  That order of 
>>>> magnitude is way easier than the 10,000 to 10,000,000 jump.  I don't
> 
>>>> buy the "registering each service is too much of a burden on the 
>>>> operator"
>>>> argument,
>>>> because it is not that much of a burden, and the network actually 
>>>> does the right thing.  Better yet, all of the corner cases work.  
>>>> Best of all, the IETF doesn't have to build an IANA Registry and 
>>>> process a thousand PS documents to register urn:service:myco:widget.
>>>>
>>>> I suppose I could live with Require:, only if you are way too lazy 
>>>> to register each application instance.  It does semantically work,
>>> because
>>>> if I
>>>> want to do PoC and you do not do PoC, I'll get a 4xx response, to
>>> which
>>>> I
>>>> could try a regular call.
>>>>
>>>>
>>> _____________________________________________________________________
>>> __
>>>> Notice:  This email message, together with any attachments, may
>>> contain
>>>> information  of  BEA Systems,  Inc.,  its subsidiaries  and
>>> affiliated
>>>> entities,  that may be confidential,  proprietary,  copyrighted
>>> and/or
>>>> legally privileged, and is intended solely for the use of the
>>> individual
>>>> or entity named in this message. If you are not the intended
>>> recipient,
>>>> and have received this message in error, please immediately return
>>> this
>>>> by email and then delete it.
>>>>
>>>> _______________________________________________
>>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>>> This list is for NEW development of the application of SIP Use 
>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>> sip@ietf.org for new developments of core SIP
>>>>
>>>>
>>>>
>>>> --------------------------------------------------------------------
>>>> - This transmission (including any attachments) may contain 
>>>> confidential
>>> information, privileged material (including material protected by the
> 
>>> solicitor-client or other applicable privileges), or constitute 
>>> non-public information. Any use of this information by anyone other 
>>> than the intended recipient is prohibited. If you have received this 
>>> transmission in error, please immediately reply to the sender and 
>>> delete this information from your system. Use, dissemination, 
>>> distribution, or reproduction of this transmission by unintended 
>>> recipients is not authorized and may be unlawful.
>>>> _______________________________________________
>>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>>> This list is for NEW development of the application of SIP Use 
>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>> sip@ietf.org for new developments of core SIP
>>>>
>>>
>>>
>>> ---------------------------------------------------------------------
>>> This transmission (including any attachments) may contain 
>>> confidential information, privileged material (including material 
>>> protected by the solicitor-client or other applicable privileges), or
> 
>>> constitute non-public information. Any use of this information by 
>>> anyone other than the intended recipient is prohibited. If you have 
>>> received this transmission in error, please immediately reply to the 
>>> sender and delete this information from your system. Use, 
>>> dissemination, distribution, or reproduction of this transmission by 
>>> unintended recipients is not authorized and may be unlawful.
>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP Use 
>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>> sip@ietf.org for new developments of core SIP
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use
> sip-implementors@cs.columbia.edu for questions on current sip Use
> sip@ietf.org for new developments of core SIP
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 10:10:46 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLhUz-0005i6-Ps; Mon, 26 Feb 2007 10:10:45 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLhUy-0005fl-04
	for sipping@ietf.org; Mon, 26 Feb 2007 10:10:44 -0500
Received: from rtp-iport-1.cisco.com ([64.102.122.148])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLhUw-00065d-MP
	for sipping@ietf.org; Mon, 26 Feb 2007 10:10:43 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by rtp-iport-1.cisco.com with ESMTP; 26 Feb 2007 10:10:43 -0500
X-IronPort-AV: i="4.14,221,1170651600"; 
	d="scan'208"; a="53684070:sNHT48084496"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1QFAgLJ011904
	for <sipping@ietf.org>; Mon, 26 Feb 2007 10:10:42 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1QFAOOk007880
	for <sipping@ietf.org>; Mon, 26 Feb 2007 10:10:42 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 10:10:03 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-201.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 10:10:03 -0500
Message-ID: <45E2F84A.4030602@cisco.com>
Date: Mon, 26 Feb 2007 10:10:02 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: IETF Sipping List <sipping@ietf.org>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 15:10:03.0088 (UTC)
	FILETIME=[30CC0500:01C759B8]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=2769; t=1172502642;
	x=1173366642; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Routing=20vs.=20Require=20in=20service=20identification
	|Sender:=20 |To:=20IETF=20Sipping=20List=20<sipping@ietf.org>;
	bh=0kOy4EQodWnmCEaDwdYJFY/PadmsUI8crAIL0gXJJQc=;
	b=dGiGJN0ZhYuGLA2ffnHrHV5N7dUTdulobk5qJgjbD6XpQxtFKIKUaH+jfLIbzW2MNMfZxtku
	Kl0aNK2voDxjQPZfJzfwjSoW/U1UG7Mk//ElzkMoTH9I+DIKbEhunPvt;
Authentication-Results: rtp-dkim-2; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 538aad3a3c4f01d8b6a6477ca4248793
Subject: [Sipping] Routing vs. Require in service identification
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

An issue that came up a lot in the threads on service identification was 
whether the problem here is just one of routing, or one of extension 
handling.

There is certainly an aspect of this issue here that has a routing 
flavor. If Joe makes a PoC call, he will want that call routed to a PoC 
application server, and then delivered to a PoC handset owned by the 
called party. So, if the called party has two handsets, one that is PoC, 
and one that isn't, then you clearly want the PoC call delivered to the 
PoC handset.

If this were JUST a routing issue, then clearly something like 
Accept-Contact might make sense, since its defined for routing.

However, routing is really a side effect of the bigger issue, which is 
that for this PoC case at least, 3gpp has introduced a non-backwards 
compatible extension to SIP. That extension must be understood in order 
to process a PoC request. That root issue brings the routing 
requirements as a side effect; you need to route to an application 
server since proper processing of the request *requires* it, and you 
want to deliver the PoC request to a PoC terminal since it will only be 
properly processed on that terminal.

When we defined RFC3841, we called these implicit caller preferences. 
RFC3841 defines these only for methods and event packages, but they 
equally well could apply to extensions to. The idea with implicit 
preferences is that other aspects of the request, like the method or 
Require option tags, clearly imply a desire to be connected to an 
endpoint which supports them. That is what is happening here. The 
problem we are trying to solve is NOT how to route to the right 
endpoint; the problem is how do we indicate that special processing is 
needed for this request, and a *side effect* of that is that we wish the 
request to be delivered to an endpoint which can support it.

As a litmus test to demonstrate this, ask yourself this. If you can 
choose putting a tag in Accept-Contact or Require, consider a use case 
where the caller and called party both had just a single device/UA, so 
that your routing decision is easy. If you put the tag in 
Accept-Contact, would the right thing happen? I would assert that it 
would not in the PoC case. This is why PoC needs Require. IF another 
service should be defined that has the same property, it needs Require 
too. But please do that sparingly.

-Jonathan R.


-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 10:22:59 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLhfw-0007VU-Qy; Mon, 26 Feb 2007 10:22:04 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLhfu-0007Ux-Ka
	for sipping@ietf.org; Mon, 26 Feb 2007 10:22:02 -0500
Received: from rtp-iport-2.cisco.com ([64.102.122.149])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLhft-0007Oi-AH
	for sipping@ietf.org; Mon, 26 Feb 2007 10:22:02 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by rtp-iport-2.cisco.com with ESMTP; 26 Feb 2007 10:22:01 -0500
X-IronPort-AV: i="4.14,221,1170651600"; 
	d="scan'208"; a="114564822:sNHT48488896"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1QFM1RI017439; 
	Mon, 26 Feb 2007 10:22:01 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1QFLXW7014954; 
	Mon, 26 Feb 2007 10:22:01 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 10:21:52 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-201.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 10:21:51 -0500
Message-ID: <45E2FB0F.6000104@cisco.com>
Date: Mon, 26 Feb 2007 10:21:51 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: Andrew Allen <aallen@rim.com>
Subject: Re: [Sipping] Service Identification Model
References: <61968779B8AC4C4BAB421D4C12F008C009463E2E@XCH47YKF.rim.net>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C009463E2E@XCH47YKF.rim.net>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 15:21:51.0991 (UTC)
	FILETIME=[D7560870:01C759B9]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=2533; t=1172503321;
	x=1173367321; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20Model
	|Sender:=20 |To:=20Andrew=20Allen=20<aallen@rim.com>;
	bh=a+nGOc02gPAdGGq+/G09GLy8m4IW7PbSevsNHfKMt2s=;
	b=yZYRFdRacVqF1uKEzo0AgzCjpoKua9d1GWEnwQZPP3TUjKg4WYXa8CFYxzyZTLYb2rHb7KGU
	qEccLnuxVrr7/iwwjH3hI0gnTdD7CPKa5cfiqgQzxDUAUvOhF8/wcsO1;
Authentication-Results: rtp-dkim-2; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 02ec665d00de228c50c93ed6b5e4fc1a
Cc: Paul Kyzivat <pkyzivat@cisco.com>, sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

inline.

Andrew Allen wrote:

> The
> following such requirements exist in the 3GPP stage 2 TS 23.228:
> 
> 
> 2.   Based on operator policy the S-CSCF or an AS shall be able to
> validate an IMS communication service identifier in a SIP request. This
> includes e.g. to check the syntactical correctness of a service
> identifier, and policing the usage of a communication service
> identifier.
> 
> 8.   Based on operator policy S-CSCF and AS shall consider the IMS
> communication service identifier for online and offline charging, e.g.
> put appropriate data into call detailed records.
> 
> 9.   The communication service identifier shall be capable of being an
> input into the policy control and charging rules.
> 
> 10. It shall be possible to use the IMS communication service identifier
> as a means to authorise whether a subscriber is allowed to initiate or
> receive request for a communication service.

I would say that I have to disagree with these requirements as stated. I 
  would propose:

2. The IMS communications service shall be determined entirely from 
fields within the SIP message that identify the functionality being 
requested, such as media types, option tags, SIP headers and so on

8. Based on operator policy, S-CSCF and AS shall consider the IMS 
communications service for online and offline charging...

9. The communications service shall be capable of being an input into 
the policy control and charging rules.

10. It shall be possible to use the IMS communications service as a 
means to authorize whether a subscriber is allowed to initiate or 
receive a request for that service

In other words, I am not disagreeing that IMS providers should be able 
to treat different services differently in every way; I just think that 
the notion of the service is entirely a function of what is already 
being said in the SIP message, not something new.


I think another way of expressing this is that the service identifier is 
really a 3gpp vehicle for DWIM - "Do What I Mean" 
(http://en.wikipedia.org/wiki/DWIM). I would prefer DWIS - "Do WHat I 
Say", where "Say" is stuff already in the SIP message.

-Jonathan R.
-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 10:27:07 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLhkm-0002UU-0Q; Mon, 26 Feb 2007 10:27:04 -0500
Received: from [10.90.34.44] (helo=chiedprmail1.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLhkk-0002Tn-2k
	for sipping@ietf.org; Mon, 26 Feb 2007 10:27:02 -0500
Received: from rtp-iport-2.cisco.com ([64.102.122.149])
	by chiedprmail1.ietf.org with esmtp (Exim 4.43) id 1HLhkh-0003sH-Oh
	for sipping@ietf.org; Mon, 26 Feb 2007 10:27:01 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by rtp-iport-2.cisco.com with ESMTP; 26 Feb 2007 10:26:59 -0500
X-IronPort-AV: i="4.14,221,1170651600"; 
	d="scan'208"; a="114565415:sNHT78844344"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1QFQxOD020091; 
	Mon, 26 Feb 2007 10:26:59 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1QFQsVf016498; 
	Mon, 26 Feb 2007 10:26:55 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 10:26:48 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-201.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 10:26:47 -0500
Message-ID: <45E2FC36.30709@cisco.com>
Date: Mon, 26 Feb 2007 10:26:46 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: Tom-PT Taylor <taylor@nortel.com>
Subject: Re: [Sipping] Service Identification Model
References: <61968779B8AC4C4BAB421D4C12F008C009463E2E@XCH47YKF.rim.net>	<FC318FB3-A497-4BED-BBC6-3B5B630FC589@estacado.net>	<45DF0451.7020209@cisco.com>
	<45E1B7A9.6070707@nortel.com>
In-Reply-To: <45E1B7A9.6070707@nortel.com>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 15:26:47.0495 (UTC)
	FILETIME=[87785D70:01C759BA]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=20015; t=1172503619;
	x=1173367619; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20Model
	|Sender:=20 |To:=20Tom-PT=20Taylor=20<taylor@nortel.com>;
	bh=Gf9NFqHs/qmgrtsdPcuwvE2MIsHEZl4MQ/YVJ4ggMCQ=;
	b=E77eHfX3oRtCynNOg5CRmXHJbX6OxKlAScIKsm/6SxKtx4TeAmyHC1uumcu2OmDBvmwqA9TU
	g1ocmqyiT3rh5JZkVN5SEdq+KH20rCNQ+szzIt8OsDD5ym1o4A1Vb4Sp;
Authentication-Results: rtp-dkim-2; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: d4a1871e876bd836d4c351e861e8720d
Cc: Andrew Allen <aallen@rim.com>, Paul Kyzivat <pkyzivat@cisco.com>,
	sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Ahh, this is a great insight.

There is a huge difference between:

1. a service identifier inserted by the network based on inspecting 
fields in the SIP message, such as SDP, Require option tags, and header 
fields

2. a service identifier inserted by endpoints which declares the 
specific service being requested


The reason is that #1 can be defined locally within a specific domain or 
service provider network without agreement between domains. As Tom says, 
it would just be a convenience so that once a server in the domain has 
performed this 'analysis', other servers don't have to redo it. You'd 
strip off this identifier prior to a request leaving the domain, and 
when it arrives in the terminating network, redo the analysis, possibly 
getting a different result. That has none of the interoperability 
problems I've been warning about.

#2 requires agreement amongst all parties on what the services are, in 
order to interoperate. That is badness.

I would not be opposed to defining a P-Asserted-Service that contains 
opaque tokens (not service URN) for use within an administrative domain, 
as long as the spec was clear that such a header was optional and always 
had its value computed by examining the rest of the sip message.

-Jonathan R.


Tom-PT Taylor wrote:

> Paul's analysis is reasonable. Basically, starting with Andrew's list, I 
> see service identification as a business issue. To optimize operation in 
> the IMS architecture, it would seem reasonable in several cases for the 
> originating S-CSCF to identify the tariffed service(s) implied by the 
> messaging from the UA and add the appropriate administrative header 
> fields so downstream elements don't have to repeat the analysis.
> 
> Paul Kyzivat wrote:
> 
>>
>> Byron Campen wrote:
>>
>>>> 1. Basic Telephony (Conversational QoS Voice only with telephony 
>>>> feature
>>>> set)
>>>> 2. Multimedia Telephony (Conversational QoS Voice plus video and other
>>>> media streams with telephony feature set)
>>
>>
>> Why are 1 & 2 different services? 1 is a proper subset of 2.
>>
>>>> 3. VoIP (Background QoS with no telephony features)
>>
>>
>> And why is 3 different from 1 and 2? Are you saying that to be able to 
>> use a feature I must decide *before* I make the call that I might want 
>> to?
>>
>> Even if that is the case, that sounds like a policy issue between a 
>> user and his own provider, rather than an e2e negotiation. (My 
>> features are provided by my UA and/or my provider.)
>>
>>>> 4. SMS
>>>> 5. MMS
>>
>>
>> Why are 4 and 5 different services? They are both datagrams. The 
>> difference in media types is irrelevant. If its delivered via MESSAGE, 
>> then it is already distinguishable from calls. This should be able to 
>> be part of the same service as IM.
>>
>>>> 6. Instant Messaging
>>
>>
>> Is just another media type. Like voice and video, these can be 
>> negotiated for the session. Why balkanize as a different service?
>>
>>>> 7. Push to Talk (PoC)
>>
>>
>> We've beaten this one to death already. Current implementation forces 
>> need for essentially an overlay network to support it. But its far 
>> from an exemplar of how to do it.
>>
>>>> 8. Network Game 1 (e.g Chess)
>>>> 9. Network Game 2 (e.g Martial arts game)
>>>> 10. Network Game 3 (e.g Dungeons & Dragons)
>>
>>
>> I tend to agree with Eric that service URIs are good for these if 
>> there is no better way. But in practice I think these could easily be 
>> distinguished in SDP using o/a.
>>
>>>> 11. Push Service 1 (e.g. Traffic information service)
>>>> 12. Push Service 2 (e.g. Weather information service)
>>>> 13. Push Service 3 (e.g  News information service)
>>>> 14. Push Service 4 (e.g  Stock information service)
>>
>>
>> There may be *something* here. I need to know more about the 
>> characteristics of these. If they consist of human intelligible media, 
>> then I think they could adequately be dealt with via callerprefs. 
>> (This may be a bit like the Actor feature tag.) Specifically, 
>> callerprefs are fine if its ok that the call end up elsewhere if the 
>> callerprefs selection doesn't happen or there isn't any target with 
>> that capability.
>>
>> If the information isn't intelligible media, then it ought to be 
>> distinguishable as a special media type in the offer.
>>
>>>> 15. Mobile TV service
>>
>>
>> AFAIK, the station doesn't call me to deliver me tv service. I would 
>> be initiating the call. And presumably it will be to a target service 
>> URI or URN. So I see no need for a service id here.
>>
>>>> So there's 15 which may all be different applications on a handset and
>>>> all those are existing services or those already envisioned.
>>
>>
>> So I only see:
>>
>> - 1 for basic telephony, multimedia telephony, SMS, MMS, IM
>>
>> - *maybe* something for games
>>
>> - *maybe* something for push services
>>
>>     Paul
>>
>>>     I only see 10 here that would need to ever send a register 
>>> request. (11-15 are strictly client-server) 4-6 will have distinct 
>>> Content-Type headers and payloads. You would need only 1 UA for 
>>> these. So we're down to 8. 8-10 will, once again, be distinct when 
>>> you look at the payload and Content-Type. So we're down to 6. There 
>>> is no reason at all that the UA that handles voice is incapable of 
>>> handling IM, and you can tell the difference by looking at the 
>>> request type, so we're down to 5. As for the voice stuff, you can 
>>> distinguish 2 from anything else through the payload. So, we could 
>>> quite easily roll 1-2 together. (4) 3 can be distinguished by 
>>> watching for any in-dialog call-control stuff. (Having a service 
>>> identifier that essentially says "I promise not to use any fancy 
>>> telephony features on this call. No, really, I promise!" is never 
>>> going to work) So we're down to 3.
>>>
>>>     SIP is more than flexible enough to handle all of this stuff in a 
>>> single UA. The UAs are perfectly capable of figuring out what 
>>> "service" the other wants to use; why are your middleboxes unable to 
>>> do the same? Adding a parallel method to make things easy on the 
>>> middleboxes is just asking for trouble, because you're conveying the 
>>> same information twice. It is ripe for abuse. Of course, the easy way 
>>> to do all this is to just let UAs send what they want, and charge per 
>>> KB, but hey, what do I know about simplicity, flexibility, 
>>> scalability, ease-of-deployment, and customer satisfaction? I'm just 
>>> a IETFer.
>>>
>>> Best regards,
>>> Byron Campen
>>>
>>>
>>>> -----Original Message-----
>>>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
>>>> Sent: Thursday, February 22, 2007 7:05 PM
>>>> To: Andrew Allen
>>>> Cc: Eric Burger; sipping@ietf.org
>>>> Subject: Re: [Sipping] Service Identification Model
>>>>
>>>>
>>>>
>>>> Andrew Allen wrote:
>>>>
>>>>> Hi Eric,
>>>>>
>>>>> It seems your prefer the approach of using URIs to identify services
>>>>> which is part of Jonathan's draft as is the use of Require for those
>>>>> services which simply do not interoperate unless the UAS understands
>>>>
>>>> the
>>>>
>>>>> service.
>>>>>
>>>>> Where I am concerned is that you seem to be suggesting that the UA
>>>>> separately register an AoR for each service it supports using
>>>>
>>>> REGISTER.
>>>>
>>>>> Certainly for cellular that wouldn't be acceptable as 10 or more
>>>>> services could easily be supported by a single device.
>>>>
>>>> Andrew, can you give me an example of 10 "services" on the device that
>>>> need to be distinguished from one another in this way - that aren't 
>>>> just
>>>>
>>>> optional behavior of a particular service? (E.g. I see no reason why
>>>> "audio" and "audio/video" need to be separate services.)
>>>>
>>>> I guess as things stand, PoC must be a separate service. And apparently
>>>> Chess. What else?
>>>>
>>>>     Paul
>>>>
>>>>> Registrations
>>>>> create a significant load in the IMS network because its not just a
>>>>
>>>> SIP
>>>>
>>>>> REGISTER but also requires a lot of interactions with other network
>>>>> elements such as the HSS and Notifications for reg-events etc. Also
>>>>> having an order of magnitude increase in multiple registrations is
>>>>
>>>> going
>>>>
>>>>> to be a big issue for the avalanche restart problem for example when a
>>>>> cell site goes down and comes back into service.
>>>>>
>>>>> I think Jonathan's approach gives us the possibility for the registrar
>>>>> to create the Service URIs based on the feature parameters included in
>>>>> the Contact of the REGISTER.
>>>>>
>>>>> I think the argument over whether the model is one UA with multiple
>>>>> services or multiple UAs each representing a separate service is
>>>>> somewhat academic. As I understand a UA can register multiple AoRs and
>>>>> can have other AoRs configured or otherwise provided by the service
>>>>> provider.
>>>>>
>>>>> Andrew
>>>>>
>>>>> -----Original Message-----
>>>>> From: Eric Burger [mailto:eburger@bea.com]
>>>>> Sent: Wednesday, February 21, 2007 1:29 PM
>>>>> To: sipping@ietf.org
>>>>> Subject: [Sipping] Service Identification Model
>>>>>
>>>>> Folks have been pestering me for a while to look at the Service
>>>>> Identification issue.  I finally have gotten around to it.
>>>>>
>>>>>
>>>>> I guess I have a fundamental problem with the principal assertion of
>>>>
>>>> the
>>>>
>>>>> service-identification draft.  Namely, the assertion that there is a
>>>>> single
>>>>> SIP Layer, separate and distinct from two services.  Here is Figure 2
>>>>> from
>>>>> the draft:
>>>>>
>>>>>                            +---------------------------------+
>>>>>                            |                                 |
>>>>>                            | +-------------+ +-------------+ |
>>>>>                            | |      UI     | |      UI     | |
>>>>>                            | +-------------+ +-------------+ |
>>>>>                            | +-------------+ +-------------+ |
>>>>>                            | |             | |             | |
>>>>>                            | |  Service 1  | |  Service 2  | |
>>>>>                            | |             | |             | |
>>>>>                            | +-------------+ +-------------+ |
>>>>>                            | +-----------------------------+ |
>>>>>                            | |                             | |
>>>>>                            | |             SIP             | |
>>>>>                            | |            Layer            | |
>>>>>                            | |                             | |
>>>>>                            | +-----------------------------+ |
>>>>>                            |                                 |
>>>>>                            +---------------------------------+
>>>>>
>>>>>                                      Physical Device
>>>>>
>>>>> I would offer this drawing confuses a common SIP Stack with a common
>>>>> User
>>>>> Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use case:
>>>>>
>>>>>> ... In addition users may have
>>>>>> multiple devices only one of which has the application that
>>>>>
>>>>> understands
>>>>>
>>>>>> the game. Other devices that the user has registered against his AoR
>>>>>
>>>>> may
>>>>>
>>>>>> handle audio for telephony and other voice applications but if Alice
>>>>>> invites Bob to play her favourite game it is not very interesting if
>>>>>
>>>>> Bob
>>>>>
>>>>>> accepts the gaming session on a device that can only accept the audio
>>>>>> stream. Also even when the terminal that has the game application
>>>>>> receives the INVITE the correct application in that terminal needs to
>>>>>
>>>>> be
>>>>>
>>>>>> identified to accept the session.
>>>>>
>>>>> I would offer that this is the clinching argument that the request is,
>>>>> in
>>>>> fact, addressing a SEPARATE USER AGENT, not a special SERVICE at a
>>>>
>>>> User
>>>>
>>>>> Agent.
>>>>>
>>>>> Same argument for PoC versus a regular voice call: One is looking for
>>>>
>>>> a
>>>>
>>>>> PoC
>>>>> User Agent, NOT the PoC service at a given User Agent.
>>>>>
>>>>> Thus, the picture really looks like the following:
>>>>>
>>>>>                            +---------------------------------+
>>>>>                            |                                 |
>>>>>                            | +-------------+ +-------------+ |
>>>>>                            | |    UI       | |    UI       | |
>>>>>                            | +-------------+ +-------------+ |
>>>>>                            | +-------------+ +-------------+ |
>>>>>                            | |             | |             | |
>>>>>                            | |    UAS 1    | |    UAS 2    | |
>>>>>                            | |             | |             | |
>>>>>                            | +-------------+ +-------------+ |
>>>>>                            | +-----------------------------+ |
>>>>>                            | |                             | |
>>>>>                            | |             SIP             | |
>>>>>                            | |            Stack            | |
>>>>>                            | |                             | |
>>>>>                            | +-----------------------------+ |
>>>>>                            |                                 |
>>>>>                            +---------------------------------+
>>>>>
>>>>>                                      Physical Device
>>>>>
>>>>>
>>>>> Another argument for different User Agents is the whole "route based
>>>>
>>>> on
>>>>
>>>>> SDP"
>>>>> argument.  As the author of RFC 3458, I think I was first to make the
>>>>> assertion that knowing the media type did NOT mean you know the
>>>>> application
>>>>> to use.  I don't want your phone or proxy or B2BUA to guess that
>>>>
>>>> because
>>>>
>>>>> I
>>>>> have Chess Media, I want the Chess Application; I want to talk with
>>>>
>>>> your
>>>>
>>>>> Chess Application.
>>>>>
>>>>> Now, is the Chess Application a Preference, Requirement, or the thing
>>>>> that I
>>>>> am addressing?  The answer drives one towards using caller prefs, the
>>>>> Require header, or the Request-URI.
>>>>>
>>>>> I would offer that it is not a preference: I really want the chess
>>>>> application; I probably do not want to talk to you.
>>>>>
>>>>> That leaves us with the Require header or a Request-URI.  My leanings
>>>>> are
>>>>> not surprisingly towards the RFC 4240 approach.  Frankly, if you are
>>>>> building an IMS (as an example) that can handle 10,000,000
>>>>
>>>> subscribers,
>>>>
>>>>> you
>>>>> should be able to handle 100,000,000 registrations.  That order of
>>>>> magnitude
>>>>> is way easier than the 10,000 to 10,000,000 jump.  I don't buy the
>>>>> "registering each service is too much of a burden on the operator"
>>>>> argument,
>>>>> because it is not that much of a burden, and the network actually does
>>>>> the
>>>>> right thing.  Better yet, all of the corner cases work.  Best of all,
>>>>> the
>>>>> IETF doesn't have to build an IANA Registry and process a thousand PS
>>>>> documents to register urn:service:myco:widget.
>>>>>
>>>>> I suppose I could live with Require:, only if you are way too lazy to
>>>>> register each application instance.  It does semantically work,
>>>>
>>>> because
>>>>
>>>>> if I
>>>>> want to do PoC and you do not do PoC, I'll get a 4xx response, to
>>>>
>>>> which
>>>>
>>>>> I
>>>>> could try a regular call.
>>>>>
>>>>>
>>>> _______________________________________________________________________
>>>>
>>>>> Notice:  This email message, together with any attachments, may
>>>>
>>>> contain
>>>>
>>>>> information  of  BEA Systems,  Inc.,  its subsidiaries  and
>>>>
>>>> affiliated
>>>>
>>>>> entities,  that may be confidential,  proprietary,  copyrighted
>>>>
>>>> and/or
>>>>
>>>>> legally privileged, and is intended solely for the use of the
>>>>
>>>> individual
>>>>
>>>>> or entity named in this message. If you are not the intended
>>>>
>>>> recipient,
>>>>
>>>>> and have received this message in error, please immediately return
>>>>
>>>> this
>>>>
>>>>> by email and then delete it.
>>>>>
>>>>> _______________________________________________
>>>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>>>> This list is for NEW development of the application of SIP
>>>>> Use sip-implementors@cs.columbia.edu for questions on current sip
>>>>> Use sip@ietf.org for new developments of core SIP
>>>>>
>>>>>
>>>>>
>>>>> ---------------------------------------------------------------------
>>>>> This transmission (including any attachments) may contain confidential
>>>>
>>>> information, privileged material (including material protected by the
>>>> solicitor-client or other applicable privileges), or constitute
>>>> non-public information. Any use of this information by anyone other 
>>>> than
>>>> the intended recipient is prohibited. If you have received this
>>>> transmission in error, please immediately reply to the sender and 
>>>> delete
>>>> this information from your system. Use, dissemination, distribution, or
>>>> reproduction of this transmission by unintended recipients is not
>>>> authorized and may be unlawful.
>>>>
>>>>> _______________________________________________
>>>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>>>> This list is for NEW development of the application of SIP
>>>>> Use sip-implementors@cs.columbia.edu for questions on current sip
>>>>> Use sip@ietf.org for new developments of core SIP
>>>>>
>>>>
>>>>
>>>> ---------------------------------------------------------------------
>>>> This transmission (including any attachments) may contain 
>>>> confidential information, privileged material (including material 
>>>> protected by the solicitor-client or other applicable privileges), 
>>>> or constitute non-public information. Any use of this information by 
>>>> anyone other than the intended recipient is prohibited. If you have 
>>>> received this transmission in error, please immediately reply to the 
>>>> sender and delete this information from your system. Use, 
>>>> dissemination, distribution, or reproduction of this transmission by 
>>>> unintended recipients is not authorized and may be unlawful.
>>>>
>>>> _______________________________________________
>>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>>> This list is for NEW development of the application of SIP
>>>> Use sip-implementors@cs.columbia.edu for questions on current sip
>>>> Use sip@ietf.org for new developments of core SIP
>>
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP
>> Use sip-implementors@cs.columbia.edu for questions on current sip
>> Use sip@ietf.org for new developments of core SIP
>>
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 10:29:29 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLhn6-0003Ah-Lq; Mon, 26 Feb 2007 10:29:28 -0500
Received: from [10.90.34.44] (helo=chiedprmail1.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLhn4-00038P-Bs
	for sipping@ietf.org; Mon, 26 Feb 2007 10:29:26 -0500
Received: from rtp-iport-2.cisco.com ([64.102.122.149])
	by chiedprmail1.ietf.org with esmtp (Exim 4.43) id 1HLhn2-0004Cu-Km
	for sipping@ietf.org; Mon, 26 Feb 2007 10:29:26 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by rtp-iport-2.cisco.com with ESMTP; 26 Feb 2007 10:29:24 -0500
X-IronPort-AV: i="4.14,221,1170651600"; 
	d="scan'208"; a="114565823:sNHT72579852"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1QFTOZo029722; 
	Mon, 26 Feb 2007 10:29:24 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1QFTFVj017446; 
	Mon, 26 Feb 2007 10:29:27 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 10:29:13 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-202.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 10:29:13 -0500
Message-ID: <45E2FCC8.60708@cisco.com>
Date: Mon, 26 Feb 2007 10:29:12 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: Henning Schulzrinne <hgs@cs.columbia.edu>
Subject: Re: [Sipping] Service Identification in SIP
References: <61968779B8AC4C4BAB421D4C12F008C0091516D0@XCH47YKF.rim.net>
	<42E60FD2-CA6C-45C4-A1E3-99C93A867A54@cs.columbia.edu>
In-Reply-To: <42E60FD2-CA6C-45C4-A1E3-99C93A867A54@cs.columbia.edu>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 15:29:13.0448 (UTC)
	FILETIME=[DE770680:01C759BA]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=11717; t=1172503764;
	x=1173367764; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20
	|To:=20Henning=20Schulzrinne=20<hgs@cs.columbia.edu>;
	bh=IVLE5VZRR48/J+eUxzSKwi/0nSDDkEMCl8bww3O1ihE=;
	b=F+3nkKk0UAO17EosKdFv0v4qrqLOzYYfN+anbfcjDZSgwZr8Nd+W6jb3EkZx8KRGY1YMpzSg
	7lenWUAvMmF0FyRSZ5PtcRhyipRHWC43Qu75J+7Syv4fNSZNbVxUHpbO;
Authentication-Results: rtp-dkim-1; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 850245b51c39701e2700a112f3032caa
Cc: Andrew Allen <aallen@rim.com>, sipping <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I'm on the fence about whether service URN, or just an arbitrary token, 
is appropriate here. I do agree that we want to do this infrequently and 
thus do not have the need for a robust and distributed allocation 
mechanism. In that sense service URN is not necessary. However, it is 
appropriate in that what is being required is understanding of some 
additional context around WHY I'm sending this SIP request; which can 
legitimately be considered the service.

-Jonathan R.

Henning Schulzrinne wrote:

> Andrew,
> 
> my concern is that I'm unsure that developing a very generic  mechanism 
> to address a problem we don't want to happen in the first  case seems 
> both misplaced effort and counterproductive. A robust and  distributed 
> allocation mechanism for identifiers makes sense if you  expect hundreds 
> of entities to independently need to allocate such  identifiers. This is 
> clearly undesirable in this case, based on the  examples presented, so 
> the IANA registration mechanism for Require  tags seems to be both 
> sufficient and sufficiently onerous to prevent  gratuitous incompatibility.
> 
> I also believe that service URNs are wholly inappropriate for this  
> particular purpose, even if you don't believe what I wrote above.
> 
> Henning
> 
> On Feb 17, 2007, at 10:15 PM, Andrew Allen wrote:
> 
>>
>> Hi Henning,
>>
>> First point on the use of Service URNs. I think the following  properties
>> of a URN are important to this work:
>>
>> a)       the guaranteed uniqueness - we need to ensure that service
>> identifiers are unique. URNs have an existing process for ensuring
>> uniqueness.
>>
>> b)       the ability to use a URN as a URI - A UA can be pre- configured
>> or through some other means obtain a URN for a given service and then
>> this can be used in the Route header or the Request-URI in order to
>> contact Application Servers in the local domain when initiating a
>> communication related to the service.
>>
>> c)       a hierarchical structure - in order to enhance the
>> interoperability URNs provide the possibility to perform best match
>> comparisons in order to determine what service should be provided.
>>
>> e.g
>> If
>>    URN:service.comm.telephony.multimedia
>> Is present in a request but only
>>    URN:service.comm.telephony.basic
>> is supported then that can be considered a good match.
>>
>>
>> Some of these communication services may also be location based for
>> example the equivalent of 1-800-PIZZA could be
>> URN:service.comm.telephony.free.restaurant.pizza.
>>
>> Linking this to the Service URN used in ecrit seems convenient but not
>> absolutely essential.
>>
>>
>> Secondly I don't have a problem with using Callee Caps and Caller
>> preferences to find the appropriate UAS however the requirements for
>> service identification are much more than just that.
>>
>> This following requirements from 3GPP TS 23.228 do not I believe fit
>> well with the semantics of Accept-Contact:
>>
>> 2.   Based on operator policy the S-CSCF or an AS shall be able to
>> validate an IMS communication service identifier in a SIP request.  This
>> includes e.g. to check the syntactical correctness of a service
>> identifier, and policing the usage of a communication service
>> identifier.
>>
>> 4.   It shall be possible for the S-CSCF to invoke appropriate service
>> logic based on the IMS communication service identifier contained in a
>> SIP request, e.g. route a SIP request containing a service identifier
>> based on initial filter criteria to the correct AS.
>>
>> 8.   Based on operator policy S-CSCF and AS shall consider the IMS
>> communication service identifier for online and offline charging, e.g.
>> put appropriate data into call detailed records.
>>
>> 9.   The communication service identifier shall be capable of being an
>> input into the policy control and charging rules.
>>
>> 10. It shall be possible to use the IMS communication service  identifier
>> as a means to authorise whether a subscriber is allowed to initiate or
>> receive request for a communication service.
>>
>> Note that the some of AS (Application Servers) that need to be invoked
>> based on requirement (4) may not have any relationship to the resource
>> identified by the Request-URI because they are Applications Servers in
>> the domain of the originating UAC.
>>
>> Not only do these requirements not semantically fit with Accept- Contact
>> but the fact that a service feature tag appears somewhere in an
>> Accept-Contact cannot be taken as a definitive indication the  initiating
>> user is only requesting that particular service as the expression may
>> involve other feature tags in a complex expression.
>>
>> In addition there needs to be a mechanism that allows UAs to obtain  the
>> appropriate service identifiers used by other UAs and in other  domains.
>> I think Jonathan's proposal to use Service URIs and Service Instance
>> URIs with Presence creates a backward compatible way for UAs to obtain
>> these identifiers.
>>
>> Andrew
>>
>>
>> -----Original Message-----
>> From: Henning Schulzrinne [mailto:hgs@cs.columbia.edu]
>> Sent: Saturday, February 17, 2007 12:27 PM
>> To: Jonathan Rosenberg
>> Cc: sipping
>> Subject: Re: [Sipping] Service Identification in SIP
>>
>> I think we're in general agreement; in particular, Require seems like
>> the right approach. (service URNs do not, in my opinion).
>>
>> Stepping back, I don't think we should generalize this problem since
>> it essentially "blesses" a particular design approach that is
>> undesirable. The right approach, from a SIP perspective, for IMS PoC
>> would have been to come up with a solution that is backwards-
>> compatible. If that's not possible, getting "branded" with Require
>> seems like the right penalty.
>>
>> In general, we have probably two requirements:
>>
>> (1) When contacting an AOR, a UAC that needs a special UAS to work
>> properly needs to get routed to that UAS only. Callee caps and caller
>> preferences seem to be the right tools for that, as it involves call
>> routing and capability matching.
>>
>> (2) If, despite doing (1), a SIP request accidentally reaches a UAS
>> that can't do anything useful with it, Requires minimizes the damage
>> (and allows a proxy to serially-fork to something else that might
>> have a chance of working).
>>
>> Henning
>>
>> On Feb 16, 2007, at 8:48 PM, Jonathan Rosenberg wrote:
>>
>>> I think its a fair point that the example is contrived. Really, the
>>> main use case being targeted is IMS PoC. So, let me talk a bit
>>> about that.
>>>
>>> One would think that PoC is just a regular SIP call that has floor
>>> control added to it as another media type, and so one could
>>> indicate that a call is a PoC call just because of the presence of
>>> this new media type in the SDP. Sadly, I wish that were true. If it
>>> was, we would not need a mechanism here.
>>>
>>> Unfortunately, PoC has layered a whole bunch of additional
>>> behaviors into the way SIP requests are handled. Dale has indicated
>>> one of them - that a 200 OK is returned on the terminating side
>>> priod to the call actually being completed. PoC also *requires* the
>>> insertion of servers into the SIP signaling path to make the system
>>> work. Consequently, though the messages that PoC uses *look* like
>>> SIP messages with a new media type, there is a hidden 'extension'
>>> in place that changes the fundamental semantics of INVITE. The fact
>>> that this extension is not visible as a specific header in a well
>>> defined way does not change the fact that PoC CHANGES THE BEHAVIOR
>>> OF SIP AND IS THUS A SIP EXTENSION.
>>>
>>> To convince yourself of this, consider what would happen if a PoC
>>> client calls a regular SIP phone. The regular SIP phone would see
>>> the PoC floor control stream and reject it as unknown. If PoC were
>>> just SIP as we've specified it, then everything would be fine, and
>>> the call would proceed but without floor control.
>>>
>>> But alas, it is not so.
>>>
>>> Because PoC makes all kinds of assumptions about the behaviors of
>>> its participants, and what the real meaning is for various SIP
>>> messages, the entire thing just won't work when a call terminates
>>> on a non-PoC device. This is because PoC is, effectively, a fairly
>>> significant and non-backwards-compatible extension to SIP.
>>>
>>> So, my proposal is, if you are going to go and DO something like
>>> that, you better use the SIP techniques that we've defined to deal
>>> with it. And when you extend SIP in a way that doesn't work with a
>>> SIP endpoint that doesn't understand an extension, you use a
>>> Require header. And that is exactly what I'm proposing that 3gpp
>>> should do. Use Require because you have made a non-backwards-
>>> compatible extension to SIP.
>>>
>>> The general problem here is that there are other places where
>>> people are layering additional 'context', as Loreto has called it,
>>> ontop of SIP signaling. This is a worrisome trend because this
>>> context is subtle and not explicitly signaled. Its just assumed.
>>> And that is dangerous, since it leads to the kind of balkanization
>>> and non-interoperability that myself, Paul, Dale and others have
>>> complained about in this thread.
>>>
>>> The current 3gpp specs, instead of using Require to signal this
>>> extension, put a tag into the Accept-Contact header field. I find
>>> this extremely worrisome, because it is really just masking the
>>> reality of what has happened. If you've got a situration where a
>>> SIP call *cannot be processed* without going through application
>>> servers, then this is not something that is signaled with Accept-
>>> Contact. Its signaled with Require or Proxy-Require as appropriate.
>>>
>>> Thanks,
>>> -Jonathan R.
>>>
>>> Henning Schulzrinne wrote:
>>>
>>
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP
>> Use sip-implementors@cs.columbia.edu for questions on current sip
>> Use sip@ietf.org for new developments of core SIP
>>
>>
>>
>> ---------------------------------------------------------------------
>> This transmission (including any attachments) may contain  
>> confidential information, privileged material (including material  
>> protected by the solicitor-client or other applicable privileges),  or 
>> constitute non-public information. Any use of this information  by 
>> anyone other than the intended recipient is prohibited. If you  have 
>> received this transmission in error, please immediately reply  to the 
>> sender and delete this information from your system. Use,  
>> dissemination, distribution, or reproduction of this transmission  by 
>> unintended recipients is not authorized and may be unlawful.
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP
>> Use sip-implementors@cs.columbia.edu for questions on current sip
>> Use sip@ietf.org for new developments of core SIP
> 
> 

-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 10:37:20 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLhuJ-0005VS-2P; Mon, 26 Feb 2007 10:36:55 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLhuH-0005VL-FB
	for sipping@ietf.org; Mon, 26 Feb 2007 10:36:53 -0500
Received: from rtp-iport-2.cisco.com ([64.102.122.149])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLhuG-0001Iq-68
	for sipping@ietf.org; Mon, 26 Feb 2007 10:36:53 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by rtp-iport-2.cisco.com with ESMTP; 26 Feb 2007 10:36:52 -0500
X-IronPort-AV: i="4.14,221,1170651600"; 
	d="scan'208"; a="114567216:sNHT49925884"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1QFapxr002696; 
	Mon, 26 Feb 2007 10:36:51 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1QFacOM016242; 
	Mon, 26 Feb 2007 10:36:47 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 10:36:41 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-201.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 10:36:41 -0500
Message-ID: <45E2FE88.5090508@cisco.com>
Date: Mon, 26 Feb 2007 10:36:40 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: "Salvatore Loreto (JO/LMF)" <salvatore.loreto@ericsson.com>
Subject: Re: [Sipping] Service Identification in SIP
References: <61968779B8AC4C4BAB421D4C12F008C009151456@XCH47YKF.rim.net>
	<1171641333.3800.34.camel@localhost.localdomain>
In-Reply-To: <1171641333.3800.34.camel@localhost.localdomain>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 15:36:41.0585 (UTC)
	FILETIME=[E9934E10:01C759BB]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=1892; t=1172504212;
	x=1173368212; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20
	|To:=20=22Salvatore=20Loreto=20(JO/LMF)=22=20<salvatore.loreto@ericsson.c
	om>; bh=Dqgrxiho7lv1K76zY8+QHHVBUFO6idDFqmK8jV0jFWs=;
	b=UKyRTAJEPySkj5+VoA2pqWGzn2VXGqJCms93nsqZqAr3vKsXmbam9BXd5MvDPG2LFtmap1XW
	JFC1p5vevcDMSngH6ZvRopOKkwB6SZOkTMLiuKoGO0I+y1bphTXooV20;
Authentication-Results: rtp-dkim-1; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a7d6aff76b15f3f56fcb94490e1052e4
Cc: ravishankar.shiroor@wipro.com, Paul Kyzivat <pkyzivat@cisco.com>,
	Gonzalo.Camarillo@ericsson.com, mary.barnes@nortel.com,
	Andrew Allen <aallen@rim.com>, sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

one clarification in line:

Salvatore Loreto (JO/LMF) wrote:


> At same time I am also curious about the mechanism proposed by Jonathan.
> It is an interesting proposal, but to tell you the truth, 
> at the present stage in both the proposal call-pref and Jonathan
> mechanism I can identify more or less the same number of issues that
> need a lot of discussion and work before reach a robust stage.
> For example the Jonathan proposal is based on the louse routing that is
> something that has not yet reached a large consensus in the sip
> community, and also the re-ordering is something that need a lot of work
> and clarification. But as I said it is an interesting proposal.

The UA loose routing is there for the following reason. Lets say user B 
publishes a presence document with PoC and telephony, and includes a 
different URI for each. When user A gets that presence document and 
invokes the PoC URI, it would be helpful for user B to know, when it got 
the request, that this URI was used. Loose routing accomplishes that.

Note that the request would still contain Require, since the originating 
app was PoC and will need to include Require since PoC is not backwards 
compatible.

In the absence of loose routing, an IMS endpoint could continue to use 
P-Called-Party-ID and have the same result.

So the loose routing bit is really orthogonal here to the rest of it. 
I'd be fine to reword it to point to loose route or other equivalent 
mechanisms so that there was not a direct dependency.

Thanks,
Jonathan R.

-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 11:01:28 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLiGt-000114-Gh; Mon, 26 Feb 2007 11:00:15 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLiGr-00010m-UT
	for sipping@ietf.org; Mon, 26 Feb 2007 11:00:13 -0500
Received: from rtp-iport-2.cisco.com ([64.102.122.149])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLiGp-0004kR-Fg
	for sipping@ietf.org; Mon, 26 Feb 2007 11:00:13 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by rtp-iport-2.cisco.com with ESMTP; 26 Feb 2007 11:00:11 -0500
X-IronPort-AV: i="4.14,221,1170651600"; 
	d="scan'208"; a="114570691:sNHT57894384"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1QG0Bmo017698; 
	Mon, 26 Feb 2007 11:00:11 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1QG02OM024389; 
	Mon, 26 Feb 2007 11:00:11 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 11:00:01 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-202.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 11:00:00 -0500
Message-ID: <45E30400.3040206@cisco.com>
Date: Mon, 26 Feb 2007 11:00:00 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: Andrew Allen <aallen@rim.com>
Subject: Re: [Sipping] Service Identification in SIP
References: <61968779B8AC4C4BAB421D4C12F008C009151469@XCH47YKF.rim.net>
In-Reply-To: <61968779B8AC4C4BAB421D4C12F008C009151469@XCH47YKF.rim.net>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 16:00:00.0698 (UTC)
	FILETIME=[2B8301A0:01C759BF]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=7099; t=1172505611;
	x=1173369611; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20 |To:=20Andrew=20Allen=20<aallen@rim.com>;
	bh=wgtWNPgNdDYlLN4c6Jhjr+0UxEXGIBrlYkp3rAxwkh4=;
	b=cfAUAS/tBu73BIq9AuRy8bO9CVV0977FPETn5dbhN3j/+CLdF7ZnYYJgpfghrWqWcLKb6TWI
	ax9RQCkOsLn1YtsdnRz8263W671X8aDhCMJIlzFidljMU0EczRrRiKrV;
Authentication-Results: rtp-dkim-1; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: c54bc2f42d02429833c0ca4b8725abd7
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Andrew Allen wrote:

> I'll have a crack at laying out examples showing the utility of service
> identification.
> 
> First a gaming example somewhat similar to Jonathan's. Take a network
> based multiplayer game (Chess, D&D, Quake, Pool, Roulette, virtual golf,
> etc.....). This game allows multiple users to participate in the game
> either as players or as observers. An audio stream is established so
> that players and observers can exchange comments on the game along with
> the media stream (either using MSRP or an application specific data
> transport that is used to exchange game commands and game position
> updates between the participants and the game server which also acts as
> a focus. The gaming session is established using SIP and SDP over a SIP
> network such as IMS that has Application Servers that perform telephony
> call features for basic voice only telephony and multimedia "rich"
> telephony calls. The multimedia telephony sessions also support combined
> audio with MSRP and other application specific data streams (such as
> whiteboards) in a call. Now gaming sessions don't require telephony
> service call feature handling in fact this isnt even desired however
> based on the SIP and SDP it isnt possible to determine whether an INVITE
> request is related to a telephony call or a gaming session and hence
> whether the request needs to be routed via an Application Server that
> performs telephony call features or not.

Sure it does. In the case of the game, there would be a media stream for 
the proprietary gaming moves. If it is using MRSP, then the content 
types that are listed in the SDP for MSRP indicate that.


> In addition users may have
> multiple devices only one of which has the application that understands
> the game. Other devices that the user has registered against his AoR may
> handle audio for telephony and other voice applications but if Alice
> invites Bob to play her favourite game it is not very interesting if Bob
> accepts the gaming session on a device that can only accept the audio
> stream.

Right, so its reasonable to apply caller prefs here too, to route the 
request to devices that support the required content types.


> Also even when the terminal that has the game application
> receives the INVITE the correct application in that terminal needs to be
> identified to accept the session. In addition service providers may
> charge different rates for telephony and for games. I also don't want to
> be charged telephony per minute call rates for playing my game.  

You can do all of that. In this particular example there are very 
specific things in the SDP that indicate that this is a game vs. text chat.

> 
> Here is another example that I think shows that service identification
> doesn't just apply to INVITE and sessions but also can apply to other
> requests such a MESSAGE as well. SIP MESSAGE method is being used for
> Instant messaging applications but it is also being used for many non
> Instant Messaging applications as well. These include as a replacement
> SIP transport for transporting encapsulated SMS (Short Message Service)
> contents (see 3GPP SMS over IP), as a transport for Group Advertisement
> documents for Push To Talk (see OMA PoC), as a notification transport
> for various Push Services (see OMA SIP Push). Each of these different
> service applications of SIP MESSAGE method requires routing to different
> Application Servers in the network, delivery to different applications
> in the terminal and the ability to be charged for separately. The only
> SIP header that differentiates each of these different useages of SIP
> MESSAGE is the Content-Type header but routing based on Content-Type is
> a very bad idea.

Why? This is widely used practice for email.

  Firstly the Content-Type doesn't identify the service
> specifically but some content related to a subset of the service.

This depends on how good a job you do at defining your content types.

If you use very targeted content types it works great. For example, if 
you define a content type specific for push for notifying a device of 
new configs, then that type indicates exactly what is happening. If you 
use text/plain, and then define some special format for that text, you 
are correct its not sufficient. This is why we have the 
Content-Disposition header field.


  Also
> as the service evolves the Content-Types used may change or new ones may
> be added and this shouldn't require updating the routing rules of the
> proxies or the internal application invocation logic in endpoints. The
> possibility of Multipart MIME Content-Types makes parsing of
> Content-Types by proxies inefficient and complex and potentially
> impossible if the message body is encrypted. 

Well, thats a different problem. But if it turns out to be one, we could 
  just define mecahnisms for signaling the internal content type and 
disposition.


> What is needed is a means
> compliant with standard SIP Routing mechanisms that allow SIP proxies to
> efficently route these different MESSAGE requests to the appropriate
> servers, endpoints and invoke applications within endpoints. I think
> Jonathan's draft is a good basis for doing this.

Require is there for when that is the only recourse; I am hoping it is 
not often needed, and in fact if its just for PoC I will be happy. In 
the examples you give above, I think existing sip mecahnisms suffice to 
identify the service.

> 
> Another example using INVITE is IPTV. IPTV may be established using
> INVITE containing SDP with both Audio and Video streams. How does the
> network determine the difference between an IPTV related INVITE and one
> related to multimedia telephony that supports video? Certainly it is not
> desired that the IPTV INVITE invoke telephony call features in an
> Application Server and you probably don't want the IPTV INVITE being
> routed to your mobile phone (unless it's a mobile TV service). It is
> likely that an SDP offer for IPTV is for a unidirectional media stream
> but this is real convoluted to use as the basis for routing and
> charging. 

In this case, the caller or called party identify the app.

FOr example, if a user uses SIP INVITE to connect to a movie, I would 
suggest an INVITE that looks like:

INVITE urn:movies:TheReturnOfTheKing SIP/2.0


In this case, the identity of the called party says it all.


> 
> In adddition it is possible that an INVITE doesn't contain SDP at all
> and then there is no basis in SDP for determining the related service.


Ahhh.

Now that is a good point. Let me make a separate post on that.

-Jonathan R.


-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 11:17:25 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLiWi-0002DE-Ag; Mon, 26 Feb 2007 11:16:36 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLiWg-0002CI-2b
	for sipping@ietf.org; Mon, 26 Feb 2007 11:16:34 -0500
Received: from rtp-iport-1.cisco.com ([64.102.122.148])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLiWe-0006em-MS
	for sipping@ietf.org; Mon, 26 Feb 2007 11:16:34 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by rtp-iport-1.cisco.com with ESMTP; 26 Feb 2007 11:16:33 -0500
X-IronPort-AV: i="4.14,221,1170651600"; 
	d="scan'208"; a="53693372:sNHT61855900"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1QGGW98027270
	for <sipping@ietf.org>; Mon, 26 Feb 2007 11:16:32 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1QGGLW1005403
	for <sipping@ietf.org>; Mon, 26 Feb 2007 11:16:35 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 11:16:26 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-202.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 11:16:25 -0500
Message-ID: <45E307D9.1090201@cisco.com>
Date: Mon, 26 Feb 2007 11:16:25 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: IETF Sipping List <sipping@ietf.org>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 16:16:25.0948 (UTC)
	FILETIME=[76C439C0:01C759C1]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=3097; t=1172506592;
	x=1173370592; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Service=20Identification=3A=20offerless=20INVITE=20case
	|Sender:=20 |To:=20IETF=20Sipping=20List=20<sipping@ietf.org>;
	bh=6HYKthhqkqzH2LO8BKcMHgZaglMSM9F15Dfo9JG42VM=;
	b=c1oAYacBWbZY2iURTgCC28uXnYp0e0PT7S/XqJwlOH1lZamQs0oHSP13b4mIZG/h9o0TuAUC
	voxDvACkBoOvLoRQpqev4QUn1bpKSsFPxDlHMcPH3bmhgCfpdiWL7T6A;
Authentication-Results: rtp-dkim-1; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: f607d15ccc2bc4eaf3ade8ffa8af02a0
Subject: [Sipping] Service Identification: offerless INVITE case
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Andrew made a great point in one of his mails which merits promotion 
into its own thread.

I and others have asserted that you should be able to identify the 
service by things naturally in the request, and in particular, SDP has 
come up as a key thing. For example, in a multiplayer game, there might 
be an m-line for the particular proprietary protocol used for game 
moves. Or, if its MSRP, you might have specific content types for game 
moves. When present, those exactly indicate what the service is.

But what do we do about offerless INVITE? In particular, there are cases 
where we need to invoke an application server to handle a request, and 
what if we don't have the SDP in the INVITE? Its too late to insert an 
application server once the INVITE has already passed to the called 
party and the 200 OK is sent.

This is a great point. I'd rather just fix this little problem, the 
define an entire notion of services and service identifiers that are not 
necessary, just because we have offerless INVITEs.

There are lots of ways to handle this problem. My preferred approach 
would be that offerless INVITE contain a capability description, using 
Flemmings new capability framework. The INVITE would contain a body that 
  looks like SDP, but has a different content type and 
content-disposition. It lists the media streams, codecs and other 
capabilities that exist. Based on this, the called party makes the offer 
in a 2xx or 18x like normal, and the answer in the ACK/PRACK as normal. 
It is completely backwards compatible.

Now, a related issue is what happens if a session starts with audio 
only, and then the user wants to turn this into a game? Its too late to 
add new application servers into the call path, which might need to be 
there for games. What can we do?

Here too, there is a simple answer. If a network provider requires app 
servers to get inserted, or a phone requires dispatch to a different 
application, it would receive a re-INVITE to add the gaming stream, and 
then 3xx it. The Contact would contain an embedded Replaces header with 
the dialog IDs for the call in progress. THe originating UA would then 
do a totally new INVITE that replaces the call in progress. This would 
allow the network and/or phone to redispatch, yet provide a seamless 
user experience. Billing can be done however the operators want, since 
the signaling is very clear on exactly what happened and when. 
Furthermore, the replaces mechansim would only get used if there was 
someone in the call path that needed to see it.

I'd rather specify things like this, than invent an entire services 
framework that would lead to balkanization and non-interoperable islands 
of SIP.

-Jonathan R.


-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 12:20:44 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLjVU-0003q0-NB; Mon, 26 Feb 2007 12:19:24 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLjVS-0003nn-P9
	for sipping@ietf.org; Mon, 26 Feb 2007 12:19:22 -0500
Received: from sj-iport-3-in.cisco.com ([171.71.176.72]
	helo=sj-iport-3.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HLjVR-0006wy-DC
	for sipping@ietf.org; Mon, 26 Feb 2007 12:19:22 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by sj-iport-3.cisco.com with ESMTP; 26 Feb 2007 09:19:21 -0800
X-IronPort-AV: i="4.14,221,1170662400"; 
	d="scan'208"; a="467118545:sNHT49315110"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1QHJKVa023919
	for <sipping@ietf.org>; Mon, 26 Feb 2007 12:19:20 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1QHJKVV027222
	for <sipping@ietf.org>; Mon, 26 Feb 2007 12:19:20 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 12:19:16 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 12:19:16 -0500
Message-ID: <45E31693.40207@cisco.com>
Date: Mon, 26 Feb 2007 12:19:15 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Jonathan Rosenberg <jdrosen@cisco.com>
Subject: Re: [Sipping] Service Identification: offerless INVITE case
References: <45E307D9.1090201@cisco.com>
In-Reply-To: <45E307D9.1090201@cisco.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 17:19:16.0109 (UTC)
	FILETIME=[3DF523D0:01C759CA]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=3319; t=1172510360;
	x=1173374360; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=3A=20offerless=2
	0INVITE=20case |Sender:=20
	|To:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>;
	bh=JqhOK3GO0fFyNkoVTXqfdz5H3mDMncgql3tZSCaAatI=;
	b=VeW/1wUWdgOQ2rQJbLrOFw4KBIKLlRTM4mKg1olESQlPvXMOZjlGRSWPmygMuUVrcyeE/cgp
	Yr482cN5ESR9vVm0q2wGCI8kyUO0ZqdtctYmx94dv3iFcgppclYHBcGg;
Authentication-Results: rtp-dkim-2; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: e8a67952aa972b528dd04570d58ad8fe
Cc: IETF Sipping List <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I too was thinking that INVITE/Replaces would be the way to get a new 
dialog that could include app servers not in the existing call, when 
that is needed. But I hadn't worked out how to know to do that. The 3xx 
does seem like a good way to handle that. The use of 3xx as a response 
to an in-dialog request is not at all well specified. But this could be 
a good motivation to clear that up.

	Paul

Jonathan Rosenberg wrote:
> Andrew made a great point in one of his mails which merits promotion 
> into its own thread.
> 
> I and others have asserted that you should be able to identify the 
> service by things naturally in the request, and in particular, SDP has 
> come up as a key thing. For example, in a multiplayer game, there might 
> be an m-line for the particular proprietary protocol used for game 
> moves. Or, if its MSRP, you might have specific content types for game 
> moves. When present, those exactly indicate what the service is.
> 
> But what do we do about offerless INVITE? In particular, there are cases 
> where we need to invoke an application server to handle a request, and 
> what if we don't have the SDP in the INVITE? Its too late to insert an 
> application server once the INVITE has already passed to the called 
> party and the 200 OK is sent.
> 
> This is a great point. I'd rather just fix this little problem, the 
> define an entire notion of services and service identifiers that are not 
> necessary, just because we have offerless INVITEs.
> 
> There are lots of ways to handle this problem. My preferred approach 
> would be that offerless INVITE contain a capability description, using 
> Flemmings new capability framework. The INVITE would contain a body that 
>  looks like SDP, but has a different content type and 
> content-disposition. It lists the media streams, codecs and other 
> capabilities that exist. Based on this, the called party makes the offer 
> in a 2xx or 18x like normal, and the answer in the ACK/PRACK as normal. 
> It is completely backwards compatible.
> 
> Now, a related issue is what happens if a session starts with audio 
> only, and then the user wants to turn this into a game? Its too late to 
> add new application servers into the call path, which might need to be 
> there for games. What can we do?
> 
> Here too, there is a simple answer. If a network provider requires app 
> servers to get inserted, or a phone requires dispatch to a different 
> application, it would receive a re-INVITE to add the gaming stream, and 
> then 3xx it. The Contact would contain an embedded Replaces header with 
> the dialog IDs for the call in progress. THe originating UA would then 
> do a totally new INVITE that replaces the call in progress. This would 
> allow the network and/or phone to redispatch, yet provide a seamless 
> user experience. Billing can be done however the operators want, since 
> the signaling is very clear on exactly what happened and when. 
> Furthermore, the replaces mechansim would only get used if there was 
> someone in the call path that needed to see it.
> 
> I'd rather specify things like this, than invent an entire services 
> framework that would lead to balkanization and non-interoperable islands 
> of SIP.
> 
> -Jonathan R.
> 
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 12:33:29 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLji1-00083W-I5; Mon, 26 Feb 2007 12:32:21 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLji0-00083R-Ph
	for sipping@ietf.org; Mon, 26 Feb 2007 12:32:20 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLjht-00089z-3Q
	for sipping@ietf.org; Mon, 26 Feb 2007 12:32:20 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	7D2BE204CA
	for <sipping@ietf.org>; Mon, 26 Feb 2007 18:32:08 +0100 (CET)
X-AuditID: c1b4fb3e-af6d2bb0000007e1-cb-45e3199833ba 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	6417A20397
	for <sipping@ietf.org>; Mon, 26 Feb 2007 18:32:08 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 18:32:08 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Mon, 26 Feb 2007 18:32:07 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD02926384@esealmw111>
In-Reply-To: <45E2F10E.3070906@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdZs/S6O+aGWZgxS9Onz8HBVIX0KQACyPBA
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 26 Feb 2007 17:32:08.0465 (UTC)
	FILETIME=[0A515010:01C759CC]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a87a9cdae4ac5d3fbeee75cd0026d632
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi,

I don't think that being able to identify services leads to any of the
mentioned "-sation"s of SIP [well maybe industralisation ;-)], except to
recognise that there are different services on SIP and that the media
does not give the service (we can also start the session without media
...).  At what point can it be said that SIP is "-ised" - or at what
point to the SIP police say that this is breaking SIP?  If it doesn't
get to all UAs?  If it doesn't go to all UAs that could do something
useful with it?, If it doesn't go through a SIP proxy? ... Sometimes
trying to deliver a service instead of rejecting the service due to
incomptabilities can lead to a bad user experience - is that the best
thing to do, or is that OK because the SIP protocol still connected? [I
think that Jeroen captured this quite well in one of his emails]. I note
that the second point raised, does handle this, leaving the question of
a middle ground as well (i.e an indication of the service without it
being "REQUIRED").

I think that it is better that the industry provides a common way to do
this, instead of "looking into the SDP in one case, looking into the
accept contact in another case" etc, that seems more complicated than
necessary .... =20

I noticed that in the BLISS group there was discussions on what
assumptions could be made - i.e. the particular example being discussed
was DND and whether it should be in the terminal or in the network.  If
something comes out of that - won't that be, by default, a service
definition or some sort.  It would be strange if that was the only
service definition allowed ...

Regards,

Steve
   =20

-----Original Message-----
From: Jonathan Rosenberg [mailto:jdrosen@cisco.com]=20
Sent: Monday, February 26, 2007 3:39 PM
To: Stephen Terrill (MI/EEM)
Cc: sipping@ietf.org
Subject: Re: [Sipping] Service Identification in SIP



Stephen Terrill (MI/EEM) wrote:

> Hi Andrew,
>=20
> I have the opposite view.  I think that if the originating network and

> the terminating network are going to interoperate, then they will have

> the same understanding of the service, hence in my view the same=20
> "identifier".  To have otherwise is problematic and complicated (the=20
> operators will have to have a higher admininstrative effort in order=20
> to understand how they should interconnect).

This is exactly what will lead to total balkanization and siloization of
IP-based communications systems.

We achieve interoperability by defining standards around how SIP works.=20
What you are basically saying above is that IETF standards are not
sufficient for operators to interwork and interconnect, and instead they
will need to all get together and discuss what services run on their
network, define identifiers for them, and agree on how those match up.=20
In what way is that not a recipe for complete balkanization and
interoperability failures? In what way is such a view not anathema to
the entire purpose of standardization?

If we have done our jobs right, all you need to interwork is to look at
the SIP message, and based on the standards we have defined, you will
know exactly what is being requested. If SDP contains audio and video,
you know that a phone is asking for an audio/video call. If it contains
MSRP then you know its asking for a text chat session. We go through a
lot of effort to standardize these things so that we have
interoperability between vendors and networks. We carefully design
fallbacks and backwards compatibility schemes so that we get
least-common-denominator service when heterogeneous devices
interconnect. I believe it is critical that we not bypass all of that
machinery by requiring that providers define and agree upon services for
interworking.

This is why I suggest that, if you have in fact defined a 'service'=20
which cannot interoperate at all unless the other side has understood
it, then you have in fact introduced a non-backwards compatible
extension to SIP. You signal that with Require. Does this make it hard
to go and standardize lots of these? ABSOLUTELY! That is what we want.=20
The bar should be quite high for defining non-backwards compatible
extensions, so that we preserve interoperability across heterogeneous
networks. I hope that there are as few of these as possible, not lots.=20
If PoC is already there, then so be it - let us define a tag for it, put
it in Require, and be done. I hope that is the first and last such
service tag we need to standardize.

-Jonathan R.


--=20
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 12:50:04 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLjyv-0003tk-KX; Mon, 26 Feb 2007 12:49:49 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLjyu-0003te-Ls
	for sipping@ietf.org; Mon, 26 Feb 2007 12:49:48 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLjyt-0001Kj-Fo
	for sipping@ietf.org; Mon, 26 Feb 2007 12:49:48 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	F0F3C2040F
	for <sipping@ietf.org>; Mon, 26 Feb 2007 18:49:44 +0100 (CET)
X-AuditID: c1b4fb3e-aeed1bb0000007e1-1f-45e31db86609 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	DD41E200B0
	for <sipping@ietf.org>; Mon, 26 Feb 2007 18:49:44 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 18:49:44 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification Model
Date: Mon, 26 Feb 2007 18:49:43 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD029263A2@esealmw111>
In-Reply-To: <45E2F759.5030302@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification Model
Thread-Index: AcdZt6YpYvd7SJApQd6PmaEV8nmXcQAFi4JA
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: <sipping@ietf.org>
X-OriginalArrivalTime: 26 Feb 2007 17:49:44.0959 (UTC)
	FILETIME=[800980F0:01C759CE]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 9361ae25efd9cff62e36f70ef966350e
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi Paul,

Doesn't this assume that if I use MSRP, then I want avoid the "alter
first" model.  If messaging was just another media (which I think it is
in some cases), then it would be natural that the alerting is applied to
that as well.

Moving on from that, I agree on your observation of the key differences
- between the "alert first model" and the "tradition IM model".
Further, my understanding is that the services applied to the model
differ in both cases as well ...

Regards,

Steve.

-----Original Message-----
From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
Sent: Monday, February 26, 2007 4:06 PM
To: Stephen Terrill (MI/EEM)
Cc: sipping@ietf.org
Subject: Re: [Sipping] Service Identification Model



Stephen Terrill (MI/EEM) wrote:
> Hi,
>=20
> Just a comment on the Messaging here;  While I fully understand the=20
> view that isntant messaging is another media, I also understand that=20
> the evolution of messaging in OMA is that instant messaging is actally

> defined as a seperate service to telephony.  SMS->MMS->Instant=20
> messaging.

There's that "S" word again!  In this case, by "different service" do
you mean "different UI", or "different billing model", or "different
implementation model"?

I can understand that user expectations will be for a different UI,
though I think providers who can support both the traditional UIs and a
unified UI will be more popular in the long run.

I can also understand "different billing model", although I am afraid
what that really means is an intent to gouge the user for more $$$ than
is justified by what the provider actually provides. But I am not
convinced anything special is needed to support whatever billing model
is desired.

Its the "different implementation model" that is most distressing. The
users don't care about the implementation model, so any history in that
regard is irrelevant.

Regarding the UI, I see the following key differences:
- telephony traditionally uses an "alert first" model, and exchanges
   media only after the recipient as accepted the incoming call.
   This is presumed to be applicable to video as well.

- in contrast, with traditional UIs for IM, presence may be
   used as a hint whether the target will accept the message,
   and the first message is sent without any *guarantee* that it
   will be accepted, or seen immediately. Depending on which IM
   model you adhere to, the message may be rejected if the recipient
   isn't ready to receive, or it may be stored for later delivery.
   With SMS/MMS and page mode SIP messaging, every message is a
   "first" message, but this may be disguised with a session oriented
   UI, or not. On the receiving side, there may be an implicit "auto
   answer" that accepts the message and displays it without alerting.
   Or there may be an "alert" that messages are present, but viewing
   them is more like viewing emails.

- The UI for PoC is in many ways similar to IM.

It is clear that people would be upset if their familiar UI paradigms
were changed. The question is whether that has anything to do with the
signaling. Ideally this would be strictly a local issue for a UA to
implement.

In the case of page mode SIP messaging, using MESSAGE out of dialog,
this is straightforward. MESSAGE can also be used in-dialog, though the
occasions when this is appropriate are not so clear. From a UI
perspective, having a separate UI for page mode IM vs voice is obviously
straightforward. Some innovation in UI design is required to handle page
mode IM within a dialog set up for other reasons - i.e. voice. But no
"service" is needed in either case. A UI that doesn't support
origination of page mode messages in-dialog could nevertheless support
them on receipt simply by treating them the same as out-of-dialog
messages.

IMO there remains work to do on session mode IM using MSRP, in terms of
mapping it to expected UIs. Mapping to the model where there separate
alerting before the first message is delivered is not straightforward.=20
It could be done in at least a couple of ways:
- the UI could be set up to "auto answer" INVITEs with MSRP,
   and then present the first message in the traditional way.
   The sender would collect the first message before sending
   the invite, and then send it as soon as an MSRP session is
   established.

- the first message could be sent as "early media", and that
   early media could be displayed in lieu of some other form of
   alerting. Sending a response from the UI could be treated as
   an indication to answer the incoming call.

My point here is that there is work to be done, but I don't see why
there is need for "service identification" to make it work. In fact, the
similarity between PoC and IM on one hand, and PoC and voice on the
other hand, suggests that there is a lot of value in figuring out how to
unify these rather than partitioning them.

After all, who here hasn't been having an IM session and wanted to
expand it to a voice session as well? This is natural in SIP if the two
media are not partitioned into different "services".

	Paul


When received, the display of the message
   serves as its own "alert".

> Cheers,
>=20
> ..//Steve
>=20
> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
> Sent: Friday, February 23, 2007 4:12 PM
> To: Byron Campen
> Cc: Andrew Allen; sipping@ietf.org
> Subject: Re: [Sipping] Service Identification Model
>=20
>=20
>=20
> Byron Campen wrote:
>>> 1. Basic Telephony (Conversational QoS Voice only with telephony=20
>>> feature
>>> set)
>>> 2. Multimedia Telephony (Conversational QoS Voice plus video and=20
>>> other media streams with telephony feature set)
>=20
> Why are 1 & 2 different services? 1 is a proper subset of 2.
>=20
>>> 3. VoIP (Background QoS with no telephony features)
>=20
> And why is 3 different from 1 and 2? Are you saying that to be able to

> use a feature I must decide *before* I make the call that I might want

> to?
>=20
> Even if that is the case, that sounds like a policy issue between a=20
> user and his own provider, rather than an e2e negotiation. (My=20
> features are provided by my UA and/or my provider.)
>=20
>>> 4. SMS
>>> 5. MMS
>=20
> Why are 4 and 5 different services? They are both datagrams. The=20
> difference in media types is irrelevant. If its delivered via MESSAGE,

> then it is already distinguishable from calls. This should be able to=20
> be part of the same service as IM.
>=20
>>> 6. Instant Messaging
>=20
> Is just another media type. Like voice and video, these can be=20
> negotiated for the session. Why balkanize as a different service?
>=20
>>> 7. Push to Talk (PoC)
>=20
> We've beaten this one to death already. Current implementation forces=20
> need for essentially an overlay network to support it. But its far=20
> from an exemplar of how to do it.
>=20
>>> 8. Network Game 1 (e.g Chess)
>>> 9. Network Game 2 (e.g Martial arts game) 10. Network Game 3 (e.g=20
>>> Dungeons & Dragons)
>=20
> I tend to agree with Eric that service URIs are good for these if=20
> there is no better way. But in practice I think these could easily be=20
> distinguished in SDP using o/a.
>=20
>>> 11. Push Service 1 (e.g. Traffic information service) 12. Push=20
>>> Service 2 (e.g. Weather information service) 13. Push Service 3 (e.g
>=20
>>> News information service) 14. Push Service 4 (e.g  Stock information
>>> service)
>=20
> There may be *something* here. I need to know more about the=20
> characteristics of these. If they consist of human intelligible media,

> then I think they could adequately be dealt with via callerprefs.=20
> (This may be a bit like the Actor feature tag.) Specifically,=20
> callerprefs are fine if its ok that the call end up elsewhere if the=20
> callerprefs selection doesn't happen or there isn't any target with
that capability.
>=20
> If the information isn't intelligible media, then it ought to be=20
> distinguishable as a special media type in the offer.
>=20
>>> 15. Mobile TV service
>=20
> AFAIK, the station doesn't call me to deliver me tv service. I would=20
> be initiating the call. And presumably it will be to a target service=20
> URI or URN. So I see no need for a service id here.
>=20
>>> So there's 15 which may all be different applications on a handset=20
>>> and all those are existing services or those already envisioned.
>=20
> So I only see:
>=20
> - 1 for basic telephony, multimedia telephony, SMS, MMS, IM
>=20
> - *maybe* something for games
>=20
> - *maybe* something for push services
>=20
> 	Paul
>=20
>>     I only see 10 here that would need to ever send a register
> request.=20
>> (11-15 are strictly client-server) 4-6 will have distinct=20
>> Content-Type
>=20
>> headers and payloads. You would need only 1 UA for these. So we're=20
>> down to 8. 8-10 will, once again, be distinct when you look at the=20
>> payload and Content-Type. So we're down to 6. There is no reason at=20
>> all that the UA that handles voice is incapable of handling IM, and=20
>> you can tell the difference by looking at the request type, so we're=20
>> down to 5. As for the voice stuff, you can distinguish 2 from=20
>> anything
>=20
>> else through the payload. So, we could quite easily roll 1-2
together.
>=20
>> (4) 3 can be distinguished by watching for any in-dialog call-control

>> stuff. (Having a service identifier that essentially says "I promise=20
>> not to use any fancy telephony features on this call. No, really, I=20
>> promise!" is never going to work) So we're down to 3.
>>
>>     SIP is more than flexible enough to handle all of this stuff in a

>> single UA. The UAs are perfectly capable of figuring out what
> "service"
>> the other wants to use; why are your middleboxes unable to do the
> same?=20
>> Adding a parallel method to make things easy on the middleboxes is=20
>> just asking for trouble, because you're conveying the same=20
>> information
> twice.
>> It is ripe for abuse. Of course, the easy way to do all this is to=20
>> just let UAs send what they want, and charge per KB, but hey, what do

>> I know about simplicity, flexibility, scalability,=20
>> ease-of-deployment,
>=20
>> and customer satisfaction? I'm just a IETFer.
>>
>> Best regards,
>> Byron Campen
>>
>>
>>> -----Original Message-----
>>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
>>> Sent: Thursday, February 22, 2007 7:05 PM
>>> To: Andrew Allen
>>> Cc: Eric Burger; sipping@ietf.org
>>> Subject: Re: [Sipping] Service Identification Model
>>>
>>>
>>>
>>> Andrew Allen wrote:
>>>> Hi Eric,
>>>>
>>>> It seems your prefer the approach of using URIs to identify=20
>>>> services
>=20
>>>> which is part of Jonathan's draft as is the use of Require for=20
>>>> those
>=20
>>>> services which simply do not interoperate unless the UAS=20
>>>> understands
>>> the
>>>> service.
>>>>
>>>> Where I am concerned is that you seem to be suggesting that the UA=20
>>>> separately register an AoR for each service it supports using
>>> REGISTER.
>>>> Certainly for cellular that wouldn't be acceptable as 10 or more=20
>>>> services could easily be supported by a single device.
>>> Andrew, can you give me an example of 10 "services" on the device=20
>>> that need to be distinguished from one another in this way - that=20
>>> aren't just
>>>
>>> optional behavior of a particular service? (E.g. I see no reason why

>>> "audio" and "audio/video" need to be separate services.)
>>>
>>> I guess as things stand, PoC must be a separate service. And=20
>>> apparently Chess. What else?
>>>
>>>     Paul
>>>
>>>> Registrations
>>>> create a significant load in the IMS network because its not just a
>>> SIP
>>>> REGISTER but also requires a lot of interactions with other network

>>>> elements such as the HSS and Notifications for reg-events etc. Also

>>>> having an order of magnitude increase in multiple registrations is
>>> going
>>>> to be a big issue for the avalanche restart problem for example=20
>>>> when
>=20
>>>> a cell site goes down and comes back into service.
>>>>
>>>> I think Jonathan's approach gives us the possibility for the=20
>>>> registrar to create the Service URIs based on the feature=20
>>>> parameters
>=20
>>>> included in the Contact of the REGISTER.
>>>>
>>>> I think the argument over whether the model is one UA with multiple

>>>> services or multiple UAs each representing a separate service is=20
>>>> somewhat academic. As I understand a UA can register multiple AoRs=20
>>>> and can have other AoRs configured or otherwise provided by the=20
>>>> service provider.
>>>>
>>>> Andrew
>>>>
>>>> -----Original Message-----
>>>> From: Eric Burger [mailto:eburger@bea.com]
>>>> Sent: Wednesday, February 21, 2007 1:29 PM
>>>> To: sipping@ietf.org
>>>> Subject: [Sipping] Service Identification Model
>>>>
>>>> Folks have been pestering me for a while to look at the Service=20
>>>> Identification issue.  I finally have gotten around to it.
>>>>
>>>>
>>>> I guess I have a fundamental problem with the principal assertion=20
>>>> of
>>> the
>>>> service-identification draft.  Namely, the assertion that there is=20
>>>> a
>=20
>>>> single SIP Layer, separate and distinct from two services.  Here is

>>>> Figure 2 from the draft:
>>>>
>>>>                            +---------------------------------+
>>>>                            |                                 |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | |      UI     | |      UI     | |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | |             | |             | |
>>>>                            | |  Service 1  | |  Service 2  | |
>>>>                            | |             | |             | |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | +-----------------------------+ |
>>>>                            | |                             | |
>>>>                            | |             SIP             | |
>>>>                            | |            Layer            | |
>>>>                            | |                             | |
>>>>                            | +-----------------------------+ |
>>>>                            |                                 |
>>>>                            +---------------------------------+
>>>>
>>>>                                      Physical Device
>>>>
>>>> I would offer this drawing confuses a common SIP Stack with a=20
>>>> common
>=20
>>>> User Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use
>>>> case:
>>>>
>>>>> ... In addition users may have
>>>>> multiple devices only one of which has the application that
>>>> understands
>>>>> the game. Other devices that the user has registered against his=20
>>>>> AoR
>>>> may
>>>>> handle audio for telephony and other voice applications but if=20
>>>>> Alice invites Bob to play her favourite game it is not very=20
>>>>> interesting if
>>>> Bob
>>>>> accepts the gaming session on a device that can only accept the=20
>>>>> audio stream. Also even when the terminal that has the game=20
>>>>> application receives the INVITE the correct application in that=20
>>>>> terminal needs to
>>>> be
>>>>> identified to accept the session.
>>>> I would offer that this is the clinching argument that the request=20
>>>> is, in fact, addressing a SEPARATE USER AGENT, not a special=20
>>>> SERVICE
>=20
>>>> at a
>>> User
>>>> Agent.
>>>>
>>>> Same argument for PoC versus a regular voice call: One is looking=20
>>>> for
>>> a
>>>> PoC
>>>> User Agent, NOT the PoC service at a given User Agent.
>>>>
>>>> Thus, the picture really looks like the following:
>>>>
>>>>                            +---------------------------------+
>>>>                            |                                 |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | |    UI       | |    UI       | |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | |             | |             | |
>>>>                            | |    UAS 1    | |    UAS 2    | |
>>>>                            | |             | |             | |
>>>>                            | +-------------+ +-------------+ |
>>>>                            | +-----------------------------+ |
>>>>                            | |                             | |
>>>>                            | |             SIP             | |
>>>>                            | |            Stack            | |
>>>>                            | |                             | |
>>>>                            | +-----------------------------+ |
>>>>                            |                                 |
>>>>                            +---------------------------------+
>>>>
>>>>                                      Physical Device
>>>>
>>>>
>>>> Another argument for different User Agents is the whole "route=20
>>>> based
>>> on
>>>> SDP"
>>>> argument.  As the author of RFC 3458, I think I was first to make=20
>>>> the assertion that knowing the media type did NOT mean you know the

>>>> application to use.  I don't want your phone or proxy or B2BUA to=20
>>>> guess that
>>> because
>>>> I
>>>> have Chess Media, I want the Chess Application; I want to talk with
>>> your
>>>> Chess Application.
>>>>
>>>> Now, is the Chess Application a Preference, Requirement, or the=20
>>>> thing that I am addressing?  The answer drives one towards using=20
>>>> caller prefs, the Require header, or the Request-URI.
>>>>
>>>> I would offer that it is not a preference: I really want the chess=20
>>>> application; I probably do not want to talk to you.
>>>>
>>>> That leaves us with the Require header or a Request-URI.  My=20
>>>> leanings are not surprisingly towards the RFC 4240 approach.
>>>> Frankly, if you are building an IMS (as an example) that can handle

>>>> 10,000,000
>>> subscribers,
>>>> you
>>>> should be able to handle 100,000,000 registrations.  That order of=20
>>>> magnitude is way easier than the 10,000 to 10,000,000 jump.  I=20
>>>> don't
>=20
>>>> buy the "registering each service is too much of a burden on the=20
>>>> operator"
>>>> argument,
>>>> because it is not that much of a burden, and the network actually=20
>>>> does the right thing.  Better yet, all of the corner cases work.
>>>> Best of all, the IETF doesn't have to build an IANA Registry and=20
>>>> process a thousand PS documents to register
urn:service:myco:widget.
>>>>
>>>> I suppose I could live with Require:, only if you are way too lazy=20
>>>> to register each application instance.  It does semantically work,
>>> because
>>>> if I
>>>> want to do PoC and you do not do PoC, I'll get a 4xx response, to
>>> which
>>>> I
>>>> could try a regular call.
>>>>
>>>>
>>> ____________________________________________________________________
>>> _
>>> __
>>>> Notice:  This email message, together with any attachments, may
>>> contain
>>>> information  of  BEA Systems,  Inc.,  its subsidiaries  and
>>> affiliated
>>>> entities,  that may be confidential,  proprietary,  copyrighted
>>> and/or
>>>> legally privileged, and is intended solely for the use of the
>>> individual
>>>> or entity named in this message. If you are not the intended
>>> recipient,
>>>> and have received this message in error, please immediately return
>>> this
>>>> by email and then delete it.
>>>>
>>>> _______________________________________________
>>>> Sipping mailing list =20
>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>> This list is for NEW development of the application of SIP Use=20
>>>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
>>>> sip@ietf.org for new developments of core SIP
>>>>
>>>>
>>>>
>>>> -------------------------------------------------------------------
>>>> -
>>>> - This transmission (including any attachments) may contain=20
>>>> confidential
>>> information, privileged material (including material protected by=20
>>> the
>=20
>>> solicitor-client or other applicable privileges), or constitute=20
>>> non-public information. Any use of this information by anyone other=20
>>> than the intended recipient is prohibited. If you have received this

>>> transmission in error, please immediately reply to the sender and=20
>>> delete this information from your system. Use, dissemination,=20
>>> distribution, or reproduction of this transmission by unintended=20
>>> recipients is not authorized and may be unlawful.
>>>> _______________________________________________
>>>> Sipping mailing list =20
>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>> This list is for NEW development of the application of SIP Use=20
>>>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
>>>> sip@ietf.org for new developments of core SIP
>>>>
>>>
>>>
>>> --------------------------------------------------------------------
>>> - This transmission (including any attachments) may contain=20
>>> confidential information, privileged material (including material=20
>>> protected by the solicitor-client or other applicable privileges),=20
>>> or
>=20
>>> constitute non-public information. Any use of this information by=20
>>> anyone other than the intended recipient is prohibited. If you have=20
>>> received this transmission in error, please immediately reply to the

>>> sender and delete this information from your system. Use,=20
>>> dissemination, distribution, or reproduction of this transmission by

>>> unintended recipients is not authorized and may be unlawful.
>>>
>>> _______________________________________________
>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP Use=20
>>> sip-implementors@cs.columbia.edu for questions on current sip Use=20
>>> sip@ietf.org for new developments of core SIP
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use=20
> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 13:06:50 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLkEQ-000057-MZ; Mon, 26 Feb 2007 13:05:50 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLkEO-000051-UE
	for sipping@ietf.org; Mon, 26 Feb 2007 13:05:48 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLkEI-00033E-8q
	for sipping@ietf.org; Mon, 26 Feb 2007 13:05:48 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	7B2BA2056E; Mon, 26 Feb 2007 19:05:37 +0100 (CET)
X-AuditID: c1b4fb3c-b17cbbb0000007de-22-45e321712cd2 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	59635203F6; Mon, 26 Feb 2007 19:05:37 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 19:05:37 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification: offerless INVITE case
Date: Mon, 26 Feb 2007 19:05:34 +0100
Message-ID: <7374777208BDC7449D5620EF94232567024653BB@esealmw113.eemea.ericsson.se>
In-Reply-To: <45E31693.40207@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification: offerless INVITE case
Thread-Index: AcdZy08DeR74S1MtR1eht5SvP4h/pQABFjKw
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>,
	"Jonathan Rosenberg" <jdrosen@cisco.com>
X-OriginalArrivalTime: 26 Feb 2007 18:05:37.0260 (UTC)
	FILETIME=[B7A71AC0:01C759D0]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 6ffdee8af20de249c24731d8414917d3
Cc: IETF Sipping List <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,=20

>I too was thinking that INVITE/Replaces would be the way to=20
>get a new dialog that could include app servers not in the=20
>existing call, when that is needed. But I hadn't worked out=20
>how to know to do that. The 3xx does seem like a good way to=20
>handle that. The use of 3xx as a response to an in-dialog=20
>request is not at all well specified. But this could be a=20
>good motivation to clear that up.

I agree that the re-INVITE 3xx response would require some thinking.

For example, an initial INVITE 3xx response may not traverse all the way
back to the UA. But, I guess a re-INVITE 3xx would have to be sent all
the way back to the UA, since the UA would create a new dialog.

The question is also what happens to the original dialog. Does the UA
have to send BYE on the original dialog, or is it terminated
"automatically"?

If terminated automatically, what happens to the nodes situated AFTER
the proxy that sent 3xx? They have no clue about the 3xx being sent, and
there is no guarantee the re-INVITE will traverse the same nodes (or
even reach the same remote UA).

Etc etc etc.

Regards,

Christer












>=20
> 	Paul
>=20
> Jonathan Rosenberg wrote:
> > Andrew made a great point in one of his mails which merits=20
> promotion=20
> > into its own thread.
> >=20
> > I and others have asserted that you should be able to identify the=20
> > service by things naturally in the request, and in=20
> particular, SDP has=20
> > come up as a key thing. For example, in a multiplayer game, there=20
> > might be an m-line for the particular proprietary protocol used for=20
> > game moves. Or, if its MSRP, you might have specific=20
> content types for=20
> > game moves. When present, those exactly indicate what the=20
> service is.
> >=20
> > But what do we do about offerless INVITE? In particular, there are=20
> > cases where we need to invoke an application server to handle a=20
> > request, and what if we don't have the SDP in the INVITE?=20
> Its too late=20
> > to insert an application server once the INVITE has already=20
> passed to=20
> > the called party and the 200 OK is sent.
> >=20
> > This is a great point. I'd rather just fix this little problem, the=20
> > define an entire notion of services and service identifiers=20
> that are=20
> > not necessary, just because we have offerless INVITEs.
> >=20
> > There are lots of ways to handle this problem. My preferred=20
> approach=20
> > would be that offerless INVITE contain a capability=20
> description, using=20
> > Flemmings new capability framework. The INVITE would contain a body=20
> > that  looks like SDP, but has a different content type and=20
> > content-disposition. It lists the media streams, codecs and other=20
> > capabilities that exist. Based on this, the called party makes the=20
> > offer in a 2xx or 18x like normal, and the answer in the=20
> ACK/PRACK as normal.
> > It is completely backwards compatible.
> >=20
> > Now, a related issue is what happens if a session starts with audio=20
> > only, and then the user wants to turn this into a game? Its=20
> too late=20
> > to add new application servers into the call path, which=20
> might need to=20
> > be there for games. What can we do?
> >=20
> > Here too, there is a simple answer. If a network provider=20
> requires app=20
> > servers to get inserted, or a phone requires dispatch to a=20
> different=20
> > application, it would receive a re-INVITE to add the gaming stream,=20
> > and then 3xx it. The Contact would contain an embedded=20
> Replaces header=20
> > with the dialog IDs for the call in progress. THe=20
> originating UA would=20
> > then do a totally new INVITE that replaces the call in=20
> progress. This=20
> > would allow the network and/or phone to redispatch, yet provide a=20
> > seamless user experience. Billing can be done however the operators=20
> > want, since the signaling is very clear on exactly what=20
> happened and when.
> > Furthermore, the replaces mechansim would only get used if=20
> there was=20
> > someone in the call path that needed to see it.
> >=20
> > I'd rather specify things like this, than invent an entire services=20
> > framework that would lead to balkanization and non-interoperable=20
> > islands of SIP.
> >=20
> > -Jonathan R.
> >=20
> >=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 13:12:42 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLkL3-0001fK-VW; Mon, 26 Feb 2007 13:12:41 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLkL2-0001fB-7o
	for sipping@ietf.org; Mon, 26 Feb 2007 13:12:40 -0500
Received: from sj-iport-2-in.cisco.com ([171.71.176.71]
	helo=sj-iport-2.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HLkKz-0003vL-2f
	for sipping@ietf.org; Mon, 26 Feb 2007 13:12:40 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by sj-iport-2.cisco.com with ESMTP; 26 Feb 2007 10:12:23 -0800
X-IronPort-AV: i="4.14,221,1170662400"; 
	d="scan'208"; a="362969937:sNHT3429569596"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1QICNbB026196; 
	Mon, 26 Feb 2007 13:12:23 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1QICIVV013703; 
	Mon, 26 Feb 2007 13:12:19 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 13:12:18 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 13:12:18 -0500
Message-ID: <45E32300.7030500@cisco.com>
Date: Mon, 26 Feb 2007 13:12:16 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Stephen Terrill (MI/EEM)" <stephen.terrill@ericsson.com>
Subject: Re: [Sipping] Service Identification Model
References: <989763567E51AB48BE015DF7BB960CAD029263A2@esealmw111>
In-Reply-To: <989763567E51AB48BE015DF7BB960CAD029263A2@esealmw111>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 18:12:18.0541 (UTC)
	FILETIME=[A6D5B9D0:01C759D1]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=25109; t=1172513543;
	x=1173377543; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20Model
	|Sender:=20
	|To:=20=22Stephen=20Terrill=20(MI/EEM)=22=20<stephen.terrill@ericsson.com
	>; bh=ytmWuDn/ucSlnu3DLDtTIux1IyX3PRo0gizzG/6j68Y=;
	b=QgkLR16BiARu8+AN8PGY89RD1d8YQYM1Akhh+csZlKZNoxL5NEP4o8oFjC+Gw1ssnqQI9+cA
	YgHy3PjyjCvGeJCZTvAbSmp4znjVdy4xz3FGK2ihrs7SIsrCRbAKvYxC;
Authentication-Results: rtp-dkim-1; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 1ed37b243475b9c4ffb6a3f90050819d
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Stephen Terrill (MI/EEM) wrote:
> Hi Paul,
> 
> Doesn't this assume that if I use MSRP, then I want avoid the "alter
I don't think you meant "alter". Maybe "alert" or "answer"?
> first" model. 

I think there is a challenge here to make the decision independent on 
the two ends. You, or the UI you choose, may prefer to just display the 
first message as the alert. I may choose to be alerted first, so I can 
reject before being exposed to potentially offensive content.

PoC has managed to allow both.

> If messaging was just another media (which I think it is
> in some cases), then it would be natural that the alerting is applied to
> that as well.

It seems to me that the alerting behavior is only loosely connected to 
the media available, and more tightly coupled to the capabilities of the 
device and the preferred UI style of the user of the device.

I think a better, and more unified, result may come from considering 
this more broadly, rather than just splitting into separate services and 
attempting to make a choice for each.

> Moving on from that, I agree on your observation of the key differences
> - between the "alert first model" and the "tradition IM model".
> Further, my understanding is that the services applied to the model
> differ in both cases as well ...

I don't know about that.

Consider message taking when there is no other UA willing to answer. It 
seems to me that capturing IM for someone not online, and recording VM 
for someone not online are really two aspects of the same service. (In 
3840 this was called msg-taker in order to make it media independent.)

Transfer and hold and conference all apply uniformly. As does call 
screening.

And in any case, those are not "services" in the same sense.

	Paul

> Regards,
> 
> Steve.
> 
> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
> Sent: Monday, February 26, 2007 4:06 PM
> To: Stephen Terrill (MI/EEM)
> Cc: sipping@ietf.org
> Subject: Re: [Sipping] Service Identification Model
> 
> 
> 
> Stephen Terrill (MI/EEM) wrote:
>> Hi,
>>
>> Just a comment on the Messaging here;  While I fully understand the 
>> view that isntant messaging is another media, I also understand that 
>> the evolution of messaging in OMA is that instant messaging is actally
> 
>> defined as a seperate service to telephony.  SMS->MMS->Instant 
>> messaging.
> 
> There's that "S" word again!  In this case, by "different service" do
> you mean "different UI", or "different billing model", or "different
> implementation model"?
> 
> I can understand that user expectations will be for a different UI,
> though I think providers who can support both the traditional UIs and a
> unified UI will be more popular in the long run.
> 
> I can also understand "different billing model", although I am afraid
> what that really means is an intent to gouge the user for more $$$ than
> is justified by what the provider actually provides. But I am not
> convinced anything special is needed to support whatever billing model
> is desired.
> 
> Its the "different implementation model" that is most distressing. The
> users don't care about the implementation model, so any history in that
> regard is irrelevant.
> 
> Regarding the UI, I see the following key differences:
> - telephony traditionally uses an "alert first" model, and exchanges
>    media only after the recipient as accepted the incoming call.
>    This is presumed to be applicable to video as well.
> 
> - in contrast, with traditional UIs for IM, presence may be
>    used as a hint whether the target will accept the message,
>    and the first message is sent without any *guarantee* that it
>    will be accepted, or seen immediately. Depending on which IM
>    model you adhere to, the message may be rejected if the recipient
>    isn't ready to receive, or it may be stored for later delivery.
>    With SMS/MMS and page mode SIP messaging, every message is a
>    "first" message, but this may be disguised with a session oriented
>    UI, or not. On the receiving side, there may be an implicit "auto
>    answer" that accepts the message and displays it without alerting.
>    Or there may be an "alert" that messages are present, but viewing
>    them is more like viewing emails.
> 
> - The UI for PoC is in many ways similar to IM.
> 
> It is clear that people would be upset if their familiar UI paradigms
> were changed. The question is whether that has anything to do with the
> signaling. Ideally this would be strictly a local issue for a UA to
> implement.
> 
> In the case of page mode SIP messaging, using MESSAGE out of dialog,
> this is straightforward. MESSAGE can also be used in-dialog, though the
> occasions when this is appropriate are not so clear. From a UI
> perspective, having a separate UI for page mode IM vs voice is obviously
> straightforward. Some innovation in UI design is required to handle page
> mode IM within a dialog set up for other reasons - i.e. voice. But no
> "service" is needed in either case. A UI that doesn't support
> origination of page mode messages in-dialog could nevertheless support
> them on receipt simply by treating them the same as out-of-dialog
> messages.
> 
> IMO there remains work to do on session mode IM using MSRP, in terms of
> mapping it to expected UIs. Mapping to the model where there separate
> alerting before the first message is delivered is not straightforward. 
> It could be done in at least a couple of ways:
> - the UI could be set up to "auto answer" INVITEs with MSRP,
>    and then present the first message in the traditional way.
>    The sender would collect the first message before sending
>    the invite, and then send it as soon as an MSRP session is
>    established.
> 
> - the first message could be sent as "early media", and that
>    early media could be displayed in lieu of some other form of
>    alerting. Sending a response from the UI could be treated as
>    an indication to answer the incoming call.
> 
> My point here is that there is work to be done, but I don't see why
> there is need for "service identification" to make it work. In fact, the
> similarity between PoC and IM on one hand, and PoC and voice on the
> other hand, suggests that there is a lot of value in figuring out how to
> unify these rather than partitioning them.
> 
> After all, who here hasn't been having an IM session and wanted to
> expand it to a voice session as well? This is natural in SIP if the two
> media are not partitioned into different "services".
> 
> 	Paul
> 
> 
> When received, the display of the message
>    serves as its own "alert".
> 
>> Cheers,
>>
>> ..//Steve
>>
>> -----Original Message-----
>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
>> Sent: Friday, February 23, 2007 4:12 PM
>> To: Byron Campen
>> Cc: Andrew Allen; sipping@ietf.org
>> Subject: Re: [Sipping] Service Identification Model
>>
>>
>>
>> Byron Campen wrote:
>>>> 1. Basic Telephony (Conversational QoS Voice only with telephony 
>>>> feature
>>>> set)
>>>> 2. Multimedia Telephony (Conversational QoS Voice plus video and 
>>>> other media streams with telephony feature set)
>> Why are 1 & 2 different services? 1 is a proper subset of 2.
>>
>>>> 3. VoIP (Background QoS with no telephony features)
>> And why is 3 different from 1 and 2? Are you saying that to be able to
> 
>> use a feature I must decide *before* I make the call that I might want
> 
>> to?
>>
>> Even if that is the case, that sounds like a policy issue between a 
>> user and his own provider, rather than an e2e negotiation. (My 
>> features are provided by my UA and/or my provider.)
>>
>>>> 4. SMS
>>>> 5. MMS
>> Why are 4 and 5 different services? They are both datagrams. The 
>> difference in media types is irrelevant. If its delivered via MESSAGE,
> 
>> then it is already distinguishable from calls. This should be able to 
>> be part of the same service as IM.
>>
>>>> 6. Instant Messaging
>> Is just another media type. Like voice and video, these can be 
>> negotiated for the session. Why balkanize as a different service?
>>
>>>> 7. Push to Talk (PoC)
>> We've beaten this one to death already. Current implementation forces 
>> need for essentially an overlay network to support it. But its far 
>> from an exemplar of how to do it.
>>
>>>> 8. Network Game 1 (e.g Chess)
>>>> 9. Network Game 2 (e.g Martial arts game) 10. Network Game 3 (e.g 
>>>> Dungeons & Dragons)
>> I tend to agree with Eric that service URIs are good for these if 
>> there is no better way. But in practice I think these could easily be 
>> distinguished in SDP using o/a.
>>
>>>> 11. Push Service 1 (e.g. Traffic information service) 12. Push 
>>>> Service 2 (e.g. Weather information service) 13. Push Service 3 (e.g
>>>> News information service) 14. Push Service 4 (e.g  Stock information
>>>> service)
>> There may be *something* here. I need to know more about the 
>> characteristics of these. If they consist of human intelligible media,
> 
>> then I think they could adequately be dealt with via callerprefs. 
>> (This may be a bit like the Actor feature tag.) Specifically, 
>> callerprefs are fine if its ok that the call end up elsewhere if the 
>> callerprefs selection doesn't happen or there isn't any target with
> that capability.
>> If the information isn't intelligible media, then it ought to be 
>> distinguishable as a special media type in the offer.
>>
>>>> 15. Mobile TV service
>> AFAIK, the station doesn't call me to deliver me tv service. I would 
>> be initiating the call. And presumably it will be to a target service 
>> URI or URN. So I see no need for a service id here.
>>
>>>> So there's 15 which may all be different applications on a handset 
>>>> and all those are existing services or those already envisioned.
>> So I only see:
>>
>> - 1 for basic telephony, multimedia telephony, SMS, MMS, IM
>>
>> - *maybe* something for games
>>
>> - *maybe* something for push services
>>
>> 	Paul
>>
>>>     I only see 10 here that would need to ever send a register
>> request. 
>>> (11-15 are strictly client-server) 4-6 will have distinct 
>>> Content-Type
>>> headers and payloads. You would need only 1 UA for these. So we're 
>>> down to 8. 8-10 will, once again, be distinct when you look at the 
>>> payload and Content-Type. So we're down to 6. There is no reason at 
>>> all that the UA that handles voice is incapable of handling IM, and 
>>> you can tell the difference by looking at the request type, so we're 
>>> down to 5. As for the voice stuff, you can distinguish 2 from 
>>> anything
>>> else through the payload. So, we could quite easily roll 1-2
> together.
>>> (4) 3 can be distinguished by watching for any in-dialog call-control
> 
>>> stuff. (Having a service identifier that essentially says "I promise 
>>> not to use any fancy telephony features on this call. No, really, I 
>>> promise!" is never going to work) So we're down to 3.
>>>
>>>     SIP is more than flexible enough to handle all of this stuff in a
> 
>>> single UA. The UAs are perfectly capable of figuring out what
>> "service"
>>> the other wants to use; why are your middleboxes unable to do the
>> same? 
>>> Adding a parallel method to make things easy on the middleboxes is 
>>> just asking for trouble, because you're conveying the same 
>>> information
>> twice.
>>> It is ripe for abuse. Of course, the easy way to do all this is to 
>>> just let UAs send what they want, and charge per KB, but hey, what do
> 
>>> I know about simplicity, flexibility, scalability, 
>>> ease-of-deployment,
>>> and customer satisfaction? I'm just a IETFer.
>>>
>>> Best regards,
>>> Byron Campen
>>>
>>>
>>>> -----Original Message-----
>>>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]
>>>> Sent: Thursday, February 22, 2007 7:05 PM
>>>> To: Andrew Allen
>>>> Cc: Eric Burger; sipping@ietf.org
>>>> Subject: Re: [Sipping] Service Identification Model
>>>>
>>>>
>>>>
>>>> Andrew Allen wrote:
>>>>> Hi Eric,
>>>>>
>>>>> It seems your prefer the approach of using URIs to identify 
>>>>> services
>>>>> which is part of Jonathan's draft as is the use of Require for 
>>>>> those
>>>>> services which simply do not interoperate unless the UAS 
>>>>> understands
>>>> the
>>>>> service.
>>>>>
>>>>> Where I am concerned is that you seem to be suggesting that the UA 
>>>>> separately register an AoR for each service it supports using
>>>> REGISTER.
>>>>> Certainly for cellular that wouldn't be acceptable as 10 or more 
>>>>> services could easily be supported by a single device.
>>>> Andrew, can you give me an example of 10 "services" on the device 
>>>> that need to be distinguished from one another in this way - that 
>>>> aren't just
>>>>
>>>> optional behavior of a particular service? (E.g. I see no reason why
> 
>>>> "audio" and "audio/video" need to be separate services.)
>>>>
>>>> I guess as things stand, PoC must be a separate service. And 
>>>> apparently Chess. What else?
>>>>
>>>>     Paul
>>>>
>>>>> Registrations
>>>>> create a significant load in the IMS network because its not just a
>>>> SIP
>>>>> REGISTER but also requires a lot of interactions with other network
> 
>>>>> elements such as the HSS and Notifications for reg-events etc. Also
> 
>>>>> having an order of magnitude increase in multiple registrations is
>>>> going
>>>>> to be a big issue for the avalanche restart problem for example 
>>>>> when
>>>>> a cell site goes down and comes back into service.
>>>>>
>>>>> I think Jonathan's approach gives us the possibility for the 
>>>>> registrar to create the Service URIs based on the feature 
>>>>> parameters
>>>>> included in the Contact of the REGISTER.
>>>>>
>>>>> I think the argument over whether the model is one UA with multiple
> 
>>>>> services or multiple UAs each representing a separate service is 
>>>>> somewhat academic. As I understand a UA can register multiple AoRs 
>>>>> and can have other AoRs configured or otherwise provided by the 
>>>>> service provider.
>>>>>
>>>>> Andrew
>>>>>
>>>>> -----Original Message-----
>>>>> From: Eric Burger [mailto:eburger@bea.com]
>>>>> Sent: Wednesday, February 21, 2007 1:29 PM
>>>>> To: sipping@ietf.org
>>>>> Subject: [Sipping] Service Identification Model
>>>>>
>>>>> Folks have been pestering me for a while to look at the Service 
>>>>> Identification issue.  I finally have gotten around to it.
>>>>>
>>>>>
>>>>> I guess I have a fundamental problem with the principal assertion 
>>>>> of
>>>> the
>>>>> service-identification draft.  Namely, the assertion that there is 
>>>>> a
>>>>> single SIP Layer, separate and distinct from two services.  Here is
> 
>>>>> Figure 2 from the draft:
>>>>>
>>>>>                            +---------------------------------+
>>>>>                            |                                 |
>>>>>                            | +-------------+ +-------------+ |
>>>>>                            | |      UI     | |      UI     | |
>>>>>                            | +-------------+ +-------------+ |
>>>>>                            | +-------------+ +-------------+ |
>>>>>                            | |             | |             | |
>>>>>                            | |  Service 1  | |  Service 2  | |
>>>>>                            | |             | |             | |
>>>>>                            | +-------------+ +-------------+ |
>>>>>                            | +-----------------------------+ |
>>>>>                            | |                             | |
>>>>>                            | |             SIP             | |
>>>>>                            | |            Layer            | |
>>>>>                            | |                             | |
>>>>>                            | +-----------------------------+ |
>>>>>                            |                                 |
>>>>>                            +---------------------------------+
>>>>>
>>>>>                                      Physical Device
>>>>>
>>>>> I would offer this drawing confuses a common SIP Stack with a 
>>>>> common
>>>>> User Agent.  To wit, consider Andrew Allen's (2/16/07 4:26 AM ) use
>>>>> case:
>>>>>
>>>>>> ... In addition users may have
>>>>>> multiple devices only one of which has the application that
>>>>> understands
>>>>>> the game. Other devices that the user has registered against his 
>>>>>> AoR
>>>>> may
>>>>>> handle audio for telephony and other voice applications but if 
>>>>>> Alice invites Bob to play her favourite game it is not very 
>>>>>> interesting if
>>>>> Bob
>>>>>> accepts the gaming session on a device that can only accept the 
>>>>>> audio stream. Also even when the terminal that has the game 
>>>>>> application receives the INVITE the correct application in that 
>>>>>> terminal needs to
>>>>> be
>>>>>> identified to accept the session.
>>>>> I would offer that this is the clinching argument that the request 
>>>>> is, in fact, addressing a SEPARATE USER AGENT, not a special 
>>>>> SERVICE
>>>>> at a
>>>> User
>>>>> Agent.
>>>>>
>>>>> Same argument for PoC versus a regular voice call: One is looking 
>>>>> for
>>>> a
>>>>> PoC
>>>>> User Agent, NOT the PoC service at a given User Agent.
>>>>>
>>>>> Thus, the picture really looks like the following:
>>>>>
>>>>>                            +---------------------------------+
>>>>>                            |                                 |
>>>>>                            | +-------------+ +-------------+ |
>>>>>                            | |    UI       | |    UI       | |
>>>>>                            | +-------------+ +-------------+ |
>>>>>                            | +-------------+ +-------------+ |
>>>>>                            | |             | |             | |
>>>>>                            | |    UAS 1    | |    UAS 2    | |
>>>>>                            | |             | |             | |
>>>>>                            | +-------------+ +-------------+ |
>>>>>                            | +-----------------------------+ |
>>>>>                            | |                             | |
>>>>>                            | |             SIP             | |
>>>>>                            | |            Stack            | |
>>>>>                            | |                             | |
>>>>>                            | +-----------------------------+ |
>>>>>                            |                                 |
>>>>>                            +---------------------------------+
>>>>>
>>>>>                                      Physical Device
>>>>>
>>>>>
>>>>> Another argument for different User Agents is the whole "route 
>>>>> based
>>>> on
>>>>> SDP"
>>>>> argument.  As the author of RFC 3458, I think I was first to make 
>>>>> the assertion that knowing the media type did NOT mean you know the
> 
>>>>> application to use.  I don't want your phone or proxy or B2BUA to 
>>>>> guess that
>>>> because
>>>>> I
>>>>> have Chess Media, I want the Chess Application; I want to talk with
>>>> your
>>>>> Chess Application.
>>>>>
>>>>> Now, is the Chess Application a Preference, Requirement, or the 
>>>>> thing that I am addressing?  The answer drives one towards using 
>>>>> caller prefs, the Require header, or the Request-URI.
>>>>>
>>>>> I would offer that it is not a preference: I really want the chess 
>>>>> application; I probably do not want to talk to you.
>>>>>
>>>>> That leaves us with the Require header or a Request-URI.  My 
>>>>> leanings are not surprisingly towards the RFC 4240 approach.
>>>>> Frankly, if you are building an IMS (as an example) that can handle
> 
>>>>> 10,000,000
>>>> subscribers,
>>>>> you
>>>>> should be able to handle 100,000,000 registrations.  That order of 
>>>>> magnitude is way easier than the 10,000 to 10,000,000 jump.  I 
>>>>> don't
>>>>> buy the "registering each service is too much of a burden on the 
>>>>> operator"
>>>>> argument,
>>>>> because it is not that much of a burden, and the network actually 
>>>>> does the right thing.  Better yet, all of the corner cases work.
>>>>> Best of all, the IETF doesn't have to build an IANA Registry and 
>>>>> process a thousand PS documents to register
> urn:service:myco:widget.
>>>>> I suppose I could live with Require:, only if you are way too lazy 
>>>>> to register each application instance.  It does semantically work,
>>>> because
>>>>> if I
>>>>> want to do PoC and you do not do PoC, I'll get a 4xx response, to
>>>> which
>>>>> I
>>>>> could try a regular call.
>>>>>
>>>>>
>>>> ____________________________________________________________________
>>>> _
>>>> __
>>>>> Notice:  This email message, together with any attachments, may
>>>> contain
>>>>> information  of  BEA Systems,  Inc.,  its subsidiaries  and
>>>> affiliated
>>>>> entities,  that may be confidential,  proprietary,  copyrighted
>>>> and/or
>>>>> legally privileged, and is intended solely for the use of the
>>>> individual
>>>>> or entity named in this message. If you are not the intended
>>>> recipient,
>>>>> and have received this message in error, please immediately return
>>>> this
>>>>> by email and then delete it.
>>>>>
>>>>> _______________________________________________
>>>>> Sipping mailing list  
>>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>> This list is for NEW development of the application of SIP Use 
>>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>>> sip@ietf.org for new developments of core SIP
>>>>>
>>>>>
>>>>>
>>>>> -------------------------------------------------------------------
>>>>> -
>>>>> - This transmission (including any attachments) may contain 
>>>>> confidential
>>>> information, privileged material (including material protected by 
>>>> the
>>>> solicitor-client or other applicable privileges), or constitute 
>>>> non-public information. Any use of this information by anyone other 
>>>> than the intended recipient is prohibited. If you have received this
> 
>>>> transmission in error, please immediately reply to the sender and 
>>>> delete this information from your system. Use, dissemination, 
>>>> distribution, or reproduction of this transmission by unintended 
>>>> recipients is not authorized and may be unlawful.
>>>>> _______________________________________________
>>>>> Sipping mailing list  
>>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>> This list is for NEW development of the application of SIP Use 
>>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>>> sip@ietf.org for new developments of core SIP
>>>>>
>>>>
>>>> --------------------------------------------------------------------
>>>> - This transmission (including any attachments) may contain 
>>>> confidential information, privileged material (including material 
>>>> protected by the solicitor-client or other applicable privileges), 
>>>> or
>>>> constitute non-public information. Any use of this information by 
>>>> anyone other than the intended recipient is prohibited. If you have 
>>>> received this transmission in error, please immediately reply to the
> 
>>>> sender and delete this information from your system. Use, 
>>>> dissemination, distribution, or reproduction of this transmission by
> 
>>>> unintended recipients is not authorized and may be unlawful.
>>>>
>>>> _______________________________________________
>>>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>>> This list is for NEW development of the application of SIP Use 
>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>> sip@ietf.org for new developments of core SIP
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP Use 
>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>> sip@ietf.org for new developments of core SIP
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP Use 
>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>> sip@ietf.org for new developments of core SIP
>>
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 13:18:07 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLkQ5-0005tm-Uv; Mon, 26 Feb 2007 13:17:53 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLkQ3-0005tN-EC
	for sipping@ietf.org; Mon, 26 Feb 2007 13:17:51 -0500
Received: from ondar.cablelabs.com ([192.160.73.61])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLkPx-0004gi-Jc
	for sipping@ietf.org; Mon, 26 Feb 2007 13:17:51 -0500
Received: from kyzyl.cablelabs.com (kyzyl.cablelabs.com [10.253.0.7])
	by ondar.cablelabs.com (8.13.8/8.13.8) with ESMTP id l1QIHiTo007193
	for <sipping@ietf.org>; Mon, 26 Feb 2007 11:17:44 -0700 (MST)
Received: from srvxchg.cablelabs.com (10.5.0.20)
	by kyzyl.cablelabs.com (F-Secure/fsigk_smtp/488/kyzyl.cablelabs.com);
	Mon, 26 Feb 2007 11:17:44 -0700 (MST)
X-Virus-Status: clean(F-Secure/fsigk_smtp/488/kyzyl.cablelabs.com)
X-MimeOLE: Produced By Microsoft Exchange V6.0.6603.0
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Date: Mon, 26 Feb 2007 11:17:43 -0700
Message-ID: <CD6CE349CFD30D40BF5E13B3E0D8480402270C7A@srvxchg.cablelabs.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: SIPPING-CONFIG on the agenda?
Thread-Index: AcdZ0mh1ZxtnTAQnR3+wS2z6Uy3smw==
From: "Sumanth Channabasappa" <sumanth@cablelabs.com>
To: "IETF Sipping List" <sipping@ietf.org>
X-Approved: ondar
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 7aefe408d50e9c7c47615841cb314bed
Subject: [Sipping] SIPPING-CONFIG on the agenda?
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Chairs,

Just a quick check to make sure that we will be able to discuss the SIP
Configuration Framework at IETF#68 (i.e., as an agenda item).

Thanks!
- S

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 13:18:22 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLkQY-00065G-2E; Mon, 26 Feb 2007 13:18:22 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLkQW-00062d-Hg
	for sipping@ietf.org; Mon, 26 Feb 2007 13:18:20 -0500
Received: from sj-iport-3-in.cisco.com ([171.71.176.72]
	helo=sj-iport-3.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HLkQU-0004jP-Sn
	for sipping@ietf.org; Mon, 26 Feb 2007 13:18:20 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by sj-iport-3.cisco.com with ESMTP; 26 Feb 2007 10:18:18 -0800
X-IronPort-AV: i="4.14,221,1170662400"; 
	d="scan'208"; a="467131551:sNHT56015776"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1QIIHBv021648; 
	Mon, 26 Feb 2007 13:18:17 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1QII9OI008073; 
	Mon, 26 Feb 2007 13:18:17 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 13:18:17 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 13:18:16 -0500
Message-ID: <45E32467.70303@cisco.com>
Date: Mon, 26 Feb 2007 13:18:15 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Christer Holmberg (JO/LMF)" <christer.holmberg@ericsson.com>
Subject: Re: [Sipping] Service Identification: offerless INVITE case
References: <7374777208BDC7449D5620EF94232567024653BB@esealmw113.eemea.ericsson.se>
In-Reply-To: <7374777208BDC7449D5620EF94232567024653BB@esealmw113.eemea.ericsson.se>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 18:18:16.0854 (UTC)
	FILETIME=[7C67F360:01C759D2]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=5195; t=1172513897;
	x=1173377897; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=3A=20offerless=2
	0INVITE=20case |Sender:=20
	|To:=20=22Christer=20Holmberg=20(JO/LMF)=22=20<christer.holmberg@ericsson
	.com>; bh=1UsFPfcDnQbak7/j2s01VOzplRbfel4yajslqocn9Nw=;
	b=tgO5zVYwkC9b7WrHzDa6/yQRJiWffCuP+agLOPSAs+kpycyrYaqEKuVyPnUqkeaghTGCSGIh
	xav47qlcW5x0uhpt5N116GyiLWVZ4iRyATsI7VTS9c5mIanBQ91AUHQH;
Authentication-Results: rtp-dkim-2; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 93e7fb8fef2e780414389440f367c879
Cc: IETF Sipping List <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Christer Holmberg (JO/LMF) wrote:
> Hi, 
> 
>> I too was thinking that INVITE/Replaces would be the way to 
>> get a new dialog that could include app servers not in the 
>> existing call, when that is needed. But I hadn't worked out 
>> how to know to do that. The 3xx does seem like a good way to 
>> handle that. The use of 3xx as a response to an in-dialog 
>> request is not at all well specified. But this could be a 
>> good motivation to clear that up.
> 
> I agree that the re-INVITE 3xx response would require some thinking.
> 
> For example, an initial INVITE 3xx response may not traverse all the way
> back to the UA. But, I guess a re-INVITE 3xx would have to be sent all
> the way back to the UA, since the UA would create a new dialog.

Needs clarification, but yes that seems to be a corollary.

> The question is also what happens to the original dialog. Does the UA
> have to send BYE on the original dialog, or is it terminated
> "automatically"?

If the invite is sent w/Replaces, then presumably the receiving end will 
send a BYE on the replaced dialog.

> If terminated automatically, what happens to the nodes situated AFTER
> the proxy that sent 3xx? They have no clue about the 3xx being sent, and
> there is no guarantee the re-INVITE will traverse the same nodes (or
> even reach the same remote UA).

This is a general issue for INVITE/Replaces. The 3xx is simply a way to 
get the UAC to do it. The UAC could decide to do it even without a 3xx. 
In the normal course of routing this any servers that are needed on the 
path will have to be discovered based on the content of the request.

	Paul

> Etc etc etc.
> 
> Regards,
> 
> Christer
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
>> 	Paul
>>
>> Jonathan Rosenberg wrote:
>>> Andrew made a great point in one of his mails which merits 
>> promotion 
>>> into its own thread.
>>>
>>> I and others have asserted that you should be able to identify the 
>>> service by things naturally in the request, and in 
>> particular, SDP has 
>>> come up as a key thing. For example, in a multiplayer game, there 
>>> might be an m-line for the particular proprietary protocol used for 
>>> game moves. Or, if its MSRP, you might have specific 
>> content types for 
>>> game moves. When present, those exactly indicate what the 
>> service is.
>>> But what do we do about offerless INVITE? In particular, there are 
>>> cases where we need to invoke an application server to handle a 
>>> request, and what if we don't have the SDP in the INVITE? 
>> Its too late 
>>> to insert an application server once the INVITE has already 
>> passed to 
>>> the called party and the 200 OK is sent.
>>>
>>> This is a great point. I'd rather just fix this little problem, the 
>>> define an entire notion of services and service identifiers 
>> that are 
>>> not necessary, just because we have offerless INVITEs.
>>>
>>> There are lots of ways to handle this problem. My preferred 
>> approach 
>>> would be that offerless INVITE contain a capability 
>> description, using 
>>> Flemmings new capability framework. The INVITE would contain a body 
>>> that  looks like SDP, but has a different content type and 
>>> content-disposition. It lists the media streams, codecs and other 
>>> capabilities that exist. Based on this, the called party makes the 
>>> offer in a 2xx or 18x like normal, and the answer in the 
>> ACK/PRACK as normal.
>>> It is completely backwards compatible.
>>>
>>> Now, a related issue is what happens if a session starts with audio 
>>> only, and then the user wants to turn this into a game? Its 
>> too late 
>>> to add new application servers into the call path, which 
>> might need to 
>>> be there for games. What can we do?
>>>
>>> Here too, there is a simple answer. If a network provider 
>> requires app 
>>> servers to get inserted, or a phone requires dispatch to a 
>> different 
>>> application, it would receive a re-INVITE to add the gaming stream, 
>>> and then 3xx it. The Contact would contain an embedded 
>> Replaces header 
>>> with the dialog IDs for the call in progress. THe 
>> originating UA would 
>>> then do a totally new INVITE that replaces the call in 
>> progress. This 
>>> would allow the network and/or phone to redispatch, yet provide a 
>>> seamless user experience. Billing can be done however the operators 
>>> want, since the signaling is very clear on exactly what 
>> happened and when.
>>> Furthermore, the replaces mechansim would only get used if 
>> there was 
>>> someone in the call path that needed to see it.
>>>
>>> I'd rather specify things like this, than invent an entire services 
>>> framework that would lead to balkanization and non-interoperable 
>>> islands of SIP.
>>>
>>> -Jonathan R.
>>>
>>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP 
>> Use sip-implementors@cs.columbia.edu for questions on current 
>> sip Use sip@ietf.org for new developments of core SIP
>>
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 13:30:09 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLkaI-0006V9-G0; Mon, 26 Feb 2007 13:28:26 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLkaH-0006Ux-93
	for sipping@ietf.org; Mon, 26 Feb 2007 13:28:25 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLkaA-0005oW-Mj
	for sipping@ietf.org; Mon, 26 Feb 2007 13:28:25 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	0364B200CB; Mon, 26 Feb 2007 19:28:12 +0100 (CET)
X-AuditID: c1b4fb3e-b0ed5bb0000007e1-95-45e326bb436c 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	D48B5200A2; Mon, 26 Feb 2007 19:28:11 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 19:28:11 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification: offerless INVITE case
Date: Mon, 26 Feb 2007 19:28:10 +0100
Message-ID: <7374777208BDC7449D5620EF9423256702465433@esealmw113.eemea.ericsson.se>
In-Reply-To: <45E32467.70303@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification: offerless INVITE case
Thread-Index: AcdZ0n6XUf9Dr6//S0GDKDdIi8udKQAAIufg
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>
X-OriginalArrivalTime: 26 Feb 2007 18:28:11.0944 (UTC)
	FILETIME=[DF1B7A80:01C759D3]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 3d7f2f6612d734db849efa86ea692407
Cc: IETF Sipping List <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,=20

>>If terminated automatically, what happens to the nodes situated AFTER=20
>>the proxy that sent 3xx? They have no clue about the 3xx being sent,=20
>>and there is no guarantee the re-INVITE will traverse the same nodes=20
>>(or even reach the same remote UA).
>=20
>This is a general issue for INVITE/Replaces. The 3xx is=20
>simply a way to get the UAC to do it. The UAC could decide to=20
>do it even without a 3xx.=20
>In the normal course of routing this any servers that are=20
>needed on the path will have to be discovered based on the=20
>content of the request.

My issue was how the proxies in the original dialog will know that a 3xx
has been sent (see picture):

UA A ------------------- Proxy 1 ---------------- Proxy 2
------------------ UA B     UA C

-------------INVITE---------> --------INVITE--------->
--------INVITE---------->
<------------200------------- <-------200-------------
<-------200--------------
-------------re-INVITE------>
<------------3xx-------------
-------------INVITE
(new)-------------------------------------------------------------->

Now, if Proxy 2 will not be traversed by the new INVITE, it will not
know that the dialog has been replaced.

And, if the new INVITE reaches UA C, UA B will have no clue that the
dialog has been replaced.

Regards,

Christer







=20
> > Etc etc etc.
> >=20
> > Regards,
> >=20
> > Christer
> >=20
> >=20
> >=20
> >=20
> >=20
> >=20
> >=20
> >=20
> >=20
> >=20
> >=20
> >=20
> >> 	Paul
> >>
> >> Jonathan Rosenberg wrote:
> >>> Andrew made a great point in one of his mails which merits
> >> promotion
> >>> into its own thread.
> >>>
> >>> I and others have asserted that you should be able to=20
> identify the=20
> >>> service by things naturally in the request, and in
> >> particular, SDP has
> >>> come up as a key thing. For example, in a multiplayer game, there=20
> >>> might be an m-line for the particular proprietary=20
> protocol used for=20
> >>> game moves. Or, if its MSRP, you might have specific
> >> content types for
> >>> game moves. When present, those exactly indicate what the
> >> service is.
> >>> But what do we do about offerless INVITE? In particular,=20
> there are=20
> >>> cases where we need to invoke an application server to handle a=20
> >>> request, and what if we don't have the SDP in the INVITE?
> >> Its too late
> >>> to insert an application server once the INVITE has already
> >> passed to
> >>> the called party and the 200 OK is sent.
> >>>
> >>> This is a great point. I'd rather just fix this little=20
> problem, the=20
> >>> define an entire notion of services and service identifiers
> >> that are
> >>> not necessary, just because we have offerless INVITEs.
> >>>
> >>> There are lots of ways to handle this problem. My preferred
> >> approach
> >>> would be that offerless INVITE contain a capability
> >> description, using
> >>> Flemmings new capability framework. The INVITE would=20
> contain a body=20
> >>> that  looks like SDP, but has a different content type and=20
> >>> content-disposition. It lists the media streams, codecs and other=20
> >>> capabilities that exist. Based on this, the called party=20
> makes the=20
> >>> offer in a 2xx or 18x like normal, and the answer in the
> >> ACK/PRACK as normal.
> >>> It is completely backwards compatible.
> >>>
> >>> Now, a related issue is what happens if a session starts=20
> with audio=20
> >>> only, and then the user wants to turn this into a game? Its
> >> too late
> >>> to add new application servers into the call path, which
> >> might need to
> >>> be there for games. What can we do?
> >>>
> >>> Here too, there is a simple answer. If a network provider
> >> requires app
> >>> servers to get inserted, or a phone requires dispatch to a
> >> different
> >>> application, it would receive a re-INVITE to add the=20
> gaming stream,=20
> >>> and then 3xx it. The Contact would contain an embedded
> >> Replaces header
> >>> with the dialog IDs for the call in progress. THe
> >> originating UA would
> >>> then do a totally new INVITE that replaces the call in
> >> progress. This
> >>> would allow the network and/or phone to redispatch, yet provide a=20
> >>> seamless user experience. Billing can be done however the=20
> operators=20
> >>> want, since the signaling is very clear on exactly what
> >> happened and when.
> >>> Furthermore, the replaces mechansim would only get used if
> >> there was
> >>> someone in the call path that needed to see it.
> >>>
> >>> I'd rather specify things like this, than invent an=20
> entire services=20
> >>> framework that would lead to balkanization and non-interoperable=20
> >>> islands of SIP.
> >>>
> >>> -Jonathan R.
> >>>
> >>>
> >> _______________________________________________
> >> Sipping mailing list =20
> https://www1.ietf.org/mailman/listinfo/sipping
> >> This list is for NEW development of the application of SIP Use=20
> >> sip-implementors@cs.columbia.edu for questions on current sip Use=20
> >> sip@ietf.org for new developments of core SIP
> >>
> >=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 14:17:57 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLlLT-00030P-ML; Mon, 26 Feb 2007 14:17:12 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLlLR-0002v8-S2
	for sipping@ietf.org; Mon, 26 Feb 2007 14:17:09 -0500
Received: from sj-iport-2-in.cisco.com ([171.71.176.71]
	helo=sj-iport-2.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HLlJO-0007LS-VK
	for sipping@ietf.org; Mon, 26 Feb 2007 14:15:04 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by sj-iport-2.cisco.com with ESMTP; 26 Feb 2007 11:15:02 -0800
X-IronPort-AV: i="4.14,221,1170662400"; 
	d="scan'208"; a="362978000:sNHT56335152"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1QJF1fC029943; 
	Mon, 26 Feb 2007 14:15:01 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1QJF0Vb004876; 
	Mon, 26 Feb 2007 14:15:01 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 14:15:01 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-202.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 14:15:01 -0500
Message-ID: <45E331B4.4090700@cisco.com>
Date: Mon, 26 Feb 2007 14:15:00 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: "Christer Holmberg (JO/LMF)" <christer.holmberg@ericsson.com>
Subject: Re: [Sipping] Service Identification: offerless INVITE case
References: <7374777208BDC7449D5620EF9423256702465433@esealmw113.eemea.ericsson.se>
In-Reply-To: <7374777208BDC7449D5620EF9423256702465433@esealmw113.eemea.ericsson.se>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 19:15:01.0119 (UTC)
	FILETIME=[698190F0:01C759DA]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=6265; t=1172517301;
	x=1173381301; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=3A=20offerless=2
	0INVITE=20case |Sender:=20
	|To:=20=22Christer=20Holmberg=20(JO/LMF)=22=20<christer.holmberg@ericsson
	.com>; bh=lNUa9VWC+Tli1sgRygy1RblpwCjnB2IYF4WnYc4cS9E=;
	b=jAUx2tTvb3pAAxK/W8h8NdHZPXnBEWv+9aID43ouCjTIxLuRpT7c90rWbGg6upYxJ9VR5674
	vwrlkGmYRrAB8P+S2XccgvAA6+d6qxug570n/YKN6UZFhXX/BvD1dND0;
Authentication-Results: rtp-dkim-1; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: f0b5a4216bfa030ed8a6f68d1833f8ae
Cc: Paul Kyzivat <pkyzivat@cisco.com>, IETF Sipping List <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I can't follow the figure since CRLFs got inserted by mailers along the 
way. Can you send one whose width is small?

I don't follow the problem though. The 3xx response would be seen by all 
record-routing proxies upstream of the entity that sends the mid-dialog 
3xx. The new INVITE would be seen by all proxies. To make sure it goes 
to the right UA instance you use GRUU. BYE is sent by the entity 
receiving the replaces, as normal. What am I missing?

-Jonathan R.

Christer Holmberg (JO/LMF) wrote:

> Hi, 
> 
> 
>>>If terminated automatically, what happens to the nodes situated AFTER 
>>>the proxy that sent 3xx? They have no clue about the 3xx being sent, 
>>>and there is no guarantee the re-INVITE will traverse the same nodes 
>>>(or even reach the same remote UA).
>>
>>This is a general issue for INVITE/Replaces. The 3xx is 
>>simply a way to get the UAC to do it. The UAC could decide to 
>>do it even without a 3xx. 
>>In the normal course of routing this any servers that are 
>>needed on the path will have to be discovered based on the 
>>content of the request.
> 
> 
> My issue was how the proxies in the original dialog will know that a 3xx
> has been sent (see picture):
> 
> UA A ------------------- Proxy 1 ---------------- Proxy 2
> ------------------ UA B     UA C
> 
> -------------INVITE---------> --------INVITE--------->
> --------INVITE---------->
> <------------200------------- <-------200-------------
> <-------200--------------
> -------------re-INVITE------>
> <------------3xx-------------
> -------------INVITE
> (new)-------------------------------------------------------------->
> 
> Now, if Proxy 2 will not be traversed by the new INVITE, it will not
> know that the dialog has been replaced.
> 
> And, if the new INVITE reaches UA C, UA B will have no clue that the
> dialog has been replaced.
> 
> Regards,
> 
> Christer
> 
> 
> 
> 
> 
> 
> 
>  
> 
>>>Etc etc etc.
>>>
>>>Regards,
>>>
>>>Christer
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>>	Paul
>>>>
>>>>Jonathan Rosenberg wrote:
>>>>
>>>>>Andrew made a great point in one of his mails which merits
>>>>
>>>>promotion
>>>>
>>>>>into its own thread.
>>>>>
>>>>>I and others have asserted that you should be able to 
>>
>>identify the 
>>
>>>>>service by things naturally in the request, and in
>>>>
>>>>particular, SDP has
>>>>
>>>>>come up as a key thing. For example, in a multiplayer game, there 
>>>>>might be an m-line for the particular proprietary 
>>
>>protocol used for 
>>
>>>>>game moves. Or, if its MSRP, you might have specific
>>>>
>>>>content types for
>>>>
>>>>>game moves. When present, those exactly indicate what the
>>>>
>>>>service is.
>>>>
>>>>>But what do we do about offerless INVITE? In particular, 
>>
>>there are 
>>
>>>>>cases where we need to invoke an application server to handle a 
>>>>>request, and what if we don't have the SDP in the INVITE?
>>>>
>>>>Its too late
>>>>
>>>>>to insert an application server once the INVITE has already
>>>>
>>>>passed to
>>>>
>>>>>the called party and the 200 OK is sent.
>>>>>
>>>>>This is a great point. I'd rather just fix this little 
>>
>>problem, the 
>>
>>>>>define an entire notion of services and service identifiers
>>>>
>>>>that are
>>>>
>>>>>not necessary, just because we have offerless INVITEs.
>>>>>
>>>>>There are lots of ways to handle this problem. My preferred
>>>>
>>>>approach
>>>>
>>>>>would be that offerless INVITE contain a capability
>>>>
>>>>description, using
>>>>
>>>>>Flemmings new capability framework. The INVITE would 
>>
>>contain a body 
>>
>>>>>that  looks like SDP, but has a different content type and 
>>>>>content-disposition. It lists the media streams, codecs and other 
>>>>>capabilities that exist. Based on this, the called party 
>>
>>makes the 
>>
>>>>>offer in a 2xx or 18x like normal, and the answer in the
>>>>
>>>>ACK/PRACK as normal.
>>>>
>>>>>It is completely backwards compatible.
>>>>>
>>>>>Now, a related issue is what happens if a session starts 
>>
>>with audio 
>>
>>>>>only, and then the user wants to turn this into a game? Its
>>>>
>>>>too late
>>>>
>>>>>to add new application servers into the call path, which
>>>>
>>>>might need to
>>>>
>>>>>be there for games. What can we do?
>>>>>
>>>>>Here too, there is a simple answer. If a network provider
>>>>
>>>>requires app
>>>>
>>>>>servers to get inserted, or a phone requires dispatch to a
>>>>
>>>>different
>>>>
>>>>>application, it would receive a re-INVITE to add the 
>>
>>gaming stream, 
>>
>>>>>and then 3xx it. The Contact would contain an embedded
>>>>
>>>>Replaces header
>>>>
>>>>>with the dialog IDs for the call in progress. THe
>>>>
>>>>originating UA would
>>>>
>>>>>then do a totally new INVITE that replaces the call in
>>>>
>>>>progress. This
>>>>
>>>>>would allow the network and/or phone to redispatch, yet provide a 
>>>>>seamless user experience. Billing can be done however the 
>>
>>operators 
>>
>>>>>want, since the signaling is very clear on exactly what
>>>>
>>>>happened and when.
>>>>
>>>>>Furthermore, the replaces mechansim would only get used if
>>>>
>>>>there was
>>>>
>>>>>someone in the call path that needed to see it.
>>>>>
>>>>>I'd rather specify things like this, than invent an 
>>
>>entire services 
>>
>>>>>framework that would lead to balkanization and non-interoperable 
>>>>>islands of SIP.
>>>>>
>>>>>-Jonathan R.
>>>>>
>>>>>
>>>>
>>>>_______________________________________________
>>>>Sipping mailing list  
>>
>>https://www1.ietf.org/mailman/listinfo/sipping
>>
>>>>This list is for NEW development of the application of SIP Use 
>>>>sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>>sip@ietf.org for new developments of core SIP
>>>>
>>>
> 

-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 14:24:11 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLlS2-0002AA-6L; Mon, 26 Feb 2007 14:23:58 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLlS0-0002A0-TU
	for sipping@ietf.org; Mon, 26 Feb 2007 14:23:56 -0500
Received: from sj-iport-4.cisco.com ([171.68.10.86])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLlRz-0000Zd-Dx
	for sipping@ietf.org; Mon, 26 Feb 2007 14:23:56 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by sj-iport-4.cisco.com with ESMTP; 26 Feb 2007 11:23:54 -0800
X-IronPort-AV: i="4.14,221,1170662400"; 
	d="scan'208"; a="42737030:sNHT50257296"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1QJNrcf027463; 
	Mon, 26 Feb 2007 14:23:53 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1QJNhOS000273; 
	Mon, 26 Feb 2007 14:23:53 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 14:23:49 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-202.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 14:23:48 -0500
Message-ID: <45E333C4.3090700@cisco.com>
Date: Mon, 26 Feb 2007 14:23:48 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: "Stephen Terrill (MI/EEM)" <stephen.terrill@ericsson.com>
Subject: Re: [Sipping] Service Identification in SIP
References: <989763567E51AB48BE015DF7BB960CAD02926384@esealmw111>
In-Reply-To: <989763567E51AB48BE015DF7BB960CAD02926384@esealmw111>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 19:23:48.0901 (UTC)
	FILETIME=[A416B550:01C759DB]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=3785; t=1172517833;
	x=1173381833; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20
	|To:=20=22Stephen=20Terrill=20(MI/EEM)=22=20<stephen.terrill@ericsson.com
	>; bh=nJUAbNUxzG6W4ThTfwmp1cFHG5Z4cIHXEfjAsXl1VSg=;
	b=LC1YFrI1cHSqxTCwSNrzop5eRTk6Z+GUBJCefo35a8zHR4lfDhVdQ/+VHZ+j1OBjIjIm3W3Y
	L8ub9dBvjutdf8XOmK8rjDAH5FiZObks8oNGLqTOB+drci1F0aOrTkLd;
Authentication-Results: rtp-dkim-2; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 31247fb3be228bb596db9127becad0bc
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

inline.

Stephen Terrill (MI/EEM) wrote:

> Hi,
> 
> I don't think that being able to identify services leads to any of the
> mentioned "-sation"s of SIP [well maybe industralisation ;-)], except to
> recognise that there are different services on SIP and that the media
> does not give the service (we can also start the session without media
> ...).

And I don't see how you can argue that point. Merely stating it does not 
make it so.

To me, its really simple. If 3gpp goes and defines a service parameter 
for "conversational telephony" and gives it some name, and puts it into 
Accept-Contact, then the ability to receive a call from a non-3gpp 
entity is ruined, since such a call will not include this parameter. 
Thus, the right QoS treatment, billing, call processing, etc. will not 
happen.

In what way is this NOT balkanization? Please be specific.

>  At what point can it be said that SIP is "-ised" - or at what
> point to the SIP police say that this is breaking SIP?  If it doesn't
> get to all UAs?  If it doesn't go to all UAs that could do something
> useful with it?, If it doesn't go through a SIP proxy? ... Sometimes
> trying to deliver a service instead of rejecting the service due to
> incomptabilities can lead to a bad user experience - is that the best
> thing to do, or is that OK because the SIP protocol still connected? [I
> think that Jeroen captured this quite well in one of his emails]. I note
> that the second point raised, does handle this, leaving the question of
> a middle ground as well (i.e an indication of the service without it
> being "REQUIRED").

We have a mechanism for you to cause a call to fail if compatible 
extensions are not supported (Require). If business requirements say 
that this is the desired result, then use the tool. I certainly don't 
hope that this is what you want for the common case. I can see how it is 
the right thing for PoC, but thats about it.


> 
> I think that it is better that the industry provides a common way to do
> this, instead of "looking into the SDP in one case, looking into the
> accept contact in another case" etc, that seems more complicated than
> necessary ....  

I disagree completely.

THis is like saying that we should put everything that a SIP message 
might indicate into a single "indicate" header, rather than invent new 
headers, methods, put stuff in bodies, and so on. The 'service' is 
nothing more or less than a summarization of the functionality that 
results from the message being sent, and that functionality includes 
everything from the method to the body.


> 
> I noticed that in the BLISS group there was discussions on what
> assumptions could be made - i.e. the particular example being discussed
> was DND and whether it should be in the terminal or in the network.  If
> something comes out of that - won't that be, by default, a service
> definition or some sort.  It would be strange if that was the only
> service definition allowed ...

I sent another note on sipping on this topic. These are related problems 
but differ quite fundamentally, since in those cases I am not requiring 
the other party to know anything about this; its a matter of control 
between a UA and its own domain. Even in that case, we are agreeing that 
we want to be as flexible as possible to what constitutes these 
services, and not standardize what it is.

-Jonathan R.

-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 14:41:08 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLli1-0006FS-M8; Mon, 26 Feb 2007 14:40:29 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLlhz-0006Ex-5e
	for sipping@ietf.org; Mon, 26 Feb 2007 14:40:27 -0500
Received: from sj-iport-6.cisco.com ([171.71.176.117])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLlhu-0002cA-5s
	for sipping@ietf.org; Mon, 26 Feb 2007 14:40:27 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by sj-iport-6.cisco.com with ESMTP; 26 Feb 2007 11:40:21 -0800
X-IronPort-AV: i="4.14,221,1170662400"; 
	d="scan'208"; a="116271021:sNHT75134277"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1QJeKpY010857; 
	Mon, 26 Feb 2007 14:40:20 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1QJdjW3012432; 
	Mon, 26 Feb 2007 14:40:20 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 14:40:14 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-201.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 14:40:13 -0500
Message-ID: <45E3379D.9010802@cisco.com>
Date: Mon, 26 Feb 2007 14:40:13 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: "Vijay K. Gurbani" <vkg@alcatel-lucent.com>
References: <459C1D99.2010504@alcatel-lucent.com>
In-Reply-To: <459C1D99.2010504@alcatel-lucent.com>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 19:40:13.0547 (UTC)
	FILETIME=[EEFBC3B0:01C759DD]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=19136; t=1172518820;
	x=1173382820; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20WGLC=20review=20of=20draft-ietf-sipping-spam-03
	|Sender:=20
	|To:=20=22Vijay=20K.=20Gurbani=22=20<vkg@alcatel-lucent.com>;
	bh=HP2N7D8lPFN24hfbjYXTdx+hSU+IiECTAMBnKMRFTN4=;
	b=ilGSw5FJMf/MZ/rWZj43lN+QegDGw7aTEYXYxyPLnFMEOifIvpOL3hhwBfSyMFrSPh+Y4+wF
	g8Czu03TDNmuOJxaqVZPtJuEJpmBso4vn6up3b4boiYMhmfiOYGjibpr;
Authentication-Results: rtp-dkim-1; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: ba9cd4f9acda58dbe142afff7265daff
Cc: Cullen Jennings <fluffy@cisco.com>, SIPPING LIST <sipping@ietf.org>,
	Mary Barnes <mary.barnes@nortel.com>,
	Gonzalo Camarillo <Gonzalo.Camarillo@ericsson.com>
Subject: [Sipping] Re: WGLC review of draft-ietf-sipping-spam-03
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Vijay,

Thanks for the comments and thorough review. I apologize for not being 
able to respond to your comments earlier.

inline.

Vijay K. Gurbani wrote:

> Draft: draft-ietf-sipping-spam-03
> Reviewer:  Vijay K. Gurbani <vkg@alcatel-lucent.com>
> Review Date: Jan 3, 2007
> Review Deadline:  Dec 31, 2006
> Status: WGLC
> 
> Review Summary:
> This draft is on the right track but has open issues, described in
> the review.
> 
> Draft Summary: This is an excellent survey on spam in general, and
> SIP-related spam in particular.  It ties in nicely the various
> extensions to SIP (consent, identity, presence, etc.) to solve a
> particular problem, namely preventing SIP spam.  In my opinion, the
> sooner it gets to RFC status, the better it would be for the SIP
> user community.
> 
> Issues needing discussion listed below, followed by some nits.
> 
> Issues needing discussion:
> 
> * S2.1, paragraph "This number (9 deliveries ... will be more
>   effective."  I am not sure how you are characterizing the first
>   sentence in that paragraph.  Certainly, a SMTP server (probably the
>   closest equivalent to a SIP proxy) would be able to, and probably
>   already does deliver 9 or more messages a second.  For some existing
>   analysis on SMTP server performance, see [1,2], which describes
>   systems that can handle anywhere from 52 to 5000 messages/second.

Right. THis paragraph is confusing and a bit out of place. Its trying to 
say that voip spam can deliver 9 successful call attempts per second, 
and that this is lower than email, which can do more. The limit there is 
actually access bandwidth. However, those emails are often not seen. 
With voip, a successful call results in content being presented to the 
user. So its valuable.

Paragraph now says:

<t>
It is useful to compare these figures with email. VoIP can deliver
approximately 9 successful call attempts per second. Email spam can,
of course, deliver more. Assuming 1 kB per email, and an upstream
link of 500 kbps, a spammer can generate 62.5 messages per
second. This number goes down with larger messages of
course. Interestingly, spam filters delete large numbers of these
mails, so the cost per viewed message is likely to be much higher. In
that sense, call spam is much more attractive, since its content is
much more likely to be examined by a user if a call attempt is
successful.
</t>




> 
>   If however, the 9 deliveries per second refers to the MUA (closest to
>   a SIP UA), then I would agree that it is hard to pin down how many
>   messages per second an individual receives.  For instance, sending a
>   message to an enterprise reflector usually always elicits "On
>   vacation" responses, and many times such responses are in numbers
>   greater than 9 messages arriving per second.

Hopefully the above makes this all clear.


> 
> * S2.2, s/which only allow spam to be/which only allow IM to be
>   I think you meant that the *IM* will be delivered only if the sender
>   is on a white list, not that the *spam* will be delivered only if the
>   sender is on a white list.

right, fixed.

> 
> * In S3.1, the third paragraph is redundant; it restates what the
>   first sentence of the second paragraph alread said.  I would suggest
>   taking it out completely, and modifying the first sentence of the
>   second paragraph as follows:
> 
>   OLD:
>   Unfortunately, this type of spam filtering is almost completely
>   useless for call spam.
> 
>   NEW:
>   Unfortunately, this type of spam filtering, while successful for email
>   spam, is completely useless for call spam.

done.


> 
> * In S3.1, to be complete, you should also add a blurb on whether or not
>   content filtering helps for presence spam (since you mention that it
>   does not help for call spam and may help for IM spam.)
>   I would suggest the addition of the following paragraph at the end
>   of the section:
> 
>     Content filtering will not help for presence spam because by
>     definition, a request subscribing for the presence of a user will
>     be devoid of any content.
> 

done.


> * In S3.2, you may want to change spammer@domain.com and spammer.com.
>   domain.com is registered to someone in Vancouver, WA, and
>   spammer.com is a registered to someone in Houston, TX.
>   Suggest using spammer@spam.example.com for spammer@domain.com
>   and spam.example.com for spammer.com.
> 
>   spammers.com also appears in S5.

good catches, fixed.


> 
> * In S3.3, it is worth discussing that in addition to Turing tests,
>   the "introduction problem" can also be solved by tagged addresses,
>   of which the Tagged Message Delivery Agent (TDMA,
>   http://tmda.net/index.html) is a good example.  Using tagged
>   addresses, you can limit an email address by temporal interval
>   (i.e., valid only for 5 days) or sender-specific addresses.  I
>   do think they are worth discussing (for email, they work well;
>   they have limited utility in call spam, but could be used for
>   IM spam and presence spam.)  I will leave it upto you to determine
>   whether a new sub-section should be added in Section 3 to discuss
>   tagged addresses, or you can discuss them as part of S3.7 "Limited
>   Use Addresses".

I'm not sure how tagged addresses differ from limited use. Perhaps you 
mean that limited use can also include time-limits and be grouped by 
topic? I modified the first paragraph in 3.7 to broaden the scope of 
limited use addresses:

<t>
A related technique to address obfuscation is limited use
addresses. In this technique, a user has a large number of email
addresses at their disposal, each of which has constraints on its
applicability. A limited use address can be time-bound, so that it
expires after a fixed period. Or, a different email address can be
given to each correspondent. When spam arrives from that
correspondent, the limited use address they were given is
terminated. In another variation, the same limited use address is
given to multiple users that share some property; for example, all
work colleagues, all coworkers from different companies, all
retailers, and so on. Should spam begin arriving on one of the
addresses, it is invalidated, preventing communications from anyone
else that received the limited use address.
</t>


> 
> * In S3.4, first complete paragraph on page 11: it is stated in the
>   last sentence that "Fortunately, traditional content-based filtering
>   can be applied to this type of information."  I presume the phrase
>   "this type of information" refers to the SIP address
>   "sip:please-buy-my-product-at-this-website@spam.example.com".  If
>   so, then this is not content, but a header; more specifically,
>   a From header.  Would white- or black-lists not be a better match
>   then content-based filtering in this case?

Its content in the sense that it represents the useful information that 
the spammer wants to the deliver to the recipient. Header vs. body is 
just protocol details. My point is that content-filtering techniques can 
  be applied to these 'headers' just as easily as they are to bodies.

Text now reads:

The user part of the SIP
URI allows for limited freeform text, and thus could be used to convey
brief messages. One can imagine receiving consent requests with
identities like
"sip:please-buy-my-product-at-this-website@spam.example.com", for
example. Fortunately, it is possible to apply traditional content
filtering systems to the header fields in the SIP messages, thus
blocking these kinds of consent requests.  </t>

> 
> * In S3.5, I don't think I agree with your assertion that "Reputation
>   systems have not enjoyed much success outside of the instant messaging
>   space." (third paragraph in that section.)  Certainly, in the business
>   space occupied by eBay and Amazon, reputation is very intrinsic and
>   an important contributor to the smooth functioning of the enterprise.
>   Since eBay and Amazon have authority over their namespace, they can
>   warn other users of identity mining (eBay, for instance, displays a
>   special icon if you have recently changed your identity.)  Likewise,
>   anyone who has bought a used book or CD through Amazon first scans
>   the reputation of the seller before pressing the "1-click" button.
> 
>   My suggestion would be to begin the third paragraph using text
>   suggested below:
> 
>   OLD:
>   Reputation systems have not enjoyed much success outside of the
>   instant messaging space.  This is in part because few other
>   communications systems admit of the same degree of centralization and
>   monolithic control.
> 
>   NEW:
>   Reputation systems have been successful in communication systems
>   where centralization of resources (user identities, authentication,
>   etc.) and monolithic control dominate.

OK, fair point. Text now reads:

<t>
Reputation systems have been successful in systems where
centralization of resources (user identities, authentication, etc.)
and monolithic control dominate. Examples of these include the large
instant messaging providers that run closed, proprietary
networks. That control, first of all, provides a relatively strong
identity assertion for users (since all users trust a common provider,
and the common provider is the arbiter of authentication and
identity). Secondly, it provides a single place where reputation can
be managed.  </t>




> 
> * S3.8, page 15, it is stated at the top of the page that "Automata
>   cannot easily perform the image recognition needed to extract the
>   word or number sequence, but a human user usually can."  Note that
>   there exist programs (called CAPTCHAs) that can acheive with 93%
>   correct recognition rate of visual puzzles (see
>   http://www.captcha.net, and [3]).

Right, sure. This wasn't clear in that paragraph. Now it reads:

<t>
In email, Turing tests are those solutions whereby the sender of the
message is given some kind of puzzle or challenge, which only a human
can answer. These tests are also known as captchas (Completely
Automated Public Turing test to tell Computers and Humans Apart). If
the puzzle is answered correctly, the sender is placed on the user's
white list. These puzzles frequently take the form of recognizing a
word or sequence of numbers in an image with a lot of background
noise. The tests need to be designed such that automata cannot easily
perform the image recognition needed to extract the word or number
sequence, but a human user usually can. Designing such tests is not
easy, since ongoing advances in image processing an artificial
intelligence continually raise the bar. Consequently, the
effectiveness of captchas are tied to whether spammers can come up
with or obtain algorithms for automatically solving them.  Since
Turing tests rely on video or audio puzzles, they sometimes cannot be
solved by individuals with handicaps.
</t>


> 
> * S3.8, of the two problems with the voice Turing tests mentioned
>   towards the end of the section, the first one appears to *not* be
>   a problem.  Yes, voice Turing test would need to be language
>   specific, but as you note, SIP already has provisions to help here.
>   Even if it did not, the experience is no different than what happens
>   now when A calls B, and is instead redirected to B's voicemail
>   which prompts him to enter "1" for A, "2" for A's wife and so on.
>   If A is in the US, and B is in Lisbon, the instructions are rendered
>   in Portugese.  A swears a bit, and hangs up, vowing to call B
>   at a later time.
> 
>   The second problem -- cheap labor -- is indeed a problem.

OK, changed it so that the first is not so much a problem as a 
consideration, and the second is the real problem.

> 
> * In S3.9, there is no mention of Cullen's hashcash draft?  Is that
>   effort dead?

Its currently expired.

> 
> * In S3.9, one disadvantage of Payments At Risk scheme is the
>   discrepancy of currency value between the sender and recipient.
>   If the sender is in a third world country (albeit being financed by a
>   spamming corporation in a first world country) and the receiver is
>   in a first world country, then the sender really does not care about
>   not getting his money refunded.  The benefactor funding the
>   sender is able to use the currency valuation of his host country
>   to conduct spam.  The sender deposits the payment in his local
>   currency, the exchange value of which is much lower when compared to
>   that of his benefactor's country.

Right. Great point. I added a paragraph on that:

<t>
Assuming a micro-payment infrastructure exists, another problem with
payment-at-risk is that it loses effectiveness when there are strong
inequities in the value of currency between sender and recipient. For
example, a poor person in a third world country might keep the money
in each mail message, regardless if it is spam. Similarly, a poor
person might not be willing to include money in an email, even if
legitimate, for fear that the recipient might keep it. If the amount of 
money
is lowered to help handle these problems, it might become sufficiently
small that spammers can just afford to spend it.
</t>

> 
> * Isn't S3.13 similar to VoIP peering?  If domains establish a
>   peering relationship, then they could optionally add spamming fines
>   to the service level agreement.  In fact, the Circle of Trust
>   idea of S3.12 is a simpler case of what is described in S3.13.
>   Suggestion: remove S3.12, or subsume it in S3.13.

Its a variation, but an important one. Circles of trust don't scale, 
whereas this does. I'd prefer to keep it separate. I'll add a note in 
the beginning of 3.13 indicating that its a variation.

> 
> * The last statement of S3.13 is a strongly worded opinion of the
>   authors.  While in and of itself, this is perfectly okay, I feel
>   that standards documents ought to be a bit more impartial (okay,
>   beat me here if you want.)  Instead, I would suggest that the last
>   two sentences of S3.13 be reworded as follows to convey the authors
>   distate in an objective manner:
> 
>   OLD:
>   However, it puts back into SIP much of the complexity and monopolistic
>   structures that SIP promised to eliminate.  As such, it is a solution
>   that the authors find distasteful and contrary to the SIP design
>   and architecture.
> 
>   NEW:
>   However, centralized architectures as these are deliberately eschewed
>   because they put back into SIP much of the complexity and mono-
>   polistic structures that the protocol promises to eliminate.

OK.


> 
> * In S6, I would suggest that the three core recommendations be given
>   labels; something like:
> 
>     1.  Enforce SIP Identity
>     ...
> 
>     2.  Leverage White Lists
>     ...
> 
>     3.  Provide a Solution for the Introduction Problem
>     ...
> 
>   Implicitly, the last paragraph contains the fourth recommendation:
> 
>     4.  Do Not Ignore Spam
>     ...

OK. I made these hangText and added the fourth.

> 
> * There is no IANA Consideration section (I realize that this is an
>   Informational track RFC, so an IANA Consideration section may not
>   be required.  I do not remember if the IESG requires such a
>   section in all drafts, even Informationals.)

added.

> 
> References
> [1] Wyman Miles, "A High-Availability High-Performance E-Mail Cluster,"
>  Proceedings of the 30th annual ACM SIGUCCS conference on User services,
>  pp. 84-88, November 20-23, 2002.
> [2] Yasushi Saito, Brian N. Bershad, and Henry M. Levy, "Manageability,
>  availability, and performance in porcupine: a highly scalable, cluster-
>  based mail service," ACM Transactions on Computer Systems (TOCS), pp.
>  298-332, 18(3), August 2000.
> [3] Luis von Ahn, Manuel Blum, and John Langford, "Telling Humans and
>  Computers Apart Automatically," Communications of the ACM, pp. 57-60,
>  47(2), February 2004.
> 
> Nits

I address these unless otherwise noted.

> * S2.1, page 6,
>   s/addresses, in that, in concert/addresses, such that in concert/
> 
> * S3.4, page 11,
>   s/but to make the permission is extremely limited/but to ensure
>   that the permission is extremely limited
> 
> * S3.5, first paragraph, second sentence:
>   s/and they attempt/and A attempts to/
> 
> * S3.5, first paragraph, third sentence:
>   s/from A, a reputation score/from A, and along with the consent, a
>   reputation score/
> 
> * S3.5, regarding "reputation mafias" in the fourth paragraph, I am
>   curious: is there any published reference?  It seems that such a
>   mafia would neccessarily be a large, distributed, self-organizing
>   system that shares a common goal, although the victims change over
>   time.

not that i know of

> 
> * S3.8, last paragraph:
>   s/turing tests/Turing tests
> 
> * In S3.9, I am not sure what the last sentence of the section
>   is supposed to convey.  Any research is subject to patent and IP
>   constraints, I suppose; SIP spam is no exception.  Best to simply take
>   that sentence out.

removed

> 
> * S3.10, first paragraph:
>   s/A sends to user B/A sends email to user B/
> 
> * S3.10, third paragraph:
>   s/for small transactions, for transaction as small as one cent./
>   for small transactions, as small as one cent.
> 
> * S3.11, second sentence of second parargraph: I am not sure I
>   can parse what it says.  Maybe you would consider, instead,
>   to simply say:
> 
>       Whether they are, or are not effective remains to be seen.
> 
> * S 3.11, last sentence of the section: suggest rewrite as follows:
> 
>   OLD:
>   For example, Habeas <http://www.habeas.com> inserts material in the
>   header that, if a spammer inserted it without an appropriate license,
>   allegedly causes the spammer to be violating US copyright and
>   trademark laws, possibly reciprocal laws, and similar laws in many
>   countries.
> 
>   NEW:
>   For example, Habeas <http://www.habeas.com> inserts material in the
>   header that, if it was inserted by a spammer without an appropriate
>   license, would allegedly causes the spammer to violate US copyright
>   and trademark laws, possibly reciprocal laws, and similar laws in
>   many countries.
> 
> * S4.1, expand MTA and MUA on their first use.
> 
> * Reference 24, the terminating comma should appear before Third
>   Annual Workshop...

these I can't fix; they come from the xml.resource.org bibliography.

Thanks!

-Jonathan R.

-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 14:46:27 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLlna-0002Js-OK; Mon, 26 Feb 2007 14:46:14 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLlnZ-0002HY-26
	for sipping@ietf.org; Mon, 26 Feb 2007 14:46:13 -0500
Received: from sj-iport-1-in.cisco.com ([171.71.176.70]
	helo=sj-iport-1.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HLlnX-00039e-LG
	for sipping@ietf.org; Mon, 26 Feb 2007 14:46:13 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by sj-iport-1.cisco.com with ESMTP; 26 Feb 2007 11:46:11 -0800
X-IronPort-AV: i="4.14,221,1170662400"; 
	d="scan'208"; a="764552489:sNHT46511606"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1QJkABU005501; 
	Mon, 26 Feb 2007 14:46:10 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1QJjjW1013965; 
	Mon, 26 Feb 2007 14:46:10 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 14:46:08 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-201.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 14:46:07 -0500
Message-ID: <45E338FF.4000203@cisco.com>
Date: Mon, 26 Feb 2007 14:46:07 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: Dan Wing <dwing@cisco.com>
Subject: Re: [Sipping] WGLC review of draft-ietf-sipping-spam-03
References: <057401c735d5$45c39420$20a36b80@amer.cisco.com>
In-Reply-To: <057401c735d5$45c39420$20a36b80@amer.cisco.com>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 19:46:07.0708 (UTC)
	FILETIME=[C21471C0:01C759DE]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=2570; t=1172519170;
	x=1173383170; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20WGLC=20review=20of=20draft-ietf-sipping-s
	pam-03 |Sender:=20 |To:=20Dan=20Wing=20<dwing@cisco.com>;
	bh=AQo0a4QLoQCbqx/9hWU/gsJQK8zeQuP2U+uGyPq5RZc=;
	b=mB5NcshzOjkLL+6gNG43VnP70VCDtOM6YeHwyyROQ5XscQqvT2AcPrAnbgvpZ49h8MKWOZs8
	rXbdVMqR9at0cZLgCdx/ScxKCwgEATSqVVd2yY1kHLr3ivCfwaYenpOx;
Authentication-Results: rtp-dkim-2; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 52f7a77164458f8c7b36b66787c853da
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I think the wording I put into sipping-spam-04 captures the conclusions 
from this thread:

<t>
Assuming a micro-payment infrastructure exists, another problem with
payment-at-risk is that it loses effectiveness when there are strong
inequities in the value of currency between sender and recipient. For
example, a poor person in a third world country might keep the money
in each mail message, regardless if it is spam. Similarly, a poor
person might not be willing to include money in an email, even if
legitimate, for fear that the recipient might keep it. If the amount of 
money
is lowered to help handle these problems, it might become sufficiently
small that spammers can just afford to spend it.
</t>


Note that its about inequities in VALUE of money.

-Jonathan R.

Dan Wing wrote:

>>>Could this be used to generate "inverse spam", where 
>>>someone posts an
>>>attractive public solicitation.  Thousands of people send 
>>>introductory
>>>e-mail, enclosing their Payment at Risk of (say) USD 0.05.  Then the
>>>recipient just keeps the money.
>>
>>Oh, neat.  I like this inverse spam idea.
> 
> 
> Such an 'attractive solicitation' could appear to be an important 
> message from your employer, your bank, your employer's healthcare 
> billing company, etc., but it would actually belong to someone 
> in a far-away country.  Like today's spam, rotating through 
> identities (to thwart reputation systems and black lists)
> and a "small" 0.01% take rate can be sufficient to make a 
> profit.
> 
> It seems the escrowing company would need to detect such 
> activity and police it or prevent it -- it's in their business
> interest to do so, or else nobody will trust them.  
> 
> Similarly, the users have an interest in not falling for such
> phishing emails -- they'll get charged money (costs them money)
> or have to contest the charge (costs them time).
> 
> -d
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 15:04:23 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLm4g-0007Al-26; Mon, 26 Feb 2007 15:03:54 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLm4e-00077b-Dx
	for sipping@ietf.org; Mon, 26 Feb 2007 15:03:52 -0500
Received: from sj-iport-3-in.cisco.com ([171.71.176.72]
	helo=sj-iport-3.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HLm4c-0005Yh-38
	for sipping@ietf.org; Mon, 26 Feb 2007 15:03:52 -0500
Received: from sj-dkim-7.cisco.com ([171.68.10.88])
	by sj-iport-3.cisco.com with ESMTP; 26 Feb 2007 12:03:49 -0800
X-IronPort-AV: i="4.14,221,1170662400"; 
	d="scan'208"; a="467155454:sNHT41415944"
Received: from sj-core-3.cisco.com (sj-core-3.cisco.com [171.68.223.137])
	by sj-dkim-7.cisco.com (8.12.11/8.12.11) with ESMTP id l1QK3n81028331
	for <sipping@ietf.org>; Mon, 26 Feb 2007 12:03:49 -0800
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by sj-core-3.cisco.com (8.12.10/8.12.6) with ESMTP id l1QK3ji4013478
	for <sipping@ietf.org>; Mon, 26 Feb 2007 12:03:49 -0800 (PST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 15:03:46 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-202.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 15:03:46 -0500
Message-ID: <45E33D22.6070004@cisco.com>
Date: Mon, 26 Feb 2007 15:03:46 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: IETF Sipping List <sipping@ietf.org>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 20:03:46.0307 (UTC)
	FILETIME=[390DD530:01C759E1]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=526; t=1172520229;
	x=1173384229; c=relaxed/simple; s=sjdkim7002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20updated=20SIP=20and=20Spam=20draft |Sender:=20;
	bh=PNrYKI26ZiorOV3CjdsWSyIOkymKk2ODp70yEmaCTyg=;
	b=YulUKucEt/iMHKkK7+bfQbsyzFPlRN6u4D6XszuYgyCnRZNg3MMykntL2K6ipFiRatWf6VSu
	u4R1c2P2XktIfi5v9+vb9aMHXqjM8PDamqJteduhJkDaTTiQtWuZd3gw;
Authentication-Results: sj-dkim-7; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim7002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 1ac7cc0a4cd376402b85bc1961a86ac2
Subject: [Sipping] updated SIP and Spam draft
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I've updated the SIP and Spam draft based on WGLC reviews and comments. 
Until it appears, you can pick it up at:

http://www.jdrosen.net/papers/draft-ietf-sipping-spam-04.txt

Thanks,
Jonathan R.
-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 15:09:27 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLmA2-0004UD-Ty; Mon, 26 Feb 2007 15:09:26 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLmA1-0004U2-KM
	for sipping@ietf.org; Mon, 26 Feb 2007 15:09:25 -0500
Received: from sj-iport-2-in.cisco.com ([171.71.176.71]
	helo=sj-iport-2.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HLm9z-0006g7-MF
	for sipping@ietf.org; Mon, 26 Feb 2007 15:09:25 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by sj-iport-2.cisco.com with ESMTP; 26 Feb 2007 12:09:06 -0800
X-IronPort-AV: i="4.14,221,1170662400"; 
	d="scan'208"; a="362992042:sNHT45057927440"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1QK96U0024080
	for <sipping@ietf.org>; Mon, 26 Feb 2007 15:09:06 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1QK95OE012484
	for <sipping@ietf.org>; Mon, 26 Feb 2007 15:09:06 -0500 (EST)
Received: from xmb-rtp-20b.amer.cisco.com ([64.102.31.53]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 15:08:17 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification: offerless INVITE case
Date: Mon, 26 Feb 2007 15:08:17 -0500
Message-ID: <072C5B76F7CEAB488172C6F64B30B5E302A8D996@xmb-rtp-20b.amer.cisco.com>
In-Reply-To: <45E307D9.1090201@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification: offerless INVITE case
Thread-Index: AcdZweZ+6mfs+2xBTv6zyelJ7h6QTgAH9thQ
References: <45E307D9.1090201@cisco.com>
From: "Michael Hammer \(mhammer\)" <mhammer@cisco.com>
To: "Jonathan Rosenberg \(jdrosen\)" <jdrosen@cisco.com>,
	"IETF Sipping List" <sipping@ietf.org>
X-OriginalArrivalTime: 26 Feb 2007 20:08:17.0803 (UTC)
	FILETIME=[DAE0D5B0:01C759E1]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=3968; t=1172520546;
	x=1173384546; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=mhammer@cisco.com;
	z=From:=20=22Michael=20Hammer=20\(mhammer\)=22=20<mhammer@cisco.com>
	|Subject:=20RE=3A=20[Sipping]=20Service=20Identification=3A=20offerless=2
	0INVITE=20case |Sender:=20
	|To:=20=22Jonathan=20Rosenberg=20\(jdrosen\)=22=20<jdrosen@cisco.com>,
	=0A
	=20=20=20=20=20=20=20=20=22IETF=20Sipping=20List=22=20<sipping@ietf.org>;
	bh=gnmcVI85aK+5vKdqfMDXpAdu0+O3fPNmg7ud8BskLv0=;
	b=Svjl1eid76iENqP8u+0X8+Y4/NeiCqtkmEkR6vExjUCw4/1b+/nEXeuyb2TRzgZV4flxLCGp
	DlT+Q6SdKKy7CbWTpa9A+Sh2M5/omWWBZxOsNLmr3zQEhe5TsETnM1mZ;
Authentication-Results: rtp-dkim-1; header.From=mhammer@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 3002fc2e661cd7f114cb6bae92fe88f1
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Is it not possible that in some cases SDP could also be encrypted?

Mike
=20

> -----Original Message-----
> From: Jonathan Rosenberg (jdrosen)=20
> Sent: Monday, February 26, 2007 11:16 AM
> To: IETF Sipping List
> Subject: [Sipping] Service Identification: offerless INVITE case
>=20
> Andrew made a great point in one of his mails which merits=20
> promotion into its own thread.
>=20
> I and others have asserted that you should be able to=20
> identify the service by things naturally in the request, and=20
> in particular, SDP has come up as a key thing. For example,=20
> in a multiplayer game, there might be an m-line for the=20
> particular proprietary protocol used for game moves. Or, if=20
> its MSRP, you might have specific content types for game=20
> moves. When present, those exactly indicate what the service is.
>=20
> But what do we do about offerless INVITE? In particular,=20
> there are cases where we need to invoke an application server=20
> to handle a request, and what if we don't have the SDP in the=20
> INVITE? Its too late to insert an application server once the=20
> INVITE has already passed to the called party and the 200 OK is sent.
>=20
> This is a great point. I'd rather just fix this little=20
> problem, the define an entire notion of services and service=20
> identifiers that are not necessary, just because we have=20
> offerless INVITEs.
>=20
> There are lots of ways to handle this problem. My preferred=20
> approach would be that offerless INVITE contain a capability=20
> description, using Flemmings new capability framework. The=20
> INVITE would contain a body that
>   looks like SDP, but has a different content type and=20
> content-disposition. It lists the media streams, codecs and=20
> other capabilities that exist. Based on this, the called=20
> party makes the offer in a 2xx or 18x like normal, and the=20
> answer in the ACK/PRACK as normal.=20
> It is completely backwards compatible.
>=20
> Now, a related issue is what happens if a session starts with=20
> audio only, and then the user wants to turn this into a game?=20
> Its too late to add new application servers into the call=20
> path, which might need to be there for games. What can we do?
>=20
> Here too, there is a simple answer. If a network provider=20
> requires app servers to get inserted, or a phone requires=20
> dispatch to a different application, it would receive a=20
> re-INVITE to add the gaming stream, and then 3xx it. The=20
> Contact would contain an embedded Replaces header with the=20
> dialog IDs for the call in progress. THe originating UA would=20
> then do a totally new INVITE that replaces the call in=20
> progress. This would allow the network and/or phone to=20
> redispatch, yet provide a seamless user experience. Billing=20
> can be done however the operators want, since the signaling=20
> is very clear on exactly what happened and when.=20
> Furthermore, the replaces mechansim would only get used if=20
> there was someone in the call path that needed to see it.
>=20
> I'd rather specify things like this, than invent an entire=20
> services framework that would lead to balkanization and=20
> non-interoperable islands of SIP.
>=20
> -Jonathan R.
>=20
>=20
> --=20
> Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
> Cisco Fellow                                   Parsippany, NJ=20
> 07054-2711
> Cisco Systems
> jdrosen@cisco.com                              FAX:   (973) 952-5050
> http://www.jdrosen.net                         PHONE: (973) 952-5000
> http://www.cisco.com
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 17:04:42 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLnwJ-00040G-Bc; Mon, 26 Feb 2007 17:03:23 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLnr4-0000GH-CS
	for sipping@ietf.org; Mon, 26 Feb 2007 16:57:58 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLnAC-0000b6-KO
	for sipping@ietf.org; Mon, 26 Feb 2007 16:13:57 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	C890020421; Mon, 26 Feb 2007 22:13:31 +0100 (CET)
X-AuditID: c1b4fb3c-b1fccbb0000007de-b6-45e34d7b74af 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	ADDDB20019; Mon, 26 Feb 2007 22:13:31 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 22:13:31 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification: offerless INVITE case
Date: Mon, 26 Feb 2007 22:13:28 +0100
Message-ID: <7374777208BDC7449D5620EF94232567024655D2@esealmw113.eemea.ericsson.se>
In-Reply-To: <45E331B4.4090700@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification: offerless INVITE case
Thread-Index: AcdZ2nHBW2t5ZR+eSJiRSdgDVlqGvQADqngg
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Jonathan Rosenberg" <jdrosen@cisco.com>
X-OriginalArrivalTime: 26 Feb 2007 21:13:31.0610 (UTC)
	FILETIME=[F7B063A0:01C759EA]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: f6ef73100908d67495ce675c3fe8f472
Cc: Paul Kyzivat <pkyzivat@cisco.com>, IETF Sipping List <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,=20

>I can't follow the figure since CRLFs got inserted by mailers=20
>along the way. Can you send one whose width is small?

I'll try again :)

UAa    P1    P2    UAb   UAc
 -INV-> -INV-> -INV->
 <-200- <-200- <-200-
 -INV->
 <-3xx-
 -INV--------------------->

>I don't follow the problem though. The 3xx response would be=20
>seen by all record-routing proxies upstream of the entity=20
>that sends the mid-dialog 3xx.

Yes, but what about all record-routing proxies downstream?

>The new INVITE would be seen by all proxies.

Assuming the new INVITE traverses the same set of proxies (if
Service-Route is used, that set of proxies will be the same), yes, but I
don't think we can assume that.

>To make sure it goes to the right UA instance you use GRUU.=20
>
>BYE is sent by the entity receiving the replaces, as normal. What am I
missing?

What if the new UA is NOT the same as the original? What if the original
UA doesn't support the new capability/media being added in the
re-INVITE, and is routed to another UA? Shall such scenario not be
allowed?

Regards,

Christer







>=20
> -Jonathan R.
>=20
> Christer Holmberg (JO/LMF) wrote:
>=20
> > Hi,
> >=20
> >=20
> >>>If terminated automatically, what happens to the nodes=20
> situated AFTER=20
> >>>the proxy that sent 3xx? They have no clue about the 3xx=20
> being sent,=20
> >>>and there is no guarantee the re-INVITE will traverse the=20
> same nodes=20
> >>>(or even reach the same remote UA).
> >>
> >>This is a general issue for INVITE/Replaces. The 3xx is=20
> simply a way=20
> >>to get the UAC to do it. The UAC could decide to do it even=20
> without a=20
> >>3xx.
> >>In the normal course of routing this any servers that are needed on=20
> >>the path will have to be discovered based on the content of the=20
> >>request.
> >=20
> >=20
> > My issue was how the proxies in the original dialog will=20
> know that a=20
> > 3xx has been sent (see picture):
> >=20
> > UA A ------------------- Proxy 1 ---------------- Proxy 2
> > ------------------ UA B     UA C
> >=20
> > -------------INVITE---------> --------INVITE--------->
> > --------INVITE---------->
> > <------------200------------- <-------200-------------
> > <-------200--------------
> > -------------re-INVITE------>
> > <------------3xx-------------
> > -------------INVITE
> > (new)-------------------------------------------------------------->
> >=20
> > Now, if Proxy 2 will not be traversed by the new INVITE, it=20
> will not=20
> > know that the dialog has been replaced.
> >=20
> > And, if the new INVITE reaches UA C, UA B will have no clue=20
> that the=20
> > dialog has been replaced.
> >=20
> > Regards,
> >=20
> > Christer
> >=20
> >=20
> >=20
> >=20
> >=20
> >=20
> >=20
> > =20
> >=20
> >>>Etc etc etc.
> >>>
> >>>Regards,
> >>>
> >>>Christer
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>>	Paul
> >>>>
> >>>>Jonathan Rosenberg wrote:
> >>>>
> >>>>>Andrew made a great point in one of his mails which merits
> >>>>
> >>>>promotion
> >>>>
> >>>>>into its own thread.
> >>>>>
> >>>>>I and others have asserted that you should be able to
> >>
> >>identify the
> >>
> >>>>>service by things naturally in the request, and in
> >>>>
> >>>>particular, SDP has
> >>>>
> >>>>>come up as a key thing. For example, in a multiplayer=20
> game, there=20
> >>>>>might be an m-line for the particular proprietary
> >>
> >>protocol used for
> >>
> >>>>>game moves. Or, if its MSRP, you might have specific
> >>>>
> >>>>content types for
> >>>>
> >>>>>game moves. When present, those exactly indicate what the
> >>>>
> >>>>service is.
> >>>>
> >>>>>But what do we do about offerless INVITE? In particular,
> >>
> >>there are
> >>
> >>>>>cases where we need to invoke an application server to handle a=20
> >>>>>request, and what if we don't have the SDP in the INVITE?
> >>>>
> >>>>Its too late
> >>>>
> >>>>>to insert an application server once the INVITE has already
> >>>>
> >>>>passed to
> >>>>
> >>>>>the called party and the 200 OK is sent.
> >>>>>
> >>>>>This is a great point. I'd rather just fix this little
> >>
> >>problem, the
> >>
> >>>>>define an entire notion of services and service identifiers
> >>>>
> >>>>that are
> >>>>
> >>>>>not necessary, just because we have offerless INVITEs.
> >>>>>
> >>>>>There are lots of ways to handle this problem. My preferred
> >>>>
> >>>>approach
> >>>>
> >>>>>would be that offerless INVITE contain a capability
> >>>>
> >>>>description, using
> >>>>
> >>>>>Flemmings new capability framework. The INVITE would
> >>
> >>contain a body
> >>
> >>>>>that  looks like SDP, but has a different content type and=20
> >>>>>content-disposition. It lists the media streams, codecs=20
> and other=20
> >>>>>capabilities that exist. Based on this, the called party
> >>
> >>makes the
> >>
> >>>>>offer in a 2xx or 18x like normal, and the answer in the
> >>>>
> >>>>ACK/PRACK as normal.
> >>>>
> >>>>>It is completely backwards compatible.
> >>>>>
> >>>>>Now, a related issue is what happens if a session starts
> >>
> >>with audio
> >>
> >>>>>only, and then the user wants to turn this into a game? Its
> >>>>
> >>>>too late
> >>>>
> >>>>>to add new application servers into the call path, which
> >>>>
> >>>>might need to
> >>>>
> >>>>>be there for games. What can we do?
> >>>>>
> >>>>>Here too, there is a simple answer. If a network provider
> >>>>
> >>>>requires app
> >>>>
> >>>>>servers to get inserted, or a phone requires dispatch to a
> >>>>
> >>>>different
> >>>>
> >>>>>application, it would receive a re-INVITE to add the
> >>
> >>gaming stream,
> >>
> >>>>>and then 3xx it. The Contact would contain an embedded
> >>>>
> >>>>Replaces header
> >>>>
> >>>>>with the dialog IDs for the call in progress. THe
> >>>>
> >>>>originating UA would
> >>>>
> >>>>>then do a totally new INVITE that replaces the call in
> >>>>
> >>>>progress. This
> >>>>
> >>>>>would allow the network and/or phone to redispatch, yet=20
> provide a=20
> >>>>>seamless user experience. Billing can be done however the
> >>
> >>operators
> >>
> >>>>>want, since the signaling is very clear on exactly what
> >>>>
> >>>>happened and when.
> >>>>
> >>>>>Furthermore, the replaces mechansim would only get used if
> >>>>
> >>>>there was
> >>>>
> >>>>>someone in the call path that needed to see it.
> >>>>>
> >>>>>I'd rather specify things like this, than invent an
> >>
> >>entire services
> >>
> >>>>>framework that would lead to balkanization and non-interoperable=20
> >>>>>islands of SIP.
> >>>>>
> >>>>>-Jonathan R.
> >>>>>
> >>>>>
> >>>>
> >>>>_______________________________________________
> >>>>Sipping mailing list
> >>
> >>https://www1.ietf.org/mailman/listinfo/sipping
> >>
> >>>>This list is for NEW development of the application of SIP Use=20
> >>>>sip-implementors@cs.columbia.edu for questions on current sip Use=20
> >>>>sip@ietf.org for new developments of core SIP
> >>>>
> >>>
> >=20
>=20
> --=20
> Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
> Cisco Fellow                                   Parsippany, NJ=20
> 07054-2711
> Cisco Systems
> jdrosen@cisco.com                              FAX:   (973) 952-5050
> http://www.jdrosen.net                         PHONE: (973) 952-5000
> http://www.cisco.com
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 17:36:09 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLoRi-0003wy-Qq; Mon, 26 Feb 2007 17:35:50 -0500
Received: from [10.90.34.44] (helo=chiedprmail1.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLoRJ-0003o5-R2
	for sipping@ietf.org; Mon, 26 Feb 2007 17:35:25 -0500
Received: from sj-iport-5.cisco.com ([171.68.10.87])
	by chiedprmail1.ietf.org with esmtp (Exim 4.43) id 1HLoRH-0001U7-Cx
	for sipping@ietf.org; Mon, 26 Feb 2007 17:35:25 -0500
Received: from sj-dkim-7.cisco.com ([171.68.10.88])
	by sj-iport-5.cisco.com with ESMTP; 26 Feb 2007 14:35:23 -0800
X-IronPort-AV: i="4.14,221,1170662400"; 
	d="scan'208"; a="393856083:sNHT3043638172"
Received: from sj-core-4.cisco.com (sj-core-4.cisco.com [171.68.223.138])
	by sj-dkim-7.cisco.com (8.12.11/8.12.11) with ESMTP id l1QMZLLt018845; 
	Mon, 26 Feb 2007 14:35:21 -0800
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by sj-core-4.cisco.com (8.12.10/8.12.6) with ESMTP id l1QMZKnH018804;
	Mon, 26 Feb 2007 14:35:21 -0800 (PST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 17:35:20 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-201.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 17:35:19 -0500
Message-ID: <45E360A7.1030508@cisco.com>
Date: Mon, 26 Feb 2007 17:35:19 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: "Christer Holmberg (JO/LMF)" <christer.holmberg@ericsson.com>
Subject: Re: [Sipping] Service Identification: offerless INVITE case
References: <7374777208BDC7449D5620EF94232567024655D2@esealmw113.eemea.ericsson.se>
In-Reply-To: <7374777208BDC7449D5620EF94232567024655D2@esealmw113.eemea.ericsson.se>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 22:35:20.0030 (UTC)
	FILETIME=[6555D7E0:01C759F6]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=1718; t=1172529321;
	x=1173393321; c=relaxed/simple; s=sjdkim7002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=3A=20offerless=2
	0INVITE=20case |Sender:=20;
	bh=40kA6+ZM97YUO/8H30uSC6+648ZdUuAmhapI8PL7MSg=;
	b=HOiqHZQp7/3hjBLYyeiCj8gHRRVGJNZk9xjHHZUHFr8rHxvkuuSX3MiYlIL67x3GUZyD4RG/
	dOGnVRPbJqBZjtUDOz8MzG+TivMeqIbbkS0r7vsWC9VeFRTs4IBF9WYP;
Authentication-Results: sj-dkim-7; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim7002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 52f7a77164458f8c7b36b66787c853da
Cc: Paul Kyzivat <pkyzivat@cisco.com>, IETF Sipping List <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

inline.

Christer Holmberg (JO/LMF) wrote:

> Hi, 
> 
> 
>>I can't follow the figure since CRLFs got inserted by mailers 
>>along the way. Can you send one whose width is small?
> 
> 
> I'll try again :)
> 
> UAa    P1    P2    UAb   UAc
>  -INV-> -INV-> -INV->
>  <-200- <-200- <-200-
>  -INV->
>  <-3xx-
>  -INV--------------------->
> 
>>I don't follow the problem though. The 3xx response would be 
>>seen by all record-routing proxies upstream of the entity 
>>that sends the mid-dialog 3xx.
> 
> 
> Yes, but what about all record-routing proxies downstream?

They won't see the 3xx, but they'll see the BYE. Thats enough.


> 
> 
>>The new INVITE would be seen by all proxies.
> 
> 
> Assuming the new INVITE traverses the same set of proxies (if
> Service-Route is used, that set of proxies will be the same), yes, but I
> don't think we can assume that.

Not clear they really need to.

> 
> 
>>To make sure it goes to the right UA instance you use GRUU. 
>>
>>BYE is sent by the entity receiving the replaces, as normal. What am I
> 
> missing?
> 
> What if the new UA is NOT the same as the original? What if the original
> UA doesn't support the new capability/media being added in the
> re-INVITE, and is routed to another UA? Shall such scenario not be
> allowed?

I think you'd need to do that as a transfer.

-Jonathan R.

-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 17:38:10 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLoTx-00057U-Uv; Mon, 26 Feb 2007 17:38:09 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLoTJ-0004pt-8I
	for sipping@ietf.org; Mon, 26 Feb 2007 17:37:29 -0500
Received: from sj-iport-4.cisco.com ([171.68.10.86])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLoTG-00040i-KV
	for sipping@ietf.org; Mon, 26 Feb 2007 17:37:29 -0500
Received: from sj-dkim-7.cisco.com ([171.68.10.88])
	by sj-iport-4.cisco.com with ESMTP; 26 Feb 2007 14:37:26 -0800
X-IronPort-AV: i="4.14,221,1170662400"; 
	d="scan'208"; a="42831091:sNHT68369607"
Received: from sj-core-4.cisco.com (sj-core-4.cisco.com [171.68.223.138])
	by sj-dkim-7.cisco.com (8.12.11/8.12.11) with ESMTP id l1QMbQtP019865
	for <sipping@ietf.org>; Mon, 26 Feb 2007 14:37:26 -0800
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by sj-core-4.cisco.com (8.12.10/8.12.6) with ESMTP id l1QMbCnP019584
	for <sipping@ietf.org>; Mon, 26 Feb 2007 14:37:25 -0800 (PST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 17:36:55 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-202.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 17:36:54 -0500
Message-ID: <45E36106.2000201@cisco.com>
Date: Mon, 26 Feb 2007 17:36:54 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: "Michael Hammer (mhammer)" <mhammer@cisco.com>
Subject: Re: [Sipping] Service Identification: offerless INVITE case
References: <45E307D9.1090201@cisco.com>
	<072C5B76F7CEAB488172C6F64B30B5E302A8D996@xmb-rtp-20b.amer.cisco.com>
In-Reply-To: <072C5B76F7CEAB488172C6F64B30B5E302A8D996@xmb-rtp-20b.amer.cisco.com>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 22:36:55.0010 (UTC)
	FILETIME=[9DF2A420:01C759F6]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=4494; t=1172529446;
	x=1173393446; c=relaxed/simple; s=sjdkim7002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=3A=20offerless=2
	0INVITE=20case |Sender:=20;
	bh=tzfcM3e9T7jxrNAK16jlOBfzHI9Qm7OiryhnyzOfy2c=;
	b=eoMfHsSFhgyhqkZK58gEJlHSP/Qwd4Zc61LfgwC3jWhTryKQ/GRa1qCutGu8e1j3pJmKQeiC
	/q92hHL9EeFhCJgFCoQ0CRtlYGtIVRSm6i3YfTQlqrK4yAlCnlC6pOJp;
Authentication-Results: sj-dkim-7; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim7002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 057ebe9b96adec30a7efb2aeda4c26a4
Cc: IETF Sipping List <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Well, that depends on whether you think we'll ever see S/MIME or not. If 
it ever happens, the problem is easily remedied by including select 
portions of the actual SDP in the body. But its a problem in theory 
right now, like it or not.

-Jonathan R.

Michael Hammer (mhammer) wrote:

> Is it not possible that in some cases SDP could also be encrypted?
> 
> Mike
>  
> 
> 
>>-----Original Message-----
>>From: Jonathan Rosenberg (jdrosen) 
>>Sent: Monday, February 26, 2007 11:16 AM
>>To: IETF Sipping List
>>Subject: [Sipping] Service Identification: offerless INVITE case
>>
>>Andrew made a great point in one of his mails which merits 
>>promotion into its own thread.
>>
>>I and others have asserted that you should be able to 
>>identify the service by things naturally in the request, and 
>>in particular, SDP has come up as a key thing. For example, 
>>in a multiplayer game, there might be an m-line for the 
>>particular proprietary protocol used for game moves. Or, if 
>>its MSRP, you might have specific content types for game 
>>moves. When present, those exactly indicate what the service is.
>>
>>But what do we do about offerless INVITE? In particular, 
>>there are cases where we need to invoke an application server 
>>to handle a request, and what if we don't have the SDP in the 
>>INVITE? Its too late to insert an application server once the 
>>INVITE has already passed to the called party and the 200 OK is sent.
>>
>>This is a great point. I'd rather just fix this little 
>>problem, the define an entire notion of services and service 
>>identifiers that are not necessary, just because we have 
>>offerless INVITEs.
>>
>>There are lots of ways to handle this problem. My preferred 
>>approach would be that offerless INVITE contain a capability 
>>description, using Flemmings new capability framework. The 
>>INVITE would contain a body that
>>  looks like SDP, but has a different content type and 
>>content-disposition. It lists the media streams, codecs and 
>>other capabilities that exist. Based on this, the called 
>>party makes the offer in a 2xx or 18x like normal, and the 
>>answer in the ACK/PRACK as normal. 
>>It is completely backwards compatible.
>>
>>Now, a related issue is what happens if a session starts with 
>>audio only, and then the user wants to turn this into a game? 
>>Its too late to add new application servers into the call 
>>path, which might need to be there for games. What can we do?
>>
>>Here too, there is a simple answer. If a network provider 
>>requires app servers to get inserted, or a phone requires 
>>dispatch to a different application, it would receive a 
>>re-INVITE to add the gaming stream, and then 3xx it. The 
>>Contact would contain an embedded Replaces header with the 
>>dialog IDs for the call in progress. THe originating UA would 
>>then do a totally new INVITE that replaces the call in 
>>progress. This would allow the network and/or phone to 
>>redispatch, yet provide a seamless user experience. Billing 
>>can be done however the operators want, since the signaling 
>>is very clear on exactly what happened and when. 
>>Furthermore, the replaces mechansim would only get used if 
>>there was someone in the call path that needed to see it.
>>
>>I'd rather specify things like this, than invent an entire 
>>services framework that would lead to balkanization and 
>>non-interoperable islands of SIP.
>>
>>-Jonathan R.
>>
>>
>>-- 
>>Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
>>Cisco Fellow                                   Parsippany, NJ 
>>07054-2711
>>Cisco Systems
>>jdrosen@cisco.com                              FAX:   (973) 952-5050
>>http://www.jdrosen.net                         PHONE: (973) 952-5000
>>http://www.cisco.com
>>
>>_______________________________________________
>>Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>>This list is for NEW development of the application of SIP 
>>Use sip-implementors@cs.columbia.edu for questions on current 
>>sip Use sip@ietf.org for new developments of core SIP
>>
> 
> 

-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 18:22:35 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLpA8-0005Dy-LQ; Mon, 26 Feb 2007 18:21:44 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLp9V-0004qD-Rj
	for sipping@ietf.org; Mon, 26 Feb 2007 18:21:05 -0500
Received: from sj-iport-1-in.cisco.com ([171.71.176.70]
	helo=sj-iport-1.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HLp9T-0002oc-Ky
	for sipping@ietf.org; Mon, 26 Feb 2007 18:21:05 -0500
Received: from sj-dkim-5.cisco.com ([171.68.10.79])
	by sj-iport-1.cisco.com with ESMTP; 26 Feb 2007 15:21:03 -0800
X-IronPort-AV: i="4.14,221,1170662400"; 
	d="scan'208"; a="764592821:sNHT52876086"
Received: from sj-core-3.cisco.com (sj-core-3.cisco.com [171.68.223.137])
	by sj-dkim-5.cisco.com (8.12.11/8.12.11) with ESMTP id l1QNL3qN021222; 
	Mon, 26 Feb 2007 15:21:03 -0800
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by sj-core-3.cisco.com (8.12.10/8.12.6) with ESMTP id l1QNL2ho019442;
	Mon, 26 Feb 2007 15:21:02 -0800 (PST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 18:21:02 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 18:21:02 -0500
Message-ID: <45E36B5B.2080509@cisco.com>
Date: Mon, 26 Feb 2007 18:20:59 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Christer Holmberg (JO/LMF)" <christer.holmberg@ericsson.com>
Subject: Re: [Sipping] Service Identification: offerless INVITE case
References: <7374777208BDC7449D5620EF9423256702465433@esealmw113.eemea.ericsson.se>
In-Reply-To: <7374777208BDC7449D5620EF9423256702465433@esealmw113.eemea.ericsson.se>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 23:21:02.0026 (UTC)
	FILETIME=[C7B13AA0:01C759FC]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=6041; t=1172532063;
	x=1173396063; c=relaxed/simple; s=sjdkim5002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=3A=20offerless=2
	0INVITE=20case |Sender:=20;
	bh=mqYYEKJEQEur/4c73mxJY6T/gtbQo1cWtYaKQSyDFs8=;
	b=e0aE+Ypajqr4CNq6L2UPwaeZr1mYAWObHzUG+hSQYDRuAiVp6Hhu+6OamFM2Ejkxt1I3hMu2
	r42DYmww/0eLHtVpD4r1m+JVwPjbjwj8REp1Po3nop1SxKegoONI/qUx;
Authentication-Results: sj-dkim-5; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim5002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 995b2e24d23b953c94bac5288c432399
Cc: IETF Sipping List <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Christer Holmberg (JO/LMF) wrote:
> Hi, 
> 
>>> If terminated automatically, what happens to the nodes situated AFTER 
>>> the proxy that sent 3xx? They have no clue about the 3xx being sent, 
>>> and there is no guarantee the re-INVITE will traverse the same nodes 
>>> (or even reach the same remote UA).
>> This is a general issue for INVITE/Replaces. The 3xx is 
>> simply a way to get the UAC to do it. The UAC could decide to 
>> do it even without a 3xx. 
>> In the normal course of routing this any servers that are 
>> needed on the path will have to be discovered based on the 
>> content of the request.
> 
> My issue was how the proxies in the original dialog will know that a 3xx
> has been sent (see picture):

There should be no reason for any of the proxies to take note of the 
3xx. It is simply a way to tell the UAC to do an INVITE/Replaces. It 
could decide to do an INVITE/Replaces without receiving a 3xx, so 
anything that depends on seeing the 3xx is broken.

The new invite will need be routed to application servers in whatever 
way is appropriate for it, independent of what application servers had 
been involved in the old session. If the new invite simply adds on 
additional features to what was used before, then presumably any app 
servers needed before will be included again, and any new servers needed 
because of additions to the invite will be included as well.

If some server that had been included is no longer included on the new 
invite, then it will drop out when the BYE is sent on the old session.

	Paul

> UA A ------------------- Proxy 1 ---------------- Proxy 2
> ------------------ UA B     UA C
> 
> -------------INVITE---------> --------INVITE--------->
> --------INVITE---------->
> <------------200------------- <-------200-------------
> <-------200--------------
> -------------re-INVITE------>
> <------------3xx-------------
> -------------INVITE
> (new)-------------------------------------------------------------->
> 
> Now, if Proxy 2 will not be traversed by the new INVITE, it will not
> know that the dialog has been replaced.
> 
> And, if the new INVITE reaches UA C, UA B will have no clue that the
> dialog has been replaced.
> 
> Regards,
> 
> Christer
> 
> 
> 
> 
> 
> 
> 
>  
>>> Etc etc etc.
>>>
>>> Regards,
>>>
>>> Christer
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>> 	Paul
>>>>
>>>> Jonathan Rosenberg wrote:
>>>>> Andrew made a great point in one of his mails which merits
>>>> promotion
>>>>> into its own thread.
>>>>>
>>>>> I and others have asserted that you should be able to 
>> identify the 
>>>>> service by things naturally in the request, and in
>>>> particular, SDP has
>>>>> come up as a key thing. For example, in a multiplayer game, there 
>>>>> might be an m-line for the particular proprietary 
>> protocol used for 
>>>>> game moves. Or, if its MSRP, you might have specific
>>>> content types for
>>>>> game moves. When present, those exactly indicate what the
>>>> service is.
>>>>> But what do we do about offerless INVITE? In particular, 
>> there are 
>>>>> cases where we need to invoke an application server to handle a 
>>>>> request, and what if we don't have the SDP in the INVITE?
>>>> Its too late
>>>>> to insert an application server once the INVITE has already
>>>> passed to
>>>>> the called party and the 200 OK is sent.
>>>>>
>>>>> This is a great point. I'd rather just fix this little 
>> problem, the 
>>>>> define an entire notion of services and service identifiers
>>>> that are
>>>>> not necessary, just because we have offerless INVITEs.
>>>>>
>>>>> There are lots of ways to handle this problem. My preferred
>>>> approach
>>>>> would be that offerless INVITE contain a capability
>>>> description, using
>>>>> Flemmings new capability framework. The INVITE would 
>> contain a body 
>>>>> that  looks like SDP, but has a different content type and 
>>>>> content-disposition. It lists the media streams, codecs and other 
>>>>> capabilities that exist. Based on this, the called party 
>> makes the 
>>>>> offer in a 2xx or 18x like normal, and the answer in the
>>>> ACK/PRACK as normal.
>>>>> It is completely backwards compatible.
>>>>>
>>>>> Now, a related issue is what happens if a session starts 
>> with audio 
>>>>> only, and then the user wants to turn this into a game? Its
>>>> too late
>>>>> to add new application servers into the call path, which
>>>> might need to
>>>>> be there for games. What can we do?
>>>>>
>>>>> Here too, there is a simple answer. If a network provider
>>>> requires app
>>>>> servers to get inserted, or a phone requires dispatch to a
>>>> different
>>>>> application, it would receive a re-INVITE to add the 
>> gaming stream, 
>>>>> and then 3xx it. The Contact would contain an embedded
>>>> Replaces header
>>>>> with the dialog IDs for the call in progress. THe
>>>> originating UA would
>>>>> then do a totally new INVITE that replaces the call in
>>>> progress. This
>>>>> would allow the network and/or phone to redispatch, yet provide a 
>>>>> seamless user experience. Billing can be done however the 
>> operators 
>>>>> want, since the signaling is very clear on exactly what
>>>> happened and when.
>>>>> Furthermore, the replaces mechansim would only get used if
>>>> there was
>>>>> someone in the call path that needed to see it.
>>>>>
>>>>> I'd rather specify things like this, than invent an 
>> entire services 
>>>>> framework that would lead to balkanization and non-interoperable 
>>>>> islands of SIP.
>>>>>
>>>>> -Jonathan R.
>>>>>
>>>>>
>>>> _______________________________________________
>>>> Sipping mailing list  
>> https://www1.ietf.org/mailman/listinfo/sipping
>>>> This list is for NEW development of the application of SIP Use 
>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>> sip@ietf.org for new developments of core SIP
>>>>
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Mon Feb 26 18:23:34 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLpBs-0005x8-W2; Mon, 26 Feb 2007 18:23:32 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLpBQ-0005k6-KQ
	for sipping@ietf.org; Mon, 26 Feb 2007 18:23:04 -0500
Received: from sj-iport-4.cisco.com ([171.68.10.86])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLpBL-0003DW-4T
	for sipping@ietf.org; Mon, 26 Feb 2007 18:23:04 -0500
Received: from sj-dkim-8.cisco.com ([171.68.10.93])
	by sj-iport-4.cisco.com with ESMTP; 26 Feb 2007 15:22:58 -0800
X-IronPort-AV: i="4.14,221,1170662400"; 
	d="scan'208"; a="42842103:sNHT50770917"
Received: from sj-core-3.cisco.com (sj-core-3.cisco.com [171.68.223.137])
	by sj-dkim-8.cisco.com (8.12.11/8.12.11) with ESMTP id l1QNMw29022457
	for <sipping@ietf.org>; Mon, 26 Feb 2007 15:22:58 -0800
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by sj-core-3.cisco.com (8.12.10/8.12.6) with ESMTP id l1QNMRiO019871
	for <sipping@ietf.org>; Mon, 26 Feb 2007 15:22:58 -0800 (PST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 18:22:51 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 26 Feb 2007 18:22:51 -0500
Message-ID: <45E36BC8.8070905@cisco.com>
Date: Mon, 26 Feb 2007 18:22:48 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Michael Hammer (mhammer)" <mhammer@cisco.com>
Subject: Re: [Sipping] Service Identification: offerless INVITE case
References: <45E307D9.1090201@cisco.com>
	<072C5B76F7CEAB488172C6F64B30B5E302A8D996@xmb-rtp-20b.amer.cisco.com>
In-Reply-To: <072C5B76F7CEAB488172C6F64B30B5E302A8D996@xmb-rtp-20b.amer.cisco.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 26 Feb 2007 23:22:51.0090 (UTC)
	FILETIME=[08B31320:01C759FD]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=4447; t=1172532178;
	x=1173396178; c=relaxed/simple; s=sjdkim8002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=3A=20offerless=2
	0INVITE=20case |Sender:=20;
	bh=VLBroSaVEwNlA8XcrU1pACD4GieQhPRfiryoBrwRbnI=;
	b=aabtwGkqHF/Y/G9GWBfllMXPC0ajV0pNLQPscNicgmzOnwEQx354dSXVrnEMEtx3nu19CYhj
	nC6tHz73lFzyaH3Q0G1lOKEt9a4PHJA1jsSBfuJD7UnQdR4MmymqxTdp;
Authentication-Results: sj-dkim-8; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim8002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 6e922792024732fb1bb6f346e63517e4
Cc: IETF Sipping List <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Michael Hammer (mhammer) wrote:
> Is it not possible that in some cases SDP could also be encrypted?

If it is, then we start to get into the subject of e2m policy. But I 
think IMS already assumes the SDP is not encrypted, so its not a problem 
for them.

	Paul

> Mike
>  
> 
>> -----Original Message-----
>> From: Jonathan Rosenberg (jdrosen) 
>> Sent: Monday, February 26, 2007 11:16 AM
>> To: IETF Sipping List
>> Subject: [Sipping] Service Identification: offerless INVITE case
>>
>> Andrew made a great point in one of his mails which merits 
>> promotion into its own thread.
>>
>> I and others have asserted that you should be able to 
>> identify the service by things naturally in the request, and 
>> in particular, SDP has come up as a key thing. For example, 
>> in a multiplayer game, there might be an m-line for the 
>> particular proprietary protocol used for game moves. Or, if 
>> its MSRP, you might have specific content types for game 
>> moves. When present, those exactly indicate what the service is.
>>
>> But what do we do about offerless INVITE? In particular, 
>> there are cases where we need to invoke an application server 
>> to handle a request, and what if we don't have the SDP in the 
>> INVITE? Its too late to insert an application server once the 
>> INVITE has already passed to the called party and the 200 OK is sent.
>>
>> This is a great point. I'd rather just fix this little 
>> problem, the define an entire notion of services and service 
>> identifiers that are not necessary, just because we have 
>> offerless INVITEs.
>>
>> There are lots of ways to handle this problem. My preferred 
>> approach would be that offerless INVITE contain a capability 
>> description, using Flemmings new capability framework. The 
>> INVITE would contain a body that
>>   looks like SDP, but has a different content type and 
>> content-disposition. It lists the media streams, codecs and 
>> other capabilities that exist. Based on this, the called 
>> party makes the offer in a 2xx or 18x like normal, and the 
>> answer in the ACK/PRACK as normal. 
>> It is completely backwards compatible.
>>
>> Now, a related issue is what happens if a session starts with 
>> audio only, and then the user wants to turn this into a game? 
>> Its too late to add new application servers into the call 
>> path, which might need to be there for games. What can we do?
>>
>> Here too, there is a simple answer. If a network provider 
>> requires app servers to get inserted, or a phone requires 
>> dispatch to a different application, it would receive a 
>> re-INVITE to add the gaming stream, and then 3xx it. The 
>> Contact would contain an embedded Replaces header with the 
>> dialog IDs for the call in progress. THe originating UA would 
>> then do a totally new INVITE that replaces the call in 
>> progress. This would allow the network and/or phone to 
>> redispatch, yet provide a seamless user experience. Billing 
>> can be done however the operators want, since the signaling 
>> is very clear on exactly what happened and when. 
>> Furthermore, the replaces mechansim would only get used if 
>> there was someone in the call path that needed to see it.
>>
>> I'd rather specify things like this, than invent an entire 
>> services framework that would lead to balkanization and 
>> non-interoperable islands of SIP.
>>
>> -Jonathan R.
>>
>>
>> -- 
>> Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
>> Cisco Fellow                                   Parsippany, NJ 
>> 07054-2711
>> Cisco Systems
>> jdrosen@cisco.com                              FAX:   (973) 952-5050
>> http://www.jdrosen.net                         PHONE: (973) 952-5000
>> http://www.cisco.com
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP 
>> Use sip-implementors@cs.columbia.edu for questions on current 
>> sip Use sip@ietf.org for new developments of core SIP
>>
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 27 00:56:25 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLvIL-0004Qm-LC; Tue, 27 Feb 2007 00:54:37 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLvIJ-0004HE-99
	for sipping@ietf.org; Tue, 27 Feb 2007 00:54:35 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLvIC-00047I-ES
	for sipping@ietf.org; Tue, 27 Feb 2007 00:54:35 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	C168B21AD4; Tue, 27 Feb 2007 06:54:19 +0100 (CET)
X-AuditID: c1b4fb3e-aeed1bb0000007e1-5c-45e3c78bb755 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	A9F2D21AFE; Tue, 27 Feb 2007 06:54:19 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 27 Feb 2007 06:54:19 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification: offerless INVITE case
Date: Tue, 27 Feb 2007 06:54:18 +0100
Message-ID: <7374777208BDC7449D5620EF94232567024656BA@esealmw113.eemea.ericsson.se>
In-Reply-To: <45E36B5B.2080509@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification: offerless INVITE case
Thread-Index: AcdZ/MsUtsX4eVxEQ+CvefReMlJ/JwAM6E0Q
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Paul Kyzivat" <pkyzivat@cisco.com>
X-OriginalArrivalTime: 27 Feb 2007 05:54:19.0849 (UTC)
	FILETIME=[B9174390:01C75A33]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: b84f8c8fba0e1389e5eb998b64078964
Cc: IETF Sipping List <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Hi,

Assuming the new UA (UA_C) is the same as the original (UA_B), I agree
there should be no problem as far as the BYE is concerned.

It is important to note, however, that this solution requires the
support of both GRUU and Replaces by all UAs involved.

One question is whether a proxy can send the mid-dialog 3xx, or if it
some B2BUA behavior is required.=20

For example, there is always a chance that the remote UA (UA_B) is
sending a new offer at the same time, which may reach UA_A before the
3xx. The proxy that sent 3xx most likely doesn't have to care about this
(UA_A should handle it, and return 491), but it's just something we
would need to be aware of since the offer/3xx will not be end-to-end in
this case.

And, of course the scenario where the new INVITE is rejected (or not
even sent) - for whatever reason - be supported, and the old dialog

Another problem may be that UA_A is required (until it receives BYE on
the original dialog) to maintain resources (including codecs) for TWO
dialogs - the original dialog, and the new - which may cause problems in
some cases. From a codec perspective that could be solved by putting the
original dialog on hold (by sending a hold re-INVITE, without the new
capabilities/medias) before sending the new invite.

Regards,

Christer
=20

> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivat@cisco.com]=20
> Sent: 27. helmikuuta 2007 1:21
> To: Christer Holmberg (JO/LMF)
> Cc: Jonathan Rosenberg; IETF Sipping List
> Subject: Re: [Sipping] Service Identification: offerless INVITE case
>=20
>=20
>=20
> Christer Holmberg (JO/LMF) wrote:
> > Hi,
> >=20
> >>> If terminated automatically, what happens to the nodes situated=20
> >>> AFTER the proxy that sent 3xx? They have no clue about=20
> the 3xx being=20
> >>> sent, and there is no guarantee the re-INVITE will=20
> traverse the same=20
> >>> nodes (or even reach the same remote UA).
> >> This is a general issue for INVITE/Replaces. The 3xx is=20
> simply a way=20
> >> to get the UAC to do it. The UAC could decide to do it=20
> even without a=20
> >> 3xx.
> >> In the normal course of routing this any servers that are=20
> needed on=20
> >> the path will have to be discovered based on the content of the=20
> >> request.
> >=20
> > My issue was how the proxies in the original dialog will=20
> know that a=20
> > 3xx has been sent (see picture):
>=20
> There should be no reason for any of the proxies to take note=20
> of the 3xx. It is simply a way to tell the UAC to do an=20
> INVITE/Replaces. It could decide to do an INVITE/Replaces=20
> without receiving a 3xx, so anything that depends on seeing=20
> the 3xx is broken.
>=20
> The new invite will need be routed to application servers in=20
> whatever way is appropriate for it, independent of what=20
> application servers had been involved in the old session. If=20
> the new invite simply adds on additional features to what was=20
> used before, then presumably any app servers needed before=20
> will be included again, and any new servers needed because of=20
> additions to the invite will be included as well.
>=20
> If some server that had been included is no longer included=20
> on the new invite, then it will drop out when the BYE is sent=20
> on the old session.
>=20
> 	Paul
>=20
> > UA A ------------------- Proxy 1 ---------------- Proxy 2
> > ------------------ UA B     UA C
> >=20
> > -------------INVITE---------> --------INVITE--------->
> > --------INVITE---------->
> > <------------200------------- <-------200-------------
> > <-------200--------------
> > -------------re-INVITE------>
> > <------------3xx-------------
> > -------------INVITE
> > (new)-------------------------------------------------------------->
> >=20
> > Now, if Proxy 2 will not be traversed by the new INVITE, it=20
> will not=20
> > know that the dialog has been replaced.
> >=20
> > And, if the new INVITE reaches UA C, UA B will have no clue=20
> that the=20
> > dialog has been replaced.
> >=20
> > Regards,
> >=20
> > Christer
> >=20
> >=20
> >=20
> >=20
> >=20
> >=20
> >=20
> > =20
> >>> Etc etc etc.
> >>>
> >>> Regards,
> >>>
> >>> Christer
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>> 	Paul
> >>>>
> >>>> Jonathan Rosenberg wrote:
> >>>>> Andrew made a great point in one of his mails which merits
> >>>> promotion
> >>>>> into its own thread.
> >>>>>
> >>>>> I and others have asserted that you should be able to
> >> identify the
> >>>>> service by things naturally in the request, and in
> >>>> particular, SDP has
> >>>>> come up as a key thing. For example, in a multiplayer=20
> game, there=20
> >>>>> might be an m-line for the particular proprietary
> >> protocol used for
> >>>>> game moves. Or, if its MSRP, you might have specific
> >>>> content types for
> >>>>> game moves. When present, those exactly indicate what the
> >>>> service is.
> >>>>> But what do we do about offerless INVITE? In particular,
> >> there are
> >>>>> cases where we need to invoke an application server to handle a=20
> >>>>> request, and what if we don't have the SDP in the INVITE?
> >>>> Its too late
> >>>>> to insert an application server once the INVITE has already
> >>>> passed to
> >>>>> the called party and the 200 OK is sent.
> >>>>>
> >>>>> This is a great point. I'd rather just fix this little
> >> problem, the
> >>>>> define an entire notion of services and service identifiers
> >>>> that are
> >>>>> not necessary, just because we have offerless INVITEs.
> >>>>>
> >>>>> There are lots of ways to handle this problem. My preferred
> >>>> approach
> >>>>> would be that offerless INVITE contain a capability
> >>>> description, using
> >>>>> Flemmings new capability framework. The INVITE would
> >> contain a body
> >>>>> that  looks like SDP, but has a different content type and=20
> >>>>> content-disposition. It lists the media streams, codecs=20
> and other=20
> >>>>> capabilities that exist. Based on this, the called party
> >> makes the
> >>>>> offer in a 2xx or 18x like normal, and the answer in the
> >>>> ACK/PRACK as normal.
> >>>>> It is completely backwards compatible.
> >>>>>
> >>>>> Now, a related issue is what happens if a session starts
> >> with audio
> >>>>> only, and then the user wants to turn this into a game? Its
> >>>> too late
> >>>>> to add new application servers into the call path, which
> >>>> might need to
> >>>>> be there for games. What can we do?
> >>>>>
> >>>>> Here too, there is a simple answer. If a network provider
> >>>> requires app
> >>>>> servers to get inserted, or a phone requires dispatch to a
> >>>> different
> >>>>> application, it would receive a re-INVITE to add the
> >> gaming stream,
> >>>>> and then 3xx it. The Contact would contain an embedded
> >>>> Replaces header
> >>>>> with the dialog IDs for the call in progress. THe
> >>>> originating UA would
> >>>>> then do a totally new INVITE that replaces the call in
> >>>> progress. This
> >>>>> would allow the network and/or phone to redispatch, yet=20
> provide a=20
> >>>>> seamless user experience. Billing can be done however the
> >> operators
> >>>>> want, since the signaling is very clear on exactly what
> >>>> happened and when.
> >>>>> Furthermore, the replaces mechansim would only get used if
> >>>> there was
> >>>>> someone in the call path that needed to see it.
> >>>>>
> >>>>> I'd rather specify things like this, than invent an
> >> entire services
> >>>>> framework that would lead to balkanization and=20
> non-interoperable=20
> >>>>> islands of SIP.
> >>>>>
> >>>>> -Jonathan R.
> >>>>>
> >>>>>
> >>>> _______________________________________________
> >>>> Sipping mailing list
> >> https://www1.ietf.org/mailman/listinfo/sipping
> >>>> This list is for NEW development of the application of SIP Use=20
> >>>> sip-implementors@cs.columbia.edu for questions on=20
> current sip Use=20
> >>>> sip@ietf.org for new developments of core SIP
> >>>>
> >=20
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 27 01:19:07 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLvfd-000120-SZ; Tue, 27 Feb 2007 01:18:41 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLvfc-00011q-U4
	for sipping@ietf.org; Tue, 27 Feb 2007 01:18:40 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLvfU-00087a-0F
	for sipping@ietf.org; Tue, 27 Feb 2007 01:18:40 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	86D7C21B07
	for <sipping@ietf.org>; Tue, 27 Feb 2007 07:18:31 +0100 (CET)
X-AuditID: c1b4fb3e-b06d4bb0000007e1-be-45e3cd373401 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	37D1521AF9
	for <sipping@ietf.org>; Tue, 27 Feb 2007 07:18:31 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 27 Feb 2007 07:18:31 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification: offerless INVITE case
Date: Tue, 27 Feb 2007 07:18:29 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD02926415@esealmw111>
In-Reply-To: <7374777208BDC7449D5620EF94232567024655D2@esealmw113.eemea.ericsson.se>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification: offerless INVITE case
Thread-Index: AcdZ2nHBW2t5ZR+eSJiRSdgDVlqGvQADqnggABNQHBA=
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: "IETF Sipping List" <sipping@ietf.org>
X-OriginalArrivalTime: 27 Feb 2007 06:18:31.0099 (UTC)
	FILETIME=[1A1A70B0:01C75A37]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: fcb459c204557d9509ce9c1b55d771f1
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Hi,

On this topic, doesn't this rely on the session-id making it to the Peer
UA, in my understanding there are a number of entities along the way
that may change the session id - any B2BUA such as an SBC .... (they may
be BAD BAD BAD, but they are THERE THERE THERE).

Steve.

-----Original Message-----
From: Christer Holmberg (JO/LMF) [mailto:christer.holmberg@ericsson.com]

Sent: Monday, February 26, 2007 10:13 PM
To: Jonathan Rosenberg
Cc: Paul Kyzivat; IETF Sipping List
Subject: RE: [Sipping] Service Identification: offerless INVITE case


Hi,=20

>I can't follow the figure since CRLFs got inserted by mailers along the

>way. Can you send one whose width is small?

I'll try again :)

UAa    P1    P2    UAb   UAc
 -INV-> -INV-> -INV->
 <-200- <-200- <-200-
 -INV->
 <-3xx-
 -INV--------------------->

>I don't follow the problem though. The 3xx response would be seen by=20
>all record-routing proxies upstream of the entity that sends the=20
>mid-dialog 3xx.

Yes, but what about all record-routing proxies downstream?

>The new INVITE would be seen by all proxies.

Assuming the new INVITE traverses the same set of proxies (if
Service-Route is used, that set of proxies will be the same), yes, but I
don't think we can assume that.

>To make sure it goes to the right UA instance you use GRUU.=20
>
>BYE is sent by the entity receiving the replaces, as normal. What am I
missing?

What if the new UA is NOT the same as the original? What if the original
UA doesn't support the new capability/media being added in the
re-INVITE, and is routed to another UA? Shall such scenario not be
allowed?

Regards,

Christer







>=20
> -Jonathan R.
>=20
> Christer Holmberg (JO/LMF) wrote:
>=20
> > Hi,
> >=20
> >=20
> >>>If terminated automatically, what happens to the nodes
> situated AFTER
> >>>the proxy that sent 3xx? They have no clue about the 3xx
> being sent,
> >>>and there is no guarantee the re-INVITE will traverse the
> same nodes
> >>>(or even reach the same remote UA).
> >>
> >>This is a general issue for INVITE/Replaces. The 3xx is
> simply a way
> >>to get the UAC to do it. The UAC could decide to do it even
> without a
> >>3xx.
> >>In the normal course of routing this any servers that are needed on=20
> >>the path will have to be discovered based on the content of the=20
> >>request.
> >=20
> >=20
> > My issue was how the proxies in the original dialog will
> know that a
> > 3xx has been sent (see picture):
> >=20
> > UA A ------------------- Proxy 1 ---------------- Proxy 2
> > ------------------ UA B     UA C
> >=20
> > -------------INVITE---------> --------INVITE--------->
> > --------INVITE---------->
> > <------------200------------- <-------200-------------
> > <-------200--------------
> > -------------re-INVITE------>
> > <------------3xx-------------
> > -------------INVITE
> > (new)-------------------------------------------------------------->
> >=20
> > Now, if Proxy 2 will not be traversed by the new INVITE, it
> will not
> > know that the dialog has been replaced.
> >=20
> > And, if the new INVITE reaches UA C, UA B will have no clue
> that the
> > dialog has been replaced.
> >=20
> > Regards,
> >=20
> > Christer
> >=20
> >=20
> >=20
> >=20
> >=20
> >=20
> >=20
> > =20
> >=20
> >>>Etc etc etc.
> >>>
> >>>Regards,
> >>>
> >>>Christer
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>>	Paul
> >>>>
> >>>>Jonathan Rosenberg wrote:
> >>>>
> >>>>>Andrew made a great point in one of his mails which merits
> >>>>
> >>>>promotion
> >>>>
> >>>>>into its own thread.
> >>>>>
> >>>>>I and others have asserted that you should be able to
> >>
> >>identify the
> >>
> >>>>>service by things naturally in the request, and in
> >>>>
> >>>>particular, SDP has
> >>>>
> >>>>>come up as a key thing. For example, in a multiplayer
> game, there
> >>>>>might be an m-line for the particular proprietary
> >>
> >>protocol used for
> >>
> >>>>>game moves. Or, if its MSRP, you might have specific
> >>>>
> >>>>content types for
> >>>>
> >>>>>game moves. When present, those exactly indicate what the
> >>>>
> >>>>service is.
> >>>>
> >>>>>But what do we do about offerless INVITE? In particular,
> >>
> >>there are
> >>
> >>>>>cases where we need to invoke an application server to handle a=20
> >>>>>request, and what if we don't have the SDP in the INVITE?
> >>>>
> >>>>Its too late
> >>>>
> >>>>>to insert an application server once the INVITE has already
> >>>>
> >>>>passed to
> >>>>
> >>>>>the called party and the 200 OK is sent.
> >>>>>
> >>>>>This is a great point. I'd rather just fix this little
> >>
> >>problem, the
> >>
> >>>>>define an entire notion of services and service identifiers
> >>>>
> >>>>that are
> >>>>
> >>>>>not necessary, just because we have offerless INVITEs.
> >>>>>
> >>>>>There are lots of ways to handle this problem. My preferred
> >>>>
> >>>>approach
> >>>>
> >>>>>would be that offerless INVITE contain a capability
> >>>>
> >>>>description, using
> >>>>
> >>>>>Flemmings new capability framework. The INVITE would
> >>
> >>contain a body
> >>
> >>>>>that  looks like SDP, but has a different content type and=20
> >>>>>content-disposition. It lists the media streams, codecs
> and other
> >>>>>capabilities that exist. Based on this, the called party
> >>
> >>makes the
> >>
> >>>>>offer in a 2xx or 18x like normal, and the answer in the
> >>>>
> >>>>ACK/PRACK as normal.
> >>>>
> >>>>>It is completely backwards compatible.
> >>>>>
> >>>>>Now, a related issue is what happens if a session starts
> >>
> >>with audio
> >>
> >>>>>only, and then the user wants to turn this into a game? Its
> >>>>
> >>>>too late
> >>>>
> >>>>>to add new application servers into the call path, which
> >>>>
> >>>>might need to
> >>>>
> >>>>>be there for games. What can we do?
> >>>>>
> >>>>>Here too, there is a simple answer. If a network provider
> >>>>
> >>>>requires app
> >>>>
> >>>>>servers to get inserted, or a phone requires dispatch to a
> >>>>
> >>>>different
> >>>>
> >>>>>application, it would receive a re-INVITE to add the
> >>
> >>gaming stream,
> >>
> >>>>>and then 3xx it. The Contact would contain an embedded
> >>>>
> >>>>Replaces header
> >>>>
> >>>>>with the dialog IDs for the call in progress. THe
> >>>>
> >>>>originating UA would
> >>>>
> >>>>>then do a totally new INVITE that replaces the call in
> >>>>
> >>>>progress. This
> >>>>
> >>>>>would allow the network and/or phone to redispatch, yet
> provide a
> >>>>>seamless user experience. Billing can be done however the
> >>
> >>operators
> >>
> >>>>>want, since the signaling is very clear on exactly what
> >>>>
> >>>>happened and when.
> >>>>
> >>>>>Furthermore, the replaces mechansim would only get used if
> >>>>
> >>>>there was
> >>>>
> >>>>>someone in the call path that needed to see it.
> >>>>>
> >>>>>I'd rather specify things like this, than invent an
> >>
> >>entire services
> >>
> >>>>>framework that would lead to balkanization and non-interoperable=20
> >>>>>islands of SIP.
> >>>>>
> >>>>>-Jonathan R.
> >>>>>
> >>>>>
> >>>>
> >>>>_______________________________________________
> >>>>Sipping mailing list
> >>
> >>https://www1.ietf.org/mailman/listinfo/sipping
> >>
> >>>>This list is for NEW development of the application of SIP Use=20
> >>>>sip-implementors@cs.columbia.edu for questions on current sip Use=20
> >>>>sip@ietf.org for new developments of core SIP
> >>>>
> >>>
> >=20
>=20
> --=20
> Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
> Cisco Fellow                                   Parsippany, NJ=20
> 07054-2711
> Cisco Systems
> jdrosen@cisco.com                              FAX:   (973) 952-5050
> http://www.jdrosen.net                         PHONE: (973) 952-5000
> http://www.cisco.com
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP Use
sip-implementors@cs.columbia.edu for questions on current sip Use
sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 27 01:52:26 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLwBr-0002BZ-1d; Tue, 27 Feb 2007 01:51:59 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLwBq-0002BQ-9x
	for sipping@ietf.org; Tue, 27 Feb 2007 01:51:58 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLwBj-0004XH-K3
	for sipping@ietf.org; Tue, 27 Feb 2007 01:51:58 -0500
Received: from mailgw4.ericsson.se (unknown [127.0.0.1])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	C08AC21B1A; Tue, 27 Feb 2007 07:51:38 +0100 (CET)
X-AuditID: c1b4fb3e-b1ed7bb0000007e1-4e-45e3d4fa3f0f 
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id
	A265F20570; Tue, 27 Feb 2007 07:51:38 +0100 (CET)
Received: from esealmw111.eemea.ericsson.se ([130.100.184.156]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 27 Feb 2007 07:51:38 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification in SIP
Date: Tue, 27 Feb 2007 07:51:36 +0100
Message-ID: <989763567E51AB48BE015DF7BB960CAD02926429@esealmw111>
In-Reply-To: <45E333C4.3090700@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification in SIP
Thread-Index: AcdZ26lPH6ZRYNNxSf+Q88khLuJLhgAXNlGw
From: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>
To: "Jonathan Rosenberg" <jdrosen@cisco.com>
X-OriginalArrivalTime: 27 Feb 2007 06:51:38.0522 (UTC)
	FILETIME=[BAB31FA0:01C75A3B]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 2086112c730e13d5955355df27e3074b
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Inline.=20

-----Original Message-----
From: Jonathan Rosenberg [mailto:jdrosen@cisco.com]=20
Sent: Monday, February 26, 2007 8:24 PM
To: Stephen Terrill (MI/EEM)
Cc: sipping@ietf.org
Subject: Re: [Sipping] Service Identification in SIP

inline.

Stephen Terrill (MI/EEM) wrote:

> Hi,
>=20
> I don't think that being able to identify services leads to any of the

> mentioned "-sation"s of SIP [well maybe industralisation ;-)], except=20
> to recognise that there are different services on SIP and that the=20
> media does not give the service (we can also start the session without

> media ...).

And I don't see how you can argue that point. Merely stating it does not
make it so.

To me, its really simple. If 3gpp goes and defines a service parameter
for "conversational telephony" and gives it some name, and puts it into
Accept-Contact, then the ability to receive a call from a non-3gpp
entity is ruined, since such a call will not include this parameter.=20
Thus, the right QoS treatment, billing, call processing, etc. will not
happen.

In what way is this NOT balkanization? Please be specific.

<Steve>
... Finally some motivation as to why it is said that the service
identifier is balkanization (uptill now has been "Merely stating it"),
however I don't see this in the same way.    The "balkanization effect"
will come from whether or not an operator wants to interconnect to the
general internet or not, which has nothing to do with being able to
identify the requested service.  An operator supporting the services,
will also be ability to make a "guess" at what the incoming requested
communication is.  Having made that guess, it will assume the rest of
the communication to follow according (and reject attempts to the
opposite).  Such interworking is nothing new.  As such, I do not see
that this reduces interworking at all.
</Steve>


>  At what point can it be said that SIP is "-ised" - or at what point=20
> to the SIP police say that this is breaking SIP?  If it doesn't get to

> all UAs?  If it doesn't go to all UAs that could do something useful=20
> with it?, If it doesn't go through a SIP proxy? ... Sometimes trying=20
> to deliver a service instead of rejecting the service due to=20
> incomptabilities can lead to a bad user experience - is that the best=20
> thing to do, or is that OK because the SIP protocol still connected?=20
> [I think that Jeroen captured this quite well in one of his emails]. I

> note that the second point raised, does handle this, leaving the=20
> question of a middle ground as well (i.e an indication of the service=20
> without it being "REQUIRED").

We have a mechanism for you to cause a call to fail if compatible
extensions are not supported (Require). If business requirements say
that this is the desired result, then use the tool. I certainly don't
hope that this is what you want for the common case. I can see how it is
the right thing for PoC, but thats about it.


>=20
> I think that it is better that the industry provides a common way to=20
> do this, instead of "looking into the SDP in one case, looking into=20
> the accept contact in another case" etc, that seems more complicated=20
> than necessary ....

I disagree completely.

THis is like saying that we should put everything that a SIP message
might indicate into a single "indicate" header, rather than invent new
headers, methods, put stuff in bodies, and so on. The 'service' is
nothing more or less than a summarization of the functionality that
results from the message being sent, and that functionality includes
everything from the method to the body.

<Steve>
I see this differently - I think that using different means in different
cases to identify the requested services is more parrallel to indicate
redirection of "moved perminatly" with a 301 response, by "moved
temporarily" through a completely different means, like introducing a
new response header ... Than it is to saying that we put all indications
into an "indicate" header.

Of course the full capabilities of SIP should remain - such as using SDP
to identify the requested media.
=20
</Steve>


>=20
> I noticed that in the BLISS group there was discussions on what=20
> assumptions could be made - i.e. the particular example being=20
> discussed was DND and whether it should be in the terminal or in the=20
> network.  If something comes out of that - won't that be, by default,=20
> a service definition or some sort.  It would be strange if that was=20
> the only service definition allowed ...

I sent another note on sipping on this topic. These are related problems
but differ quite fundamentally, since in those cases I am not requiring
the other party to know anything about this; its a matter of control
between a UA and its own domain. Even in that case, we are agreeing that
we want to be as flexible as possible to what constitutes these
services, and not standardize what it is.

-Jonathan R.

--=20
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 27 02:32:46 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLwoU-0005EU-TN; Tue, 27 Feb 2007 02:31:54 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLwoT-0005ED-3R
	for sipping@ietf.org; Tue, 27 Feb 2007 02:31:53 -0500
Received: from mailgw3.ericsson.se ([193.180.251.60])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLwoM-0002Id-6j
	for sipping@ietf.org; Tue, 27 Feb 2007 02:31:53 -0500
Received: from mailgw3.ericsson.se (unknown [127.0.0.1])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	86C9321113
	for <sipping@ietf.org>; Tue, 27 Feb 2007 08:31:41 +0100 (CET)
X-AuditID: c1b4fb3c-af7c7bb0000007de-c1-45e3de5d014b 
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw3.ericsson.se (Symantec Mail Security) with ESMTP id
	6DEE420B4C
	for <sipping@ietf.org>; Tue, 27 Feb 2007 08:31:41 +0100 (CET)
Received: from esealmw113.eemea.ericsson.se ([153.88.200.4]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 27 Feb 2007 08:31:41 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [Sipping] Service Identification: offerless INVITE case
Date: Tue, 27 Feb 2007 08:31:38 +0100
Message-ID: <7374777208BDC7449D5620EF942325670246592E@esealmw113.eemea.ericsson.se>
In-Reply-To: <989763567E51AB48BE015DF7BB960CAD02926415@esealmw111>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [Sipping] Service Identification: offerless INVITE case
Thread-Index: AcdZ2nHBW2t5ZR+eSJiRSdgDVlqGvQADqnggABNQHBAAAq1vIA==
From: "Christer Holmberg \(JO/LMF\)" <christer.holmberg@ericsson.com>
To: "Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>,
	"IETF Sipping List" <sipping@ietf.org>
X-OriginalArrivalTime: 27 Feb 2007 07:31:41.0383 (UTC)
	FILETIME=[52EA9D70:01C75A41]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 73f7847c44628482de9d5f1018acf469
Cc: 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


And, if you don't take Stephen's WORD WORD WORD for it, have a look at
the slides from the San Diego SPEERMINT session. There was one slide
with so many SBC boxes init that there was no room for anything else at
the slide... :)

Regards,

Christer=20


> -----Original Message-----
> From: Stephen Terrill (MI/EEM) [mailto:stephen.terrill@ericsson.com]=20
> Sent: 27. helmikuuta 2007 8:18
> To: IETF Sipping List
> Subject: RE: [Sipping] Service Identification: offerless INVITE case
>=20
> Hi,
>=20
> On this topic, doesn't this rely on the session-id making it=20
> to the Peer UA, in my understanding there are a number of=20
> entities along the way that may change the session id - any=20
> B2BUA such as an SBC .... (they may be BAD BAD BAD, but they=20
> are THERE THERE THERE).
>=20
> Steve.
>=20
> -----Original Message-----
> From: Christer Holmberg (JO/LMF)=20
> [mailto:christer.holmberg@ericsson.com]
>=20
> Sent: Monday, February 26, 2007 10:13 PM
> To: Jonathan Rosenberg
> Cc: Paul Kyzivat; IETF Sipping List
> Subject: RE: [Sipping] Service Identification: offerless INVITE case
>=20
>=20
> Hi,=20
>=20
> >I can't follow the figure since CRLFs got inserted by=20
> mailers along the
>=20
> >way. Can you send one whose width is small?
>=20
> I'll try again :)
>=20
> UAa    P1    P2    UAb   UAc
>  -INV-> -INV-> -INV->
>  <-200- <-200- <-200-
>  -INV->
>  <-3xx-
>  -INV--------------------->
>=20
> >I don't follow the problem though. The 3xx response would be seen by=20
> >all record-routing proxies upstream of the entity that sends the=20
> >mid-dialog 3xx.
>=20
> Yes, but what about all record-routing proxies downstream?
>=20
> >The new INVITE would be seen by all proxies.
>=20
> Assuming the new INVITE traverses the same set of proxies (if=20
> Service-Route is used, that set of proxies will be the same),=20
> yes, but I don't think we can assume that.
>=20
> >To make sure it goes to the right UA instance you use GRUU.=20
> >
> >BYE is sent by the entity receiving the replaces, as normal.=20
> What am I
> missing?
>=20
> What if the new UA is NOT the same as the original? What if=20
> the original UA doesn't support the new capability/media=20
> being added in the re-INVITE, and is routed to another UA?=20
> Shall such scenario not be allowed?
>=20
> Regards,
>=20
> Christer
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> >=20
> > -Jonathan R.
> >=20
> > Christer Holmberg (JO/LMF) wrote:
> >=20
> > > Hi,
> > >=20
> > >=20
> > >>>If terminated automatically, what happens to the nodes
> > situated AFTER
> > >>>the proxy that sent 3xx? They have no clue about the 3xx
> > being sent,
> > >>>and there is no guarantee the re-INVITE will traverse the
> > same nodes
> > >>>(or even reach the same remote UA).
> > >>
> > >>This is a general issue for INVITE/Replaces. The 3xx is
> > simply a way
> > >>to get the UAC to do it. The UAC could decide to do it even
> > without a
> > >>3xx.
> > >>In the normal course of routing this any servers that are=20
> needed on=20
> > >>the path will have to be discovered based on the content of the=20
> > >>request.
> > >=20
> > >=20
> > > My issue was how the proxies in the original dialog will
> > know that a
> > > 3xx has been sent (see picture):
> > >=20
> > > UA A ------------------- Proxy 1 ---------------- Proxy 2
> > > ------------------ UA B     UA C
> > >=20
> > > -------------INVITE---------> --------INVITE--------->
> > > --------INVITE---------->
> > > <------------200------------- <-------200-------------
> > > <-------200--------------
> > > -------------re-INVITE------>
> > > <------------3xx-------------
> > > -------------INVITE
> > >=20
> (new)-------------------------------------------------------------->
> > >=20
> > > Now, if Proxy 2 will not be traversed by the new INVITE, it
> > will not
> > > know that the dialog has been replaced.
> > >=20
> > > And, if the new INVITE reaches UA C, UA B will have no clue
> > that the
> > > dialog has been replaced.
> > >=20
> > > Regards,
> > >=20
> > > Christer
> > >=20
> > >=20
> > >=20
> > >=20
> > >=20
> > >=20
> > >=20
> > > =20
> > >=20
> > >>>Etc etc etc.
> > >>>
> > >>>Regards,
> > >>>
> > >>>Christer
> > >>>
> > >>>
> > >>>
> > >>>
> > >>>
> > >>>
> > >>>
> > >>>
> > >>>
> > >>>
> > >>>
> > >>>
> > >>>
> > >>>>	Paul
> > >>>>
> > >>>>Jonathan Rosenberg wrote:
> > >>>>
> > >>>>>Andrew made a great point in one of his mails which merits
> > >>>>
> > >>>>promotion
> > >>>>
> > >>>>>into its own thread.
> > >>>>>
> > >>>>>I and others have asserted that you should be able to
> > >>
> > >>identify the
> > >>
> > >>>>>service by things naturally in the request, and in
> > >>>>
> > >>>>particular, SDP has
> > >>>>
> > >>>>>come up as a key thing. For example, in a multiplayer
> > game, there
> > >>>>>might be an m-line for the particular proprietary
> > >>
> > >>protocol used for
> > >>
> > >>>>>game moves. Or, if its MSRP, you might have specific
> > >>>>
> > >>>>content types for
> > >>>>
> > >>>>>game moves. When present, those exactly indicate what the
> > >>>>
> > >>>>service is.
> > >>>>
> > >>>>>But what do we do about offerless INVITE? In particular,
> > >>
> > >>there are
> > >>
> > >>>>>cases where we need to invoke an application server to=20
> handle a=20
> > >>>>>request, and what if we don't have the SDP in the INVITE?
> > >>>>
> > >>>>Its too late
> > >>>>
> > >>>>>to insert an application server once the INVITE has already
> > >>>>
> > >>>>passed to
> > >>>>
> > >>>>>the called party and the 200 OK is sent.
> > >>>>>
> > >>>>>This is a great point. I'd rather just fix this little
> > >>
> > >>problem, the
> > >>
> > >>>>>define an entire notion of services and service identifiers
> > >>>>
> > >>>>that are
> > >>>>
> > >>>>>not necessary, just because we have offerless INVITEs.
> > >>>>>
> > >>>>>There are lots of ways to handle this problem. My preferred
> > >>>>
> > >>>>approach
> > >>>>
> > >>>>>would be that offerless INVITE contain a capability
> > >>>>
> > >>>>description, using
> > >>>>
> > >>>>>Flemmings new capability framework. The INVITE would
> > >>
> > >>contain a body
> > >>
> > >>>>>that  looks like SDP, but has a different content type and=20
> > >>>>>content-disposition. It lists the media streams, codecs
> > and other
> > >>>>>capabilities that exist. Based on this, the called party
> > >>
> > >>makes the
> > >>
> > >>>>>offer in a 2xx or 18x like normal, and the answer in the
> > >>>>
> > >>>>ACK/PRACK as normal.
> > >>>>
> > >>>>>It is completely backwards compatible.
> > >>>>>
> > >>>>>Now, a related issue is what happens if a session starts
> > >>
> > >>with audio
> > >>
> > >>>>>only, and then the user wants to turn this into a game? Its
> > >>>>
> > >>>>too late
> > >>>>
> > >>>>>to add new application servers into the call path, which
> > >>>>
> > >>>>might need to
> > >>>>
> > >>>>>be there for games. What can we do?
> > >>>>>
> > >>>>>Here too, there is a simple answer. If a network provider
> > >>>>
> > >>>>requires app
> > >>>>
> > >>>>>servers to get inserted, or a phone requires dispatch to a
> > >>>>
> > >>>>different
> > >>>>
> > >>>>>application, it would receive a re-INVITE to add the
> > >>
> > >>gaming stream,
> > >>
> > >>>>>and then 3xx it. The Contact would contain an embedded
> > >>>>
> > >>>>Replaces header
> > >>>>
> > >>>>>with the dialog IDs for the call in progress. THe
> > >>>>
> > >>>>originating UA would
> > >>>>
> > >>>>>then do a totally new INVITE that replaces the call in
> > >>>>
> > >>>>progress. This
> > >>>>
> > >>>>>would allow the network and/or phone to redispatch, yet
> > provide a
> > >>>>>seamless user experience. Billing can be done however the
> > >>
> > >>operators
> > >>
> > >>>>>want, since the signaling is very clear on exactly what
> > >>>>
> > >>>>happened and when.
> > >>>>
> > >>>>>Furthermore, the replaces mechansim would only get used if
> > >>>>
> > >>>>there was
> > >>>>
> > >>>>>someone in the call path that needed to see it.
> > >>>>>
> > >>>>>I'd rather specify things like this, than invent an
> > >>
> > >>entire services
> > >>
> > >>>>>framework that would lead to balkanization and=20
> non-interoperable=20
> > >>>>>islands of SIP.
> > >>>>>
> > >>>>>-Jonathan R.
> > >>>>>
> > >>>>>
> > >>>>
> > >>>>_______________________________________________
> > >>>>Sipping mailing list
> > >>
> > >>https://www1.ietf.org/mailman/listinfo/sipping
> > >>
> > >>>>This list is for NEW development of the application of SIP Use=20
> > >>>>sip-implementors@cs.columbia.edu for questions on=20
> current sip Use=20
> > >>>>sip@ietf.org for new developments of core SIP
> > >>>>
> > >>>
> > >=20
> >=20
> > --=20
> > Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
> > Cisco Fellow                                   Parsippany, NJ=20
> > 07054-2711
> > Cisco Systems
> > jdrosen@cisco.com                              FAX:   (973) 952-5050
> > http://www.jdrosen.net                         PHONE: (973) 952-5000
> > http://www.cisco.com
> >=20
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP=20
> Use sip-implementors@cs.columbia.edu for questions on current=20
> sip Use sip@ietf.org for new developments of core SIP
>=20

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 27 02:35:13 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HLwra-0002dI-7b; Tue, 27 Feb 2007 02:35:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HLwrZ-0002d3-5Y
	for sipping@ietf.org; Tue, 27 Feb 2007 02:35:05 -0500
Received: from lohi.tutpro.com ([192.98.100.2] helo=tutpro.com)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HLwrJ-0002op-EP
	for sipping@ietf.org; Tue, 27 Feb 2007 02:35:05 -0500
Received: from localhost (localhost [127.0.0.1])
	by tutpro.com (Postfix) with ESMTP id 87AD81EC659;
	Tue, 27 Feb 2007 09:34:48 +0200 (EET)
Received: from tutpro.com ([127.0.0.1])
	by localhost (tutpro.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id RrKdSigHUaym; Tue, 27 Feb 2007 09:34:42 +0200 (EET)
Received: from rautu (unknown [124.6.202.252])
	by tutpro.com (Postfix) with ESMTP;
	Tue, 27 Feb 2007 09:34:42 +0200 (EET)
Received: by rautu (Postfix, from userid 1000)
	id 47347F004D; Tue, 27 Feb 2007 09:34:10 +0200 (EET)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <17891.57074.222652.313631@rautu.test.fi>
Date: Tue, 27 Feb 2007 09:34:10 +0200
From: Juha Heinanen <jh@tutpro.com>
To: Jonathan Rosenberg <jdrosen@cisco.com>
Subject: Re: [Sipping] Service Identification in SIP
In-Reply-To: <45E2F2BD.9080508@cisco.com>
References: <61968779B8AC4C4BAB421D4C12F008C009152056@XCH47YKF.rim.net>
	<072C5B76F7CEAB488172C6F64B30B5E302A2F204@xmb-rtp-20b.amer.cisco.com>
	<45E2F2BD.9080508@cisco.com>
X-Mailer: VM 7.19 under Emacs 21.4.1
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 08170828343bcf1325e4a0fb4584481c
Cc: Andrew Allen <aallen@rim.com>,
	"Stephen Terrill \(MI/EEM\)" <stephen.terrill@ericsson.com>,
	sipping@ietf.org, "Michael Hammer \(mhammer\)" <mhammer@cisco.com>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Jonathan Rosenberg writes:

 > It is nothing short of complete balkanization and 
 > segmentation of communications systems into non-interoperable islands 
 > that will substantially and permanantly raise the cost and complexity of 
 > IP communications.

this is no new news.  everyone should have known from very start of IMS
project that is nothing more that a walled garden effort where stupid
customers are made to pay more for no added value.  it should be a high
time to stop dealing with IMS or 3GPP problems on this mailing list and
put a spam filter on IMS related messages.

-- juha


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 27 09:04:59 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HM2vS-0001Z5-Po; Tue, 27 Feb 2007 09:03:30 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HM2vR-0001YY-N1
	for sipping@ietf.org; Tue, 27 Feb 2007 09:03:29 -0500
Received: from rtp-iport-2.cisco.com ([64.102.122.149])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HM2vO-0002e1-QL
	for sipping@ietf.org; Tue, 27 Feb 2007 09:03:29 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by rtp-iport-2.cisco.com with ESMTP; 27 Feb 2007 09:03:24 -0500
X-IronPort-AV: i="4.14,225,1170651600"; 
	d="scan'208"; a="114613783:sNHT61530888"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1RE3OG8016971; 
	Tue, 27 Feb 2007 09:03:24 -0500
Received: from xbh-rtp-211.amer.cisco.com (xbh-rtp-211.cisco.com
	[64.102.31.102])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1RE3BVr017158; 
	Tue, 27 Feb 2007 09:03:24 -0500 (EST)
Received: from xfe-rtp-201.amer.cisco.com ([64.102.31.38]) by
	xbh-rtp-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 27 Feb 2007 09:03:22 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 27 Feb 2007 09:03:22 -0500
Message-ID: <45E43A29.4030207@cisco.com>
Date: Tue, 27 Feb 2007 09:03:21 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Christer Holmberg (JO/LMF)" <christer.holmberg@ericsson.com>
Subject: Re: [Sipping] Service Identification: offerless INVITE case
References: <7374777208BDC7449D5620EF94232567024656BA@esealmw113.eemea.ericsson.se>
In-Reply-To: <7374777208BDC7449D5620EF94232567024656BA@esealmw113.eemea.ericsson.se>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 27 Feb 2007 14:03:22.0492 (UTC)
	FILETIME=[0AAB43C0:01C75A78]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=9732; t=1172585004;
	x=1173449004; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=3A=20offerless=2
	0INVITE=20case |Sender:=20
	|To:=20=22Christer=20Holmberg=20(JO/LMF)=22=20<christer.holmberg@ericsson
	.com>; bh=bpXgRAOEzv1ON7ktaXBLAbv2ea8G1jBpCjzKxb4qNkU=;
	b=abXU1Bvzl3QPG6Cki2FITfVuO7IDCCmbNC5xpugG8OLyKLKA7rneR25kMej1aDclFTmzUkiK
	j0hVAFkN9j/FX0e/Le8uR9RJ88B/AZXr0HFMyIkBd/3OBC3aKFfE+z1+;
Authentication-Results: rtp-dkim-2; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 27ec2ff0f5c3b18b49c722f4f1748838
Cc: IETF Sipping List <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Christer Holmberg (JO/LMF) wrote:
> Hi,
> 
> Assuming the new UA (UA_C) is the same as the original (UA_B), I agree
> there should be no problem as far as the BYE is concerned.

If the new UA is *different* from the original UA then INVITE/Replaces 
will fail.

> It is important to note, however, that this solution requires the
> support of both GRUU and Replaces by all UAs involved.

Yup.

Frankly, there are many features that require these things in order to 
work properly. Why do you suppose they have been worked on for so long?

> One question is whether a proxy can send the mid-dialog 3xx, or if it
> some B2BUA behavior is required. 

Yes, there may be some issue with that. It is of course legal for a 
proxy to send a 3xx for an out of dialog request. And there is no 
technical issue in doing so for an in dialog request. Its only a matter 
of whether it is legal.

I suppose one could make an argument that a 3xx should only be returned 
by something responsible for the domain of the R-URI. But I don't think 
there is currently any such restriction.

> For example, there is always a chance that the remote UA (UA_B) is
> sending a new offer at the same time, which may reach UA_A before the
> 3xx. The proxy that sent 3xx most likely doesn't have to care about this
> (UA_A should handle it, and return 491), but it's just something we
> would need to be aware of since the offer/3xx will not be end-to-end in
> this case.

Again, I think this has nothing to do with the 3xx. It only has to do 
with the UAC and the INVITEs it sends. The 3xx just tells the UAC to do 
an INVITE/Replaces.

There is probably some behavior that has never been documented about 
what a UAC should do after it has decided (because of 3xx or whatever) 
that it wants to send an INVITE/Replaces to establish a new session 
replacing one it already has. After it has started down that path it may 
want to reject reinvites on the old session. But I think this is a 
second order issue.

> And, of course the scenario where the new INVITE is rejected (or not
> even sent) - for whatever reason - be supported, and the old dialog

Well, or course. There is nothing that requires the UAC to send a new 
invite upon receiving a 3xx to a reinvite. Just like any failed 
reinvite, it leaves the session as is. There is nothing new or 
surprising here.

> Another problem may be that UA_A is required (until it receives BYE on
> the original dialog) to maintain resources (including codecs) for TWO
> dialogs - the original dialog, and the new - which may cause problems in
> some cases. From a codec perspective that could be solved by putting the
> original dialog on hold (by sending a hold re-INVITE, without the new
> capabilities/medias) before sending the new invite.

This is just like using INVITE/Replaces for transfer. If there is need 
to put the session on hold first then of course that may be done.

	Paul

> Regards,
> 
> Christer
>  
> 
>> -----Original Message-----
>> From: Paul Kyzivat [mailto:pkyzivat@cisco.com] 
>> Sent: 27. helmikuuta 2007 1:21
>> To: Christer Holmberg (JO/LMF)
>> Cc: Jonathan Rosenberg; IETF Sipping List
>> Subject: Re: [Sipping] Service Identification: offerless INVITE case
>>
>>
>>
>> Christer Holmberg (JO/LMF) wrote:
>>> Hi,
>>>
>>>>> If terminated automatically, what happens to the nodes situated 
>>>>> AFTER the proxy that sent 3xx? They have no clue about 
>> the 3xx being 
>>>>> sent, and there is no guarantee the re-INVITE will 
>> traverse the same 
>>>>> nodes (or even reach the same remote UA).
>>>> This is a general issue for INVITE/Replaces. The 3xx is 
>> simply a way 
>>>> to get the UAC to do it. The UAC could decide to do it 
>> even without a 
>>>> 3xx.
>>>> In the normal course of routing this any servers that are 
>> needed on 
>>>> the path will have to be discovered based on the content of the 
>>>> request.
>>> My issue was how the proxies in the original dialog will 
>> know that a 
>>> 3xx has been sent (see picture):
>> There should be no reason for any of the proxies to take note 
>> of the 3xx. It is simply a way to tell the UAC to do an 
>> INVITE/Replaces. It could decide to do an INVITE/Replaces 
>> without receiving a 3xx, so anything that depends on seeing 
>> the 3xx is broken.
>>
>> The new invite will need be routed to application servers in 
>> whatever way is appropriate for it, independent of what 
>> application servers had been involved in the old session. If 
>> the new invite simply adds on additional features to what was 
>> used before, then presumably any app servers needed before 
>> will be included again, and any new servers needed because of 
>> additions to the invite will be included as well.
>>
>> If some server that had been included is no longer included 
>> on the new invite, then it will drop out when the BYE is sent 
>> on the old session.
>>
>> 	Paul
>>
>>> UA A ------------------- Proxy 1 ---------------- Proxy 2
>>> ------------------ UA B     UA C
>>>
>>> -------------INVITE---------> --------INVITE--------->
>>> --------INVITE---------->
>>> <------------200------------- <-------200-------------
>>> <-------200--------------
>>> -------------re-INVITE------>
>>> <------------3xx-------------
>>> -------------INVITE
>>> (new)-------------------------------------------------------------->
>>>
>>> Now, if Proxy 2 will not be traversed by the new INVITE, it 
>> will not 
>>> know that the dialog has been replaced.
>>>
>>> And, if the new INVITE reaches UA C, UA B will have no clue 
>> that the 
>>> dialog has been replaced.
>>>
>>> Regards,
>>>
>>> Christer
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>  
>>>>> Etc etc etc.
>>>>>
>>>>> Regards,
>>>>>
>>>>> Christer
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>> 	Paul
>>>>>>
>>>>>> Jonathan Rosenberg wrote:
>>>>>>> Andrew made a great point in one of his mails which merits
>>>>>> promotion
>>>>>>> into its own thread.
>>>>>>>
>>>>>>> I and others have asserted that you should be able to
>>>> identify the
>>>>>>> service by things naturally in the request, and in
>>>>>> particular, SDP has
>>>>>>> come up as a key thing. For example, in a multiplayer 
>> game, there 
>>>>>>> might be an m-line for the particular proprietary
>>>> protocol used for
>>>>>>> game moves. Or, if its MSRP, you might have specific
>>>>>> content types for
>>>>>>> game moves. When present, those exactly indicate what the
>>>>>> service is.
>>>>>>> But what do we do about offerless INVITE? In particular,
>>>> there are
>>>>>>> cases where we need to invoke an application server to handle a 
>>>>>>> request, and what if we don't have the SDP in the INVITE?
>>>>>> Its too late
>>>>>>> to insert an application server once the INVITE has already
>>>>>> passed to
>>>>>>> the called party and the 200 OK is sent.
>>>>>>>
>>>>>>> This is a great point. I'd rather just fix this little
>>>> problem, the
>>>>>>> define an entire notion of services and service identifiers
>>>>>> that are
>>>>>>> not necessary, just because we have offerless INVITEs.
>>>>>>>
>>>>>>> There are lots of ways to handle this problem. My preferred
>>>>>> approach
>>>>>>> would be that offerless INVITE contain a capability
>>>>>> description, using
>>>>>>> Flemmings new capability framework. The INVITE would
>>>> contain a body
>>>>>>> that  looks like SDP, but has a different content type and 
>>>>>>> content-disposition. It lists the media streams, codecs 
>> and other 
>>>>>>> capabilities that exist. Based on this, the called party
>>>> makes the
>>>>>>> offer in a 2xx or 18x like normal, and the answer in the
>>>>>> ACK/PRACK as normal.
>>>>>>> It is completely backwards compatible.
>>>>>>>
>>>>>>> Now, a related issue is what happens if a session starts
>>>> with audio
>>>>>>> only, and then the user wants to turn this into a game? Its
>>>>>> too late
>>>>>>> to add new application servers into the call path, which
>>>>>> might need to
>>>>>>> be there for games. What can we do?
>>>>>>>
>>>>>>> Here too, there is a simple answer. If a network provider
>>>>>> requires app
>>>>>>> servers to get inserted, or a phone requires dispatch to a
>>>>>> different
>>>>>>> application, it would receive a re-INVITE to add the
>>>> gaming stream,
>>>>>>> and then 3xx it. The Contact would contain an embedded
>>>>>> Replaces header
>>>>>>> with the dialog IDs for the call in progress. THe
>>>>>> originating UA would
>>>>>>> then do a totally new INVITE that replaces the call in
>>>>>> progress. This
>>>>>>> would allow the network and/or phone to redispatch, yet 
>> provide a 
>>>>>>> seamless user experience. Billing can be done however the
>>>> operators
>>>>>>> want, since the signaling is very clear on exactly what
>>>>>> happened and when.
>>>>>>> Furthermore, the replaces mechansim would only get used if
>>>>>> there was
>>>>>>> someone in the call path that needed to see it.
>>>>>>>
>>>>>>> I'd rather specify things like this, than invent an
>>>> entire services
>>>>>>> framework that would lead to balkanization and 
>> non-interoperable 
>>>>>>> islands of SIP.
>>>>>>>
>>>>>>> -Jonathan R.
>>>>>>>
>>>>>>>
>>>>>> _______________________________________________
>>>>>> Sipping mailing list
>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>>> This list is for NEW development of the application of SIP Use 
>>>>>> sip-implementors@cs.columbia.edu for questions on 
>> current sip Use 
>>>>>> sip@ietf.org for new developments of core SIP
>>>>>>
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 27 09:10:02 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HM31a-0001pG-9r; Tue, 27 Feb 2007 09:09:50 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HM31Y-0001ei-Ji
	for sipping@ietf.org; Tue, 27 Feb 2007 09:09:48 -0500
Received: from sj-iport-6.cisco.com ([171.71.176.117])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HM31W-0003Na-Sw
	for sipping@ietf.org; Tue, 27 Feb 2007 09:09:48 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by sj-iport-6.cisco.com with ESMTP; 27 Feb 2007 06:09:46 -0800
X-IronPort-AV: i="4.14,225,1170662400"; 
	d="scan'208"; a="116476544:sNHT55539684"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1RE9j6H026161; 
	Tue, 27 Feb 2007 09:09:45 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1RE9jOE026550; 
	Tue, 27 Feb 2007 09:09:45 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 27 Feb 2007 09:09:44 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 27 Feb 2007 09:09:44 -0500
Message-ID: <45E43BA6.6050109@cisco.com>
Date: Tue, 27 Feb 2007 09:09:42 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: "Stephen Terrill (MI/EEM)" <stephen.terrill@ericsson.com>
Subject: Re: [Sipping] Service Identification: offerless INVITE case
References: <989763567E51AB48BE015DF7BB960CAD02926415@esealmw111>
In-Reply-To: <989763567E51AB48BE015DF7BB960CAD02926415@esealmw111>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 27 Feb 2007 14:09:44.0455 (UTC)
	FILETIME=[EE563170:01C75A78]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=9215; t=1172585385;
	x=1173449385; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=3A=20offerless=2
	0INVITE=20case |Sender:=20
	|To:=20=22Stephen=20Terrill=20(MI/EEM)=22=20<stephen.terrill@ericsson.com
	>; bh=5rBFM7qAq3e0R/l5oXSNX5XibgBEx5JuWnf7noIP11c=;
	b=AlwkW1uEjZHb9G+94JE1X6ghb49B7vZdafarJWqkMmrTbfkoXmJVkV1P/QW49OXOQh/mwoUL
	RAMCDd9mOVV/74dxGh1Vcdx25ez9jf0RYGruMuWEG7jtBhPhrDlFCi21;
Authentication-Results: rtp-dkim-1; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 7e267523e0685e5aa2dbbdde4b659686
Cc: IETF Sipping List <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Stephen Terrill (MI/EEM) wrote:
> Hi,
> 
> On this topic, doesn't this rely on the session-id making it to the Peer
> UA, in my understanding there are a number of entities along the way
> that may change the session id - any B2BUA such as an SBC .... (they may
> be BAD BAD BAD, but they are THERE THERE THERE).

I presume you mean dialog-id.

One would hope that a B2BUA isn't trying to totally break SIP. 
INVITE/Replaces if fundamental to many features. If some middlebox is 
going to change the dialog-id for the dialog, then it has the 
responsibility to patch of the damage that causes. If it can't do that 
then it has no business being there.

This is analogous to a NAT that screws up address and port, and thus 
breaks things that reference those. We don't want to go down that path 
again.

Nor, IMO, do we want to reduce ourselves to features that can work in 
the presence of such damage.

	Paul

> Steve.
> 
> -----Original Message-----
> From: Christer Holmberg (JO/LMF) [mailto:christer.holmberg@ericsson.com]
> 
> Sent: Monday, February 26, 2007 10:13 PM
> To: Jonathan Rosenberg
> Cc: Paul Kyzivat; IETF Sipping List
> Subject: RE: [Sipping] Service Identification: offerless INVITE case
> 
> 
> Hi, 
> 
>> I can't follow the figure since CRLFs got inserted by mailers along the
> 
>> way. Can you send one whose width is small?
> 
> I'll try again :)
> 
> UAa    P1    P2    UAb   UAc
>  -INV-> -INV-> -INV->
>  <-200- <-200- <-200-
>  -INV->
>  <-3xx-
>  -INV--------------------->
> 
>> I don't follow the problem though. The 3xx response would be seen by 
>> all record-routing proxies upstream of the entity that sends the 
>> mid-dialog 3xx.
> 
> Yes, but what about all record-routing proxies downstream?
> 
>> The new INVITE would be seen by all proxies.
> 
> Assuming the new INVITE traverses the same set of proxies (if
> Service-Route is used, that set of proxies will be the same), yes, but I
> don't think we can assume that.
> 
>> To make sure it goes to the right UA instance you use GRUU. 
>>
>> BYE is sent by the entity receiving the replaces, as normal. What am I
> missing?
> 
> What if the new UA is NOT the same as the original? What if the original
> UA doesn't support the new capability/media being added in the
> re-INVITE, and is routed to another UA? Shall such scenario not be
> allowed?
> 
> Regards,
> 
> Christer
> 
> 
> 
> 
> 
> 
> 
>> -Jonathan R.
>>
>> Christer Holmberg (JO/LMF) wrote:
>>
>>> Hi,
>>>
>>>
>>>>> If terminated automatically, what happens to the nodes
>> situated AFTER
>>>>> the proxy that sent 3xx? They have no clue about the 3xx
>> being sent,
>>>>> and there is no guarantee the re-INVITE will traverse the
>> same nodes
>>>>> (or even reach the same remote UA).
>>>> This is a general issue for INVITE/Replaces. The 3xx is
>> simply a way
>>>> to get the UAC to do it. The UAC could decide to do it even
>> without a
>>>> 3xx.
>>>> In the normal course of routing this any servers that are needed on 
>>>> the path will have to be discovered based on the content of the 
>>>> request.
>>>
>>> My issue was how the proxies in the original dialog will
>> know that a
>>> 3xx has been sent (see picture):
>>>
>>> UA A ------------------- Proxy 1 ---------------- Proxy 2
>>> ------------------ UA B     UA C
>>>
>>> -------------INVITE---------> --------INVITE--------->
>>> --------INVITE---------->
>>> <------------200------------- <-------200-------------
>>> <-------200--------------
>>> -------------re-INVITE------>
>>> <------------3xx-------------
>>> -------------INVITE
>>> (new)-------------------------------------------------------------->
>>>
>>> Now, if Proxy 2 will not be traversed by the new INVITE, it
>> will not
>>> know that the dialog has been replaced.
>>>
>>> And, if the new INVITE reaches UA C, UA B will have no clue
>> that the
>>> dialog has been replaced.
>>>
>>> Regards,
>>>
>>> Christer
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>  
>>>
>>>>> Etc etc etc.
>>>>>
>>>>> Regards,
>>>>>
>>>>> Christer
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>> 	Paul
>>>>>>
>>>>>> Jonathan Rosenberg wrote:
>>>>>>
>>>>>>> Andrew made a great point in one of his mails which merits
>>>>>> promotion
>>>>>>
>>>>>>> into its own thread.
>>>>>>>
>>>>>>> I and others have asserted that you should be able to
>>>> identify the
>>>>
>>>>>>> service by things naturally in the request, and in
>>>>>> particular, SDP has
>>>>>>
>>>>>>> come up as a key thing. For example, in a multiplayer
>> game, there
>>>>>>> might be an m-line for the particular proprietary
>>>> protocol used for
>>>>
>>>>>>> game moves. Or, if its MSRP, you might have specific
>>>>>> content types for
>>>>>>
>>>>>>> game moves. When present, those exactly indicate what the
>>>>>> service is.
>>>>>>
>>>>>>> But what do we do about offerless INVITE? In particular,
>>>> there are
>>>>
>>>>>>> cases where we need to invoke an application server to handle a 
>>>>>>> request, and what if we don't have the SDP in the INVITE?
>>>>>> Its too late
>>>>>>
>>>>>>> to insert an application server once the INVITE has already
>>>>>> passed to
>>>>>>
>>>>>>> the called party and the 200 OK is sent.
>>>>>>>
>>>>>>> This is a great point. I'd rather just fix this little
>>>> problem, the
>>>>
>>>>>>> define an entire notion of services and service identifiers
>>>>>> that are
>>>>>>
>>>>>>> not necessary, just because we have offerless INVITEs.
>>>>>>>
>>>>>>> There are lots of ways to handle this problem. My preferred
>>>>>> approach
>>>>>>
>>>>>>> would be that offerless INVITE contain a capability
>>>>>> description, using
>>>>>>
>>>>>>> Flemmings new capability framework. The INVITE would
>>>> contain a body
>>>>
>>>>>>> that  looks like SDP, but has a different content type and 
>>>>>>> content-disposition. It lists the media streams, codecs
>> and other
>>>>>>> capabilities that exist. Based on this, the called party
>>>> makes the
>>>>
>>>>>>> offer in a 2xx or 18x like normal, and the answer in the
>>>>>> ACK/PRACK as normal.
>>>>>>
>>>>>>> It is completely backwards compatible.
>>>>>>>
>>>>>>> Now, a related issue is what happens if a session starts
>>>> with audio
>>>>
>>>>>>> only, and then the user wants to turn this into a game? Its
>>>>>> too late
>>>>>>
>>>>>>> to add new application servers into the call path, which
>>>>>> might need to
>>>>>>
>>>>>>> be there for games. What can we do?
>>>>>>>
>>>>>>> Here too, there is a simple answer. If a network provider
>>>>>> requires app
>>>>>>
>>>>>>> servers to get inserted, or a phone requires dispatch to a
>>>>>> different
>>>>>>
>>>>>>> application, it would receive a re-INVITE to add the
>>>> gaming stream,
>>>>
>>>>>>> and then 3xx it. The Contact would contain an embedded
>>>>>> Replaces header
>>>>>>
>>>>>>> with the dialog IDs for the call in progress. THe
>>>>>> originating UA would
>>>>>>
>>>>>>> then do a totally new INVITE that replaces the call in
>>>>>> progress. This
>>>>>>
>>>>>>> would allow the network and/or phone to redispatch, yet
>> provide a
>>>>>>> seamless user experience. Billing can be done however the
>>>> operators
>>>>
>>>>>>> want, since the signaling is very clear on exactly what
>>>>>> happened and when.
>>>>>>
>>>>>>> Furthermore, the replaces mechansim would only get used if
>>>>>> there was
>>>>>>
>>>>>>> someone in the call path that needed to see it.
>>>>>>>
>>>>>>> I'd rather specify things like this, than invent an
>>>> entire services
>>>>
>>>>>>> framework that would lead to balkanization and non-interoperable 
>>>>>>> islands of SIP.
>>>>>>>
>>>>>>> -Jonathan R.
>>>>>>>
>>>>>>>
>>>>>> _______________________________________________
>>>>>> Sipping mailing list
>>>> https://www1.ietf.org/mailman/listinfo/sipping
>>>>
>>>>>> This list is for NEW development of the application of SIP Use 
>>>>>> sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>>>> sip@ietf.org for new developments of core SIP
>>>>>>
>> -- 
>> Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
>> Cisco Fellow                                   Parsippany, NJ 
>> 07054-2711
>> Cisco Systems
>> jdrosen@cisco.com                              FAX:   (973) 952-5050
>> http://www.jdrosen.net                         PHONE: (973) 952-5000
>> http://www.cisco.com
>>
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use
> sip-implementors@cs.columbia.edu for questions on current sip Use
> sip@ietf.org for new developments of core SIP
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 27 10:05:50 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HM3se-0004tx-KM; Tue, 27 Feb 2007 10:04:40 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HM3sd-0004ny-8G
	for sipping@ietf.org; Tue, 27 Feb 2007 10:04:39 -0500
Received: from rtp-iport-1.cisco.com ([64.102.122.148])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HM3sa-0000if-Md
	for sipping@ietf.org; Tue, 27 Feb 2007 10:04:39 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by rtp-iport-1.cisco.com with ESMTP; 27 Feb 2007 10:04:35 -0500
X-IronPort-AV: i="4.14,226,1170651600"; 
	d="scan'208"; a="53737629:sNHT62228904"
Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1RF4YSl013987; 
	Tue, 27 Feb 2007 10:04:34 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id l1RF3rVZ004130; 
	Tue, 27 Feb 2007 10:04:34 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 27 Feb 2007 10:04:18 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-202.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 27 Feb 2007 10:04:17 -0500
Message-ID: <45E44871.10002@cisco.com>
Date: Tue, 27 Feb 2007 10:04:17 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: "Stephen Terrill (MI/EEM)" <stephen.terrill@ericsson.com>
Subject: Re: [Sipping] Service Identification: offerless INVITE case
References: <989763567E51AB48BE015DF7BB960CAD02926415@esealmw111>
In-Reply-To: <989763567E51AB48BE015DF7BB960CAD02926415@esealmw111>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 27 Feb 2007 15:04:17.0940 (UTC)
	FILETIME=[8D7C5D40:01C75A80]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=10118; t=1172588674;
	x=1173452674; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=3A=20offerless=2
	0INVITE=20case |Sender:=20
	|To:=20=22Stephen=20Terrill=20(MI/EEM)=22=20<stephen.terrill@ericsson.com
	>; bh=Fp5Zkp27+OVovDTvynXLy6EKj2KwCM/MzPUx82SM9ZI=;
	b=S5ycQi/KBe6bdXnv4wels5WcPfNiL8f0RmfcSlHLgCS8vUxsVEZZlQOxUz085qlpCS5oorW8
	udsvaUq3KwayhoLgmhAroowDhPqQ7u6PcfStd+Ooxn0whr+eX8LVJvAA;
Authentication-Results: rtp-dkim-2; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 4ec3642ae9025e273a4a263d640f3300
Cc: IETF Sipping List <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

If they are present, they would need to properly support handling of 
Replaces, and patch up the dialog IDs that are in them. Whether they do 
that for Replaces embedded in 3xx Contacts, I don't know (probably not), 
but this is just a matter of making sure they do if you want to support 
those flows.

As Paul said, we should not stop using or specifying new SIP extensions 
because someone might have an SBC deployed that won't work with it. Its 
a consideration, of course, but this issue with SBCs is exactly why so 
many in the community have complained about them, and the burden falls 
on the deployers and implementors to get them upgraded to support the 
functions they need.

-Jonathan R.

Stephen Terrill (MI/EEM) wrote:

> Hi,
> 
> On this topic, doesn't this rely on the session-id making it to the Peer
> UA, in my understanding there are a number of entities along the way
> that may change the session id - any B2BUA such as an SBC .... (they may
> be BAD BAD BAD, but they are THERE THERE THERE).
> 
> Steve.
> 
> -----Original Message-----
> From: Christer Holmberg (JO/LMF) [mailto:christer.holmberg@ericsson.com]
> 
> Sent: Monday, February 26, 2007 10:13 PM
> To: Jonathan Rosenberg
> Cc: Paul Kyzivat; IETF Sipping List
> Subject: RE: [Sipping] Service Identification: offerless INVITE case
> 
> 
> Hi, 
> 
> 
>>I can't follow the figure since CRLFs got inserted by mailers along the
> 
> 
>>way. Can you send one whose width is small?
> 
> 
> I'll try again :)
> 
> UAa    P1    P2    UAb   UAc
>  -INV-> -INV-> -INV->
>  <-200- <-200- <-200-
>  -INV->
>  <-3xx-
>  -INV--------------------->
> 
>>I don't follow the problem though. The 3xx response would be seen by 
>>all record-routing proxies upstream of the entity that sends the 
>>mid-dialog 3xx.
> 
> 
> Yes, but what about all record-routing proxies downstream?
> 
> 
>>The new INVITE would be seen by all proxies.
> 
> 
> Assuming the new INVITE traverses the same set of proxies (if
> Service-Route is used, that set of proxies will be the same), yes, but I
> don't think we can assume that.
> 
> 
>>To make sure it goes to the right UA instance you use GRUU. 
>>
>>BYE is sent by the entity receiving the replaces, as normal. What am I
> 
> missing?
> 
> What if the new UA is NOT the same as the original? What if the original
> UA doesn't support the new capability/media being added in the
> re-INVITE, and is routed to another UA? Shall such scenario not be
> allowed?
> 
> Regards,
> 
> Christer
> 
> 
> 
> 
> 
> 
> 
> 
>>-Jonathan R.
>>
>>Christer Holmberg (JO/LMF) wrote:
>>
>>
>>>Hi,
>>>
>>>
>>>
>>>>>If terminated automatically, what happens to the nodes
>>
>>situated AFTER
>>
>>>>>the proxy that sent 3xx? They have no clue about the 3xx
>>
>>being sent,
>>
>>>>>and there is no guarantee the re-INVITE will traverse the
>>
>>same nodes
>>
>>>>>(or even reach the same remote UA).
>>>>
>>>>This is a general issue for INVITE/Replaces. The 3xx is
>>
>>simply a way
>>
>>>>to get the UAC to do it. The UAC could decide to do it even
>>
>>without a
>>
>>>>3xx.
>>>>In the normal course of routing this any servers that are needed on 
>>>>the path will have to be discovered based on the content of the 
>>>>request.
>>>
>>>
>>>My issue was how the proxies in the original dialog will
>>
>>know that a
>>
>>>3xx has been sent (see picture):
>>>
>>>UA A ------------------- Proxy 1 ---------------- Proxy 2
>>>------------------ UA B     UA C
>>>
>>>-------------INVITE---------> --------INVITE--------->
>>>--------INVITE---------->
>>><------------200------------- <-------200-------------
>>><-------200--------------
>>>-------------re-INVITE------>
>>><------------3xx-------------
>>>-------------INVITE
>>>(new)-------------------------------------------------------------->
>>>
>>>Now, if Proxy 2 will not be traversed by the new INVITE, it
>>
>>will not
>>
>>>know that the dialog has been replaced.
>>>
>>>And, if the new INVITE reaches UA C, UA B will have no clue
>>
>>that the
>>
>>>dialog has been replaced.
>>>
>>>Regards,
>>>
>>>Christer
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> 
>>>
>>>
>>>>>Etc etc etc.
>>>>>
>>>>>Regards,
>>>>>
>>>>>Christer
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>>	Paul
>>>>>>
>>>>>>Jonathan Rosenberg wrote:
>>>>>>
>>>>>>
>>>>>>>Andrew made a great point in one of his mails which merits
>>>>>>
>>>>>>promotion
>>>>>>
>>>>>>
>>>>>>>into its own thread.
>>>>>>>
>>>>>>>I and others have asserted that you should be able to
>>>>
>>>>identify the
>>>>
>>>>
>>>>>>>service by things naturally in the request, and in
>>>>>>
>>>>>>particular, SDP has
>>>>>>
>>>>>>
>>>>>>>come up as a key thing. For example, in a multiplayer
>>
>>game, there
>>
>>>>>>>might be an m-line for the particular proprietary
>>>>
>>>>protocol used for
>>>>
>>>>
>>>>>>>game moves. Or, if its MSRP, you might have specific
>>>>>>
>>>>>>content types for
>>>>>>
>>>>>>
>>>>>>>game moves. When present, those exactly indicate what the
>>>>>>
>>>>>>service is.
>>>>>>
>>>>>>
>>>>>>>But what do we do about offerless INVITE? In particular,
>>>>
>>>>there are
>>>>
>>>>
>>>>>>>cases where we need to invoke an application server to handle a 
>>>>>>>request, and what if we don't have the SDP in the INVITE?
>>>>>>
>>>>>>Its too late
>>>>>>
>>>>>>
>>>>>>>to insert an application server once the INVITE has already
>>>>>>
>>>>>>passed to
>>>>>>
>>>>>>
>>>>>>>the called party and the 200 OK is sent.
>>>>>>>
>>>>>>>This is a great point. I'd rather just fix this little
>>>>
>>>>problem, the
>>>>
>>>>
>>>>>>>define an entire notion of services and service identifiers
>>>>>>
>>>>>>that are
>>>>>>
>>>>>>
>>>>>>>not necessary, just because we have offerless INVITEs.
>>>>>>>
>>>>>>>There are lots of ways to handle this problem. My preferred
>>>>>>
>>>>>>approach
>>>>>>
>>>>>>
>>>>>>>would be that offerless INVITE contain a capability
>>>>>>
>>>>>>description, using
>>>>>>
>>>>>>
>>>>>>>Flemmings new capability framework. The INVITE would
>>>>
>>>>contain a body
>>>>
>>>>
>>>>>>>that  looks like SDP, but has a different content type and 
>>>>>>>content-disposition. It lists the media streams, codecs
>>
>>and other
>>
>>>>>>>capabilities that exist. Based on this, the called party
>>>>
>>>>makes the
>>>>
>>>>
>>>>>>>offer in a 2xx or 18x like normal, and the answer in the
>>>>>>
>>>>>>ACK/PRACK as normal.
>>>>>>
>>>>>>
>>>>>>>It is completely backwards compatible.
>>>>>>>
>>>>>>>Now, a related issue is what happens if a session starts
>>>>
>>>>with audio
>>>>
>>>>
>>>>>>>only, and then the user wants to turn this into a game? Its
>>>>>>
>>>>>>too late
>>>>>>
>>>>>>
>>>>>>>to add new application servers into the call path, which
>>>>>>
>>>>>>might need to
>>>>>>
>>>>>>
>>>>>>>be there for games. What can we do?
>>>>>>>
>>>>>>>Here too, there is a simple answer. If a network provider
>>>>>>
>>>>>>requires app
>>>>>>
>>>>>>
>>>>>>>servers to get inserted, or a phone requires dispatch to a
>>>>>>
>>>>>>different
>>>>>>
>>>>>>
>>>>>>>application, it would receive a re-INVITE to add the
>>>>
>>>>gaming stream,
>>>>
>>>>
>>>>>>>and then 3xx it. The Contact would contain an embedded
>>>>>>
>>>>>>Replaces header
>>>>>>
>>>>>>
>>>>>>>with the dialog IDs for the call in progress. THe
>>>>>>
>>>>>>originating UA would
>>>>>>
>>>>>>
>>>>>>>then do a totally new INVITE that replaces the call in
>>>>>>
>>>>>>progress. This
>>>>>>
>>>>>>
>>>>>>>would allow the network and/or phone to redispatch, yet
>>
>>provide a
>>
>>>>>>>seamless user experience. Billing can be done however the
>>>>
>>>>operators
>>>>
>>>>
>>>>>>>want, since the signaling is very clear on exactly what
>>>>>>
>>>>>>happened and when.
>>>>>>
>>>>>>
>>>>>>>Furthermore, the replaces mechansim would only get used if
>>>>>>
>>>>>>there was
>>>>>>
>>>>>>
>>>>>>>someone in the call path that needed to see it.
>>>>>>>
>>>>>>>I'd rather specify things like this, than invent an
>>>>
>>>>entire services
>>>>
>>>>
>>>>>>>framework that would lead to balkanization and non-interoperable 
>>>>>>>islands of SIP.
>>>>>>>
>>>>>>>-Jonathan R.
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>_______________________________________________
>>>>>>Sipping mailing list
>>>>
>>>>https://www1.ietf.org/mailman/listinfo/sipping
>>>>
>>>>
>>>>>>This list is for NEW development of the application of SIP Use 
>>>>>>sip-implementors@cs.columbia.edu for questions on current sip Use 
>>>>>>sip@ietf.org for new developments of core SIP
>>>>>>
>>>>>
>>-- 
>>Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
>>Cisco Fellow                                   Parsippany, NJ 
>>07054-2711
>>Cisco Systems
>>jdrosen@cisco.com                              FAX:   (973) 952-5050
>>http://www.jdrosen.net                         PHONE: (973) 952-5000
>>http://www.cisco.com
>>
> 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP Use
> sip-implementors@cs.columbia.edu for questions on current sip Use
> sip@ietf.org for new developments of core SIP
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 27 10:15:55 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HM43H-0007R6-Tz; Tue, 27 Feb 2007 10:15:39 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HM43H-0007R1-16
	for sipping@ietf.org; Tue, 27 Feb 2007 10:15:39 -0500
Received: from sj-iport-2-in.cisco.com ([171.71.176.71]
	helo=sj-iport-2.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HM43E-0001ZN-ID
	for sipping@ietf.org; Tue, 27 Feb 2007 10:15:39 -0500
Received: from rtp-dkim-1.cisco.com ([64.102.121.158])
	by sj-iport-2.cisco.com with ESMTP; 27 Feb 2007 07:15:35 -0800
X-IronPort-AV: i="4.14,226,1170662400"; 
	d="scan'208"; a="363133604:sNHT50118840"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id l1RFFZfq026376; 
	Tue, 27 Feb 2007 10:15:35 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1RFFPOQ014640; 
	Tue, 27 Feb 2007 10:15:35 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 27 Feb 2007 10:15:28 -0500
Received: from [192.168.1.104] ([10.86.242.23]) by xfe-rtp-202.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 27 Feb 2007 10:15:27 -0500
Message-ID: <45E44B0F.8020504@cisco.com>
Date: Tue, 27 Feb 2007 10:15:27 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: "Stephen Terrill (MI/EEM)" <stephen.terrill@ericsson.com>
Subject: Re: [Sipping] Service Identification in SIP
References: <989763567E51AB48BE015DF7BB960CAD02926429@esealmw111>
In-Reply-To: <989763567E51AB48BE015DF7BB960CAD02926429@esealmw111>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 27 Feb 2007 15:15:27.0940 (UTC)
	FILETIME=[1CD64040:01C75A82]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=3634; t=1172589335;
	x=1173453335; c=relaxed/simple; s=rtpdkim1001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20in=20SIP
	|Sender:=20
	|To:=20=22Stephen=20Terrill=20(MI/EEM)=22=20<stephen.terrill@ericsson.com
	>; bh=/CRCEB6dOibYZw0IIjHukEH3y+O3/T9OMQrfmEiZxQQ=;
	b=mtjAaZ6dzjIs9XbVYaP2oBwXYQ1aUEQ9syS3WTJ+KVBKnsbi/9T32KKNoj/V+jlMBHOMkiTA
	3hrdm1i5fB8+OSTBbrKFxLFfQO+cMdkc8tkYQfQUkXB6+qdYtc5pT3vF;
Authentication-Results: rtp-dkim-1; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim1001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 0fa76816851382eb71b0a882ccdc29ac
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

inline.

Stephen Terrill (MI/EEM) wrote:

>
> And I don't see how you can argue that point. Merely stating it does not
> make it so.
> 
> To me, its really simple. If 3gpp goes and defines a service parameter
> for "conversational telephony" and gives it some name, and puts it into
> Accept-Contact, then the ability to receive a call from a non-3gpp
> entity is ruined, since such a call will not include this parameter. 
> Thus, the right QoS treatment, billing, call processing, etc. will not
> happen.
> 
> In what way is this NOT balkanization? Please be specific.
> 
> <Steve>
> ... Finally some motivation as to why it is said that the service
> identifier is balkanization (uptill now has been "Merely stating it"),
> however I don't see this in the same way.    The "balkanization effect"
> will come from whether or not an operator wants to interconnect to the
> general internet or not, which has nothing to do with being able to
> identify the requested service.  An operator supporting the services,
> will also be ability to make a "guess" at what the incoming requested
> communication is.  Having made that guess, it will assume the rest of
> the communication to follow according (and reject attempts to the
> opposite).  Such interworking is nothing new.  As such, I do not see
> that this reduces interworking at all.
> </Steve>

Lets talk about how one goes about doing this 'guessing'.

It would do guessing by looking at the SDP, SIP headers, option tags in 
Require and so on, and figure out what service was intended. Now, there 
are two cases. In one case, the information in the SIP/SDP is sufficient 
to match exactly to the operator definition of services, and the overall 
user experience remains good. In the other case, the information is not 
sufficient, and the user experience of calls inter-provider will 
fundamentally and forever be worse than intra-provider. I assert, we 
don't want to go down the latter route. That means that you must always 
be able to accurately determine the service from inspection of the SIP/SDP.

And this is my point.

Let me also add that this balkanization comes from sevearl specific 
problems which have been explicitly stated, despite your claim above 
that they have not. They are:

1. if caller and called party support services which overlap but are not 
the same (for example, both endpoints do video but one does IM too and 
the other doesn't), you will not be able to properly deal with this, 
since you are deriving service definitions from this one header, rather 
than inspecting the individual SDP and SIP headers that are negotiated 
to determine the actual session contents.

2. it will not work with non-3gpp providers that have no notion of what 
these services are - this is the discussion above

3. for providers that do have a notion of these service identifiers, it 
will require global agreement on what the services are and what they 
mean in order to properly interoperate. THis stifles innovation by 
moving us back to the old model where everyone had to agree on the 
services in every network before they could be deployed. One of the 
stated goals of IMS is to enable feature velocity and innovation. You 
are going to kill that.


-Jonathan R.

-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 27 13:54:28 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HM7S7-00008g-3R; Tue, 27 Feb 2007 13:53:31 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HM7S5-00008Z-9Z
	for sipping@ietf.org; Tue, 27 Feb 2007 13:53:29 -0500
Received: from usremg01.bea.com ([66.248.192.21])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HM7S3-0004LF-Pn
	for sipping@ietf.org; Tue, 27 Feb 2007 13:53:29 -0500
Received: from usremr02.bea.com (usremr02.bea.com [10.160.29.92])
	by usremg01.bea.com (Switch-3.2.2/Switch-3.2.2) with ESMTP id
	l1RIqwXm005733
	(version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=FAIL);
	Tue, 27 Feb 2007 10:53:13 -0800
Received: from repbex01.amer.bea.com (repbex01.bea.com [10.160.26.98])
	by usremr02.bea.com (Switch-3.2.2/Switch-3.2.2) with ESMTP id
	l1RIqv7R001216; Tue, 27 Feb 2007 10:52:57 -0800
Received: from 172.24.29.89 ([172.24.29.89]) by repbex01.amer.bea.com
	([10.160.26.98]) with Microsoft Exchange Server HTTP-DAV ; 
	Tue, 27 Feb 2007 18:52:56 +0000
User-Agent: Microsoft-Entourage/11.3.3.061214
Date: Sun, 25 Feb 2007 19:11:23 -0800
From: Eric Burger <eburger@bea.com>
To: Byron Campen <bcampen@estacado.net>
Message-ID: <C2078FDB.23E9%eburger@bea.com>
Thread-Topic: Billing Models, Protocols, and a Rant
Thread-Index: AcdZOqXr5EbyCsUtEdukwwAWy4mm/w==
In-Reply-To: <681D70B6-F649-4CAF-8E43-72118FCE8575@estacado.net>
Mime-version: 1.0
Content-type: text/plain;
	charset="US-ASCII"
Content-transfer-encoding: 7bit
X-PMX-Version: 4.7.1.128075, Antispam-Engine: 2.5.0.283055,
	Antispam-Data: 2007.2.21.151434
X-Spam-Score: 0.3 (/)
X-Scan-Signature: 8de5f93cb2b4e3bee75302e9eacc33db
Cc: sipping@ietf.org
Subject: [Sipping] Billing Models, Protocols, and a Rant
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

I would offer that it is very much NOT our place to dictate pricing models
on service providers or anyone else, for that matter.

For example, I am composing this message on an airplane.  I have a
polarizing security filter to reduce the ability of someone sitting next to
me from reading my screen.  Why do I have such a filter?  Because once I was
sitting next to a distributor for a very, very, very large, public, network
equipment company.  He happened to be reviewing that company's sales
forecast for the Eastern region, including sales prospects.  At first, I
thought, "Wow - that's cool!  If I were so inclined, I could mine his
prospects."  Then I realized, if I can read his super-secret stuff, he could
read mine.

That's when I went out and bought the filter.

What is the point of the story?  I would be surprised if the filter costs
more than USD 3 to manufacture.  Cost-based pricing would place the ASP
between $10 - $20 each.  However, they go for $50.  Why?  Because the VALUE
of the product far outweighs what the cost-based price would indicate.

Same goes for telecommunication services.   Back in the day, I had a
platform that, from a single number, would find the subscriber, let the
subscriber know *really* who called, and let the subscriber control ultimate
call treatment (connect call, send to voice mail, offer a polite Get Lost,
etc.).  If you called the offering a single-number service, you could hope
to charge $25/month + $0.50/min.  However, if you organized it to be a
personal assistant, given that you were replacing a (in those days)
$35,000/year assistant, you could easily charge a flat $1,000/month.  I'm
happy, because I'm making way better than 65% gross margin.  Subscriber is
happy, because they are saving 60% of their costs, and getting a better
service.

If I were only allowed to create business models on cost, the world would be
worse off.

THAT SAID, part of our special sauce was that we did NOT rely on the
telephone network for our billing.  Price is something entirely in the
service provider's control (barring governmental intervention, like
tariffs).  We could do our cool services BECAUSE we did not confuse
signaling with billing.

That has been the IETF lesson for at least 20 years.  Thus, I will push back
when one group says, "we have to put in this feature into a session protocol
for billing" -or- when another group says, "service providers must use my
model for billing."


On 2/25/07 12:11 PM, "Byron Campen" <bcampen@estacado.net> wrote:

> This is more palatable than having the UAs provide these
> identifiers. I can still see interop problems down the road where UAs
> could confuse (or deliberately mislead) the S-CSCF by trying to do
> something with which it is not familiar. It seems to me that
> approaching this from the angle of "we have a fixed set of services,
> with individual pricing" is fundamentally and unavoidably at odds
> with the basic structure of SIP. Pricing should be based on the
> amount of work that the provider has to do. This is extremely easy to
> figure out. Bandwidth costs x, location service (ie, call
> establishment) costs y, other functionality (that your servers have
> to do work to accomplish) will cost whatever you feel is appropriate,
> etc. Putting in all the extra infrastructure/logic to simulate the
> charging for old 2G networks, or even the PSTN, seems pointless to
> me. You end up wasting money, time(ie, money), and effort(ie, more
> money). Why not use a charging system that isn't a total pain to
> implement? If you base your charging on the actual work you're doing,
> there is less to exploit in your system, and you can make _more_
> money because you don't have all the overhead from the PSTN billing
> emulation layer! Adopting this strategy does not impede your ability
> to set your pricing to any level you want. It in no way impedes your
> ability to make money.
> 
> Best regards,
> Byron Campen
> 
>> Paul's analysis is reasonable. Basically, starting with Andrew's
>> list, I see service identification as a business issue. To optimize
>> operation in the IMS architecture, it would seem reasonable in
>> several cases for the originating S-CSCF to identify the tariffed
>> service(s) implied by the messaging from the UA and add the
>> appropriate administrative header fields so downstream elements
>> don't have to repeat the analysis.
>> 
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP

_______________________________________________________________________
Notice:  This email message, together with any attachments, may contain
information  of  BEA Systems,  Inc.,  its subsidiaries  and  affiliated
entities,  that may be confidential,  proprietary,  copyrighted  and/or
legally privileged, and is intended solely for the use of the individual
or entity named in this message. If you are not the intended recipient,
and have received this message in error, please immediately return this
by email and then delete it.

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 27 15:07:38 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HM8av-0005Ec-NW; Tue, 27 Feb 2007 15:06:41 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HM8aj-0004e8-AL
	for sipping@ietf.org; Tue, 27 Feb 2007 15:06:29 -0500
Received: from dsl001-129-069.dfw1.dsl.speakeasy.net ([72.1.129.69]
	helo=vicuna.estacado.net) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HM8XP-0001LY-BN
	for sipping@ietf.org; Tue, 27 Feb 2007 15:03:04 -0500
Received: from [172.17.2.60] ([172.17.2.60]) (authenticated bits=0)
	by vicuna.estacado.net (8.13.8/8.13.8) with ESMTP id l1RK2rwl017759
	(version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO);
	Tue, 27 Feb 2007 14:02:53 -0600 (CST)
	(envelope-from bcampen@estacado.net)
In-Reply-To: <C2078FDB.23E9%eburger@bea.com>
References: <C2078FDB.23E9%eburger@bea.com>
Mime-Version: 1.0 (Apple Message framework v752.3)
Message-Id: <C5751F23-78F2-4FCD-866F-9477865C96D4@estacado.net>
From: Byron Campen <bcampen@estacado.net>
Date: Tue, 27 Feb 2007 14:02:48 -0600
To: Eric Burger <eburger@bea.com>
X-Mailer: Apple Mail (2.752.3)
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 7268a2980febc47a9fa732aba2b737ba
Cc: sipping@ietf.org
Subject: [Sipping] Re: Billing Models, Protocols, and a Rant
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1310570600=="
Errors-To: sipping-bounces@ietf.org


--===============1310570600==
Content-Type: multipart/signed; micalg=sha1; boundary=Apple-Mail-4-1025843765;
	protocol="application/pkcs7-signature"


--Apple-Mail-4-1025843765
Content-Transfer-Encoding: 7bit
Content-Type: text/plain;
	charset=US-ASCII;
	delsp=yes;
	format=flowed

	I was not trying to say that telcos must ignore value when making  
their charging decisions. None of my message is normative in any way;  
I was just giving an example of how simple/easy things really become  
when you treat bandwidth and the purpose the bandwidth is being used  
for as orthogonal issues.

	That said, I do think that the pattern of charging a special price  
on say, SMS, and then getting bent out of shape when someone puts a  
third-party IM client on their smartphone so they can do IM for  
cheap, seems dangerous to me. The fact is, in the world of software,  
the stuff with high value (to individuals) almost always ends up  
being free. It is also true that every time providers use this  
pattern, they are forcing themselves into a situation where they will  
have to disable all third-party applications by locking down their  
network, because their alternative is to try competing with a free  
product. This is just how things are. There is nothing that they (or  
the IETF) can do to prevent it. And when they take the obvious choice  
of locking down their network, everyone loses; the customers lose  
useful third-party applications for which there is no substitute, and  
the provider loses customers and profits from data plans, since their  
customers aren't using the nifty third-party apps anymore.

	Once again, there is not a single normative statement in here. I am  
just pointing out the reality of the situation. Take it for what you  
will.

Best regards,
Byron Campen


> I would offer that it is very much NOT our place to dictate pricing  
> models
> on service providers or anyone else, for that matter.
>
> For example, I am composing this message on an airplane.  I have a
> polarizing security filter to reduce the ability of someone sitting  
> next to
> me from reading my screen.  Why do I have such a filter?  Because  
> once I was
> sitting next to a distributor for a very, very, very large, public,  
> network
> equipment company.  He happened to be reviewing that company's sales
> forecast for the Eastern region, including sales prospects.  At  
> first, I
> thought, "Wow - that's cool!  If I were so inclined, I could mine his
> prospects."  Then I realized, if I can read his super-secret stuff,  
> he could
> read mine.
>
> That's when I went out and bought the filter.
>
> What is the point of the story?  I would be surprised if the filter  
> costs
> more than USD 3 to manufacture.  Cost-based pricing would place the  
> ASP
> between $10 - $20 each.  However, they go for $50.  Why?  Because  
> the VALUE
> of the product far outweighs what the cost-based price would indicate.
>
> Same goes for telecommunication services.   Back in the day, I had a
> platform that, from a single number, would find the subscriber, let  
> the
> subscriber know *really* who called, and let the subscriber control  
> ultimate
> call treatment (connect call, send to voice mail, offer a polite  
> Get Lost,
> etc.).  If you called the offering a single-number service, you  
> could hope
> to charge $25/month + $0.50/min.  However, if you organized it to be a
> personal assistant, given that you were replacing a (in those days)
> $35,000/year assistant, you could easily charge a flat $1,000/ 
> month.  I'm
> happy, because I'm making way better than 65% gross margin.   
> Subscriber is
> happy, because they are saving 60% of their costs, and getting a  
> better
> service.
>
> If I were only allowed to create business models on cost, the world  
> would be
> worse off.
>
> THAT SAID, part of our special sauce was that we did NOT rely on the
> telephone network for our billing.  Price is something entirely in the
> service provider's control (barring governmental intervention, like
> tariffs).  We could do our cool services BECAUSE we did not confuse
> signaling with billing.
>
> That has been the IETF lesson for at least 20 years.  Thus, I will  
> push back
> when one group says, "we have to put in this feature into a session  
> protocol
> for billing" -or- when another group says, "service providers must  
> use my
> model for billing."
>
>
> On 2/25/07 12:11 PM, "Byron Campen" <bcampen@estacado.net> wrote:
>
>> This is more palatable than having the UAs provide these
>> identifiers. I can still see interop problems down the road where UAs
>> could confuse (or deliberately mislead) the S-CSCF by trying to do
>> something with which it is not familiar. It seems to me that
>> approaching this from the angle of "we have a fixed set of services,
>> with individual pricing" is fundamentally and unavoidably at odds
>> with the basic structure of SIP. Pricing should be based on the
>> amount of work that the provider has to do. This is extremely easy to
>> figure out. Bandwidth costs x, location service (ie, call
>> establishment) costs y, other functionality (that your servers have
>> to do work to accomplish) will cost whatever you feel is appropriate,
>> etc. Putting in all the extra infrastructure/logic to simulate the
>> charging for old 2G networks, or even the PSTN, seems pointless to
>> me. You end up wasting money, time(ie, money), and effort(ie, more
>> money). Why not use a charging system that isn't a total pain to
>> implement? If you base your charging on the actual work you're doing,
>> there is less to exploit in your system, and you can make _more_
>> money because you don't have all the overhead from the PSTN billing
>> emulation layer! Adopting this strategy does not impede your ability
>> to set your pricing to any level you want. It in no way impedes your
>> ability to make money.
>>
>> Best regards,
>> Byron Campen
>>
>>> Paul's analysis is reasonable. Basically, starting with Andrew's
>>> list, I see service identification as a business issue. To optimize
>>> operation in the IMS architecture, it would seem reasonable in
>>> several cases for the originating S-CSCF to identify the tariffed
>>> service(s) implied by the messaging from the UA and add the
>>> appropriate administrative header fields so downstream elements
>>> don't have to repeat the analysis.
>>>
>>
>> _______________________________________________
>> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP
>> Use sip-implementors@cs.columbia.edu for questions on current sip
>> Use sip@ietf.org for new developments of core SIP
>
> ______________________________________________________________________ 
> _
> Notice:  This email message, together with any attachments, may  
> contain
> information  of  BEA Systems,  Inc.,  its subsidiaries  and   
> affiliated
> entities,  that may be confidential,  proprietary,  copyrighted   
> and/or
> legally privileged, and is intended solely for the use of the  
> individual
> or entity named in this message. If you are not the intended  
> recipient,
> and have received this message in error, please immediately return  
> this
> by email and then delete it.


--Apple-Mail-4-1025843765
Content-Transfer-Encoding: base64
Content-Type: application/pkcs7-signature;
	name=smime.p7s
Content-Disposition: attachment;
	filename=smime.p7s
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--Apple-Mail-4-1025843765--


--===============1310570600==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1310570600==--




From sipping-bounces@ietf.org Tue Feb 27 17:02:13 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMANj-0005jj-7Z; Tue, 27 Feb 2007 17:01:11 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HMANh-0005jT-Lm
	for sipping@ietf.org; Tue, 27 Feb 2007 17:01:10 -0500
Received: from alnrmhc16.comcast.net ([206.18.177.56])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HMABl-0000Z7-1M
	for sipping@ietf.org; Tue, 27 Feb 2007 16:48:51 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (alnrmhc16) with ESMTP
	id <20070227214848b1600ao0fne>; Tue, 27 Feb 2007 21:48:48 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1RLmlbV017453
	for <sipping@ietf.org>; Tue, 27 Feb 2007 16:48:47 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1RLmlIJ017449;
	Tue, 27 Feb 2007 16:48:47 -0500
Date: Tue, 27 Feb 2007 16:48:47 -0500
Message-Id: <200702272148.l1RLmlIJ017449@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <C2078FDB.23E9%eburger@bea.com>
Subject: Re: [Sipping] Billing Models, Protocols, and a Rant
References: <C2078FDB.23E9%eburger@bea.com>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 02ec665d00de228c50c93ed6b5e4fc1a
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Let me try to compose a version of this rant, which I think will be
helpful in (1) getting IETF folks to understand where the telco folks
are coming from, and (2) getting the telco folks to see that they have
been bundling several concepts together.  I am making some guesses as
to how the telcos do things, not based on insider knowledge, but by
connecting a bunch of dots that I've seen over the years.
Nonetheless, it assembles a picture that needs to be examined more
carefully than people have been.

I think the crux of the situation is that the telcos look at what
their network does as a series of "silos", services, that actually are
largely independent and parallel to each other.  That is, each of the
following is separate for each service:

- user agent facilities
- application servers
- network transmission facilities (if the recent BT announcement is typical)
- price structure
- return-on-investment analysis
- regulatory approval
- vendor of the equipment
- perhaps even organizational departments

That is, the fundamental unit of, well, *everything* is the "service".

On the other hand, from the IETF point of view, "services" are just
specialized functionality in the user agents, there are no application
servers, and the network transmission facilities treat all packets the
same under a simple billing scheme.  Silo-ization is an evil to be
fought.

But the IETF model does break down in a lot of real-world cases:

- If the UA cannot support all the processing needed, or the last-mile
  link is bandwidth-constrained, you need application servers.

- Network transmission doesn't have a simple price structure if you
  have high utilization and you do real QoS.  Real competition tends
  to make prices match costs, but when QoS matters, costs are nowhere
  near uniform, and capacity management and billing become
  "interesting" problems.

- In the real world there are regulators when you're rolling out
  services for the masses, and regulators worry about things in a way
  that does not neatly correspond to network engineering.  (I've long
  suspected that the real difference between PoC and "telephony" is
  that by not being full-duplex, PoC escapes some mandated charge that
  ordinary cellular is subject to, and the savings are in part passed
  on to the customers.)

>From the IETF side, we're going to have to start thinking about the
implications of these facts, and ensure that SIP copes with them.

>From the telco side, people should realize that even if the world is
partitioned into "services" in one respect, that it isn't necessarily
a good thing to partition it into services in another respect.  (In
particular, we've discovered that dividing the long-haul network
protocol is not a good thing -- there is only one TCP, for instance,
and the meaning of the data stream is imposed only by the endpoints.)

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 27 17:19:39 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMAfS-0004iU-MZ; Tue, 27 Feb 2007 17:19:30 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HMAfQ-0004i4-EV
	for sipping@ietf.org; Tue, 27 Feb 2007 17:19:28 -0500
Received: from rwcrmhc14.comcast.net ([204.127.192.84])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HMAfP-0001Vc-6T
	for sipping@ietf.org; Tue, 27 Feb 2007 17:19:28 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (rwcrmhc14) with ESMTP
	id <20070227221925m1400ponuhe>; Tue, 27 Feb 2007 22:19:25 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1RMJObV019686
	for <sipping@ietf.org>; Tue, 27 Feb 2007 17:19:24 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1RMJOEa019682;
	Tue, 27 Feb 2007 17:19:24 -0500
Date: Tue, 27 Feb 2007 17:19:24 -0500
Message-Id: <200702272219.l1RMJOEa019682@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <45E44871.10002@cisco.com> (jdrosen@cisco.com)
Subject: Re: [Sipping] Service Identification: offerless INVITE case
References: <989763567E51AB48BE015DF7BB960CAD02926415@esealmw111>
	<45E44871.10002@cisco.com>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: ea4ac80f790299f943f0a53be7e1a21a
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

Most of this is good, but the use of an in-dialog 3xx is a really bad
idea for re-routing a dialog.  Primarily because it's not specified
how such a thing is processed, and also because it has to be inserted
in-dialog (whereas the controller of this little dance may not be in
the signalling path).

Let me suggest some alternatives:

You can send a REFER to a UA, which will cause it to replace an
existing dialog with a new one.  If you phrase it correctly, it
replaces both ends:

	REFER UA-A-gruu
	Target-Dialog: call-id...
	Refer-To: UA-B-gruu?Replaces=call-id...?Route=new-app-server

A B2BUA exploit INVITE/Join to insert itself into an existing dialog;

	send:
	INVITE UA-A-gruu
	Join: call-id...

	then send:
	INVITE UA-B-gruu
	Replaces: call-id...

Since these mechanisms are already defined, we know that cleaning up
the original dialog is properly specified, etc.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Tue Feb 27 17:55:35 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMBDu-0008OA-PA; Tue, 27 Feb 2007 17:55:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HMBDq-0008LM-Md
	for sipping@ietf.org; Tue, 27 Feb 2007 17:55:03 -0500
Received: from dsl001-129-069.dfw1.dsl.speakeasy.net ([72.1.129.69]
	helo=vicuna.estacado.net) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HMBD3-00084V-RZ
	for sipping@ietf.org; Tue, 27 Feb 2007 17:54:35 -0500
Received: from [172.17.2.60] ([172.17.2.60]) (authenticated bits=0)
	by vicuna.estacado.net (8.13.8/8.13.8) with ESMTP id l1RMsAj6034798
	(version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO);
	Tue, 27 Feb 2007 16:54:10 -0600 (CST)
	(envelope-from bcampen@estacado.net)
In-Reply-To: <200702272148.l1RLmlIJ017449@dragon.ariadne.com>
References: <C2078FDB.23E9%eburger@bea.com>
	<200702272148.l1RLmlIJ017449@dragon.ariadne.com>
Mime-Version: 1.0 (Apple Message framework v752.3)
Message-Id: <D4D6C7F4-96AB-42D5-B8B0-AB9F7925A067@estacado.net>
From: Byron Campen <bcampen@estacado.net>
Subject: Re: [Sipping] Billing Models, Protocols, and a Rant
Date: Tue, 27 Feb 2007 16:54:04 -0600
To: Dale.Worley@comcast.net
X-Mailer: Apple Mail (2.752.3)
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 03169bfe4792634a390035a01a6c6d2f
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1134827445=="
Errors-To: sipping-bounces@ietf.org


--===============1134827445==
Content-Type: multipart/signed; micalg=sha1; boundary=Apple-Mail-2-1036120816;
	protocol="application/pkcs7-signature"


--Apple-Mail-2-1036120816
Content-Transfer-Encoding: 7bit
Content-Type: text/plain;
	charset=US-ASCII;
	delsp=yes;
	format=flowed

	Alright, I do want to make it clear that I have no problem with  
service differentiation that reflects differing responsibilities of  
the provider. The example cases you mention all entail additional/ 
different responsibilities on the part of the provider, and so I have  
no problem with classifying these as distinct services. A call with  
QoS is a very different animal than one without, and a provider is  
completely justified in charging extra for the QoS. Also, cases where  
an application server needs to do extra work are perfectly  
justifiable as separate services.

	What I object to is the cases where there is service differentiation  
that is not motivated by differing provider responsibilities. If  
service A requires you to forward packets P, and service B requires  
you to forward similar packets Q in exactly the same way, A and B  
probably shouldn't be different services. By breaking these up, at  
the very least you invite attacks where sneaky UAs will pretend to  
use the cheaper service A (wink wink, nudge nudge), but really end up  
using service B (imagine a case where, in the SDP offer, a  
proprietary attribute would indicate 'and by "A", I mean "B"'). By  
keying your service differentiation on whether your responsibilities  
differ, your infrastructure becomes much harder to trick.

Best regards,
Byron Campen

> Let me try to compose a version of this rant, which I think will be
> helpful in (1) getting IETF folks to understand where the telco folks
> are coming from, and (2) getting the telco folks to see that they have
> been bundling several concepts together.  I am making some guesses as
> to how the telcos do things, not based on insider knowledge, but by
> connecting a bunch of dots that I've seen over the years.
> Nonetheless, it assembles a picture that needs to be examined more
> carefully than people have been.
>
> I think the crux of the situation is that the telcos look at what
> their network does as a series of "silos", services, that actually are
> largely independent and parallel to each other.  That is, each of the
> following is separate for each service:
>
> - user agent facilities
> - application servers
> - network transmission facilities (if the recent BT announcement is  
> typical)
> - price structure
> - return-on-investment analysis
> - regulatory approval
> - vendor of the equipment
> - perhaps even organizational departments
>
> That is, the fundamental unit of, well, *everything* is the "service".
>
> On the other hand, from the IETF point of view, "services" are just
> specialized functionality in the user agents, there are no application
> servers, and the network transmission facilities treat all packets the
> same under a simple billing scheme.  Silo-ization is an evil to be
> fought.
>
> But the IETF model does break down in a lot of real-world cases:
>
> - If the UA cannot support all the processing needed, or the last-mile
>   link is bandwidth-constrained, you need application servers.
>
> - Network transmission doesn't have a simple price structure if you
>   have high utilization and you do real QoS.  Real competition tends
>   to make prices match costs, but when QoS matters, costs are nowhere
>   near uniform, and capacity management and billing become
>   "interesting" problems.
>
> - In the real world there are regulators when you're rolling out
>   services for the masses, and regulators worry about things in a way
>   that does not neatly correspond to network engineering.  (I've long
>   suspected that the real difference between PoC and "telephony" is
>   that by not being full-duplex, PoC escapes some mandated charge that
>   ordinary cellular is subject to, and the savings are in part passed
>   on to the customers.)
>
>> From the IETF side, we're going to have to start thinking about the
> implications of these facts, and ensure that SIP copes with them.
>
>> From the telco side, people should realize that even if the world is
> partitioned into "services" in one respect, that it isn't necessarily
> a good thing to partition it into services in another respect.  (In
> particular, we've discovered that dividing the long-haul network
> protocol is not a good thing -- there is only one TCP, for instance,
> and the meaning of the data stream is imposed only by the endpoints.)
>
> Dale
>
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP


--Apple-Mail-2-1036120816
Content-Transfer-Encoding: base64
Content-Type: application/pkcs7-signature;
	name=smime.p7s
Content-Disposition: attachment;
	filename=smime.p7s
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--Apple-Mail-2-1036120816--


--===============1134827445==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1134827445==--




From sipping-bounces@ietf.org Wed Feb 28 02:07:55 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMIsi-0000mx-6W; Wed, 28 Feb 2007 02:05:44 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HMIsg-0000iw-WE
	for sipping@ietf.org; Wed, 28 Feb 2007 02:05:43 -0500
Received: from mailgate.siemenscomms.co.uk ([195.171.110.225]
	helo=bemg01.siemenscomms.co.uk)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HMIsd-0000Cr-MR
	for sipping@ietf.org; Wed, 28 Feb 2007 02:05:42 -0500
Received: from ntht207e.uksgcs.siemenscomms.co.uk ([137.223.247.82])
	by siemenscomms.co.uk (PMDF V6.0-24 #40642)
	with ESMTP id <0JE500C0ZVP2ED@siemenscomms.co.uk> for sipping@ietf.org;
	Wed, 28 Feb 2007 07:05:26 +0000 (GMT)
Received: by ntht207e.uksgcs.siemenscomms.co.uk with Internet Mail Service
	(5.5.2657.72)	id <FF5Y4M99>; Wed, 28 Feb 2007 07:05:26 +0000
Content-return: allowed
Date: Wed, 28 Feb 2007 07:05:26 +0000
From: "Elwell, John" <john.elwell@siemens.com>
To: sipping@ietf.org
Message-id: <50B1CBA96870A34799A506B2313F26670B28F0A8@ntht201e.siemenscomms.co.uk>
MIME-version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
Content-type: text/plain
Content-transfer-encoding: 7BIT
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 79899194edc4f33a41f49410777972f8
Subject: [Sipping] FW: New Version Notification -
	draft-ietf-sip-connected-identity- 05.txt
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

The new version just corrects a few nits that emerged during gen-art review.

John 

> -----Original Message-----
> From: ID Tracker [mailto:internet-drafts-reply@ietf.org] 
> Sent: 27 February 2007 20:50
> To: sip-chairs@tools.ietf.org; Elwell, John; fluffy@cisco.com
> Subject: New Version Notification - 
> draft-ietf-sip-connected-identity-05.txt
> 
> New version (-05) has been submitted for 
> draft-ietf-sip-connected-identity-05.txt.
>
http://www.ietf.org/internet-drafts/draft-ietf-sip-connected-identity-05.txt
> 
> Sub state has been changed to AD Follow up from New Id Needed
> 
> IETF Secretariat.
> 

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 28 10:11:36 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMQQj-0004cW-1d; Wed, 28 Feb 2007 10:09:21 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMQQZ-0004Kd-H2; Wed, 28 Feb 2007 10:09:12 -0500
Received: from expint.psytechnics.com ([217.33.106.234])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HMQKJ-0008Jy-1D; Wed, 28 Feb 2007 10:02:44 -0500
Content-Class: urn:content-classes:message
Importance: normal
Priority: normal
MIME-Version: 1.0
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1896
Date: Wed, 28 Feb 2007 15:03:26 -0000
Message-ID: <257B18BFF522214E92EB52F3478B014E094688@psy01exch.intra.psytechnics.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Perceptual objective media quality reporting
thread-index: AcdbSZlLw9qRfvsQSU61gayfTs285Q==
From: "Ravi Raviraj" <ravi.raviraj@psytechnics.com>
To: <sipping@ietf.org>,
	<sip@ietf.org>,
	<avt@ietf.org>
X-Spam-Score: 0.4 (/)
X-Scan-Signature: 8b6657e60309a1317174c9db2ae5f227
Cc: 
Subject: [Sipping] Perceptual objective media quality reporting
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1620746469=="
Errors-To: sipping-bounces@ietf.org

This is a multi-part message in MIME format.

--===============1620746469==
Content-Class: urn:content-classes:message
Content-Transfer-Encoding: 7bit
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C75B49.9952ED7C"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C75B49.9952ED7C
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

Hi all,

Over the last few years, a number of perceptual media quality objective
measures and assessment techniques have been standardized by ITU-T. For
example, MOS-LQO, MOS-CQO, etc. Each one of these parameters could have
number of variations like narrow-band/wide-band, electrical/acoustical,
full reference/no-reference, audio/video, etc. When applied within a
network end-point, each media direction (ingress/egress) could have one
or more of these values associated with it. We need a way to report
these values efficiently without any ambiguity. These MOS values could
be sent/distributed as perceptual objective quality SIP events. In this
way, all the advantages of SIP like mobility, etc. will be pulled in.

=20

I would like to propose a work/discussion item on an event package for
this purpose. Even define/specify a quality reporting framework. Is
there any interest out there?

=20

I presume that, an event package work will be within SIPPING WG, where
should we do the framework work? SIP/AVT?

=20

Thanks.

=20

Ravi.

=20





This email and attachment are confidential. If you have received this =
message in error, please notify the sender immediately and delete the =
message from your computer. Psytechnics Ltd has taken steps to ensure =
that this email is virus free but the recipient should take necessary =
steps to make certain.




------_=_NextPart_001_01C75B49.9952ED7C
Content-Type: text/html;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

<html xmlns:o=3D"urn:schemas-microsoft-com:office:office" =
xmlns:w=3D"urn:schemas-microsoft-com:office:word" =
xmlns:st1=3D"urn:schemas-microsoft-com:office:smarttags" =
xmlns=3D"http://www.w3.org/TR/REC-html40">

<head>
<meta http-equiv=3DContent-Type content=3D"text/html; =
charset=3Dus-ascii">
<meta name=3DGenerator content=3D"Microsoft Word 11 (filtered medium)">
<o:SmartTagType =
namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"
 name=3D"place"/>
<!--[if !mso]>
<style>
st1\:*{behavior:url(#default#ieooui) }
</style>
<![endif]-->
<style>
<!--
 /* Style Definitions */
 p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0cm;
	margin-bottom:.0001pt;
	font-size:12.0pt;
	font-family:"Times New Roman";}
a:link, span.MsoHyperlink
	{color:blue;
	text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
	{color:purple;
	text-decoration:underline;}
span.EmailStyle17
	{mso-style-type:personal-compose;
	font-family:Arial;
	color:windowtext;}
@page Section1
	{size:612.0pt 792.0pt;
	margin:72.0pt 90.0pt 72.0pt 90.0pt;}
div.Section1
	{page:Section1;}
-->
</style>

</head>

<body lang=3DEN-US link=3Dblue vlink=3Dpurple>

<div class=3DSection1>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>Hi all,<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>Over the last few years, a number of =
perceptual media
quality objective measures and assessment techniques have been =
standardized by
ITU-T. For example, MOS-LQO, MOS-CQO, etc. Each one of these parameters =
could
have number of variations like narrow-band/wide-band, =
electrical/acoustical,
full reference/no-reference, audio/video, etc. When applied within a =
network
end-point, each media direction (ingress/egress) could have one or more =
of
these values associated with it. We need a way to report these values
efficiently without any ambiguity. These MOS values could be =
sent/distributed as
perceptual objective quality SIP events. In this way, all the advantages =
of SIP
like mobility, etc. will be pulled in.<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>I would like to propose a work/discussion item =
on an
event package for this purpose. Even define/specify a quality reporting
framework. Is there any interest out there?<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>I presume that, an event package work will be =
within
SIPPING WG, where should we do the framework work? =
SIP/AVT?<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>Thanks.<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><st1:place w:st=3D"on"><font size=3D2 =
face=3DArial><span
 lang=3DEN-GB =
style=3D'font-size:10.0pt;font-family:Arial'>Ravi</span></font></st1:plac=
e><font
size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:10.0pt;font-family:Arial'>.<o:p></o:p></span></font></=
p>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'><o:p>&nbsp;</o:p></span></font></p>

</div>

</body>

</html>




This email and attachment are confidential. If you have received this =
message in error, please notify the sender immediately and delete the =
message from your computer. Psytechnics Ltd has taken steps to ensure =
that this email is virus free but the recipient should take necessary =
steps to make certain.




------_=_NextPart_001_01C75B49.9952ED7C--


--===============1620746469==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1620746469==--




From sipping-bounces@ietf.org Wed Feb 28 10:41:17 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMQvS-000452-JN; Wed, 28 Feb 2007 10:41:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HMQvQ-00043w-TU
	for sipping@ietf.org; Wed, 28 Feb 2007 10:41:04 -0500
Received: from shaman.nostrum.com ([72.232.15.10] helo=nostrum.com)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HMQvO-0005Ts-DF
	for sipping@ietf.org; Wed, 28 Feb 2007 10:41:04 -0500
Received: from [172.17.1.65] (vicuna-alt.estacado.net [75.53.54.121])
	(authenticated bits=0)
	by nostrum.com (8.13.8/8.13.8) with ESMTP id l1SFex9D042482
	(version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO);
	Wed, 28 Feb 2007 09:40:59 -0600 (CST)
	(envelope-from rjsparks@nostrum.com)
In-Reply-To: <257B18BFF522214E92EB52F3478B014E094688@psy01exch.intra.psytechnics.com>
References: <257B18BFF522214E92EB52F3478B014E094688@psy01exch.intra.psytechnics.com>
Mime-Version: 1.0 (Apple Message framework v752.3)
Message-Id: <7AD8B91F-605F-4E71-A57F-38DEA0C93697@nostrum.com>
From: Robert Sparks <rjsparks@nostrum.com>
Subject: Re: [Sipping] Perceptual objective media quality reporting
Date: Wed, 28 Feb 2007 09:40:57 -0600
To: "Ravi Raviraj" <ravi.raviraj@psytechnics.com>
X-Mailer: Apple Mail (2.752.3)
Received-SPF: pass (nostrum.com: 75.53.54.121 is authenticated by a trusted
	mechanism)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 29dc808194f5fb921c09d0040806d6eb
Cc: sipping LIST <sipping@ietf.org>
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1525721529=="
Errors-To: sipping-bounces@ietf.org


--===============1525721529==
Content-Type: multipart/alternative; boundary=Apple-Mail-31--1050954182


--Apple-Mail-31--1050954182
Content-Transfer-Encoding: 7bit
Content-Type: text/plain;
	charset=US-ASCII;
	delsp=yes;
	format=flowed

(Trimming this to sipping)

How is what you're looking for different from http://www.ietf.org/ 
internet-drafts/draft-ietf-sipping-rtcp-summary-01.txt?

RjS

On Feb 28, 2007, at 9:03 AM, Ravi Raviraj wrote:

> Hi all,
>
> Over the last few years, a number of perceptual media quality  
> objective measures and assessment techniques have been standardized  
> by ITU-T. For example, MOS-LQO, MOS-CQO, etc. Each one of these  
> parameters could have number of variations like narrow-band/wide- 
> band, electrical/acoustical, full reference/no-reference, audio/ 
> video, etc. When applied within a network end-point, each media  
> direction (ingress/egress) could have one or more of these values  
> associated with it. We need a way to report these values  
> efficiently without any ambiguity. These MOS values could be sent/ 
> distributed as perceptual objective quality SIP events. In this  
> way, all the advantages of SIP like mobility, etc. will be pulled in.
>
>
>
> I would like to propose a work/discussion item on an event package  
> for this purpose. Even define/specify a quality reporting  
> framework. Is there any interest out there?
>
>
>
> I presume that, an event package work will be within SIPPING WG,  
> where should we do the framework work? SIP/AVT?
>
>
>
> Thanks.
>
>
>
> Ravi.
>
>
>
> This email and attachment are confidential. If you have received  
> this message in error, please notify the sender immediately and  
> delete the message from your computer. Psytechnics Ltd has taken  
> steps to ensure that this email is virus free but the recipient  
> should take necessary steps to make certain.
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP


--Apple-Mail-31--1050954182
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
	charset=ISO-8859-1

<HTML><BODY style=3D"word-wrap: break-word; -khtml-nbsp-mode: space; =
-khtml-line-break: after-white-space; ">(Trimming this to =
sipping)<DIV><BR class=3D"khtml-block-placeholder"></DIV><DIV>How is =
what you're looking for different from=A0<A =
href=3D"http://www.ietf.org/internet-drafts/draft-ietf-sipping-rtcp-summar=
y-01.txt?">http://www.ietf.org/internet-drafts/draft-ietf-sipping-rtcp-sum=
mary-01.txt?</A></DIV><DIV><BR =
class=3D"khtml-block-placeholder"></DIV><DIV>RjS</DIV><DIV><BR =
class=3D"khtml-block-placeholder"></DIV><DIV><DIV><DIV>On Feb 28, 2007, =
at 9:03 AM, Ravi Raviraj wrote:</DIV><BR =
class=3D"Apple-interchange-newline"><BLOCKQUOTE =
type=3D"cite"><O:SMARTTAGTYPE =
namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags" =
name=3D"place"> <DIV class=3D"Section1"><P class=3D"MsoNormal"><FONT =
size=3D"2" face=3D"Arial"><SPAN lang=3D"EN-GB" style=3D"font-size: =
10.0pt;font-family:Arial">Hi all,<O:P></O:P></SPAN></FONT></P><P =
class=3D"MsoNormal"><FONT size=3D"2" face=3D"Arial"><SPAN lang=3D"EN-GB" =
style=3D"font-size: 10.0pt;font-family:Arial">Over the last few years, a =
number of perceptual media quality objective measures and assessment =
techniques have been standardized by ITU-T. For example, MOS-LQO, =
MOS-CQO, etc. Each one of these parameters could have number of =
variations like narrow-band/wide-band, electrical/acoustical, full =
reference/no-reference, audio/video, etc. When applied within a network =
end-point, each media direction (ingress/egress) could have one or more =
of these values associated with it. We need a way to report these values =
efficiently without any ambiguity. These MOS values could be =
sent/distributed as perceptual objective quality SIP events. In this =
way, all the advantages of SIP like mobility, etc. will be pulled =
in.<O:P></O:P></SPAN></FONT></P><P class=3D"MsoNormal"><FONT size=3D"2" =
face=3D"Arial"><SPAN lang=3D"EN-GB" style=3D"font-size: =
10.0pt;font-family:Arial"><O:P>=A0</O:P></SPAN></FONT></P><P =
class=3D"MsoNormal"><FONT size=3D"2" face=3D"Arial"><SPAN lang=3D"EN-GB" =
style=3D"font-size: 10.0pt;font-family:Arial">I would like to propose a =
work/discussion item on an event package for this purpose. Even =
define/specify a quality reporting framework. Is there any interest out =
there?<O:P></O:P></SPAN></FONT></P><P class=3D"MsoNormal"><FONT size=3D"2"=
 face=3D"Arial"><SPAN lang=3D"EN-GB" style=3D"font-size: =
10.0pt;font-family:Arial"><O:P>=A0</O:P></SPAN></FONT></P><P =
class=3D"MsoNormal"><FONT size=3D"2" face=3D"Arial"><SPAN lang=3D"EN-GB" =
style=3D"font-size: 10.0pt;font-family:Arial">I presume that, an event =
package work will be within SIPPING WG, where should we do the framework =
work? SIP/AVT?<O:P></O:P></SPAN></FONT></P><P class=3D"MsoNormal"><FONT =
size=3D"2" face=3D"Arial"><SPAN lang=3D"EN-GB" style=3D"font-size: =
10.0pt;font-family:Arial"><O:P>=A0</O:P></SPAN></FONT></P><P =
class=3D"MsoNormal"><FONT size=3D"2" face=3D"Arial"><SPAN lang=3D"EN-GB" =
style=3D"font-size: =
10.0pt;font-family:Arial">Thanks.<O:P></O:P></SPAN></FONT></P><P =
class=3D"MsoNormal"><FONT size=3D"2" face=3D"Arial"><SPAN lang=3D"EN-GB" =
style=3D"font-size: =
10.0pt;font-family:Arial"><O:P>=A0</O:P></SPAN></FONT></P><P =
class=3D"MsoNormal"><ST1:PLACE w:st=3D"on"><FONT size=3D"2" =
face=3D"Arial"><SPAN lang=3D"EN-GB" =
style=3D"font-size:10.0pt;font-family:Arial">Ravi</SPAN></FONT></ST1:PLACE=
><FONT size=3D"2" face=3D"Arial"><SPAN lang=3D"EN-GB" =
style=3D"font-size:10.0pt;font-family:Arial">.<O:P></O:P></SPAN></FONT></P=
><P class=3D"MsoNormal"><FONT size=3D"3" face=3D"Times New Roman"><SPAN =
style=3D"font-size: 12.0pt"><O:P>=A0</O:P></SPAN></FONT></P> </DIV> This =
email and attachment are confidential. If you have received this message =
in error, please notify the sender immediately and delete the message =
from your computer. Psytechnics Ltd has taken steps to ensure that this =
email is virus free but the recipient should take necessary steps to =
make certain. </O:SMARTTAGTYPE><DIV style=3D"margin-top: 0px; =
margin-right: 0px; margin-bottom: 0px; margin-left: 0px; =
">_______________________________________________</DIV><DIV =
style=3D"margin-top: 0px; margin-right: 0px; margin-bottom: 0px; =
margin-left: 0px; ">Sipping mailing list<SPAN =
class=3D"Apple-converted-space">=A0 </SPAN><A =
href=3D"https://www1.ietf.org/mailman/listinfo/sipping">https://www1.ietf.=
org/mailman/listinfo/sipping</A></DIV><DIV style=3D"margin-top: 0px; =
margin-right: 0px; margin-bottom: 0px; margin-left: 0px; ">This list is =
for NEW development of the application of SIP</DIV><DIV =
style=3D"margin-top: 0px; margin-right: 0px; margin-bottom: 0px; =
margin-left: 0px; ">Use <A =
href=3D"mailto:sip-implementors@cs.columbia.edu">sip-implementors@cs.colum=
bia.edu</A> for questions on current sip</DIV><DIV style=3D"margin-top: =
0px; margin-right: 0px; margin-bottom: 0px; margin-left: 0px; ">Use <A =
href=3D"mailto:sip@ietf.org">sip@ietf.org</A> for new developments of =
core SIP</DIV> </BLOCKQUOTE></DIV><BR></DIV></BODY></HTML>=

--Apple-Mail-31--1050954182--


--===============1525721529==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1525721529==--




From sipping-bounces@ietf.org Wed Feb 28 11:04:36 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMRI3-0002dW-V3; Wed, 28 Feb 2007 11:04:27 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMR6L-00017W-C7; Wed, 28 Feb 2007 10:52:21 -0500
Received: from mx.cbeyond.com ([66.180.96.58])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HMR6J-0006eM-RC; Wed, 28 Feb 2007 10:52:21 -0500
Received: from [69.15.238.196] (port=3427 helo=TELCHADCPC)
	by mx.cbeyond.com with esmtpa (Exim 4.62)
	(envelope-from <alan.d.clark@telchemy.com>)
	id 1HMS3S-0000M7-AQ; Wed, 28 Feb 2007 11:53:26 -0500
From: "Alan Clark" <alan.d.clark@telchemy.com>
To: "'Ravi Raviraj'" <ravi.raviraj@psytechnics.com>, <sipping@ietf.org>,
	<sip@ietf.org>, <avt@ietf.org>
Date: Wed, 28 Feb 2007 10:52:13 -0500
MIME-Version: 1.0
X-Mailer: Microsoft Office Outlook, Build 11.0.5510
Thread-Index: AcdbSZlLw9qRfvsQSU61gayfTs285QABkZRA
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <257B18BFF522214E92EB52F3478B014E094688@psy01exch.intra.psytechnics.com>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: f0b5a4216bfa030ed8a6f68d1833f8ae
X-Mailman-Approved-At: Wed, 28 Feb 2007 11:04:25 -0500
Cc: 
Subject: [Sipping] RE: [AVT] Perceptual objective media quality reporting
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============0309631818=="
Errors-To: sipping-bounces@ietf.org
Message-Id: <E1HMRI3-0002dW-V3@megatron.ietf.org>

This is a multi-part message in MIME format.

--===============0309631818==
Content-Type: multipart/alternative;
	boundary="----=_NextPart_000_0039_01C75B26.819B6C40"

This is a multi-part message in MIME format.

------=_NextPart_000_0039_01C75B26.819B6C40
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit

Hi Ravi
 
The industry has been doing this for some time - there is an existing draft
in SIPPING - for the SIP RTCP Summary Report that does exactly what you are
asking for, also RTCP XR (RFC3611), RTCP HR (current AVT WG draft), and a
laundry list of other protocols for MGCP, Megaco.......  In some cases (e.g.
RTCP HR, SIP) you can indicate the algorithm used to estimate quality, and
hence if you wanted to use PESQ in an active test then you could report
using the SIP RTCP Summary Report and indicate that the MOS estimation
algorithm was P.862.n
 
Regards
 
Alan
 

  _____  

From: Ravi Raviraj [mailto:ravi.raviraj@psytechnics.com] 
Sent: Wednesday, February 28, 2007 10:03 AM
To: sipping@ietf.org; sip@ietf.org; avt@ietf.org
Subject: [AVT] Perceptual objective media quality reporting



Hi all,

Over the last few years, a number of perceptual media quality objective
measures and assessment techniques have been standardized by ITU-T. For
example, MOS-LQO, MOS-CQO, etc. Each one of these parameters could have
number of variations like narrow-band/wide-band, electrical/acoustical, full
reference/no-reference, audio/video, etc. When applied within a network
end-point, each media direction (ingress/egress) could have one or more of
these values associated with it. We need a way to report these values
efficiently without any ambiguity. These MOS values could be
sent/distributed as perceptual objective quality SIP events. In this way,
all the advantages of SIP like mobility, etc. will be pulled in.

 

I would like to propose a work/discussion item on an event package for this
purpose. Even define/specify a quality reporting framework. Is there any
interest out there?

 

I presume that, an event package work will be within SIPPING WG, where
should we do the framework work? SIP/AVT?

 

Thanks.

 

Ravi.

 

This email and attachment are confidential. If you have received this
message in error, please notify the sender immediately and delete the
message from your computer. Psytechnics Ltd has taken steps to ensure that
this email is virus free but the recipient should take necessary steps to
make certain. 

------=_NextPart_000_0039_01C75B26.819B6C40
Content-Type: text/html;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML xmlns=3D"http://www.w3.org/TR/REC-html40" xmlns:o =3D=20
"urn:schemas-microsoft-com:office:office" xmlns:w =3D=20
"urn:schemas-microsoft-com:office:word" xmlns:st1 =3D=20
"urn:schemas-microsoft-com:office:smarttags"><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; =
charset=3Dus-ascii">
<META content=3D"MSHTML 6.00.5730.11" name=3DGENERATOR><o:SmartTagType =
name=3D"place"=20
namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"></o:SmartTagT=
ype><!--[if !mso]>
<STYLE>st1\:* {
	BEHAVIOR: url(#default#ieooui)
}
</STYLE>
<![endif]-->
<STYLE>@page Section1 {size: 612.0pt 792.0pt; margin: 72.0pt 90.0pt =
72.0pt 90.0pt; }
P.MsoNormal {
	FONT-SIZE: 12pt; MARGIN: 0cm 0cm 0pt; FONT-FAMILY: "Times New Roman"
}
LI.MsoNormal {
	FONT-SIZE: 12pt; MARGIN: 0cm 0cm 0pt; FONT-FAMILY: "Times New Roman"
}
DIV.MsoNormal {
	FONT-SIZE: 12pt; MARGIN: 0cm 0cm 0pt; FONT-FAMILY: "Times New Roman"
}
A:link {
	COLOR: blue; TEXT-DECORATION: underline
}
SPAN.MsoHyperlink {
	COLOR: blue; TEXT-DECORATION: underline
}
A:visited {
	COLOR: purple; TEXT-DECORATION: underline
}
SPAN.MsoHyperlinkFollowed {
	COLOR: purple; TEXT-DECORATION: underline
}
SPAN.EmailStyle17 {
	COLOR: windowtext; FONT-FAMILY: Arial; mso-style-type: personal-compose
}
DIV.Section1 {
	page: Section1
}
</STYLE>
</HEAD>
<BODY lang=3DEN-US vLink=3Dpurple link=3Dblue>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D546214815-28022007><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>Hi Ravi</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D546214815-28022007><FONT =
color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D546214815-28022007><FONT =
color=3D#0000ff=20
size=3D2>The industry&nbsp;has been doing this for some time - there is =
an=20
existing draft in SIPPING - for the SIP RTCP Summary Report that does =
exactly=20
what you are asking for, also RTCP XR (RFC3611), RTCP HR (current AVT WG =
draft),=20
and a laundry list of other protocols for MGCP, Megaco.......&nbsp; In =
some=20
cases (e.g. RTCP HR, SIP) you can indicate the algorithm used to =
estimate=20
quality, and hence if you wanted to use PESQ in an active test then you =
could=20
report using the SIP RTCP Summary Report and indicate that the MOS =
estimation=20
algorithm was P.862.n</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D546214815-28022007><FONT =
color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D546214815-28022007><FONT =
color=3D#0000ff=20
size=3D2>Regards</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D546214815-28022007><FONT =
color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D546214815-28022007><FONT =
color=3D#0000ff=20
size=3D2>Alan</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D546214815-28022007><FONT =
color=3D#0000ff=20
size=3D2></FONT></SPAN>&nbsp;</DIV><BR>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr align=3Dleft>
<HR tabIndex=3D-1>
<FONT face=3DTahoma size=3D2><B>From:</B> Ravi Raviraj=20
[mailto:ravi.raviraj@psytechnics.com] <BR><B>Sent:</B> Wednesday, =
February 28,=20
2007 10:03 AM<BR><B>To:</B> sipping@ietf.org; sip@ietf.org;=20
avt@ietf.org<BR><B>Subject:</B> [AVT] Perceptual objective media quality =

reporting<BR></FONT><BR></DIV>
<DIV></DIV>
<DIV class=3DSection1>
<P class=3DMsoNormal><FONT face=3DArial size=3D2><SPAN lang=3DEN-GB=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Hi =
all,<o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal><FONT face=3DArial size=3D2><SPAN lang=3DEN-GB=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Over the last few years, a =
number of=20
perceptual media quality objective measures and assessment techniques =
have been=20
standardized by ITU-T. For example, MOS-LQO, MOS-CQO, etc. Each one of =
these=20
parameters could have number of variations like narrow-band/wide-band,=20
electrical/acoustical, full reference/no-reference, audio/video, etc. =
When=20
applied within a network end-point, each media direction =
(ingress/egress) could=20
have one or more of these values associated with it. We need a way to =
report=20
these values efficiently without any ambiguity. These MOS values could =
be=20
sent/distributed as perceptual objective quality SIP events. In this =
way, all=20
the advantages of SIP like mobility, etc. will be pulled=20
in.<o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal><FONT face=3DArial size=3D2><SPAN lang=3DEN-GB=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial"><o:p>&nbsp;</o:p></SPAN></FONT></P>
<P class=3DMsoNormal><FONT face=3DArial size=3D2><SPAN lang=3DEN-GB=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">I would like to propose a=20
work/discussion item on an event package for this purpose. Even =
define/specify a=20
quality reporting framework. Is there any interest out=20
there?<o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal><FONT face=3DArial size=3D2><SPAN lang=3DEN-GB=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial"><o:p>&nbsp;</o:p></SPAN></FONT></P>
<P class=3DMsoNormal><FONT face=3DArial size=3D2><SPAN lang=3DEN-GB=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">I presume that, an event =
package=20
work will be within SIPPING WG, where should we do the framework work?=20
SIP/AVT?<o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal><FONT face=3DArial size=3D2><SPAN lang=3DEN-GB=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial"><o:p>&nbsp;</o:p></SPAN></FONT></P>
<P class=3DMsoNormal><FONT face=3DArial size=3D2><SPAN lang=3DEN-GB=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">Thanks.<o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal><FONT face=3DArial size=3D2><SPAN lang=3DEN-GB=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial"><o:p>&nbsp;</o:p></SPAN></FONT></P>
<P class=3DMsoNormal><st1:place w:st=3D"on"><FONT face=3DArial =
size=3D2><SPAN lang=3DEN-GB=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">Ravi</SPAN></FONT></st1:place><FONT=20
face=3DArial size=3D2><SPAN lang=3DEN-GB=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">.<o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal><FONT face=3D"Times New Roman" size=3D3><SPAN=20
style=3D"FONT-SIZE: 12pt"><o:p>&nbsp;</o:p></SPAN></FONT></P></DIV>This =
email and=20
attachment are confidential. If you have received this message in error, =
please=20
notify the sender immediately and delete the message from your computer. =

Psytechnics Ltd has taken steps to ensure that this email is virus free =
but the=20
recipient should take necessary steps to make certain. </BODY></HTML>

------=_NextPart_000_0039_01C75B26.819B6C40--



--===============0309631818==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============0309631818==--





From sipping-bounces@ietf.org Wed Feb 28 11:14:40 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMRQn-0002yD-1z; Wed, 28 Feb 2007 11:13:29 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HMRQm-0002xr-CB
	for sipping@ietf.org; Wed, 28 Feb 2007 11:13:28 -0500
Received: from rwcrmhc12.comcast.net ([216.148.227.152])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HMRQj-0000Oy-4e
	for sipping@ietf.org; Wed, 28 Feb 2007 11:13:28 -0500
Received: from dragon.ariadne.com (failure[24.34.79.42])
	by comcast.net (rwcrmhc12) with ESMTP
	id <20070228161304m1200j9qb5e>; Wed, 28 Feb 2007 16:13:24 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1SGC8Wc025155
	for <sipping@ietf.org>; Wed, 28 Feb 2007 11:12:08 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1SGC80m025151;
	Wed, 28 Feb 2007 11:12:08 -0500
Date: Wed, 28 Feb 2007 11:12:08 -0500
Message-Id: <200702281612.l1SGC80m025151@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <CD6CE349CFD30D40BF5E13B3E0D8480402270C7A@srvxchg.cablelabs.com>
	(sumanth@cablelabs.com)
Subject: Re: [Sipping] SIPPING-CONFIG on the agenda?
References: <CD6CE349CFD30D40BF5E13B3E0D8480402270C7A@srvxchg.cablelabs.com>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 08e48e05374109708c00c6208b534009
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   From: "Sumanth Channabasappa" <sumanth@cablelabs.com>

   Just a quick check to make sure that we will be able to discuss the SIP
   Configuration Framework at IETF#68 (i.e., as an agenda item).

Yes, I do hope it's on the agenda.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 28 11:24:23 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMRb6-0002jx-RN; Wed, 28 Feb 2007 11:24:08 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HMRb5-0002fR-2K
	for sipping@ietf.org; Wed, 28 Feb 2007 11:24:07 -0500
Received: from alnrmhc12.comcast.net ([206.18.177.52])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HMRaw-0001rl-Pr
	for sipping@ietf.org; Wed, 28 Feb 2007 11:24:07 -0500
Received: from dragon.ariadne.com (dworley.hsd1.ma.comcast.net[24.34.79.42])
	by comcast.net (alnrmhc12) with ESMTP
	id <20070228162337b12009f3vje>; Wed, 28 Feb 2007 16:23:57 +0000
Received: from dragon.ariadne.com (dragon.ariadne.com [127.0.0.1])
	by dragon.ariadne.com (8.12.8/8.12.8) with ESMTP id l1SGNaWc025971
	for <sipping@ietf.org>; Wed, 28 Feb 2007 11:23:36 -0500
Received: (from worley@localhost)
	by dragon.ariadne.com (8.12.8/8.12.8/Submit) id l1SGNalY025967;
	Wed, 28 Feb 2007 11:23:36 -0500
Date: Wed, 28 Feb 2007 11:23:36 -0500
Message-Id: <200702281623.l1SGNalY025967@dragon.ariadne.com>
To: sipping@ietf.org
From: Dale.Worley@comcast.net
In-reply-to: <FC318FB3-A497-4BED-BBC6-3B5B630FC589@estacado.net>
	(bcampen@estacado.net)
Subject: Re: [Sipping] Service Identification Model
References: <61968779B8AC4C4BAB421D4C12F008C009463E2E@XCH47YKF.rim.net>
	<FC318FB3-A497-4BED-BBC6-3B5B630FC589@estacado.net>
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 2409bba43e9c8d580670fda8b695204a
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

   > 1. Basic Telephony (Conversational QoS Voice only with telephony  
   > feature set)
   > 2. Multimedia Telephony (Conversational QoS Voice plus video and other
   > media streams with telephony feature set)
   > 3. VoIP (Background QoS with no telephony features)

As Byron said, it's hard to see how 1 and 2 are different.  But 3
vs. 1 is an interesting matter -- the user can request different
grades of QoS and pay appropriately.

(My position on QoS is that in truly congested networks, QoS matters,
because different users, depending on what sort of service they
demand, put different pressures on the infrastructure, and they should
be charged accordingly.)

   > 11. Push Service 1 (e.g. Traffic information service)
   > 12. Push Service 2 (e.g. Weather information service)
   > 13. Push Service 3 (e.g  News information service)
   > 14. Push Service 4 (e.g  Stock information service)

Why would you consider these different?  Perhaps that need to be
routed to different UI elements at the very end (though 2 and 3 can
almost certainly be displayed by the same application), but all of
this can be done by tagging the destination address -- there seems to
be no reason that the *network* would distinguish them.

Dale

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 28 11:46:44 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMRwo-0007vm-PD; Wed, 28 Feb 2007 11:46:34 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HMRwm-0007vC-El
	for sipping@ietf.org; Wed, 28 Feb 2007 11:46:32 -0500
Received: from expint.psytechnics.com ([217.33.106.234])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HMRwl-0004At-5I
	for sipping@ietf.org; Wed, 28 Feb 2007 11:46:32 -0500
Content-Class: urn:content-classes:message
Importance: normal
Priority: normal
MIME-Version: 1.0
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1896
Date: Wed, 28 Feb 2007 16:47:14 -0000
Message-ID: <257B18BFF522214E92EB52F3478B014E09468C@psy01exch.intra.psytechnics.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Perceptual objective media quality reporting
thread-index: AcdbTvbY1a1DTHjhRUKksya6mZvapAAAndLg
From: "Ravi Raviraj" <ravi.raviraj@psytechnics.com>
To: <sipping@ietf.org>
X-Spam-Score: 0.4 (/)
X-Scan-Signature: 78a8240bd7a9c0d7033035fffd7b84c6
Cc: 
Subject: [Sipping] Perceptual objective media quality reporting
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1063299996=="
Errors-To: sipping-bounces@ietf.org

This is a multi-part message in MIME format.

--===============1063299996==
Content-Class: urn:content-classes:message
Content-Transfer-Encoding: 7bit
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C75B58.1966159E"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C75B58.1966159E
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

Robert,

The event package defined in "draft-ietf-sipping-rtcp-summary-01.txt"
and RTCP-XR/HR techniques are mainly for RTP/RTCP based media transport
statistics.

There are non RTP/RTCP scenarios like MPEG-TS over UDP.

An example for clarity requirement is reference point for MOS measure.
Is it acoustical, electrical or packet level?

These are two examples of work required.

=20

Ideally, we need an efficient event package with enough flexibility to
support different possibilities without any ambiguity.

=20

=20

Ravi.

=20

=20

  _____ =20

From: Robert Sparks [mailto:rjsparks@nostrum.com]=20
Sent: 28 February 2007 15:41
To: Ravi Raviraj
Cc: sipping LIST
Subject: Re: [Sipping] Perceptual objective media quality
reporting[Scanned]

=20

(Trimming this to sipping)

=20

How is what you're looking for different from
http://www.ietf.org/internet-drafts/draft-ietf-sipping-rtcp-summary-01.t
xt?

=20

RjS

=20

On Feb 28, 2007, at 9:03 AM, Ravi Raviraj wrote:





Hi all,

Over the last few years, a number of perceptual media quality objective
measures and assessment techniques have been standardized by ITU-T. For
example, MOS-LQO, MOS-CQO, etc. Each one of these parameters could have
number of variations like narrow-band/wide-band, electrical/acoustical,
full reference/no-reference, audio/video, etc. When applied within a
network end-point, each media direction (ingress/egress) could have one
or more of these values associated with it. We need a way to report
these values efficiently without any ambiguity. These MOS values could
be sent/distributed as perceptual objective quality SIP events. In this
way, all the advantages of SIP like mobility, etc. will be pulled in.

=20

I would like to propose a work/discussion item on an event package for
this purpose. Even define/specify a quality reporting framework. Is
there any interest out there?

=20

I presume that, an event package work will be within SIPPING WG, where
should we do the framework work? SIP/AVT?

=20

Thanks.

=20

Ravi.

=20

This email and attachment are confidential. If you have received this
message in error, please notify the sender immediately and delete the
message from your computer. Psytechnics Ltd has taken steps to ensure
that this email is virus free but the recipient should take necessary
steps to make certain.=20

_______________________________________________

Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping

This list is for NEW development of the application of SIP

Use sip-implementors@cs.columbia.edu for questions on current sip

Use sip@ietf.org for new developments of core SIP

=20





This email and attachment are confidential. If you have received this =
message in error, please notify the sender immediately and delete the =
message from your computer. Psytechnics Ltd has taken steps to ensure =
that this email is virus free but the recipient should take necessary =
steps to make certain.




------_=_NextPart_001_01C75B58.1966159E
Content-Type: text/html;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

<html xmlns:v=3D"urn:schemas-microsoft-com:vml" =
xmlns:o=3D"urn:schemas-microsoft-com:office:office" =
xmlns:w=3D"urn:schemas-microsoft-com:office:word" =
xmlns:st1=3D"urn:schemas-microsoft-com:office:smarttags" =
xmlns=3D"http://www.w3.org/TR/REC-html40">

<head>
<meta http-equiv=3DContent-Type content=3D"text/html; =
charset=3Dus-ascii">
<meta name=3DGenerator content=3D"Microsoft Word 11 (filtered medium)">
<!--[if !mso]>
<style>
v\:* {behavior:url(#default#VML);}
o\:* {behavior:url(#default#VML);}
w\:* {behavior:url(#default#VML);}
.shape {behavior:url(#default#VML);}
</style>
<![endif]--><o:SmartTagType
 namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags" =
name=3D"place"/>
<o:SmartTagType =
namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"
 name=3D"PersonName"/>
<!--[if !mso]>
<style>
st1\:*{behavior:url(#default#ieooui) }
</style>
<![endif]-->
<style>
<!--
 /* Font Definitions */
 @font-face
	{font-family:Tahoma;
	panose-1:2 11 6 4 3 5 4 4 2 4;}
 /* Style Definitions */
 p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0cm;
	margin-bottom:.0001pt;
	font-size:12.0pt;
	font-family:"Times New Roman";}
a:link, span.MsoHyperlink
	{color:blue;
	text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
	{color:blue;
	text-decoration:underline;}
span.EmailStyle18
	{mso-style-type:personal-reply;
	font-family:Arial;
	color:navy;}
@page Section1
	{size:612.0pt 792.0pt;
	margin:72.0pt 90.0pt 72.0pt 90.0pt;}
div.Section1
	{page:Section1;}
-->
</style>

</head>

<body lang=3DEN-US link=3Dblue vlink=3Dblue style=3D'word-wrap: =
break-word;-khtml-nbsp-mode: space;
-khtml-line-break: after-white-space'>

<div class=3DSection1>

<p class=3DMsoNormal><font size=3D2 color=3Dnavy face=3DArial><span =
style=3D'font-size:
10.0pt;font-family:Arial;color:navy'>Robert,<o:p></o:p></span></font></p>=


<p class=3DMsoNormal><font size=3D2 color=3Dnavy face=3DArial><span =
style=3D'font-size:
10.0pt;font-family:Arial;color:navy'>The event package defined in =
&#8220;draft-ietf-sipping-rtcp-summary-01.txt&#8221;
and RTCP-XR/HR techniques are mainly for RTP/RTCP based media transport
statistics.<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 color=3Dnavy face=3DArial><span =
style=3D'font-size:
10.0pt;font-family:Arial;color:navy'>There are non RTP/RTCP scenarios =
like
MPEG-TS over UDP.<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 color=3Dnavy face=3DArial><span =
style=3D'font-size:
10.0pt;font-family:Arial;color:navy'>An example for clarity requirement =
is
reference point for MOS measure. Is it acoustical, electrical or packet =
level?<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 color=3Dnavy face=3DArial><span =
style=3D'font-size:
10.0pt;font-family:Arial;color:navy'>These are two examples of work =
required.<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 color=3Dnavy face=3DArial><span =
style=3D'font-size:
10.0pt;font-family:Arial;color:navy'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 color=3Dnavy face=3DArial><span =
style=3D'font-size:
10.0pt;font-family:Arial;color:navy'>Ideally, we need an efficient event =
package
with enough flexibility to support different possibilities without any =
ambiguity.<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 color=3Dnavy face=3DArial><span =
style=3D'font-size:
10.0pt;font-family:Arial;color:navy'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 color=3Dnavy face=3DArial><span =
style=3D'font-size:
10.0pt;font-family:Arial;color:navy'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><st1:place w:st=3D"on"><font size=3D2 color=3Dnavy =
face=3DArial><span
 =
style=3D'font-size:10.0pt;font-family:Arial;color:navy'>Ravi</span></font=
></st1:place><font
size=3D2 color=3Dnavy face=3DArial><span =
style=3D'font-size:10.0pt;font-family:Arial;
color:navy'>.<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 color=3Dnavy face=3DArial><span =
style=3D'font-size:
10.0pt;font-family:Arial;color:navy'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 color=3Dnavy face=3DArial><span =
style=3D'font-size:
10.0pt;font-family:Arial;color:navy'><o:p>&nbsp;</o:p></span></font></p>

<div>

<div class=3DMsoNormal align=3Dcenter style=3D'text-align:center'><font =
size=3D3
face=3D"Times New Roman"><span style=3D'font-size:12.0pt'>

<hr size=3D2 width=3D"100%" align=3Dcenter tabindex=3D-1>

</span></font></div>

<p class=3DMsoNormal><b><font size=3D2 face=3DTahoma><span =
style=3D'font-size:10.0pt;
font-family:Tahoma;font-weight:bold'>From:</span></font></b><font =
size=3D2
face=3DTahoma><span style=3D'font-size:10.0pt;font-family:Tahoma'> =
Robert Sparks
[mailto:rjsparks@nostrum.com] <br>
<b><span style=3D'font-weight:bold'>Sent:</span></b> 28 February 2007 =
15:41<br>
<b><span style=3D'font-weight:bold'>To:</span></b> <st1:PersonName =
w:st=3D"on">Ravi
 Raviraj</st1:PersonName><br>
<b><span style=3D'font-weight:bold'>Cc:</span></b> sipping LIST<br>
<b><span style=3D'font-weight:bold'>Subject:</span></b> Re: [Sipping] =
Perceptual
objective media quality reporting[Scanned]</span></font><o:p></o:p></p>

</div>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'>(Trimming this to sipping)<o:p></o:p></span></font></p>

<div>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'><o:p>&nbsp;</o:p></span></font></p>

</div>

<div>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'>How is what you're looking for different from&nbsp;<a
href=3D"http://www.ietf.org/internet-drafts/draft-ietf-sipping-rtcp-summa=
ry-01.txt?">http://www.ietf.org/internet-drafts/draft-ietf-sipping-rtcp-s=
ummary-01.txt?</a><o:p></o:p></span></font></p>

</div>

<div>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'><o:p>&nbsp;</o:p></span></font></p>

</div>

<div>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'>RjS<o:p></o:p></span></font></p>

</div>

<div>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'><o:p>&nbsp;</o:p></span></font></p>

</div>

<div>

<div>

<div>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'>On Feb 28, 2007, at 9:03 AM, <st1:PersonName w:st=3D"on">Ravi =
Raviraj</st1:PersonName>
wrote:<o:p></o:p></span></font></p>

</div>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'><br>
<br>
<o:p></o:p></span></font></p>

<div>

<p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><font
size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:10.0pt;font-family:Arial'><O:SMARTTAGTYPE =
namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags" =
name=3D"place">Hi
all,<O:P></O:P></span></font><o:p></o:p></p>

<p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><font
size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:10.0pt;font-family:Arial'>Over
the last few years, a number of perceptual media quality objective =
measures and
assessment techniques have been standardized by ITU-T. For example, =
MOS-LQO,
MOS-CQO, etc. Each one of these parameters could have number of =
variations like
narrow-band/wide-band, electrical/acoustical, full =
reference/no-reference,
audio/video, etc. When applied within a network end-point, each media =
direction
(ingress/egress) could have one or more of these values associated with =
it. We
need a way to report these values efficiently without any ambiguity. =
These MOS
values could be sent/distributed as perceptual objective quality SIP =
events. In
this way, all the advantages of SIP like mobility, etc. will be pulled =
in.<O:P></O:P></span></font><o:p></o:p></p>

<p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><font
size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:10.0pt;font-family:Arial'><O:P>&nbsp;</O:P></span></fo=
nt><o:p></o:p></p>

<p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><font
size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:10.0pt;font-family:Arial'>I
would like to propose a work/discussion item on an event package for =
this
purpose. Even define/specify a quality reporting framework. Is there any
interest out there?<O:P></O:P></span></font><o:p></o:p></p>

<p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><font
size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:10.0pt;font-family:Arial'><O:P>&nbsp;</O:P></span></fo=
nt><o:p></o:p></p>

<p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><font
size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:10.0pt;font-family:Arial'>I
presume that, an event package work will be within SIPPING WG, where =
should we
do the framework work? SIP/AVT?<O:P></O:P></span></font><o:p></o:p></p>

<p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><font
size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:10.0pt;font-family:Arial'><O:P>&nbsp;</O:P></span></fo=
nt><o:p></o:p></p>

<p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><font
size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:10.0pt;font-family:Arial'>Thanks.<O:P></O:P></span></f=
ont><o:p></o:p></p>

<p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><font
size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:10.0pt;font-family:Arial'><O:P>&nbsp;</O:P></span></fo=
nt><o:p></o:p></p>

<p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><ST1:PLACE =
u1:st=3D"on"><st1:place
w:st=3D"on"><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:10.0pt;
 font-family:Arial'>Ravi</span></font></st1:place></ST1:PLACE><font =
size=3D2
face=3DArial><span lang=3DEN-GB =
style=3D'font-size:10.0pt;font-family:Arial'>.<O:P></O:P></span></font><o=
:p></o:p></p>

<p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><font
size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:12.0pt'><O:P>&nbsp;</O:P><o:p></o:p></span></font></p>=


</div>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'>This email and attachment are confidential. If you have received =
this
message in error, please notify the sender immediately and delete the =
message
from your computer. Psytechnics Ltd has taken steps to ensure that this =
email
is virus free but the recipient should take necessary steps to make =
certain. <o:p></o:p></span></font></p>

</O:SMARTTAGTYPE>

<div>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'>_______________________________________________<o:p></o:p></span>=
</font></p>

</div>

<div>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'>Sipping mailing list<span class=3Dapple-converted-space>&nbsp; =
</span><a
href=3D"https://www1.ietf.org/mailman/listinfo/sipping">https://www1.ietf=
.org/mailman/listinfo/sipping</a><o:p></o:p></span></font></p>

</div>

<div>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'>This list is for NEW development of the application of =
SIP<o:p></o:p></span></font></p>

</div>

<div>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'>Use <a =
href=3D"mailto:sip-implementors@cs.columbia.edu">sip-implementors@cs.colu=
mbia.edu</a>
for questions on current sip<o:p></o:p></span></font></p>

</div>

<div>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'>Use <a href=3D"mailto:sip@ietf.org">sip@ietf.org</a> for new =
developments
of core SIP<o:p></o:p></span></font></p>

</div>

</div>

<p class=3DMsoNormal><font size=3D3 face=3D"Times New Roman"><span =
style=3D'font-size:
12.0pt'><o:p>&nbsp;</o:p></span></font></p>

</div>

</div>

</body>

</html>




This email and attachment are confidential. If you have received this =
message in error, please notify the sender immediately and delete the =
message from your computer. Psytechnics Ltd has taken steps to ensure =
that this email is virus free but the recipient should take necessary =
steps to make certain.




------_=_NextPart_001_01C75B58.1966159E--


--===============1063299996==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1063299996==--




From sipping-bounces@ietf.org Wed Feb 28 12:44:41 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMSpC-0005qP-Fh; Wed, 28 Feb 2007 12:42:46 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HMSp9-0005el-Tz
	for sipping@ietf.org; Wed, 28 Feb 2007 12:42:43 -0500
Received: from sj-iport-1-in.cisco.com ([171.71.176.70]
	helo=sj-iport-1.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HMSp8-00069W-HH
	for sipping@ietf.org; Wed, 28 Feb 2007 12:42:43 -0500
Received: from rtp-dkim-2.cisco.com ([64.102.121.159])
	by sj-iport-1.cisco.com with ESMTP; 28 Feb 2007 09:42:41 -0800
X-IronPort-AV: i="4.14,231,1170662400"; 
	d="scan'208"; a="764885087:sNHT48263220"
Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13])
	by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1SHgfTB001928; 
	Wed, 28 Feb 2007 12:42:41 -0500
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by rtp-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1SHgNOi014833; 
	Wed, 28 Feb 2007 12:42:41 -0500 (EST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 28 Feb 2007 12:42:40 -0500
Received: from [192.168.1.104] ([10.86.241.21]) by xfe-rtp-202.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 28 Feb 2007 12:42:40 -0500
Message-ID: <45E5BF0F.1050506@cisco.com>
Date: Wed, 28 Feb 2007 12:42:39 -0500
From: Jonathan Rosenberg <jdrosen@cisco.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.7.8) Gecko/20050511
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: Dale.Worley@comcast.net
Subject: Re: [Sipping] Service Identification: offerless INVITE case
References: <989763567E51AB48BE015DF7BB960CAD02926415@esealmw111>	<45E44871.10002@cisco.com>
	<200702272219.l1RMJOEa019682@dragon.ariadne.com>
In-Reply-To: <200702272219.l1RMJOEa019682@dragon.ariadne.com>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 28 Feb 2007 17:42:40.0291 (UTC)
	FILETIME=[D7BDCF30:01C75B5F]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=2593; t=1172684561;
	x=1173548561; c=relaxed/simple; s=rtpdkim2001;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=jdrosen@cisco.com;
	z=From:=20Jonathan=20Rosenberg=20<jdrosen@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=3A=20offerless=2
	0INVITE=20case |Sender:=20 |To:=20Dale.Worley@comcast.net;
	bh=o94+89GECP0u+eQerOiWoA9B7GR56CLYJNpzPlTMPgo=;
	b=oSf4TttZfUW0fclRzVsxwl+tmNUzrV8NZsGtPmISMcJuOFQ+bdIs2WsSlpene6EoRgXInpnl
	yvfolGet8VujMGAerjSLqo04sTZH/fPQO7T2OFC5AD+JryLLmrPK1TSV;
Authentication-Results: rtp-dkim-2; header.From=jdrosen@cisco.com; dkim=pass (
	sig from cisco.com/rtpdkim2001 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: b4a0a5f5992e2a4954405484e7717d8c
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

The 3xx came up because the trigger for this is that the UAC sends a 
re-INVITE, say, one that adds a new media stream, and a proxy in the 
terminating side decides that a new element needs to be on the dialog path.

So, the proxy could reject the 4xx and then send a REFER, but the UAC is 
not going to know how to construct that INVITE properly. Namely, it 
won't know that this REFER-triggered INVITE is really a retry of the 
just-failed request, and should look the same more or less (besides 
being a new dialog forming request). Add another to your list of REFER 
uses, with a big missing semantic here.

So, while I agree that mid-dialog 3xx is kind of new territory (note it 
is partially treated in 
http://www.ietf.org/internet-drafts/draft-rosenberg-sip-ua-loose-route-00.txt), 
its less new territory than the use of REFER for the problem at hand. I 
think 3xx is much closer to the desired semantic.

-Jonathan R.

Dale.Worley@comcast.net wrote:

> Most of this is good, but the use of an in-dialog 3xx is a really bad
> idea for re-routing a dialog.  Primarily because it's not specified
> how such a thing is processed, and also because it has to be inserted
> in-dialog (whereas the controller of this little dance may not be in
> the signalling path).
> 
> Let me suggest some alternatives:
> 
> You can send a REFER to a UA, which will cause it to replace an
> existing dialog with a new one.  If you phrase it correctly, it
> replaces both ends:
> 
> 	REFER UA-A-gruu
> 	Target-Dialog: call-id...
> 	Refer-To: UA-B-gruu?Replaces=call-id...?Route=new-app-server
> 
> A B2BUA exploit INVITE/Join to insert itself into an existing dialog;
> 
> 	send:
> 	INVITE UA-A-gruu
> 	Join: call-id...
> 
> 	then send:
> 	INVITE UA-B-gruu
> 	Replaces: call-id...
> 
> Since these mechanisms are already defined, we know that cleaning up
> the original dialog is properly specified, etc.
> 
> Dale
> 
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP
> 

-- 
Jonathan D. Rosenberg, Ph.D.                   600 Lanidex Plaza
Cisco Fellow                                   Parsippany, NJ 07054-2711
Cisco Systems
jdrosen@cisco.com                              FAX:   (973) 952-5050
http://www.jdrosen.net                         PHONE: (973) 952-5000
http://www.cisco.com

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 28 13:38:09 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMTg0-00061o-Rd; Wed, 28 Feb 2007 13:37:20 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1HMTfz-000615-HZ
	for sipping@ietf.org; Wed, 28 Feb 2007 13:37:19 -0500
Received: from sj-iport-4.cisco.com ([171.68.10.86])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HMTfx-0004iy-57
	for sipping@ietf.org; Wed, 28 Feb 2007 13:37:19 -0500
Received: from sj-dkim-5.cisco.com ([171.68.10.79])
	by sj-iport-4.cisco.com with ESMTP; 28 Feb 2007 10:37:15 -0800
X-IronPort-AV: i="4.14,231,1170662400"; 
	d="scan'208"; a="43727493:sNHT1078496505"
Received: from sj-core-4.cisco.com (sj-core-4.cisco.com [171.68.223.138])
	by sj-dkim-5.cisco.com (8.12.11/8.12.11) with ESMTP id l1SIbEEV029956; 
	Wed, 28 Feb 2007 10:37:14 -0800
Received: from xbh-rtp-201.amer.cisco.com (xbh-rtp-201.cisco.com
	[64.102.31.12])
	by sj-core-4.cisco.com (8.12.10/8.12.6) with ESMTP id l1SIatnZ029174;
	Wed, 28 Feb 2007 10:37:14 -0800 (PST)
Received: from xfe-rtp-202.amer.cisco.com ([64.102.31.21]) by
	xbh-rtp-201.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 28 Feb 2007 13:37:04 -0500
Received: from [161.44.174.118] ([161.44.174.118]) by
	xfe-rtp-202.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 28 Feb 2007 13:37:03 -0500
Message-ID: <45E5CBD8.2070703@cisco.com>
Date: Wed, 28 Feb 2007 13:37:12 -0500
From: Paul Kyzivat <pkyzivat@cisco.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: Dale.Worley@comcast.net
Subject: Re: [Sipping] Service Identification Model
References: <61968779B8AC4C4BAB421D4C12F008C009463E2E@XCH47YKF.rim.net>	<FC318FB3-A497-4BED-BBC6-3B5B630FC589@estacado.net>
	<200702281623.l1SGNalY025967@dragon.ariadne.com>
In-Reply-To: <200702281623.l1SGNalY025967@dragon.ariadne.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 28 Feb 2007 18:37:03.0775 (UTC)
	FILETIME=[70EDF2F0:01C75B67]
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=842; t=1172687834;
	x=1173551834; c=relaxed/simple; s=sjdkim5002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=pkyzivat@cisco.com;
	z=From:=20Paul=20Kyzivat=20<pkyzivat@cisco.com>
	|Subject:=20Re=3A=20[Sipping]=20Service=20Identification=20Model
	|Sender:=20; bh=arBr/hbIbpVVgTb4hBq9AMIUAveXHk01Nu6ICoUhlM4=;
	b=cwbqEyycFR5gzeUpZGhcMYBverMQ01ude5sTgVy2tTxcYSVgfiYk25nIXU2W2KEgCi6XKVAE
	XJmF6s5nuy6Rw1r+vqr+ngdrDT5GS3Q4TMqmN5wZUl7z2mSwy0OXm7dg;
Authentication-Results: sj-dkim-5; header.From=pkyzivat@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim5002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 79899194edc4f33a41f49410777972f8
Cc: sipping@ietf.org
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org



Dale.Worley@comcast.net wrote:
>    > 1. Basic Telephony (Conversational QoS Voice only with telephony  
>    > feature set)
>    > 2. Multimedia Telephony (Conversational QoS Voice plus video and other
>    > media streams with telephony feature set)
>    > 3. VoIP (Background QoS with no telephony features)
> 
> As Byron said, it's hard to see how 1 and 2 are different.  But 3
> vs. 1 is an interesting matter -- the user can request different
> grades of QoS and pay appropriately.

Sure, QoS can matter, and differ from call to call. But calling each QoS 
option a different service is folly. QoS can be *negotiated* rather than 
just declared. And when you can have multiple media streams in the call, 
with differing QoS needs, it would be insane to assign a different 
service name to each combination.

	Paul

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 28 13:51:29 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMTtG-0003kk-7x; Wed, 28 Feb 2007 13:51:02 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMTtE-0003ka-2Z; Wed, 28 Feb 2007 13:51:00 -0500
Received: from sj-iport-3-in.cisco.com ([171.71.176.72]
	helo=sj-iport-3.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HMTtB-0006hC-Q3; Wed, 28 Feb 2007 13:51:00 -0500
Received: from sj-dkim-2.cisco.com ([171.71.179.186])
	by sj-iport-3.cisco.com with ESMTP; 28 Feb 2007 10:50:58 -0800
X-IronPort-AV: i="4.14,231,1170662400"; 
	d="scan'208"; a="467617126:sNHT43176216"
Received: from sj-core-2.cisco.com (sj-core-2.cisco.com [171.71.177.254])
	by sj-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id l1SIovXA020317; 
	Wed, 28 Feb 2007 10:50:57 -0800
Received: from [10.150.154.224] (sjc-vpn-hwcore-153.cisco.com [10.21.152.153])
	by sj-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id l1SIotGk016205;
	Wed, 28 Feb 2007 10:50:55 -0800 (PST)
Mime-Version: 1.0 (Apple Message framework v752.3)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <3EE44EC6-21E1-40B9-9FFE-B3935529CD0C@cisco.com>
Content-Transfer-Encoding: 7bit
From: Cullen Jennings <fluffy@cisco.com>
Date: Wed, 28 Feb 2007 10:50:48 -0800
To: sipping <sipping@ietf.org>, ECRIT <ecrit@ietf.org>
X-Mailer: Apple Mail (2.752.3)
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=749; t=1172688657;
	x=1173552657; c=relaxed/simple; s=sjdkim2002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=fluffy@cisco.com;
	z=From:=20Cullen=20Jennings=20<fluffy@cisco.com>
	|Subject:=20Updated=20on=20agenda=20time |Sender:=20;
	bh=jBljW3qmeqiEN/PMw4vBDvGkMOOsZmlX5kGVf0751zU=;
	b=e64+3maxCM4I0AHGW5Wmz+7u0Nroj5AgrJj3wZtNkCxjHoUqunahOztyNAgmZy1f3jqFnyq6
	7puvJ2jZsL4d+r2ggAcqa1XtSnFPJvBAhjY9vIcFgmEN+DsWzk5W2/2k;
Authentication-Results: sj-dkim-2; header.From=fluffy@cisco.com; dkim=pass (
	sig from cisco.com/sjdkim2002 verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: ffa9dfbbe7cc58b3fa6b8ae3e57b0aa3
Cc: Mary Barnes <mary.barnes@nortel.com>,
	Jon Peterson <jon.peterson@neustar.biz>,
	Gonzalo Camarillo <gonzalo.camarillo@ericsson.com>
Subject: [Sipping] Updated on agenda time
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org


Well we have another update on all this - hopefully the last one :-)

Hannes and Mary worked to figure out schedules and we think we can  
steal some time from sipping for ECRIT. So the plan will be to use  
the last hour of the Friday morning sipping session for ecrit WG  
related items.

So ecrit will have
THURSDAY 1510-1610 Afternoon Session II
and
FRIDAY 1030-1130 Morning Session I
and
it will also have an ad-hoc meeting to present IEEE material at a  
time TBD (likely lunch some day)


And sipping will have
Monday 0900-1130 Morning Session I
and
Friday 0900-1025 Morning Session I


I am working with secretary to try and figure out how to represent  
this on the agenda and I have no idea how that will end up.


_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP



From sipping-bounces@ietf.org Wed Feb 28 17:36:06 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMXOS-0001a0-TR; Wed, 28 Feb 2007 17:35:28 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMXOR-0001W5-Mt; Wed, 28 Feb 2007 17:35:27 -0500
Received: from dsl001-129-069.dfw1.dsl.speakeasy.net ([72.1.129.69]
	helo=vicuna.estacado.net) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HMXOP-0001zJ-M2; Wed, 28 Feb 2007 17:35:27 -0500
Received: from [172.17.2.60] ([172.17.2.60]) (authenticated bits=0)
	by vicuna.estacado.net (8.13.8/8.13.8) with ESMTP id l1SMZGDL019113
	(version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO);
	Wed, 28 Feb 2007 16:35:16 -0600 (CST)
	(envelope-from bcampen@estacado.net)
In-Reply-To: <OFDE647A19.552BB4D6-ON6525728F.0055428B-6525728F.00551FD6@flextronicssoftware.com>
References: <OFDE647A19.552BB4D6-ON6525728F.0055428B-6525728F.00551FD6@flextronicssoftware.com>
Mime-Version: 1.0 (Apple Message framework v752.3)
Message-Id: <0B635C81-5339-4527-9210-6A7DCC964C77@estacado.net>
From: Byron Campen <bcampen@estacado.net>
Date: Wed, 28 Feb 2007 16:35:05 -0600
To: Vishnu.Yadav@aricent.com
X-Mailer: Apple Mail (2.752.3)
X-Spam-Score: 0.3 (/)
X-Scan-Signature: 578c2c9d0cb01ffe6e1ca36540edd070
Cc: sip@ietf.org, sipping@ietf.org, kamalakar.mergu@wipro.com
Subject: [Sipping] Re: [Sip] Knowing status of sent message
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1588199258=="
Errors-To: sipping-bounces@ietf.org


--===============1588199258==
Content-Type: multipart/signed; micalg=sha1; boundary=Apple-Mail-2--1026102060;
	protocol="application/pkcs7-signature"


--Apple-Mail-2--1026102060
Content-Type: multipart/alternative;
	boundary=Apple-Mail-1--1026106800


--Apple-Mail-1--1026106800
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=WINDOWS-1252;
	delsp=yes;
	format=flowed

	Just a note; a proxy, as defined by 3261, is not allowed to do =
this. =20
This is strictly a B2BUA thing.

Best regards,
Byron Campen

>
> intermediate proxies can send 202 .
> In MESSAGE case
> 200 -> implies that the Msg has reached to destination
> 202 -> that  proxy has recvd it & will frwrded it towards final =20
> destination
>
> vishnu
>
>
> <kamalakar.mergu@wipro.com>
> 02/20/2007 03:29 PM
>
> To
> <ranjit@motorola.com>, <sipping@ietf.org>, <sip@ietf.org>
> cc
> Subject
> RE: [Sip] Knowing status of sent message
>
>
>
>
>
> Hi Ranjit,
> Every MESSAGE method will be acknowledged by 200 OK success response.
>
> That means,
> If =93A=94 sends MESSAGE to =93B=94, upon receiving this MESSAGE B =
sends =20
> 200 OK response to A.
>
> Regards
> Kamal
>
>
> From: Avasarala Ranjit-A20990 [mailto:ranjit@motorola.com]
> Sent: Tuesday, February 20, 2007 3:21 PM
> To: sipping@ietf.org; sip@ietf.org
> Subject: [Sip] Knowing status of sent message
> Importance: High
>
> Hi
>
> Is there a mechanism in SIP to know the status of sent instant =20
> message using MESSAGE?
>
> Thanks
>
> Regards
> Ranjit _______________________________________________
> Sip mailing list  https://www1.ietf.org/mailman/listinfo/sip
> This list is for NEW development of the core SIP Protocol
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sipping@ietf.org for new developments on the application of sip
>
>
>
> ***********************  Aricent-Private   ***********************
> "DISCLAIMER: This message is proprietary to Aricent and is intended =20=

> solely for the use of
> the individual to whom it is addressed. It may contain privileged =20
> or confidential information and should not be
> circulated or used for any purpose other than for what it is =20
> intended. If you have received this message in error,
> please notify the originator immediately. If you are not the =20
> intended recipient, you are notified that you are strictly
> prohibited from using, copying, altering, or disclosing the =20
> contents of this message. Aricent accepts no responsibility for
> loss or damage arising from the use of the information transmitted =20
> by this email including damage from virus."
>
> _______________________________________________
> Sip mailing list  https://www1.ietf.org/mailman/listinfo/sip
> This list is for NEW development of the core SIP Protocol
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sipping@ietf.org for new developments on the application of sip


--Apple-Mail-1--1026106800
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
	charset=WINDOWS-1252

<HTML><BODY style=3D"word-wrap: break-word; -khtml-nbsp-mode: space; =
-khtml-line-break: after-white-space; "><DIV><SPAN =
class=3D"Apple-tab-span" style=3D"white-space:pre">	</SPAN>Just a =
note; a proxy, as defined by 3261, is not allowed to do this. This is =
strictly a B2BUA thing.</DIV><DIV><BR =
class=3D"khtml-block-placeholder"></DIV><DIV>Best =
regards,</DIV><DIV>Byron Campen</DIV><DIV><BR =
class=3D"Apple-interchange-newline"><BLOCKQUOTE type=3D"cite"><BR><FONT =
size=3D"2" face=3D"sans-serif">intermediate proxies can send 202 =
.</FONT> <BR><FONT size=3D"2" face=3D"sans-serif">In MESSAGE case =
</FONT> <BR><FONT size=3D"2" face=3D"sans-serif">200 -&gt; implies that =
the Msg has reached to destination</FONT> <BR><FONT size=3D"2" =
face=3D"sans-serif">202 -&gt; that =A0proxy has recvd it &amp; will =
frwrded it towards final destination</FONT> <BR> <BR><FONT size=3D"2" =
face=3D"sans-serif">vishnu</FONT> <BR> <BR> <BR> <TABLE width=3D"100%"> =
<TBODY><TR valign=3D"top"><TD width=3D"40%"><FONT size=3D"1" =
face=3D"sans-serif"><B>&lt;<A =
href=3D"mailto:kamalakar.mergu@wipro.com">kamalakar.mergu@wipro.com</A>&gt=
;</B> </FONT><P><FONT size=3D"1" face=3D"sans-serif">02/20/2007 03:29 =
PM</FONT> <BR> </P></TD><TD width=3D"59%"> <TABLE width=3D"100%"> =
<TBODY><TR><TD> <DIV align=3D"right"><FONT size=3D"1" =
face=3D"sans-serif">To</FONT></DIV> </TD><TD valign=3D"top"><FONT =
size=3D"1" face=3D"sans-serif">&lt;<A =
href=3D"mailto:ranjit@motorola.com">ranjit@motorola.com</A>&gt;, &lt;<A =
href=3D"mailto:sipping@ietf.org">sipping@ietf.org</A>&gt;, &lt;<A =
href=3D"mailto:sip@ietf.org">sip@ietf.org</A>&gt;</FONT> =
</TD></TR><TR><TD> <DIV align=3D"right"><FONT size=3D"1" =
face=3D"sans-serif">cc</FONT></DIV> </TD><TD valign=3D"top"> =
</TD></TR><TR><TD> <DIV align=3D"right"><FONT size=3D"1" =
face=3D"sans-serif">Subject</FONT></DIV> </TD><TD valign=3D"top"><FONT =
size=3D"1" face=3D"sans-serif">RE: [Sip] Knowing status of sent =
message</FONT></TD></TR></TBODY></TABLE> <BR> <TABLE> <TBODY><TR =
valign=3D"top"><TD> </TD><TD></TD></TR></TBODY></TABLE> =
<BR></TD></TR></TBODY></TABLE> <BR> <BR> <BR><FONT size=3D"2" =
color=3D"blue" face=3D"Trebuchet MS">Hi Ranjit,</FONT> <BR><FONT =
size=3D"2" color=3D"blue" face=3D"Trebuchet MS">Every MESSAGE method =
will be acknowledged by 200 OK success response.</FONT> <BR><FONT =
size=3D"2" color=3D"blue" face=3D"Trebuchet MS">=A0</FONT> <BR><FONT =
size=3D"2" color=3D"blue" face=3D"Trebuchet MS">That means, </FONT> =
<BR><FONT size=3D"2" color=3D"blue" face=3D"Trebuchet MS">If =93A=94 =
sends MESSAGE to =93B=94, upon receiving this MESSAGE B sends 200 OK =
response to A.</FONT> <BR><FONT size=3D"2" color=3D"blue" =
face=3D"Trebuchet MS">=A0</FONT> <BR><FONT size=3D"2" color=3D"blue" =
face=3D"Trebuchet MS">Regards</FONT> <BR><FONT size=3D"2" color=3D"blue" =
face=3D"Trebuchet MS">Kamal</FONT> <DIV align=3D"center"> <BR> =
<HR></DIV> <BR><FONT size=3D"2" face=3D"Tahoma"><B>From:</B> Avasarala =
Ranjit-A20990 [<A =
href=3D"mailto:ranjit@motorola.com">mailto:ranjit@motorola.com</A>] =
<B><BR> Sent:</B> Tuesday, February 20, 2007 3:21 PM<B><BR> To:</B> <A =
href=3D"mailto:sipping@ietf.org">sipping@ietf.org</A>; <A =
href=3D"mailto:sip@ietf.org">sip@ietf.org</A><B><BR> Subject:</B> [Sip] =
Knowing status of sent message<B><BR> Importance:</B> High</FONT> =
<BR><FONT size=3D"3" face=3D"Times New Roman">=A0</FONT><P><FONT =
size=3D"3" color=3D"blue" face=3D"Times New Roman">Hi</FONT><FONT =
size=3D"3" face=3D"Times New Roman"> </FONT> </P><P><FONT size=3D"3" =
color=3D"blue" face=3D"Times New Roman">Is there a mechanism in SIP to =
know the status of sent instant message using MESSAGE?</FONT><FONT =
size=3D"3" face=3D"Times New Roman"> </FONT> </P><P><FONT size=3D"3" =
color=3D"blue" face=3D"Times New Roman">Thanks</FONT><FONT size=3D"3" =
face=3D"Times New Roman"> </FONT> </P><P><FONT size=3D"3" color=3D"blue" =
face=3D"Monotype Corsiva">Regards</FONT><FONT size=3D"3" face=3D"Times =
New Roman"> </FONT><FONT size=3D"3" color=3D"blue" face=3D"Monotype =
Corsiva"><BR> Ranjit</FONT><FONT size=3D"3" face=3D"Times New Roman"> =
</FONT><TT><FONT =
size=3D"2">_______________________________________________<BR> Sip =
mailing list =A0<A =
href=3D"https://www1.ietf.org/mailman/listinfo/sip">https://www1.ietf.org/=
mailman/listinfo/sip</A><BR> This list is for NEW development of the =
core SIP Protocol<BR> Use <A =
href=3D"mailto:sip-implementors@cs.columbia.edu">sip-implementors@cs.colum=
bia.edu</A> for questions on current sip<BR> Use <A =
href=3D"mailto:sipping@ietf.org">sipping@ietf.org</A> for new =
developments on the application of sip</FONT></TT> </P><P><FONT size=3D"2"=
 face=3D"sans-serif"><BR> <BR> *********************** =A0Aricent-Private =
=A0 ***********************</FONT> <TABLE><TBODY><TR><TD =
bgcolor=3D"#ffffff"><FONT color=3D"#000000"><PRE>"DISCLAIMER: This =
message is proprietary to Aricent and is intended solely for the use of=20=

the individual to whom it is addressed. It may contain privileged or =
confidential information and should not be=20
circulated or used for any purpose other than for what it is intended. =
If you have received this message in error,=20
please notify the originator immediately. If you are not the intended =
recipient, you are notified that you are strictly
prohibited from using, copying, altering, or disclosing the contents of =
this message. Aricent accepts no responsibility for=20
loss or damage arising from the use of the information transmitted by =
this email including damage from virus."
</PRE></FONT></TD></TR></TBODY></TABLE></P><DIV style=3D"margin-top: =
0px; margin-right: 0px; margin-bottom: 0px; margin-left: 0px; =
">_______________________________________________</DIV><DIV =
style=3D"margin-top: 0px; margin-right: 0px; margin-bottom: 0px; =
margin-left: 0px; ">Sip mailing list<SPAN =
class=3D"Apple-converted-space">=A0 </SPAN><A =
href=3D"https://www1.ietf.org/mailman/listinfo/sip">https://www1.ietf.org/=
mailman/listinfo/sip</A></DIV><DIV style=3D"margin-top: 0px; =
margin-right: 0px; margin-bottom: 0px; margin-left: 0px; ">This list is =
for NEW development of the core SIP Protocol</DIV><DIV =
style=3D"margin-top: 0px; margin-right: 0px; margin-bottom: 0px; =
margin-left: 0px; ">Use <A =
href=3D"mailto:sip-implementors@cs.columbia.edu">sip-implementors@cs.colum=
bia.edu</A> for questions on current sip</DIV><DIV style=3D"margin-top: =
0px; margin-right: 0px; margin-bottom: 0px; margin-left: 0px; ">Use <A =
href=3D"mailto:sipping@ietf.org">sipping@ietf.org</A> for new =
developments on the application of sip</DIV> =
</BLOCKQUOTE></DIV><BR></BODY></HTML>=

--Apple-Mail-1--1026106800--

--Apple-Mail-2--1026102060
Content-Transfer-Encoding: base64
Content-Type: application/pkcs7-signature;
	name=smime.p7s
Content-Disposition: attachment;
	filename=smime.p7s

MIAGCSqGSIb3DQEHAqCAMIACAQExCzAJBgUrDgMCGgUAMIAGCSqGSIb3DQEHAQAAoIIGKTCCAuIw
ggJLoAMCAQICEAda7Ce+sudhzrOd9CS5wJQwDQYJKoZIhvcNAQEFBQAwYjELMAkGA1UEBhMCWkEx
JTAjBgNVBAoTHFRoYXd0ZSBDb25zdWx0aW5nIChQdHkpIEx0ZC4xLDAqBgNVBAMTI1RoYXd0ZSBQ
ZXJzb25hbCBGcmVlbWFpbCBJc3N1aW5nIENBMB4XDTA2MDkyNTE0NDUzNloXDTA3MDkyNTE0NDUz
NlowRjEfMB0GA1UEAxMWVGhhd3RlIEZyZWVtYWlsIE1lbWJlcjEjMCEGCSqGSIb3DQEJARYUYmNh
bXBlbkBlc3RhY2Fkby5uZXQwggEiMA0GCSqGSIb3DQEBAQUAA4IBDwAwggEKAoIBAQDWxjwykU0M
R1yTjA3db2wzExKsMhLa/mc2ekFkBP9BVXi4OHacKSZM3a9T0bWvPcw97ycBfDZX3q2prMRsPUep
svJWA0wbiWyOtLzuQGWaQQ6rYbvYlRxil/teAJb1fAKdfFXIXXsiqcdHkhbFn/J7oSn1BCz6v9/v
M1zgGNeMPIR34bo9A0hW+QZJxwkSg4KZHdaQ50WcKXtdHdBg46s+0bvDqNmmt7eP91UJk9aiyxyw
ZdWVNxzF1oj+nfsnsaHv1py6l62Je82mHsvIZwdSAic1PtXVpU8QmkLJHPmkBEVtN/uHMQ95xgxi
eFO5hbYuAT9f7V0ZGe7pUJBR3KwdAgMBAAGjMTAvMB8GA1UdEQQYMBaBFGJjYW1wZW5AZXN0YWNh
ZG8ubmV0MAwGA1UdEwEB/wQCMAAwDQYJKoZIhvcNAQEFBQADgYEAeJC75lz4KrtNdPOovb3fkqzD
i2HXuOtYHCTOPMc5xJGLdqSF5sc8bwd4mIfevi9Ji4Ccj85A6qINQM/v4OjAvxPkwRwXey/YN6kl
19Te3Eh/t9l4zo2wPdbHTipyc8YJ0GAtT4/fwWiboW3+aquk8DGc7ityr9cdAYY1jAud7d0wggM/
MIICqKADAgECAgENMA0GCSqGSIb3DQEBBQUAMIHRMQswCQYDVQQGEwJaQTEVMBMGA1UECBMMV2Vz
dGVybiBDYXBlMRIwEAYDVQQHEwlDYXBlIFRvd24xGjAYBgNVBAoTEVRoYXd0ZSBDb25zdWx0aW5n
MSgwJgYDVQQLEx9DZXJ0aWZpY2F0aW9uIFNlcnZpY2VzIERpdmlzaW9uMSQwIgYDVQQDExtUaGF3
dGUgUGVyc29uYWwgRnJlZW1haWwgQ0ExKzApBgkqhkiG9w0BCQEWHHBlcnNvbmFsLWZyZWVtYWls
QHRoYXd0ZS5jb20wHhcNMDMwNzE3MDAwMDAwWhcNMTMwNzE2MjM1OTU5WjBiMQswCQYDVQQGEwJa
QTElMCMGA1UEChMcVGhhd3RlIENvbnN1bHRpbmcgKFB0eSkgTHRkLjEsMCoGA1UEAxMjVGhhd3Rl
IFBlcnNvbmFsIEZyZWVtYWlsIElzc3VpbmcgQ0EwgZ8wDQYJKoZIhvcNAQEBBQADgY0AMIGJAoGB
AMSmPFVzVftOucqZWh5owHUEcJ3f6f+jHuy9zfVb8hp2vX8MOmHyv1HOAdTlUAow1wJjWiyJFXCO
3cnwK4Vaqj9xVsuvPAsH5/EfkTYkKhPPK9Xzgnc9A74r/rsYPge/QIACZNenprufZdHFKlSFD0gE
f6e20TxhBEAeZBlyYLf7AgMBAAGjgZQwgZEwEgYDVR0TAQH/BAgwBgEB/wIBADBDBgNVHR8EPDA6
MDigNqA0hjJodHRwOi8vY3JsLnRoYXd0ZS5jb20vVGhhd3RlUGVyc29uYWxGcmVlbWFpbENBLmNy
bDALBgNVHQ8EBAMCAQYwKQYDVR0RBCIwIKQeMBwxGjAYBgNVBAMTEVByaXZhdGVMYWJlbDItMTM4
MA0GCSqGSIb3DQEBBQUAA4GBAEiM0VCD6gsuzA2jZqxnD3+vrL7CF6FDlpSdf0whuPg2H6otnzYv
wPQcUCCTcDz9reFhYsPZOhl+hLGZGwDFGguCdJ4lUJRix9sncVcljd2pnDmOjCBPZV+V2vf3h9bG
CE6u9uo05RAaWzVNd+NWIXiC3CEZNd4ksdMdRv9dX2VPMYIDEDCCAwwCAQEwdjBiMQswCQYDVQQG
EwJaQTElMCMGA1UEChMcVGhhd3RlIENvbnN1bHRpbmcgKFB0eSkgTHRkLjEsMCoGA1UEAxMjVGhh
d3RlIFBlcnNvbmFsIEZyZWVtYWlsIElzc3VpbmcgQ0ECEAda7Ce+sudhzrOd9CS5wJQwCQYFKw4D
AhoFAKCCAW8wGAYJKoZIhvcNAQkDMQsGCSqGSIb3DQEHATAcBgkqhkiG9w0BCQUxDxcNMDcwMjI4
MjIzNTA1WjAjBgkqhkiG9w0BCQQxFgQUiJiNRO6/fj8mCpRzHDpK6B9QrT8wgYUGCSsGAQQBgjcQ
BDF4MHYwYjELMAkGA1UEBhMCWkExJTAjBgNVBAoTHFRoYXd0ZSBDb25zdWx0aW5nIChQdHkpIEx0
ZC4xLDAqBgNVBAMTI1RoYXd0ZSBQZXJzb25hbCBGcmVlbWFpbCBJc3N1aW5nIENBAhAHWuwnvrLn
Yc6znfQkucCUMIGHBgsqhkiG9w0BCRACCzF4oHYwYjELMAkGA1UEBhMCWkExJTAjBgNVBAoTHFRo
YXd0ZSBDb25zdWx0aW5nIChQdHkpIEx0ZC4xLDAqBgNVBAMTI1RoYXd0ZSBQZXJzb25hbCBGcmVl
bWFpbCBJc3N1aW5nIENBAhAHWuwnvrLnYc6znfQkucCUMA0GCSqGSIb3DQEBAQUABIIBAAtMSogg
ZzBiuNpoGwhN0MpNg61f9iYbthsopf+cWZpGFZYeiWtJR+/sEby/SbhmIrOiVHSKedQqe2ruw4Ba
86KT8hG36xdjSdS6a+CKt/IQDfgca3dbucjmlTknGu1zRV8Ye9SK3O5YIEz82+rN5S2mdCt/IS0h
iQiSgTScVSvwj90+mIHAiIHZwm/kEXPuwDaKpaZRHJXuPaRdSDZVfEKfIZd+5tq1zhANezbrO23B
bGUgZyUuj4M6aiYgHRsT17Y0T/mKLv4KtbHQlCSKGJ3X0mDb6FDoqnQ7uiNhlQVIgLMqCd2ysHfU
vZaqYtS/r2TYaycNUAgg3JpxKCTMBuQAAAAAAAA=

--Apple-Mail-2--1026102060--


--===============1588199258==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--===============1588199258==--




From sipping-bounces@ietf.org Wed Feb 28 18:51:22 2007
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMYZ0-00019m-NA; Wed, 28 Feb 2007 18:50:26 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1HMYYe-000152-AC; Wed, 28 Feb 2007 18:50:04 -0500
Received: from ns0.neustar.com ([156.154.16.158])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1HMYYd-0002iQ-Fo; Wed, 28 Feb 2007 18:50:04 -0500
Received: from stiedprstage1.ietf.org (stiedprstage1.va.neustar.com
	[10.31.47.10]) by ns0.neustar.com (Postfix) with ESMTP id 47D0A3296B;
	Wed, 28 Feb 2007 23:50:03 +0000 (GMT)
Received: from ietf by stiedprstage1.ietf.org with local (Exim 4.43)
	id 1HMYYd-000455-4O; Wed, 28 Feb 2007 18:50:03 -0500
Content-Type: Multipart/Mixed; Boundary="NextPart"
Mime-Version: 1.0
To: i-d-announce@ietf.org
From: Internet-Drafts@ietf.org
Message-Id: <E1HMYYd-000455-4O@stiedprstage1.ietf.org>
Date: Wed, 28 Feb 2007 18:50:03 -0500
X-Spam-Score: -2.5 (--)
X-Scan-Signature: f66b12316365a3fe519e75911daf28a8
Cc: sipping@ietf.org
Subject: [Sipping] I-D ACTION:draft-ietf-sipping-spam-04.txt 
X-BeenThere: sipping@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SIPPING Working Group \(applications of SIP\)" <sipping.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:sipping@ietf.org>
List-Help: <mailto:sipping-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/sipping>,
	<mailto:sipping-request@ietf.org?subject=subscribe>
Errors-To: sipping-bounces@ietf.org

--NextPart

A New Internet-Draft is available from the on-line Internet-Drafts 
directories.
This draft is a work item of the Session Initiation Proposal Investigation Working Group of the IETF.

	Title		: The Session Initiation Protocol (SIP) and Spam
	Author(s)	: C. Jennings, J. Rosenberg
	Filename	: draft-ietf-sipping-spam-04.txt
	Pages		: 28
	Date		: 2007-2-28
	
Spam, defined as the transmission of bulk unsolicited messages, has
   plagued Internet email.  Unfortunately, spam is not limited to email.
   It can affect any system that enables user to user communications.
   The Session Initiation Protocol (SIP) defines a system for user to
   user multimedia communications.  Therefore, it is susceptible to
   spam, just as email is.  In this document, we analyze the problem of
   spam in SIP.  We first identify the ways in which the problem is the
   same and the ways in which it is different from email.  We then
   examine the various possible solutions that have been discussed for
   email and consider their applicability to SIP.

A URL for this Internet-Draft is:
http://www.ietf.org/internet-drafts/draft-ietf-sipping-spam-04.txt

To remove yourself from the I-D Announcement list, send a message to 
i-d-announce-request@ietf.org with the word unsubscribe in the body of 
the message. 
You can also visit https://www1.ietf.org/mailman/listinfo/I-D-announce 
to change your subscription settings.

Internet-Drafts are also available by anonymous FTP. Login with the 
username "anonymous" and a password of your e-mail address. After 
logging in, type "cd internet-drafts" and then 
"get draft-ietf-sipping-spam-04.txt".

A list of Internet-Drafts directories can be found in
http://www.ietf.org/shadow.html 
or ftp://ftp.ietf.org/ietf/1shadow-sites.txt

Internet-Drafts can also be obtained by e-mail.

Send a message to:
	mailserv@ietf.org.
In the body type:
	"FILE /internet-drafts/draft-ietf-sipping-spam-04.txt".
	
NOTE:	The mail server at ietf.org can return the document in
	MIME-encoded form by using the "mpack" utility.  To use this
	feature, insert the command "ENCODING mime" before the "FILE"
	command.  To decode the response(s), you will need "munpack" or
	a MIME-compliant mail reader.  Different MIME-compliant mail readers
	exhibit different behavior, especially when dealing with
	"multipart" MIME messages (i.e. documents which have been split
	up into multiple messages), so check your local documentation on
	how to manipulate these messages.

Below is the data which will enable a MIME compliant mail reader
implementation to automatically retrieve the ASCII version of the
Internet-Draft.

--NextPart
Content-Type: Multipart/Alternative; Boundary="OtherAccess"

--OtherAccess
Content-Type: Message/External-body; access-type="mail-server";
	server="mailserv@ietf.org"

Content-Type: text/plain
Content-ID: <2007-2-28163108.I-D@ietf.org>

ENCODING mime
FILE /internet-drafts/draft-ietf-sipping-spam-04.txt

--OtherAccess
Content-Type: Message/External-body; name="draft-ietf-sipping-spam-04.txt";
	site="ftp.ietf.org"; access-type="anon-ftp";
	directory="internet-drafts"

Content-Type: text/plain
Content-ID: <2007-2-28163108.I-D@ietf.org>


--OtherAccess--

--NextPart
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP
--NextPart--





