
From tyoshino@google.com  Mon Jan  6 23:36:48 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 91D2C1AE477 for <hybi@ietfa.amsl.com>; Mon,  6 Jan 2014 23:36:48 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 9cSI7ypCbs-T for <hybi@ietfa.amsl.com>; Mon,  6 Jan 2014 23:36:46 -0800 (PST)
Received: from mail-we0-x22a.google.com (mail-we0-x22a.google.com [IPv6:2a00:1450:400c:c03::22a]) by ietfa.amsl.com (Postfix) with ESMTP id 693101AE46C for <hybi@ietf.org>; Mon,  6 Jan 2014 23:36:46 -0800 (PST)
Received: by mail-we0-f170.google.com with SMTP id w61so16869249wes.1 for <hybi@ietf.org>; Mon, 06 Jan 2014 23:36:37 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:from:date:message-id:subject:to:content-type; bh=QFu88o0Lrq2LKpJxkOiSlvO+q/YUglMLA0OK4/Wo36Y=; b=jHmyexQq+OpBkuddTeqZSV5/TZ5OiXAjQsy7+vCuMzo3o5nccq2DYuMWDqHADKPeYD ukx5ISuc/CoUtPASjnnNq0itaX6h+dEfxl1l4chKLfC+Q+GzcCM7jJzev4M7885kYHEd 2+KPr5oYeOmkSfODvdXsH/JQLS6ceQTCWonmCXnnHm/Rs+ILHicj9+ZqpYw5XJuKnHer xI4dbM2CHO1yDow+vby0oHip8Fv9oMBXRb8yvMIl3bGUd7CEagR8Yakh121P+DgxnkVZ 91VB4vmUr3mzJvaA5RTn7Y1rUMDd8rLZgE/KhQigjZ8gdBMIREuFYm6LIRMae6Dfb8eU 5Ctg==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:from:date:message-id:subject:to :content-type; bh=QFu88o0Lrq2LKpJxkOiSlvO+q/YUglMLA0OK4/Wo36Y=; b=Mq6F9eNLCW9XkEXycUM8tO5XiQJlyXU9S3k080cNytSMPcaXqRwxgJPOl84nryZ0lG Yk6L+aYrDa5D6sfmhvRPRSSuIE/WaKL7TUfmDGdWZ3NNMmzTCZo0o649jVHFyiS5uioA jRCgMWyyzmgapbo3sXwV8Ilxos/Q6pw8odbfEjGlpk9/zmG6MJxC56xRCswktLllSnvf rZ8Ov0qcLadFiF8EyrKmlhGqzLEzux+yb3+PMcOXHOtFvnrf5hx2p3kuRNFDJE92w1wV 4TqC/pqsZluGb3vxNNiTOmQtFSnVpaDrm02hAspeFcsI1ivmSDfZ8DVoO1kRUDuFedW9 jpdA==
X-Gm-Message-State: ALoCoQmi09ZLU8gbjd1PLlv73jbPN1he+tzhHTSXNm7ElDnquqopPmMvsIdCEAFHNSf36bV6/0nRdJ230q+koO8/YZmLQLXHQnT28NbcfZW+e9RhGeclsYcGF4YllFPq7r/eWbDdlgg92RZqT4yk7f5CQUtIsxxmGEHnHenDlLcNW381lMxnqYFmCWJ2uxzMuFobEddGWIq9
X-Received: by 10.194.78.179 with SMTP id c19mr3674021wjx.84.1389080197072; Mon, 06 Jan 2014 23:36:37 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Mon, 6 Jan 2014 23:36:17 -0800 (PST)
From: Takeshi Yoshino <tyoshino@google.com>
Date: Tue, 7 Jan 2014 16:36:17 +0900
Message-ID: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com>
To: "hybi@ietf.org" <hybi@ietf.org>
Content-Type: multipart/alternative; boundary=047d7bfcf6a2b6134204ef5c6f0e
Subject: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 07 Jan 2014 07:36:48 -0000

--047d7bfcf6a2b6134204ef5c6f0e
Content-Type: text/plain; charset=UTF-8

Hi all,

We've been working on mux extension to make WebSocket scalable for long
time, but as SPDY got widely supported and adopted quickly, now there seems
to be no longer any good reason to continue this work which is WebSocket
dedicated.

Scalability is still important issue to address. From now, I'd like to
focus on WS/SPDY effort which we've been working on in parallel with the
mux extension.

Please take a look at the latest proposal in this post.
https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc

If you really want the dedicated mux extension than one over SPDY, sorry
but please take over the work.

--047d7bfcf6a2b6134204ef5c6f0e
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Hi all,
<div><br></div><div>We&#39;ve been working on mux extension to make WebSock=
et scalable for long time, but as SPDY got widely supported and adopted qui=
ckly, now there seems to be no longer any good reason to continue this work=
 which is WebSocket dedicated.</div>

<div><br></div><div>Scalability is still important issue to address. From n=
ow, I&#39;d like to focus on WS/SPDY effort which we&#39;ve been working on=
 in parallel with the mux extension.</div><div><br></div><div>Please take a=
 look at the latest proposal in this post.</div>

<div><a href=3D"https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3gl=
c">https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc</a><br></di=
v><div><br></div><div>If you really want the dedicated mux extension than o=
ne over SPDY, sorry but please take over the work.</div>

</div>

--047d7bfcf6a2b6134204ef5c6f0e--

From duerst@it.aoyama.ac.jp  Tue Jan  7 00:11:12 2014
Return-Path: <duerst@it.aoyama.ac.jp>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 372D71ADFCD for <hybi@ietfa.amsl.com>; Tue,  7 Jan 2014 00:11:12 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.929
X-Spam-Level: 
X-Spam-Status: No, score=-1.929 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HELO_EQ_JP=1.244, HOST_EQ_JP=1.265, MIME_8BIT_HEADER=0.3, RCVD_IN_DNSWL_MED=-2.3, RP_MATCHES_RCVD=-0.538] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id JmANWC1cQALA for <hybi@ietfa.amsl.com>; Tue,  7 Jan 2014 00:11:09 -0800 (PST)
Received: from scintmta02.scbb.aoyama.ac.jp (scintmta02.scbb.aoyama.ac.jp [133.2.253.34]) by ietfa.amsl.com (Postfix) with ESMTP id 99E011AD7BE for <hybi@ietf.org>; Tue,  7 Jan 2014 00:11:08 -0800 (PST)
Received: from scmse02.scbb.aoyama.ac.jp ([133.2.253.231]) by scintmta02.scbb.aoyama.ac.jp (secret/secret) with SMTP id s078Audu029707; Tue, 7 Jan 2014 17:10:56 +0900
Received: from (unknown [133.2.206.134]) by scmse02.scbb.aoyama.ac.jp with smtp id 04f2_a998_3bf2d626_7773_11e3_b16d_001e6722eec2; Tue, 07 Jan 2014 17:10:55 +0900
Received: from [IPv6:::1] (unknown [133.2.210.1]) by itmail2.it.aoyama.ac.jp (Postfix) with ESMTP id 8738CBFBC2; Tue,  7 Jan 2014 17:10:55 +0900 (JST)
Message-ID: <52CBB683.7090107@it.aoyama.ac.jp>
Date: Tue, 07 Jan 2014 17:10:43 +0900
From: =?UTF-8?B?Ik1hcnRpbiBKLiBEw7xyc3Qi?= <duerst@it.aoyama.ac.jp>
Organization: Aoyama Gakuin University
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.1.9) Gecko/20100722 Eudora/3.0.4
MIME-Version: 1.0
To: Takeshi Yoshino <tyoshino@google.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com>
In-Reply-To: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 07 Jan 2014 08:11:12 -0000

Hello Takeshi,

On 2014/01/07 16:36, Takeshi Yoshino wrote:
> Hi all,
>
> We've been working on mux extension to make WebSocket scalable for long
> time, but as SPDY got widely supported and adopted quickly, now there seems
> to be no longer any good reason to continue this work which is WebSocket
> dedicated.
>
> Scalability is still important issue to address. From now, I'd like to
> focus on WS/SPDY effort which we've been working on in parallel with the
> mux extension.

I think not having a WebSocket-specific mux extension is the right way 
to go. But in the end, I'm much more interested in WS/HTTP2 than in 
WS/SPDY. Do you have a plan with respect to WS/HTTP2?

Regards,   Martin.

> Please take a look at the latest proposal in this post.
> https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc
>
> If you really want the dedicated mux extension than one over SPDY, sorry
> but please take over the work.

From tyoshino@google.com  Tue Jan  7 00:28:18 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 2BEF51ADFF7 for <hybi@ietfa.amsl.com>; Tue,  7 Jan 2014 00:28:18 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id P9wlvGReLoUj for <hybi@ietfa.amsl.com>; Tue,  7 Jan 2014 00:28:16 -0800 (PST)
Received: from mail-we0-x229.google.com (mail-we0-x229.google.com [IPv6:2a00:1450:400c:c03::229]) by ietfa.amsl.com (Postfix) with ESMTP id EF5AD1AD687 for <hybi@ietf.org>; Tue,  7 Jan 2014 00:28:15 -0800 (PST)
Received: by mail-we0-f169.google.com with SMTP id w61so16960551wes.14 for <hybi@ietf.org>; Tue, 07 Jan 2014 00:28:06 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=Gca7NxGrHQzwgCjhJ2S9VWC6jU+jFFfmOODF5UFWfeY=; b=eb3aeRcqalT8hTKhWB8NIrxFBrxq0+6f+XlNmHjP8rDkgMhgVq6pA/L+QApkvtlBwh tEoYVGefPWYzxRZJab7cjVb623+5P9T6htQBt1Rdw97h93mmXuid25bZS5yVLRxAWrb3 xX9Nyb1X1FyxmTSEMuS3JIOEbvwiAcobzuZBDMHH26Tx1bea4KxDjNpUkJmB3rms5dJK FOmezqSVNheJwriI6KARbPYeX28JnE1+7G6WfuOP9drVp4sINiYDXw+qXQBuPlL/EfxE TQ4IxSZMy6MJWFNHq8b3Wm1HX5zYsTCx2HnSwZySSWUz147ypBFlRzXkdmjwWwNfMlbG rqaA==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=Gca7NxGrHQzwgCjhJ2S9VWC6jU+jFFfmOODF5UFWfeY=; b=fIfHN8+qChRYA8AHOVaN8VMc7/0mIS0T3noXHsrZjRvVGl6DfQGdRoFPUjZLyrOiQP aNLbDLrFOJ7bkS8sZqVmXLlDbAkpn5XoLXUImXaY8GDy80iLCfEcF+YC/MVnaned14oM +p+0P+jt/PEDMuiUVWxGjgPBxFszBBtSavVCqWQsABnmAaDXxjW4qFbRXuy089pfO2Rt 2vdDUX78/sqw+MfiRGOF1L1DVfYDV4GBu7oFT5FDsl6knmeKMkALRrTohqu5ad88Sxg4 EtEtZKxhsj8WENbkp+57/52lAx9A+z5O7sxFxSsgpEoL8poLFYHtgBxEe/ekZiMG4IRB f+KA==
X-Gm-Message-State: ALoCoQmozVM69D7lGX232BefbBjTb4I3m8ou4ezK8cir6gUCSBUmDN03SnKRQ9ZuYtlilJh1Xpgn+j33MGJ0KQOxxbvAYNr55g7K9Y0vd96a3BiskkUFRwl3juZqjhnzNBG+B95cvc6gveyvgabLWd4oFU+Jt17CPonLkLNlBZExe5CCjiDItnv84utqvDF8T2xoPgLtXWHG
X-Received: by 10.180.212.98 with SMTP id nj2mr15765708wic.52.1389083286538; Tue, 07 Jan 2014 00:28:06 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Tue, 7 Jan 2014 00:27:46 -0800 (PST)
In-Reply-To: <CABihn6EYJ5BbMsCHsatmHbLxdqEfR7CBRGU4tEHZHUbQD1z-Jg@mail.gmail.com>
References: <CABihn6ErtW-ewspJ9LAVU-kpf9iJuPhbap+0KjSqfZFgWCy+nw@mail.gmail.com> <CAG4zZZBtjyFXMWthh3XwDaOBxc94GTBsFjXqS7=y-_1+913FKA@mail.gmail.com> <CABihn6EYJ5BbMsCHsatmHbLxdqEfR7CBRGU4tEHZHUbQD1z-Jg@mail.gmail.com>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Tue, 7 Jan 2014 17:27:46 +0900
Message-ID: <CAH9hSJZQGtr9jn81erVdO2pj+Q_GmzBVAEmuHqrC6uONPZsN3g@mail.gmail.com>
To: Yutaka Hirano <yhirano@chromium.org>
Content-Type: multipart/alternative; boundary=001a11c223dedb9ddb04ef5d2702
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] was_clean and _Fail the WebSocket connection_
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 07 Jan 2014 08:28:18 -0000

--001a11c223dedb9ddb04ef5d2702
Content-Type: text/plain; charset=UTF-8

Thanks, Joakim for comments with the concrete cases listed.

As Jetty is doing, it's fine not to send any close frame for these cases as
the peer is broken.

The point we'd like to clarify is whether to dispatch an error event or
not. IIRC, there is no reason to require closing only for this case. So, I
think it's just a bug in the spec that clients are required not to fail but
close on receiving a masked frame from the server.

----

"Closed _cleanly_" is defined in [1]. Completion of closing handshake is
not explicitly defined in any normative section, but it should mean that
the endpoint has received a valid close frame and has sent a close frame.
[2] says that unexpected TCP closure also requires run of fail algorithm.
So, "unclean => (failed || buffer full on send())" stands, I think.

[1] http://tools.ietf.org/html/rfc6455#section-7.1.4
[2] http://tools.ietf.org/html/rfc6455#section-7.2.1

--001a11c223dedb9ddb04ef5d2702
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra">Thanks, Joakim for comments wit=
h the concrete cases listed.</div><div class=3D"gmail_extra"><br></div><div=
 class=3D"gmail_extra">As Jetty is doing, it&#39;s fine not to send any clo=
se frame for these cases as the peer is broken.<br>

</div><div class=3D"gmail_extra"><br></div><div class=3D"gmail_extra">The p=
oint we&#39;d like to clarify is whether to dispatch an error event or not.=
 IIRC, there is no reason to require closing only for this case. So, I thin=
k it&#39;s just a bug in the spec that clients are required not to fail but=
 close on receiving a masked frame from the server.</div>

<div class=3D"gmail_extra"><br></div><div class=3D"gmail_extra">----</div><=
div class=3D"gmail_extra"><br></div><div class=3D"gmail_extra">&quot;Closed=
 _cleanly_&quot; is defined in [1]. Completion of closing handshake is not =
explicitly defined in any normative section, but it should mean that the en=
dpoint has received a valid close frame and has sent a close frame. [2] say=
s that unexpected TCP closure also requires run of fail algorithm. So, &quo=
t;unclean =3D&gt; (failed || buffer full on send())&quot; stands, I think.<=
/div>

<div class=3D"gmail_extra"><br></div><div class=3D"gmail_extra">[1] <a href=
=3D"http://tools.ietf.org/html/rfc6455#section-7.1.4">http://tools.ietf.org=
/html/rfc6455#section-7.1.4</a></div><div class=3D"gmail_extra">[2]=C2=A0<a=
 href=3D"http://tools.ietf.org/html/rfc6455#section-7.2.1">http://tools.iet=
f.org/html/rfc6455#section-7.2.1</a></div>

</div>

--001a11c223dedb9ddb04ef5d2702--

From tyoshino@google.com  Tue Jan  7 00:37:59 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 070861AE4AD for <hybi@ietfa.amsl.com>; Tue,  7 Jan 2014 00:37:59 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.616
X-Spam-Level: 
X-Spam-Status: No, score=-1.616 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, MIME_8BIT_HEADER=0.3, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 8Mi0IwFfX8f7 for <hybi@ietfa.amsl.com>; Tue,  7 Jan 2014 00:37:57 -0800 (PST)
Received: from mail-wi0-x232.google.com (mail-wi0-x232.google.com [IPv6:2a00:1450:400c:c05::232]) by ietfa.amsl.com (Postfix) with ESMTP id 206611AE48D for <hybi@ietf.org>; Tue,  7 Jan 2014 00:37:56 -0800 (PST)
Received: by mail-wi0-f178.google.com with SMTP id bz8so381946wib.17 for <hybi@ietf.org>; Tue, 07 Jan 2014 00:37:48 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=FrgxpxTaJNscZaytEE1QPccweAu6ZoajyMBU3g5X76c=; b=Oxmgnj/2/SNJbjYbDhlfm+ihZc6JtyCgLMuF5nU06ye6+jT4IH/dN2nUhJ0bDlQJnh C156Me3hNW9XS4NLdDA5OnGhMNaH3NNVB0EBA3ItY4TR38kEOrzGalXiMFrkPPlIybgm hemVUHG3NMArYnYh0+8i3eSMiHysU8ZJXw2E58vB/IrUBRnEFdrEeZNpbEqe4gw7JY7p CIl7q5mzevJrMmKDa7lxaZ/wZDmCqODtVruAJzPIf0zjcB+grlMHOKjyxQKR4onobazZ 8hfjfElWcWMmpwhpcv0IaPiDmASFMobeQLfYz9t362jCI/qhDQ/21DjBho7P5/aUSyjo mHeQ==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=FrgxpxTaJNscZaytEE1QPccweAu6ZoajyMBU3g5X76c=; b=ZJ65njusb1ooEv1PPPBOaTCIGXDUXxJcsmSC905hvGUGmmXmUwy3f+08zy+RX5e4xa c0bNUU3Be5NjRkcijEEH1cG+kY9WFliN6eieMxfGCWEQNPSnb//UFj7np+8dO5c9Ygrt PQOFBTqFrTSLZZwAgc90XNJE4wHLhrnnQOvwEpEpqt0xW5Xg4jzhT6fRG70P4g8Xm0sP 1WMI+YOOeS0CJjwtRVVp2RnZYvCQkGi4aZXeXpkth84qJd1tnpg3FenPx6w7pimTzDsx AOzAmGpKv4cf7N8/qN1g0uIMD/1RGL6rJComeppu2PNhyfqCIh6ZG4DtuDwl2NzH0eg2 xkIQ==
X-Gm-Message-State: ALoCoQmA93uKYP0fMg2jlCJzU6/PVS23B+4iktIftoXA8WeswL9/gP4Ovn42t3voYakf3uTdmHsT3hPi+5yRv1Mb3x1AXqspQ3XMWJdcOc4MBzX+dK4k683+Wpufl6xohbMVrOuh8gjAfhU17UhyjJ0cFlTP7DPIUgqLKLIAeksVyDoHnnBb6VgL6Mztsbm1KaU8JBeFQoLZ
X-Received: by 10.180.211.49 with SMTP id mz17mr4770485wic.30.1389083867969; Tue, 07 Jan 2014 00:37:47 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Tue, 7 Jan 2014 00:37:27 -0800 (PST)
In-Reply-To: <52CBB683.7090107@it.aoyama.ac.jp>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <52CBB683.7090107@it.aoyama.ac.jp>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Tue, 7 Jan 2014 17:37:27 +0900
Message-ID: <CAH9hSJb-BHEGn280pRDcnzUe+qGvs8_bwur5dct90Pey5OBHDw@mail.gmail.com>
To: =?UTF-8?Q?Martin_J=2E_D=C3=BCrst?= <duerst@it.aoyama.ac.jp>
Content-Type: multipart/alternative; boundary=001a11c25f8c8391ef04ef5d4aab
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 07 Jan 2014 08:37:59 -0000

--001a11c25f8c8391ef04ef5d4aab
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Hi Martin,

On Tue, Jan 7, 2014 at 5:10 PM, "Martin J. D=C3=BCrst" <duerst@it.aoyama.ac=
.jp>wrote:

> Hello Takeshi,
>
>
> On 2014/01/07 16:36, Takeshi Yoshino wrote:
>
>> Hi all,
>>
>> We've been working on mux extension to make WebSocket scalable for long
>> time, but as SPDY got widely supported and adopted quickly, now there
>> seems
>> to be no longer any good reason to continue this work which is WebSocket
>> dedicated.
>>
>> Scalability is still important issue to address. From now, I'd like to
>> focus on WS/SPDY effort which we've been working on in parallel with the
>> mux extension.
>>
>
> I think not having a WebSocket-specific mux extension is the right way to
> go.


If there's real need to continue the work, I'll support the work. But I'd
like to ask someone else to take over or co-edit the spec.


> But in the end, I'm much more interested in WS/HTTP2 than in WS/SPDY. Do
> you have a plan with respect to WS/HTTP2?
>
>
Yes. We'd like to standardize WS/HTTP/2.0 eventually. It depends on how
quickly HTTP/2.0 is standardized and adopted and how SPDY is going to
evolve, which protocol is stable, ready and good for running experiment,
etc. We plan to coordinate with httpbis and SPDY team to figure out how to
drive standardization and implementation.


> Regards,   Martin.
>
>
>  Please take a look at the latest proposal in this post.
>> https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc
>>
>> If you really want the dedicated mux extension than one over SPDY, sorry
>> but please take over the work.
>>
>

--001a11c25f8c8391ef04ef5d4aab
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra">Hi Martin,<br><br><div class=3D=
"gmail_quote">On Tue, Jan 7, 2014 at 5:10 PM, &quot;Martin J. D=C3=BCrst&qu=
ot; <span dir=3D"ltr">&lt;<a href=3D"mailto:duerst@it.aoyama.ac.jp" target=
=3D"_blank">duerst@it.aoyama.ac.jp</a>&gt;</span> wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">Hello Takeshi,<div class=3D"im"><br>
<br>
On 2014/01/07 16:36, Takeshi Yoshino wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">
Hi all,<br>
<br>
We&#39;ve been working on mux extension to make WebSocket scalable for long=
<br>
time, but as SPDY got widely supported and adopted quickly, now there seems=
<br>
to be no longer any good reason to continue this work which is WebSocket<br=
>
dedicated.<br>
<br>
Scalability is still important issue to address. From now, I&#39;d like to<=
br>
focus on WS/SPDY effort which we&#39;ve been working on in parallel with th=
e<br>
mux extension.<br>
</blockquote>
<br></div>
I think not having a WebSocket-specific mux extension is the right way to g=
o.</blockquote><div><br></div><div>If there&#39;s real need to continue the=
 work, I&#39;ll support the work. But I&#39;d like to ask someone else to t=
ake over or co-edit the spec.</div>

<div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8=
ex;border-left:1px #ccc solid;padding-left:1ex"> But in the end, I&#39;m mu=
ch more interested in WS/HTTP2 than in WS/SPDY. Do you have a plan with res=
pect to WS/HTTP2?<br>


<br></blockquote><div><br></div><div>Yes. We&#39;d like to standardize WS/H=
TTP/2.0 eventually. It depends on how quickly HTTP/2.0 is standardized and =
adopted and how SPDY is going to evolve, which protocol is stable, ready an=
d good for running experiment, etc. We plan to coordinate with httpbis and =
SPDY team to figure out how to drive standardization and implementation.</d=
iv>

<div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8=
ex;border-left:1px #ccc solid;padding-left:1ex">
Regards, =C2=A0 Martin.<div class=3D"HOEnZb"><div class=3D"h5"><br>
<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">
Please take a look at the latest proposal in this post.<br>
<a href=3D"https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc" ta=
rget=3D"_blank">https://groups.google.com/<u></u>forum/#!topic/spdy-dev/<u>=
</u>aJut2ax3glc</a><br>
<br>
If you really want the dedicated mux extension than one over SPDY, sorry<br=
>
but please take over the work.<br>
</blockquote>
</div></div></blockquote></div><br></div></div>

--001a11c25f8c8391ef04ef5d4aab--

From tobias.oberstein@tavendo.de  Thu Jan  9 01:11:22 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 6ECA81AE063 for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 01:11:22 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.001
X-Spam-Level: 
X-Spam-Status: No, score=-0.001 tagged_above=-999 required=5 tests=[BAYES_20=-0.001, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id nd3E9OLteGoI for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 01:11:21 -0800 (PST)
Received: from EXHUB020-5.exch020.serverdata.net (exhub020-5.exch020.serverdata.net [206.225.164.32]) by ietfa.amsl.com (Postfix) with ESMTP id 0BBB91ADF5A for <hybi@ietf.org>; Thu,  9 Jan 2014 01:11:21 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-5.exch020.serverdata.net ([206.225.164.32]) with mapi; Thu, 9 Jan 2014 01:11:11 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Takeshi Yoshino <tyoshino@google.com>, "hybi@ietf.org" <hybi@ietf.org>
Date: Thu, 9 Jan 2014 01:11:06 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8LezfCoM4V94vlQsauRUb5fEzOMwBnpxXQ
Message-ID: <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com>
In-Reply-To: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 09 Jan 2014 09:11:22 -0000
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From yhirano@google.com  Thu Jan  9 01:36:50 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 07FAD1AE068 for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 01:36:50 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id B2xJ-n5XdvUa for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 01:36:46 -0800 (PST)
Received: from mail-qc0-x22e.google.com (mail-qc0-x22e.google.com [IPv6:2607:f8b0:400d:c01::22e]) by ietfa.amsl.com (Postfix) with ESMTP id 87E5B1AC7F1 for <hybi@ietf.org>; Thu,  9 Jan 2014 01:36:46 -0800 (PST)
Received: by mail-qc0-f174.google.com with SMTP id x13so1410385qcv.19 for <hybi@ietf.org>; Thu, 09 Jan 2014 01:36:36 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=rDg2z9Tpgu8LjQ90rpsekCMg3g7KFPjxaIJ6F6KJ9e8=; b=EFUNqfC2eAEYRq/20pHgw/nwXRyZJ7r6kSBie6vSKoM5B8gHWEx4eR3EzzEv2WKSGf H2JMIYI1spNPmoqdckFvvkpgPToTWdPhgmkhT6JrMgtXVi122D7FaB8SHpmsyqTIXmjA a810vAhvx8koDY8J7Dpc/AKZHZs7W5Sw8qC/yYsQkRRYh43zPUPpEAC+Wurky1JdCd+m uem9qAXQx7tGBXU4C9qmVf7+10hftyvYVWHnL1VhavRIyDMyGVD2dzwbbAWTwh+LhrV4 0SkIwhTgco80Z+Z4mPlNZ9zNDzcGnPuyQhaWUEi78DPnmZhkvF7adqk1lo0/dYHMjGOt T2RA==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=rDg2z9Tpgu8LjQ90rpsekCMg3g7KFPjxaIJ6F6KJ9e8=; b=msgciKnkD4m51wnrObIQpHHKIaQE66CU/oUBMQTaItJfB6VhmVYdFuaMNgrcmv/MmT hYLdmhHT410Da/I/tFwPmaoW11RtRLX0BQx0A31OLB9H5euKDhnw8FmthEvcat7PMOvM e2VhsFWwMKb3/t5ENZHxVFmZ7Q4z6RiVykDy8=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=rDg2z9Tpgu8LjQ90rpsekCMg3g7KFPjxaIJ6F6KJ9e8=; b=APFuc2y+SGjTrtHlt88WAwFBxS4rJ4qMnPB2NL10kRsZDL3jP7k8c3xXuhYXbxVeW8 Qrc5ZfhbLAHfi4sVMXIcDxLzZ40LYUjG545HaMxQC2v0s0DG5dEbzoW+qHYdNSUITkdG 4ukAy6wRjMni8/bZlpgVb/QkRWeZI4TXASr6/oxwrHLwL86gxJFwQ27wvBHlH8pqeUnN DZOlpLqL0NnqlbTYDb9Yd0KRvk5y6qzE6FyJfn8GLCBDSzRi5oCEfeXxP/ITQohwdLFS ggPnuwDteobii+tJnLM0VU4SYunk+pFpRnRO6X1kBHD6Ztcsq+XBVIFVKieIq5/9NofP AaUg==
X-Gm-Message-State: ALoCoQnEqlQfzCa9QokX3cAXKGd/uA0CL4xaOqXhfY+mDNYxVMaHXfND5TIwAaXAScWzz6ZANk40tHDol3aLQcEXkRNvuXC/eo/ffqV0hRMDEizPRPgoFYwNn3YPmf2zNCZFbGY1BsDc27FAckk5IvuyWswQuNsQeUYYqse865pFLI77MVg+t2gIOdLRnSx2GuyNJY7O5v/d
MIME-Version: 1.0
X-Received: by 10.49.63.164 with SMTP id h4mr5362968qes.62.1389260196830; Thu, 09 Jan 2014 01:36:36 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.140.109.202 with HTTP; Thu, 9 Jan 2014 01:36:36 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net>
Date: Thu, 9 Jan 2014 18:36:36 +0900
X-Google-Sender-Auth: fz8s8uKdEO9xf9IUEn5nksISWZY
Message-ID: <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=047d7bd75c9488760c04ef86580d
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 09 Jan 2014 09:36:50 -0000

--047d7bd75c9488760c04ef86580d
Content-Type: text/plain; charset=UTF-8

Hi Tobias,
Thank you for the comments.

1) Frame Mapping
> """
> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
> following plural SPDY data frames. The SPDY HEADERS frame must contain
> WebSocket frame fields and following SPDY data frames contain payload data.
> These data frames are free to be reframed.
> """
>
> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
> So reframing of WebSocket frames (the payload) into multiple SPDY data
> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
> is free to choose the boundaries within WS frame payload at which to
> reframe into different SPDY data frames as it likes?

A WebSocket frame MAY be split to multiple SPDY DATA frames.
Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA frame
into multiple DATA frames.
Is this answering your question?


> 2) End of WS Frame
> The end of a WS message is signaled with a SPDY data frame having MSG_DONE
> set.
> How is the end of a WS frame signaled, given that the WS frame may have
> been reframed into multiple SPDY data frames?

When you receive a HEADERS frame, you can recognize that the last SPDY
frame is that last frame in a WebSocket frame.
When you receive a SPDY frame with MSG_DONE, you can recognize that it is
the last SPDY frame in a WebSocket message and hence the last SPDY frame in
a WebSocket frame.


> 3) WS extension
> Are WS extensions fully supported? In particular, is "permessage-deflate"
> supported when running WS/SPDY?

Yes. We have reserved bits and we can emulate the opening handshake, I
think.

On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> Hi Takeshi,
>
> it's a pity, but I can understand the motivation to shove everything into
> HTTP2.
>
> Anyway, some questions:
>
> 1) Frame Mapping
> """
> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
> following plural SPDY data frames. The SPDY HEADERS frame must contain
> WebSocket frame fields and following SPDY data frames contain payload data.
> These data frames are free to be reframed.
> """
>
> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
>
> So reframing of WebSocket frames (the payload) into multiple SPDY data
> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
> is free to choose the boundaries within WS frame payload at which to
> reframe into different SPDY data frames as it likes?
>
> 2) End of WS Frame
>
> The end of a WS message is signaled with a SPDY data frame having MSG_DONE
> set.
>
> How is the end of a WS frame signaled, given that the WS frame may have
> been reframed into multiple SPDY data frames?
>
> 3) WS extension
>
> Are WS extensions fully supported? In particular, is "permessage-deflate"
> supported when running WS/SPDY?
>
> Thanks!
> /Tobias
>
>
> Von: hybi [mailto:hybi-bounces@ietf.org] Im Auftrag von Takeshi Yoshino
> Gesendet: Dienstag, 7. Januar 2014 08:36
> An: hybi@ietf.org
> Betreff: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
> Hi all,
>
> We've been working on mux extension to make WebSocket scalable for long
> time, but as SPDY got widely supported and adopted quickly, now there seems
> to be no longer any good reason to continue this work which is WebSocket
> dedicated.
>
> Scalability is still important issue to address. From now, I'd like to
> focus on WS/SPDY effort which we've been working on in parallel with the
> mux extension.
>
> Please take a look at the latest proposal in this post.
> https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc
>
> If you really want the dedicated mux extension than one over SPDY, sorry
> but please take over the work.
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>

--047d7bd75c9488760c04ef86580d
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Hi Tobias,<div>Thank you for the comments.<br></div><div c=
lass=3D"gmail_extra"><br><blockquote class=3D"gmail_quote" style=3D"margin:=
0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);=
border-left-style:solid;padding-left:1ex">
<span style=3D"font-family:arial,sans-serif;font-size:13px">1) Frame Mappin=
g<br></span><span style=3D"font-family:arial,sans-serif;font-size:13px">&qu=
ot;&quot;&quot;<br></span><span style=3D"font-family:arial,sans-serif;font-=
size:13px">One WebSocket frame is mapped into one preceding SPDY HEADERS fr=
ame and following plural SPDY data frames. The SPDY HEADERS frame must cont=
ain WebSocket frame fields and following SPDY data frames contain payload d=
ata. These data frames are free to be reframed.<br>
</span><span style=3D"font-family:arial,sans-serif;font-size:13px">&quot;&q=
uot;&quot;<br></span><a href=3D"https://docs.google.com/document/d/124vB5NV=
1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank" style=3D"font-=
family:arial,sans-serif;font-size:13px">https://docs.google.com/document/d/=
124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit</a><br style=3D"font-fami=
ly:arial,sans-serif;font-size:13px">
<span style=3D"font-family:arial,sans-serif;font-size:13px">So reframing of=
 WebSocket frames (the payload) into multiple SPDY data frames MAY occur, b=
ut does not need to occur? And a WS/SPDY implementation is free to choose t=
he boundaries within WS frame payload at which to reframe into different SP=
DY data frames as it likes?</span></blockquote>
<div>A WebSocket frame MAY be split to multiple SPDY DATA frames.</div><div=
>Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA frame =
into multiple DATA frames.</div><div>Is this answering your question?</div>
<div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px =
0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-l=
eft-style:solid;padding-left:1ex"><span style=3D"font-family:arial,sans-ser=
if;font-size:13px">2) End of WS Frame</span><br style=3D"font-family:arial,=
sans-serif;font-size:13px">
<span style=3D"font-family:arial,sans-serif;font-size:13px">The end of a WS=
 message is signaled with a SPDY data frame having MSG_DONE set.</span><br =
style=3D"font-family:arial,sans-serif;font-size:13px"><span style=3D"font-f=
amily:arial,sans-serif;font-size:13px">How is the end of a WS frame signale=
d, given that the WS frame may have been reframed into multiple SPDY data f=
rames?</span></blockquote>
<div>When you receive a HEADERS frame, you can recognize that the last SPDY=
 frame is that last frame in a WebSocket frame.</div><div>When you receive =
a SPDY frame with MSG_DONE, you can recognize that it is the last SPDY fram=
e in a WebSocket message and hence the last SPDY frame in a WebSocket frame=
.</div>
<div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px =
0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-l=
eft-style:solid;padding-left:1ex"><span style=3D"font-family:arial,sans-ser=
if;font-size:13px">3) WS extension</span><br style=3D"font-family:arial,san=
s-serif;font-size:13px">
<span style=3D"font-family:arial,sans-serif;font-size:13px">Are WS extensio=
ns fully supported? In particular, is &quot;permessage-deflate&quot; suppor=
ted when running WS/SPDY?</span></blockquote><div>Yes. We have reserved bit=
s and we can emulate the opening handshake, I think.</div>
<div class=3D"gmail_quote"><br></div><div class=3D"gmail_quote">On Thu, Jan=
 9, 2014 at 6:11 PM, Tobias Oberstein <span dir=3D"ltr">&lt;<a href=3D"mail=
to:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstein@tavendo.=
de</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-=
left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;p=
adding-left:1ex">Hi Takeshi,<br>
<br>
it&#39;s a pity, but I can understand the motivation to shove everything in=
to HTTP2.<br>
<br>
Anyway, some questions:<br>
<br>
1) Frame Mapping<br>
&quot;&quot;&quot;<br>
One WebSocket frame is mapped into one preceding SPDY HEADERS frame and fol=
lowing plural SPDY data frames. The SPDY HEADERS frame must contain WebSock=
et frame fields and following SPDY data frames contain payload data. These =
data frames are free to be reframed.<br>

&quot;&quot;&quot;<br>
<a href=3D"https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmC=
fc7ukw9aWzSHijk/edit" target=3D"_blank">https://docs.google.com/document/d/=
124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit</a><br>
<br>
So reframing of WebSocket frames (the payload) into multiple SPDY data fram=
es MAY occur, but does not need to occur? And a WS/SPDY implementation is f=
ree to choose the boundaries within WS frame payload at which to reframe in=
to different SPDY data frames as it likes?<br>

<br>
2) End of WS Frame<br>
<br>
The end of a WS message is signaled with a SPDY data frame having MSG_DONE =
set.<br>
<br>
How is the end of a WS frame signaled, given that the WS frame may have bee=
n reframed into multiple SPDY data frames?<br>
<br>
3) WS extension<br>
<br>
Are WS extensions fully supported? In particular, is &quot;permessage-defla=
te&quot; supported when running WS/SPDY?<br>
<br>
Thanks!<br>
/Tobias<br>
<br>
<br>
Von: hybi [mailto:<a href=3D"mailto:hybi-bounces@ietf.org">hybi-bounces@iet=
f.org</a>] Im Auftrag von Takeshi Yoshino<br>
Gesendet: Dienstag, 7. Januar 2014 08:36<br>
An: <a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br>
Betreff: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2=
.0<br>
<div class=3D""><div class=3D"h5"><br>
Hi all,<br>
<br>
We&#39;ve been working on mux extension to make WebSocket scalable for long=
 time, but as SPDY got widely supported and adopted quickly, now there seem=
s to be no longer any good reason to continue this work which is WebSocket =
dedicated.<br>

<br>
Scalability is still important issue to address. From now, I&#39;d like to =
focus on WS/SPDY effort which we&#39;ve been working on in parallel with th=
e mux extension.<br>
<br>
Please take a look at the latest proposal in this post.<br>
<a href=3D"https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc" ta=
rget=3D"_blank">https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3gl=
c</a><br>
<br>
If you really want the dedicated mux extension than one over SPDY, sorry bu=
t please take over the work.<br>
</div></div><div class=3D""><div class=3D"h5">_____________________________=
__________________<br>
hybi mailing list<br>
<a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br>
<a href=3D"https://www.ietf.org/mailman/listinfo/hybi" target=3D"_blank">ht=
tps://www.ietf.org/mailman/listinfo/hybi</a><br>
</div></div></blockquote></div><br></div></div>

--047d7bd75c9488760c04ef86580d--

From tobias.oberstein@tavendo.de  Thu Jan  9 01:45:57 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 399151AE18C for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 01:45:57 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.899
X-Spam-Level: 
X-Spam-Status: No, score=-1.899 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ysbYgNDKJhiE for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 01:45:55 -0800 (PST)
Received: from EXHUB020-5.exch020.serverdata.net (exhub020-5.exch020.serverdata.net [206.225.164.32]) by ietfa.amsl.com (Postfix) with ESMTP id A09901AE175 for <hybi@ietf.org>; Thu,  9 Jan 2014 01:45:54 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-5.exch020.serverdata.net ([206.225.164.32]) with mapi; Thu, 9 Jan 2014 01:45:44 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Takeshi Yoshino <tyoshino@google.com>, "hybi@ietf.org" <hybi@ietf.org>
Date: Thu, 9 Jan 2014 01:45:41 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8LezfCoM4V94vlQsauRUb5fEzOMwBomXsA
Message-ID: <634914A010D0B943A035D226786325D4446BE403E0@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com>
In-Reply-To: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: multipart/alternative; boundary="_000_634914A010D0B943A035D226786325D4446BE403E0EXVMBX02012ex_"
MIME-Version: 1.0
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 09 Jan 2014 09:45:57 -0000

--_000_634914A010D0B943A035D226786325D4446BE403E0EXVMBX02012ex_
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_634914A010D0B943A035D226786325D4446BE403E0EXVMBX02012ex_
Content-Type: text/html; charset="utf-8"
Content-Transfer-Encoding: base64
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==

--_000_634914A010D0B943A035D226786325D4446BE403E0EXVMBX02012ex_--

From tobias.oberstein@tavendo.de  Thu Jan  9 01:54:05 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 0A2571AE1E9 for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 01:54:05 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.899
X-Spam-Level: 
X-Spam-Status: No, score=-1.899 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id M-O_YrMpqphF for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 01:54:02 -0800 (PST)
Received: from EXHUB020-1.exch020.serverdata.net (exhub020-1.exch020.serverdata.net [206.225.164.28]) by ietfa.amsl.com (Postfix) with ESMTP id 5CE7C1AE0F5 for <hybi@ietf.org>; Thu,  9 Jan 2014 01:54:02 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-1.exch020.serverdata.net ([206.225.164.28]) with mapi; Thu, 9 Jan 2014 01:53:52 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Yutaka Hirano <yhirano@chromium.org>
Date: Thu, 9 Jan 2014 01:53:49 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8NHkuvlpbbcTKaTyq/+T5Rzp4kOQAAYW9A
Message-ID: <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com>
In-Reply-To: <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: multipart/alternative; boundary="_000_634914A010D0B943A035D226786325D4446BE403E1EXVMBX02012ex_"
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 09 Jan 2014 09:54:05 -0000

--_000_634914A010D0B943A035D226786325D4446BE403E1EXVMBX02012ex_
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_634914A010D0B943A035D226786325D4446BE403E1EXVMBX02012ex_
Content-Type: text/html; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_634914A010D0B943A035D226786325D4446BE403E1EXVMBX02012ex_--

From duerst@it.aoyama.ac.jp  Thu Jan  9 03:52:22 2014
Return-Path: <duerst@it.aoyama.ac.jp>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 2B9F21AE2A0 for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 03:52:22 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.929
X-Spam-Level: 
X-Spam-Status: No, score=-1.929 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HELO_EQ_JP=1.244, HOST_EQ_JP=1.265, MIME_8BIT_HEADER=0.3, RCVD_IN_DNSWL_MED=-2.3, RP_MATCHES_RCVD=-0.538] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Fy8JlGOaYHWp for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 03:52:20 -0800 (PST)
Received: from scintmta02.scbb.aoyama.ac.jp (scintmta02.scbb.aoyama.ac.jp [133.2.253.34]) by ietfa.amsl.com (Postfix) with ESMTP id 252BB1AE29B for <hybi@ietf.org>; Thu,  9 Jan 2014 03:52:20 -0800 (PST)
Received: from scmse02.scbb.aoyama.ac.jp ([133.2.253.231]) by scintmta02.scbb.aoyama.ac.jp (secret/secret) with SMTP id s09Bq8QO004280 for <hybi@ietf.org>; Thu, 9 Jan 2014 20:52:08 +0900
Received: from (unknown [133.2.206.134]) by scmse02.scbb.aoyama.ac.jp with smtp id 2b1e_5bf7_77dd7c74_7924_11e3_82cf_001e6722eec2; Thu, 09 Jan 2014 20:52:07 +0900
Received: from [IPv6:::1] (unknown [133.2.210.1]) by itmail2.it.aoyama.ac.jp (Postfix) with ESMTP id 0054ABF548 for <hybi@ietf.org>; Thu,  9 Jan 2014 20:52:07 +0900 (JST)
Message-ID: <52CE8D59.3070802@it.aoyama.ac.jp>
Date: Thu, 09 Jan 2014 20:51:53 +0900
From: =?UTF-8?B?Ik1hcnRpbiBKLiBEw7xyc3Qi?= <duerst@it.aoyama.ac.jp>
Organization: Aoyama Gakuin University
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.1.9) Gecko/20100722 Eudora/3.0.4
MIME-Version: 1.0
To: "hybi@ietf.org" <hybi@ietf.org>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E0@EXVMBX020-12.exch020.serverdata.net>
In-Reply-To: <634914A010D0B943A035D226786325D4446BE403E0@EXVMBX020-12.exch020.serverdata.net>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Subject: [hybi] Detection of broken WS connections
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 09 Jan 2014 11:52:22 -0000

This is a problem that I have already encountered in three different 
applications, and think there must/should be a well-known solution for. 
It's not strictly standards-related, but I'm sending it to this lists 
because I think that's where the experts are.

It's about detecting broken connections on the application level. 
Setting up a WS connection and using it while it's working is very easy. 
However, what is supposed to happen when e.g. the server dies or the 
cable gets cut (directly or figuratively), or the PC is put to sleep for 
a while, and then wakes up again?

The WS API is explicit about the fact that there is no way from 
JavaScript to generate PINGs or check for PONGs. So do I just create my 
own, application level equivalent, and send a ping-like message every so 
often and assume the connection is dead when I don't get a reply 
"soonish"? Are there libraries that support such a mechanism? Or is 
every browser supported to regularly PINGPONG with the server, and call 
the onclose function when there's no answer?

Any advice or pointers appreciated!

Regards,   Martin.

From tyoshino@google.com  Thu Jan  9 04:00:09 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 968AE1AE25B for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 04:00:09 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id YrhhK2tdgYLD for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 04:00:07 -0800 (PST)
Received: from mail-wg0-x230.google.com (mail-wg0-x230.google.com [IPv6:2a00:1450:400c:c00::230]) by ietfa.amsl.com (Postfix) with ESMTP id B87F01AE1D7 for <hybi@ietf.org>; Thu,  9 Jan 2014 04:00:06 -0800 (PST)
Received: by mail-wg0-f48.google.com with SMTP id x13so2622727wgg.3 for <hybi@ietf.org>; Thu, 09 Jan 2014 03:59:56 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=YOhAZ3vIqv8PLK8+RFzUinknEfJfupOKZjXFuKgz6b8=; b=k3tWs2wW6/DzAP/2KI0M6fRiGrgk/x+4hf09Y6rYC5uqMbfY56zpDEGbf4RiCJurjq GUys25JNgUaXFLo38yY7IpH/EkaR3JEWDOwMfwJCljk7tLYmeiM9S77VRyIGin0BFxOF cJLZZuHJS6f2/QLacopd5oTju1irKDryyYckLL9gM96CWQc+tMFoe0pyxfh2v8tKAciP uGuVlfns2reMscWS7M6TaPLaiWC+A62ln03cxi99Yti2a22wjVc/6818I3tNHXpt3uaC O3YCSCo668LpLqbK7deDpddfUOSkyIELBluvdlPyt7M7799nODBSFAnHTWAuLA/XFcNJ 8Olg==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=YOhAZ3vIqv8PLK8+RFzUinknEfJfupOKZjXFuKgz6b8=; b=RenaahaBTJx9KesNscTiUXydN8Peja322S645TAJHo4CEO7UTMnP1q5UgFaMk21B7/ YB+EMbYemhze0gqRBTZrV54PYA3DRI9BkOQgbB6RLYZM+QlWSDru+k+k8B4nsx/eR7Wl 3jbllcjIzJJfV67iWmgbnuUCzttmX398IPZJ6hFKW4saoWX0WLbF2dGEFtguF/Mm7nN0 XCMQh84vDTKXiBNv6eSXgtmXduazdHxDr+K7ptlD01NCysFZ3UkgF8VC9y30H1xsfI0u f0E7yjtTtM3NivcjxeWosJmDbTV9XqyuLVM5qWwisSgThtW9ZrwV+9SK+hdPshKSM2Df B2nQ==
X-Gm-Message-State: ALoCoQn05O76P8225ck+uWVZ0HXD1hKxBZVbH3j5r+EpuSxN05fGNHEEWIy28+YWqhCwWLAFbxHKwAke3iy/hPVW/hg/2vadLJBPWlGMQ5VQ8fHc0arqOsEA7XgXpr+Br2O3e1rgiAwrrg702nh/bAQJDKKGX4IUtWMrqhTrtxZXwHl/iBGQnK2dWZpqy29TFdDQd/v6B2uf
X-Received: by 10.194.175.133 with SMTP id ca5mr2540479wjc.19.1389268796797; Thu, 09 Jan 2014 03:59:56 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Thu, 9 Jan 2014 03:59:36 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BE403E0@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E0@EXVMBX020-12.exch020.serverdata.net>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Thu, 9 Jan 2014 20:59:36 +0900
Message-ID: <CAH9hSJZfBMEL1zWx+Eh9+rTJpTYV=q3Be0ZeW-WQQ7NXszharA@mail.gmail.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=089e01493c5a21980a04ef885939
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 09 Jan 2014 12:00:09 -0000

--089e01493c5a21980a04ef885939
Content-Type: text/plain; charset=UTF-8

On Thu, Jan 9, 2014 at 6:45 PM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> Two more thoughts:
>
>
>
> 1) WS/SPDY
>
>
>
> WS essentially provides a reliable, ordered, bidirectional/full-duplex
> message based channel.
>
>
>
> WS/SPDY is layering WS over SPDY.
>
>
>
> Are there efforts to implement a WebSocket-like semantics (like above)
> _directly_ using SPDY facilities?
>
>
>

I don't know for now. As we already have WS API, there's no reason to
re-invent something else and ignore WS API.


>
>
> 2) WS Mux
>
>
>
> As it stands (as far as I understand), SPDY MUX and the WS/MUX both
> roughly target the same semantics:
>
>
>
> multiplex different channels over one underlying transport
>
>
>
> The channels are independent. There is no ordering implied across data
> sent over _different_ logical channels.
>
>
>
> In the case of WS/SPDY, that means, there is no ordering implied between
> WS messages sent over different SPDY-multiplexed channels.
>
>
>
> Now, I can see a need for a different semantics (see my old mails/comments
> wrt to WS/MUX):
>
>
>
> Have a per-message priority that allows (weak) reordering of respective
> messages on a _single_ channel:
>
>
>
> Two WS messages (and frames for that) MAY be reordered ("prioritized") by
> the transport channel as follows:
>
>
>
> Msg-A MAY overtake Msg-B, if and only if, Priority(Msg-A) > Priority(Msg-B)
>
>
>
> Use case / scenario: chat (high prio) and file transfer (low prio) both
> running over 1 WS for 1 app
>
>
>
> Is there interest in above?
>
>
>
> Would it make sense to start an effort for that (retargeting the WS/MUX
> goals)?
>
>
>

Possible but no ideas. Do you want messages with the same priority never be
reordered?


>  /Tobias
>
>
>
>
>
>
>
>
>
> *Von:* hybi [mailto:hybi-bounces@ietf.org] *Im Auftrag von *Takeshi
> Yoshino
> *Gesendet:* Dienstag, 7. Januar 2014 08:36
> *An:* hybi@ietf.org
> *Betreff:* [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
>
> Hi all,
>
>
>
> We've been working on mux extension to make WebSocket scalable for long
> time, but as SPDY got widely supported and adopted quickly, now there seems
> to be no longer any good reason to continue this work which is WebSocket
> dedicated.
>
>
>
> Scalability is still important issue to address. From now, I'd like to
> focus on WS/SPDY effort which we've been working on in parallel with the
> mux extension.
>
>
>
> Please take a look at the latest proposal in this post.
>
> https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc
>
>
>
> If you really want the dedicated mux extension than one over SPDY, sorry
> but please take over the work.
>

--089e01493c5a21980a04ef885939
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On T=
hu, Jan 9, 2014 at 6:45 PM, Tobias Oberstein <span dir=3D"ltr">&lt;<a href=
=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstein@=
tavendo.de</a>&gt;</span> wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div lang=3D"DE" link=3D"blue" vlink=3D"purp=
le"><div><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family=
:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Two more thought=
s:<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">1) WS/SPDY<u></u><u=
></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">WS essentially prov=
ides a reliable, ordered, bidirectional/full-duplex message based channel.<=
u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">WS/SPDY is layering=
 WS over SPDY.<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Are there efforts t=
o implement a WebSocket-like semantics (like above) _directly_ using SPDY f=
acilities?<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0</span></p><=
/div></div></blockquote><div><br></div><div>I don&#39;t know for now. As we=
 already have WS API, there&#39;s no reason to re-invent something else and=
 ignore WS API.</div>

<div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8=
ex;border-left:1px #ccc solid;padding-left:1ex"><div lang=3D"DE" link=3D"bl=
ue" vlink=3D"purple"><div><p class=3D"MsoNormal"><span style=3D"font-size:1=
1.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"=
><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">2) WS Mux<u></u><u>=
</u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">As it stands (as fa=
r as I understand), SPDY MUX and the WS/MUX both roughly target the same se=
mantics:<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">multiplex different=
 channels over one underlying transport<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">The channels are in=
dependent. There is no ordering implied across data sent over _different_ l=
ogical channels.<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">In the case of WS/S=
PDY, that means, there is no ordering implied between WS messages sent over=
 different SPDY-multiplexed channels.<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Now, I can see a ne=
ed for a different semantics (see my old mails/comments wrt to WS/MUX):<u><=
/u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Have a per-message =
priority that allows (weak) reordering of respective messages on a _single_=
 channel:<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Two WS messages (an=
d frames for that) MAY be reordered (&quot;prioritized&quot;) by the transp=
ort channel as follows:<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Msg-A MAY overtake =
Msg-B, if and only if, Priority(Msg-A) &gt; Priority(Msg-B)<u></u><u></u></=
span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Use case / scenario=
: chat (high prio) and file transfer (low prio) both running over 1 WS for =
1 app<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Is there interest i=
n above?<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Would it make sense=
 to start an effort for that (retargeting the WS/MUX goals)?<u></u><u></u><=
/span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0</span></p><=
/div></div></blockquote><div><br></div><div>Possible but no ideas. Do you w=
ant messages with the same priority never be reordered?</div>

<div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8=
ex;border-left:1px #ccc solid;padding-left:1ex"><div lang=3D"DE" link=3D"bl=
ue" vlink=3D"purple"><div><p class=3D"MsoNormal"><span style=3D"font-size:1=
1.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"=
><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">/Tobias<u></u><u></u></sp=
an></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&=
quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></=
u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u=
></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><div style=3D"border:none;border-left:solid blue 1.5pt;padding:0cm 0c=
m 0cm 4.0pt">

<div class=3D"im"><div><div style=3D"border:none;border-top:solid #e1e1e1 1=
.0pt;padding:3.0pt 0cm 0cm 0cm"><p class=3D"MsoNormal"><b><span style=3D"fo=
nt-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;">Von:=
</span></b><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,=
&quot;sans-serif&quot;"> hybi [mailto:<a href=3D"mailto:hybi-bounces@ietf.o=
rg" target=3D"_blank">hybi-bounces@ietf.org</a>] <b>Im Auftrag von </b>Take=
shi Yoshino<br>

<b>Gesendet:</b> Dienstag, 7. Januar 2014 08:36<br><b>An:</b> <a href=3D"ma=
ilto:hybi@ietf.org" target=3D"_blank">hybi@ietf.org</a><br><b>Betreff:</b> =
[hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0<u></u>=
<u></u></span></p>

</div></div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div><p cl=
ass=3D"MsoNormal">Hi all, <u></u><u></u></p><div><div class=3D"h5"><div><p =
class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div><p class=3D"MsoNorma=
l">We&#39;ve been working on mux extension to make WebSocket scalable for l=
ong time, but as SPDY got widely supported and adopted quickly, now there s=
eems to be no longer any good reason to continue this work which is WebSock=
et dedicated.<u></u><u></u></p>

</div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div><p cla=
ss=3D"MsoNormal">Scalability is still important issue to address. From now,=
 I&#39;d like to focus on WS/SPDY effort which we&#39;ve been working on in=
 parallel with the mux extension.<u></u><u></u></p>

</div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div><p cla=
ss=3D"MsoNormal">Please take a look at the latest proposal in this post.<u>=
</u><u></u></p></div><div><p class=3D"MsoNormal"><a href=3D"https://groups.=
google.com/forum/#!topic/spdy-dev/aJut2ax3glc" target=3D"_blank">https://gr=
oups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc</a><u></u><u></u></p>

</div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div><p cla=
ss=3D"MsoNormal">If you really want the dedicated mux extension than one ov=
er SPDY, sorry but please take over the work.<u></u><u></u></p></div></div>=
</div>

</div></div></div></div></blockquote></div><br></div></div>

--089e01493c5a21980a04ef885939--

From tobias.oberstein@tavendo.de  Thu Jan  9 04:11:46 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 1B5D51ADF52 for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 04:11:46 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.899
X-Spam-Level: 
X-Spam-Status: No, score=-1.899 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 70o03Y2uRvZy for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 04:11:44 -0800 (PST)
Received: from EXHUB020-3.exch020.serverdata.net (exhub020-3.exch020.serverdata.net [206.225.164.30]) by ietfa.amsl.com (Postfix) with ESMTP id B52F91AE212 for <hybi@ietf.org>; Thu,  9 Jan 2014 04:11:41 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-3.exch020.serverdata.net ([206.225.164.30]) with mapi; Thu, 9 Jan 2014 04:11:31 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Takeshi Yoshino <tyoshino@google.com>
Date: Thu, 9 Jan 2014 04:11:28 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8NMlGvT3ZMC3mpR3CphOZufoGRUQAAHPHw
Message-ID: <634914A010D0B943A035D226786325D4446BE403E9@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E0@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZfBMEL1zWx+Eh9+rTJpTYV=q3Be0ZeW-WQQ7NXszharA@mail.gmail.com>
In-Reply-To: <CAH9hSJZfBMEL1zWx+Eh9+rTJpTYV=q3Be0ZeW-WQQ7NXszharA@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: multipart/alternative; boundary="_000_634914A010D0B943A035D226786325D4446BE403E9EXVMBX02012ex_"
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 09 Jan 2014 12:11:46 -0000

--_000_634914A010D0B943A035D226786325D4446BE403E9EXVMBX02012ex_
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_634914A010D0B943A035D226786325D4446BE403E9EXVMBX02012ex_
Content-Type: text/html; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_634914A010D0B943A035D226786325D4446BE403E9EXVMBX02012ex_--

From tobias.oberstein@tavendo.de  Thu Jan  9 04:13:55 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id BB0301AE29C for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 04:13:55 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.6
X-Spam-Level: 
X-Spam-Status: No, score=-1.6 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, MIME_8BIT_HEADER=0.3, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id eAUvfv60cjs3 for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 04:13:54 -0800 (PST)
Received: from EXHUB020-2.exch020.serverdata.net (exhub020-2.exch020.serverdata.net [206.225.164.29]) by ietfa.amsl.com (Postfix) with ESMTP id 9AA431AE29B for <hybi@ietf.org>; Thu,  9 Jan 2014 04:13:54 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-2.exch020.serverdata.net ([206.225.164.29]) with mapi; Thu, 9 Jan 2014 04:13:44 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: =?iso-8859-1?Q?=22Martin_J=2E_D=FCrst=22?= <duerst@it.aoyama.ac.jp>, "hybi@ietf.org" <hybi@ietf.org>
Date: Thu, 9 Jan 2014 04:13:41 -0800
Thread-Topic: [hybi] Detection of broken WS connections
Thread-Index: Ac8NMT886wzRrPL5S6Kv73LVZzcw8QAArjPw
Message-ID: <634914A010D0B943A035D226786325D4446BE403EB@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E0@EXVMBX020-12.exch020.serverdata.net> <52CE8D59.3070802@it.aoyama.ac.jp>
In-Reply-To: <52CE8D59.3070802@it.aoyama.ac.jp>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
Subject: Re: [hybi] Detection of broken WS connections
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 09 Jan 2014 12:13:56 -0000

> The WS API is explicit about the fact that there is no way from JavaScrip=
t to
> generate PINGs or check for PONGs. So do I just create my own, applicatio=
n
> level equivalent, and send a ping-like message every so often and assume

You can have a WS server that does automatic ping-pong for connection keep-=
alive/loss-detection.

Or you can do it on app level from JS.

> the connection is dead when I don't get a reply "soonish"? Are there libr=
aries
> that support such a mechanism? Or is every browser supported to regularly
> PINGPONG with the server, and call the onclose function when there's no

As far as I know, there is no JS API where you could request the browser to=
 do automatic ping-pong upon opening a WS connection.

> answer?
>=20
> Any advice or pointers appreciated!
>=20
> Regards,   Martin.
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi

From tyoshino@google.com  Thu Jan  9 04:32:20 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 680B41AD8CD for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 04:32:20 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id sasTDzoib2ej for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 04:32:16 -0800 (PST)
Received: from mail-wi0-x231.google.com (mail-wi0-x231.google.com [IPv6:2a00:1450:400c:c05::231]) by ietfa.amsl.com (Postfix) with ESMTP id 40F6D1AE215 for <hybi@ietf.org>; Thu,  9 Jan 2014 04:32:12 -0800 (PST)
Received: by mail-wi0-f177.google.com with SMTP id cc10so3384241wib.4 for <hybi@ietf.org>; Thu, 09 Jan 2014 04:32:02 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=f3CJ9lmB5zqdBfROhebNkAO/KABjt6nZGbpVmkJRPAY=; b=dVpdtrOEgiRRpPUCV5eiuem5iewcVyN8ufsxFEd1ZqXQ0OBmjw0TfKOiUqdw8rFKp2 qyjP8Q5jQFf7zLwSWwdhiGCbgNTFyXJs7gmN0BsNBsLeUZVZzAJ/sKArtJW/3SS9LtTw ti7FrYR7l1KZsGI2w4TBT9FqPZ2cGwqmkSp6dh67kJrtlkbkQySxSiIos15i1wOdkKpW iTOtkCswYrvLtGNxIbTC/HKCmn+KR9fsy2NlOYRuD61ad3pleNGlsoDgqtrbHRSfzHLT 227mMk2rfjqO6ByqQKqy3KHs08+KZI0kyCh06CYxTAijtAHLl5mHGaMAQAR+9sf41b8i rz8g==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=f3CJ9lmB5zqdBfROhebNkAO/KABjt6nZGbpVmkJRPAY=; b=F6FKDCs2GwJHOcXfIMMM7X2TR9uy/6jvtFQxMQYXfqyIwy+64ST8vTKwj431ABbmBO wmr5Xd5b9jdoQh2ZhpyMZ/OUfh9Er8Gw3AVqzzelb4pz5MmLl01d54ePYC70GWgGJw2I 9SAlKmUPjPYmTJXerbtblPRNB0sjjxkkA8YtGPUl1Qjvy1f3aCT/rXhVB6o9t3VYTpU/ U00/hVToHQ+oqnRGIVUkRT7z26C0WOkoZBykQfUc3LBOHm6jAeUgLqTKZdD1BAoo5hMS oncFy9/iOKu4Ipx4xMwDHYpq+Yg0RoWXNNJI4wHFesxcqQsrjdcl5eht/ZjAYGZPiPW8 NzmQ==
X-Gm-Message-State: ALoCoQkhZ/dTr/2ZnTb4Y9X67IJMOtGEM2fDzWWPgGSmASjarVym2BYBIns9HXGALLCicNFgWfLJqu5FXO9ayQuIxSeucZUJg3SD/PKUkmg7xp+47ED+f1ll1zcW8DAUQRy/qTKbtVYngGeEheNlm0mOwoeVxB8Y9nrjSD7dX6LAtIjmd02jucv9oHAb5/0CiIuuoTYq3PtZ
X-Received: by 10.180.97.164 with SMTP id eb4mr26514901wib.11.1389270722039; Thu, 09 Jan 2014 04:32:02 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Thu, 9 Jan 2014 04:31:41 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BE403E9@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E0@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZfBMEL1zWx+Eh9+rTJpTYV=q3Be0ZeW-WQQ7NXszharA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E9@EXVMBX020-12.exch020.serverdata.net>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Thu, 9 Jan 2014 21:31:41 +0900
Message-ID: <CAH9hSJYyA+-5LPTNPH+uNLGGcHFq1XVBNbEqP=grHpBgTTxN3g@mail.gmail.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=f46d04440320e27baf04ef88cbd6
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 09 Jan 2014 12:32:20 -0000

--f46d04440320e27baf04ef88cbd6
Content-Type: text/plain; charset=UTF-8

On Thu, Jan 9, 2014 at 9:11 PM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> Hi Takeshi,
>
>
>
> 1) What I meant is keep the WS API (I suppose you mean the one in HTML5),
> but layer WS-like messaging directly on SPDY,
>

Sorry for asking for clarification repeatedly.

Do you mean layering WS protocol like messaging over HTTP/1.1 over SPDY? If
so, it has been discussed at spdy-dev@ and was in older spec draft.


> doing away with the reframing/double-framing with WS/SPDY
>

Hmm, current proposal is kinda direct use of SPDY. Please elaborate.


>
>
> 2) Yes, messages with equal priority must never be reordered.
>
>
>
> One problem with standard WS is that there cannot be more than one message
> in flight: if a large (possible fragmented) low prio message has been
> started, but not yet completed, the frames for a subsequently started
> high-prio message cannot overtake the former.
>
>
>
> Not sure yet how that would work (if at all) with the mechnism that WS
> extensions allow.
>
>
>

Is there any use cases in your mind that cannot be addressed by having WS
connections with different prio multiplexed over one TCP connection by SPDY
mechanism?


>  3) Regarding HTML5 API: will JavaScript have the option to open WS/SPDY
> connections providing channel priority? So that an app can open >1 logical
> WS channels, with different priorities.
>
>
>

Yeah, if your answer to my question above is no, then you need that kind of
API.


>  /Tobias
>
>
>
>
>
> *Von:* Takeshi Yoshino [mailto:tyoshino@google.com]
> *Gesendet:* Donnerstag, 9. Januar 2014 13:00
> *An:* Tobias Oberstein
> *Cc:* hybi@ietf.org
> *Betreff:* Re: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
>
> On Thu, Jan 9, 2014 at 6:45 PM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
>
> Two more thoughts:
>
>
>
> 1) WS/SPDY
>
>
>
> WS essentially provides a reliable, ordered, bidirectional/full-duplex
> message based channel.
>
>
>
> WS/SPDY is layering WS over SPDY.
>
>
>
> Are there efforts to implement a WebSocket-like semantics (like above)
> _directly_ using SPDY facilities?
>
>
>
>
>
> I don't know for now. As we already have WS API, there's no reason to
> re-invent something else and ignore WS API.
>
>
>
>
>
> 2) WS Mux
>
>
>
> As it stands (as far as I understand), SPDY MUX and the WS/MUX both
> roughly target the same semantics:
>
>
>
> multiplex different channels over one underlying transport
>
>
>
> The channels are independent. There is no ordering implied across data
> sent over _different_ logical channels.
>
>
>
> In the case of WS/SPDY, that means, there is no ordering implied between
> WS messages sent over different SPDY-multiplexed channels.
>
>
>
> Now, I can see a need for a different semantics (see my old mails/comments
> wrt to WS/MUX):
>
>
>
> Have a per-message priority that allows (weak) reordering of respective
> messages on a _single_ channel:
>
>
>
> Two WS messages (and frames for that) MAY be reordered ("prioritized") by
> the transport channel as follows:
>
>
>
> Msg-A MAY overtake Msg-B, if and only if, Priority(Msg-A) > Priority(Msg-B)
>
>
>
> Use case / scenario: chat (high prio) and file transfer (low prio) both
> running over 1 WS for 1 app
>
>
>
> Is there interest in above?
>
>
>
> Would it make sense to start an effort for that (retargeting the WS/MUX
> goals)?
>
>
>
>
>
> Possible but no ideas. Do you want messages with the same priority never
> be reordered?
>
>
>
> /Tobias
>
>
>
>
>
>
>
>
>
> *Von:* hybi [mailto:hybi-bounces@ietf.org] *Im Auftrag von *Takeshi
> Yoshino
> *Gesendet:* Dienstag, 7. Januar 2014 08:36
> *An:* hybi@ietf.org
> *Betreff:* [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
>
> Hi all,
>
>
>
> We've been working on mux extension to make WebSocket scalable for long
> time, but as SPDY got widely supported and adopted quickly, now there seems
> to be no longer any good reason to continue this work which is WebSocket
> dedicated.
>
>
>
> Scalability is still important issue to address. From now, I'd like to
> focus on WS/SPDY effort which we've been working on in parallel with the
> mux extension.
>
>
>
> Please take a look at the latest proposal in this post.
>
> https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc
>
>
>
> If you really want the dedicated mux extension than one over SPDY, sorry
> but please take over the work.
>
>
>

--f46d04440320e27baf04ef88cbd6
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On T=
hu, Jan 9, 2014 at 9:11 PM, Tobias Oberstein <span dir=3D"ltr">&lt;<a href=
=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstein@=
tavendo.de</a>&gt;</span> wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div lang=3D"DE" link=3D"blue" vlink=3D"purp=
le"><div><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family=
:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Hi Takeshi,<u></=
u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">1) What I meant is =
keep the WS API (I suppose you mean the one in HTML5), but layer WS-like me=
ssaging directly on SPDY,</span></p>

</div></div></blockquote><div><br></div><div>Sorry for asking for clarifica=
tion repeatedly.</div><div><br></div><div>Do you mean layering WS protocol =
like messaging over HTTP/1.1 over SPDY? If so, it has been discussed at spd=
y-dev@ and was in older spec draft.</div>

<div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8=
ex;border-left:1px #ccc solid;padding-left:1ex"><div lang=3D"DE" link=3D"bl=
ue" vlink=3D"purple"><div><p class=3D"MsoNormal"><span style=3D"font-size:1=
1.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"=
> doing away with the reframing/double-framing with WS/SPDY</span></p>

</div></div></blockquote><div><br></div><div>Hmm, current proposal is kinda=
 direct use of SPDY. Please elaborate.</div><div>=C2=A0</div><blockquote cl=
ass=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;p=
adding-left:1ex">

<div lang=3D"DE" link=3D"blue" vlink=3D"purple"><div><p class=3D"MsoNormal"=
><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans=
-serif&quot;;color:#1f497d"><u></u><u></u></span></p><p class=3D"MsoNormal"=
><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans=
-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">2) Yes, messages with equ=
al priority must never be reordered.<u></u><u></u></span></p><p class=3D"Ms=
oNormal">

<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></span></p><p class=3D"MsoNo=
rmal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot=
;sans-serif&quot;;color:#1f497d">One problem with standard WS is that there=
 cannot be more than one message in flight: if a large (possible fragmented=
) low prio message has been started, but not yet completed, the frames for =
a subsequently started high-prio message cannot overtake the former.<u></u>=
<u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Not sure yet how th=
at would work (if at all) with the mechnism that WS extensions allow.<u></u=
><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0</span></p><=
/div></div></blockquote><div><br></div><div>Is there any use cases in your =
mind that cannot be addressed by having WS connections with different prio =
multiplexed over one TCP connection by SPDY mechanism?</div>

<div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8=
ex;border-left:1px #ccc solid;padding-left:1ex"><div lang=3D"DE" link=3D"bl=
ue" vlink=3D"purple"><div><p class=3D"MsoNormal"><span style=3D"font-size:1=
1.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"=
><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">3) Regarding HTML5 API: w=
ill JavaScript have the option to open WS/SPDY connections providing channe=
l priority? So that an app can open &gt;1 logical WS channels, with differe=
nt priorities.<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0</span></p><=
/div></div></blockquote><div><br></div><div>Yeah, if your answer to my ques=
tion above is no, then you need that kind of API.</div>

<div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8=
ex;border-left:1px #ccc solid;padding-left:1ex"><div lang=3D"DE" link=3D"bl=
ue" vlink=3D"purple"><div><p class=3D"MsoNormal"><span style=3D"font-size:1=
1.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"=
><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">/Tobias<u></u><u></u></sp=
an></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&=
quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></=
u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><div style=3D"border:none;border-left:solid blue 1.5pt;padding:0cm 0c=
m 0cm 4.0pt">

<div><div style=3D"border:none;border-top:solid #e1e1e1 1.0pt;padding:3.0pt=
 0cm 0cm 0cm"><p class=3D"MsoNormal"><b><span style=3D"font-size:11.0pt;fon=
t-family:&quot;Calibri&quot;,&quot;sans-serif&quot;">Von:</span></b><span s=
tyle=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&q=
uot;"> Takeshi Yoshino [mailto:<a href=3D"mailto:tyoshino@google.com" targe=
t=3D"_blank">tyoshino@google.com</a>] <br>

<b>Gesendet:</b> Donnerstag, 9. Januar 2014 13:00<br><b>An:</b> Tobias Ober=
stein<br><b>Cc:</b> <a href=3D"mailto:hybi@ietf.org" target=3D"_blank">hybi=
@ietf.org</a><br><b>Betreff:</b> Re: [hybi] Discontinuation of mux standard=
izaton in favor of WS/HTTP/2.0<u></u><u></u></span></p>

</div></div><div><div class=3D"h5"><p class=3D"MsoNormal"><u></u>=C2=A0<u><=
/u></p><div><div><div><p class=3D"MsoNormal">On Thu, Jan 9, 2014 at 6:45 PM=
, Tobias Oberstein &lt;<a href=3D"mailto:tobias.oberstein@tavendo.de" targe=
t=3D"_blank">tobias.oberstein@tavendo.de</a>&gt; wrote:<u></u><u></u></p>

<blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0c=
m 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm"><div><div><p class=3D"M=
soNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&=
quot;sans-serif&quot;;color:#1f497d">Two more thoughts:</span><u></u><u></u=
></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">1) WS/SPDY</span><u=
></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">WS essentially prov=
ides a reliable, ordered, bidirectional/full-duplex message based channel.<=
/span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">WS/SPDY is layering=
 WS over SPDY.</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Are there efforts t=
o implement a WebSocket-like semantics (like above) _directly_ using SPDY f=
acilities?</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p></div></div></blockquote><div><p class=3D"MsoNormal"><u></u>=C2=A0<u>=
</u></p></div>

<div><p class=3D"MsoNormal">I don&#39;t know for now. As we already have WS=
 API, there&#39;s no reason to re-invent something else and ignore WS API.<=
u></u><u></u></p></div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>=
</div>

<blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0c=
m 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm"><div><div><p class=3D"M=
soNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&=
quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">2) WS Mux</span><u></u><u=
></u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family=
:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u>=
</u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">As it stands (as far as I=
 understand), SPDY MUX and the WS/MUX both roughly target the same semantic=
s:</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">multiplex different=
 channels over one underlying transport</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">The channels are in=
dependent. There is no ordering implied across data sent over _different_ l=
ogical channels.</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">In the case of WS/S=
PDY, that means, there is no ordering implied between WS messages sent over=
 different SPDY-multiplexed channels.</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Now, I can see a ne=
ed for a different semantics (see my old mails/comments wrt to WS/MUX):</sp=
an><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Have a per-message =
priority that allows (weak) reordering of respective messages on a _single_=
 channel:</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Two WS messages (an=
d frames for that) MAY be reordered (&quot;prioritized&quot;) by the transp=
ort channel as follows:</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Msg-A MAY overtake =
Msg-B, if and only if, Priority(Msg-A) &gt; Priority(Msg-B)</span><u></u><u=
></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Use case / scenario=
: chat (high prio) and file transfer (low prio) both running over 1 WS for =
1 app</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Is there interest i=
n above?</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Would it make sense=
 to start an effort for that (retargeting the WS/MUX goals)?</span><u></u><=
u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p></div></div></blockquote><div><p class=3D"MsoNormal"><u></u>=C2=A0<u>=
</u></p></div>

<div><p class=3D"MsoNormal">Possible but no ideas. Do you want messages wit=
h the same priority never be reordered?<u></u><u></u></p></div><div><p clas=
s=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><blockquote style=3D"border:n=
one;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4=
.8pt;margin-right:0cm">

<div><div><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-famil=
y:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">/Tobias</span><=
u></u><u></u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;fon=
t-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</=
span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u=
><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><div style=3D"border:none;border-left:solid blue 1.5pt;padding:0cm 0c=
m 0cm 4.0pt">

<div><div><div style=3D"border:none;border-top:solid #e1e1e1 1.0pt;padding:=
3.0pt 0cm 0cm 0cm"><p class=3D"MsoNormal"><b><span style=3D"font-size:11.0p=
t;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;">Von:</span></b><s=
pan style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-se=
rif&quot;"> hybi [mailto:<a href=3D"mailto:hybi-bounces@ietf.org" target=3D=
"_blank">hybi-bounces@ietf.org</a>] <b>Im Auftrag von </b>Takeshi Yoshino<b=
r>

<b>Gesendet:</b> Dienstag, 7. Januar 2014 08:36<br><b>An:</b> <a href=3D"ma=
ilto:hybi@ietf.org" target=3D"_blank">hybi@ietf.org</a><br><b>Betreff:</b> =
[hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0</span>=
<u></u><u></u></p>

</div></div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><div><p cl=
ass=3D"MsoNormal">Hi all, <u></u><u></u></p><div><div><div><p class=3D"MsoN=
ormal">=C2=A0<u></u><u></u></p></div><div><p class=3D"MsoNormal">We&#39;ve =
been working on mux extension to make WebSocket scalable for long time, but=
 as SPDY got widely supported and adopted quickly, now there seems to be no=
 longer any good reason to continue this work which is WebSocket dedicated.=
<u></u><u></u></p>

</div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><div><p cla=
ss=3D"MsoNormal">Scalability is still important issue to address. From now,=
 I&#39;d like to focus on WS/SPDY effort which we&#39;ve been working on in=
 parallel with the mux extension.<u></u><u></u></p>

</div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><div><p cla=
ss=3D"MsoNormal">Please take a look at the latest proposal in this post.<u>=
</u><u></u></p></div><div><p class=3D"MsoNormal"><a href=3D"https://groups.=
google.com/forum/#!topic/spdy-dev/aJut2ax3glc" target=3D"_blank">https://gr=
oups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc</a><u></u><u></u></p>

</div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><div><p cla=
ss=3D"MsoNormal">If you really want the dedicated mux extension than one ov=
er SPDY, sorry but please take over the work.<u></u><u></u></p></div></div>=
</div>

</div></div></div></div></blockquote></div><p class=3D"MsoNormal"><u></u>=
=C2=A0<u></u></p></div></div></div></div></div></div></div></blockquote></d=
iv><br></div></div>

--f46d04440320e27baf04ef88cbd6--

From arman@noemax.com  Thu Jan  9 06:46:32 2014
Return-Path: <arman@noemax.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 8CA0E1AE36F for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 06:46:32 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.138
X-Spam-Level: 
X-Spam-Status: No, score=-1.138 tagged_above=-999 required=5 tests=[BAYES_05=-0.5, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id sTXwDmsMPT_J for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 06:46:30 -0800 (PST)
Received: from mail.noemax.com (mail.noemax.com [74.208.113.37]) by ietfa.amsl.com (Postfix) with ESMTP id 196891AE338 for <hybi@ietf.org>; Thu,  9 Jan 2014 06:46:30 -0800 (PST)
DKIM-Signature: a=rsa-sha1; t=1389278723; x=1389883523; s=m2048; d=noemax.com; c=relaxed/relaxed; v=1; bh=KDHStErVoxORen6rgvJXypeK20I=; h=From:Subject:Date:Message-ID:To:Cc:MIME-Version:Content-Type:In-Reply-To:References; b=ZawccKZjz5tTGYl3kD7Opc+OI6YVrYelK1EPlLE/Sk6stFNNnwrZiCILXCqyy4/fX5ey5oFWjXAWrGug5LmoGaqJdtlTfMectL8EKt0HHGg6+T0t13k72MWnnrWWop0JP4KU5f6CF+T8H3+J/sbH+0nH/d5XO7yXDjEcdXthMzqTj+lN80CYsww2aNwBiZps2z+wndKvmSkcYwV5l7s+uvXnY1vyIIZIM1SjXg5YiX9GnwOhVEix6muZp+Tw78qcCaR/4jtFu0ALLDT8XCAHyf8eVaoYD6vUZzFm6el7FtXRxogV+TlCjaBh/D0DT1PlxB/QX84biMSGP1KL63sJZg==
Received: from mail.noemax.com by mail.noemax.com (Noemax Mail Server) with ASMTP (SSL) id 201401091645198013; Thu, 09 Jan 2014 16:45:19 +0200
From: "Arman Djusupov" <arman@noemax.com>
To: "'Takeshi Yoshino'" <tyoshino@google.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E0@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZfBMEL1zWx+Eh9+rTJpTYV=q3Be0ZeW-WQQ7NXszharA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E9@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJYyA+-5LPTNPH+uNLGGcHFq1XVBNbEqP=grHpBgTTxN3g@mail.gmail.com>
In-Reply-To: <CAH9hSJYyA+-5LPTNPH+uNLGGcHFq1XVBNbEqP=grHpBgTTxN3g@mail.gmail.com>
Date: Thu, 9 Jan 2014 16:46:03 +0200
Message-ID: <007301cf0d49$87938740$96ba95c0$@noemax.com>
MIME-Version: 1.0
Content-Type: multipart/alternative; boundary="----=_NextPart_000_0074_01CF0D5A.4B1EC840"
X-Mailer: Microsoft Outlook 14.0
Thread-Index: AQDDXzHDEplKmgqjHdYhTVNFZY+cTQGHCuO2Af0gYtsCSy0+lQIhac7anFRN/3A=
Content-Language: en-us
Cc: hybi@ietf.org
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of	WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 09 Jan 2014 14:46:32 -0000

This is a multipart message in MIME format.

------=_NextPart_000_0074_01CF0D5A.4B1EC840
Content-Type: text/plain;
	charset="utf-8"
Content-Transfer-Encoding: quoted-printable

Hello Takeshi,

=20

In the current draft, the description of the compression process states =
:

=20
   1.  Compress all the octets of the payload of the message using
       DEFLATE.

=20

This doesn=E2=80=99t take into account outbound messages of arbitrary =
size that NEED to be fragmented, for which it is more favorable to use =
fragmentation and to compress & send fragments until the end of the =
message is reached.

=20

In such cases the implementation buffers the input from its source and =
then flushes it out into a compressed fragment, repeating this process =
until the end of the source data is reached. At that point it flushes =
any remaining bytes buffered into a final frame, or flushes a final =
frame of 0 length.

=20

=20

When the implementation produces compressed fragments, it periodically =
produces frames with 0x00 0x00 0xFF 0xFF at the end of the frame due to =
flushing DEFLATE. But because of the following requirements:

=20

   2. If the resulting data does not end with an empty DEFLATE block

       with no compression (the "BTYPE" bits is set to 00), append an

       empty DEFLATE block with no compression to the tail end.

=20

   3.  Remove 4 octets (that are 0x00 0x00 0xff 0xff) from the tail end.

       After this step, the last octet of the compressed data contains

       (possibly part of) the DEFLATE header bits with the "BTYPE" bits

       set to 00.

=20

the implementation must ensure that the message ends empty blocks =
without a trailing 0x0000FFFF, so the implementation must keep track on =
what type of block was at the end of the frame that was sent last. If =
the last frame sent ends with 0x0000FFFF then the implementation cannot =
remove those 4 bytes from the wire, but it must artificially produce an =
empty DEFLATE block to send as the final frame. This is only due to the =
requirement to remove 0x0000FFFF.

=20

Wouldn=E2=80=99t it be easier to make the removal 0x00 0x00 0xFF 0xFF =
optional, while at the same time requiring that the receiving side =
appends a 0x00 0x00 0xFF 0xFF to the final frame in case when its =
missing?

=20

It is much simpler to just check the final bytes of the message instead =
of having to remember the state of the previous frame.

=20

With best regards,

Arman

=20


------=_NextPart_000_0074_01CF0D5A.4B1EC840
Content-Type: text/html;
	charset="utf-8"
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:m=3D"http://schemas.microsoft.com/office/2004/12/omml" =
xmlns=3D"http://www.w3.org/TR/REC-html40"><head><meta =
http-equiv=3DContent-Type content=3D"text/html; charset=3Dutf-8"><meta =
name=3DGenerator content=3D"Microsoft Word 14 (filtered =
medium)"><style><!--
/* Font Definitions */
@font-face
	{font-family:"Cambria Math";
	panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
	{font-family:Calibri;
	panose-1:2 15 5 2 2 2 4 3 2 4;}
@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:0in;
	margin-bottom:.0001pt;
	font-size:12.0pt;
	font-family:"Times New Roman","serif";}
a:link, span.MsoHyperlink
	{mso-style-priority:99;
	color:blue;
	text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
	{mso-style-priority:99;
	color:purple;
	text-decoration:underline;}
pre
	{mso-style-priority:99;
	mso-style-link:"HTML Preformatted Char";
	margin:0in;
	margin-bottom:.0001pt;
	font-size:10.0pt;
	font-family:"Courier New";}
p.MsoAcetate, li.MsoAcetate, div.MsoAcetate
	{mso-style-priority:99;
	mso-style-link:"Balloon Text Char";
	margin:0in;
	margin-bottom:.0001pt;
	font-size:8.0pt;
	font-family:"Tahoma","sans-serif";}
span.HTMLPreformattedChar
	{mso-style-name:"HTML Preformatted Char";
	mso-style-priority:99;
	mso-style-link:"HTML Preformatted";
	font-family:"Courier New";
	mso-ligatures:none;}
span.BalloonTextChar
	{mso-style-name:"Balloon Text Char";
	mso-style-priority:99;
	mso-style-link:"Balloon Text";
	font-family:"Tahoma","sans-serif";
	mso-ligatures:none;}
span.EmailStyle21
	{mso-style-type:personal-reply;
	font-family:"Calibri","sans-serif";
	color:#44546A;}
.MsoChpDefault
	{mso-style-type:export-only;
	font-size:10.0pt;}
@page WordSection1
	{size:8.5in 11.0in;
	margin:1.0in 1.0in 1.0in 1.0in;}
div.WordSection1
	{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext=3D"edit" spidmax=3D"1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext=3D"edit">
<o:idmap v:ext=3D"edit" data=3D"1" />
</o:shapelayout></xml><![endif]--></head><body lang=3DEN-US link=3Dblue =
vlink=3Dpurple><div class=3DWordSection1><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>Hello Takeshi,<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>In the current draft, t</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>he description of </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>compression process states =
:<o:p></o:p></span></p><pre><o:p>&nbsp;</o:p></pre><pre>&nbsp;&nbsp; =
1.&nbsp; Compress all the octets of the payload of the message =
using<o:p></o:p></pre><pre>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
DEFLATE.<o:p></o:p></pre><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>T</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>his doesn=E2=80=99t</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'> take</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> into account </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>outbound </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>messages of arbitrary size that NEED to be fragmented, </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>for which it is</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> more favorable to use fragmentation and </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>to </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>compress</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'> &amp; send</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> fragments until the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>end of the message is reached</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>.<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>In such cases the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>implementation buffers </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the in</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>put </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>from its source </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>and then flushes it out into a compressed fragment</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>, repeating this process</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> until the end of the source data </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>is </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>reached</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>. At that point </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>it </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>flushes any remaining bytes buffered into a final frame, or =
</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>flushes </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>a</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> final frame of 0 length.<o:p></o:p></span></p><p =
class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>When </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>implementation produces compressed </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>fragments,</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> it periodically produces frames with 0x00 0x00 0xFF 0xFF at the end =
of the frame</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'> due to flushing DEFLATE</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>. But </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>because of the</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> following requirements:<o:p></o:p></span></p><p =
class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:10.0pt;font-family:"Courier New"'>&nbsp;&nbsp; 2. If =
the resulting data does not end with an empty DEFLATE =
block<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:10.0pt;font-family:"Courier =
New"'>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; with no compression (the =
&quot;BTYPE&quot; bits is set to 00), append an<o:p></o:p></span></p><p =
class=3DMsoNormal><span style=3D'font-size:10.0pt;font-family:"Courier =
New"'>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; empty DEFLATE block with no =
compression to the tail end.<o:p></o:p></span></p><p =
class=3DMsoNormal><span style=3D'font-size:10.0pt;font-family:"Courier =
New"'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:10.0pt;font-family:"Courier New"'>&nbsp;&nbsp; =
3.&nbsp; Remove 4 octets (that are 0x00 0x00 0xff 0xff) from the tail =
end.<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:10.0pt;font-family:"Courier =
New"'>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; After this step, the last =
octet of the compressed data contains<o:p></o:p></span></p><p =
class=3DMsoNormal><span style=3D'font-size:10.0pt;font-family:"Courier =
New"'>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; (possibly part of) the =
DEFLATE header bits with the &quot;BTYPE&quot; =
bits<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:10.0pt;font-family:"Courier =
New"'>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; set to =
00.<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>implementation must </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>e</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>nsure that the message ends empty block</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>s</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> without </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>a trailing</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>&nbsp;0x0000FFFF, so </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>implementation must keep track on what type of block was at the end =
of the frame that was sent last. If the last frame sent ends with =
0x0000FFFF</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'> then</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>implementation cannot remove those 4 bytes from the wire</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>, but </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>it must artificially produce an empty DEFLATE block to send =
</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>as the</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> final frame. </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>This is only due to the requirement to remove </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>0x0000FFFF.<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>Wouldn=E2=80=99t it be easier to make </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>remov</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>al</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> 0x00 0x00 0xFF 0xFF optional</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>, while at the same time requiring that the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>receiving side a</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>ppends a</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> 0x00 0x00 0xFF 0xFF </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>to</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> the final frame in case when </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>its</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> missing</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>?</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>It </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>is </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>much simpler to </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>just</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> check </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>final bytes of the message </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>instead of</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>having to </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>remember the state of </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>previous frame.<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>With best regards,<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>Arman<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p>&nbsp;</o:p></span></p></div></body></html>
------=_NextPart_000_0074_01CF0D5A.4B1EC840--


From yhirano@google.com  Thu Jan  9 19:54:58 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 8CFE61ADF9B for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 19:54:58 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id JPLgeQvz_cBR for <hybi@ietfa.amsl.com>; Thu,  9 Jan 2014 19:54:55 -0800 (PST)
Received: from mail-qe0-x236.google.com (mail-qe0-x236.google.com [IPv6:2607:f8b0:400d:c02::236]) by ietfa.amsl.com (Postfix) with ESMTP id 7508C1ADFCD for <hybi@ietf.org>; Thu,  9 Jan 2014 19:54:55 -0800 (PST)
Received: by mail-qe0-f54.google.com with SMTP id cy11so4072506qeb.13 for <hybi@ietf.org>; Thu, 09 Jan 2014 19:54:45 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=CKu71y0Oa/DZe4t3fFxXQ9wW7DHKEL491hA8tIumJWQ=; b=KkTHV2hyM6KGCkjzg2395l1pc6zqWO96fx3d7DD5bMD9XFzYypFQ5hYHwrJlD4i1cC zmyj31pcx18dMuIt4Lj4sSQ3oIfyAx3+JicQs6R9WkEssehkotoy2zEI0xBSg3Q5ZXlw 7mHllMZnF8hrlFOLZrMfZp0WzY66DRQq38Ysnz8XqK9Ao7Wc1RU35zdejBEow0RWG4e2 QWzexwEuA5fw6CioZIXhJngI/7lJN3kKMgcHJdTeQJ5CybWdDYWHjnr4wMzjSaWGpX9C +ugQR2hau/h2MZIqeLC5RM6PzOs/7IH3xj/JJs1OPP6eyg3dPRszUagt/C5uladLrpUR LIHg==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=CKu71y0Oa/DZe4t3fFxXQ9wW7DHKEL491hA8tIumJWQ=; b=Vd09EBFgvBb7kZGGcRR0BNsGl4lqsXCvdo/5uahh/b68Gkeq/XqJ3tcwTPHYhsqtTp 40zLsUC/eUnDs7TgE0wNNyzOPWOGS7Zrchl6D/hwbt+xWPmNKI/o8Mhdwwnl7PWWERsx EVm1HPwMp9usIN7OY6C5QCCcwvTu8dzXQidA4=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=CKu71y0Oa/DZe4t3fFxXQ9wW7DHKEL491hA8tIumJWQ=; b=Flk9aTGDqiSPp4g5FvPQsZ1ZJR5oxWNHdfx1vtrAyKCN5EivQ7dwdLg6B+TllYI7YU gB7QuX6ahqs5Q8D3aO2EM2OIfUthCnmqwGwO7320v/sP/BqDXxJcI992mFs0dbqEey85 jJZXvDhnjeY+jOOLtwPjsxHRthiqDd+Uu+Mo0nON2PXyc4kEUV0r38Rzw2sdIcB6SAq/ XjDEmrErDQ4pD0AMEDzs8Wfh9q65WJbU73bn4k32NPJBb9dfrxslDLpaaY1FRTME9gFw rRm5W6TWhRNUzaP6ZwgOvPmtz459gr8xF3IC2s6RJ1ImgGQwg/9+xe9hNVxdl1vU/0+J 1WXg==
X-Gm-Message-State: ALoCoQnj71pgsguM3egNt6xpONKr2/3ok0foVB9JsH+k+2+ghN01lY9Kb25gOzesujMa65LmSv61Xiz7KdWOBAzbJly658kZP0xRuQzkM51WGP1DyGi5hP79veHztljQ18nhdJer7aqHrrjodWnpQn4BgQtHEZHXa/YrpQmmf/50MuPKJ3fedKfYTsMY4yT8HHuP2OrvxDU8
MIME-Version: 1.0
X-Received: by 10.49.2.132 with SMTP id 4mr2645024qeu.15.1389326085323; Thu, 09 Jan 2014 19:54:45 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.140.109.202 with HTTP; Thu, 9 Jan 2014 19:54:45 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net>
Date: Fri, 10 Jan 2014 12:54:45 +0900
X-Google-Sender-Auth: T5c3qQfStecdKGKCPXtMmgxE4P0
Message-ID: <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>, Roberto Peon <fenix@google.com>
Content-Type: multipart/alternative; boundary=047d7b6d99b0cb1f1904ef95afbb
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 10 Jan 2014 03:54:58 -0000

--047d7b6d99b0cb1f1904ef95afbb
Content-Type: text/plain; charset=UTF-8

+Roberto as an HTTP2 expert.


> 2) Ok, both WS frame and message boundaries are preserved.
>
> But: I cannot recognize the end of a WS frame until the next WS frame
> starts (for which I receive a HEADERS frame, which is the implicit signal
> that the previous WS frame ended), right?
>
You are right, you can't.


>
> And if so, why is there no FRAME_DONE bit on the SPDY frame that
> corresponds to the end of a WS frame?

Given that WebSocket frames "have no semantic meaning" as in RFC6455, I
don't think the immediate recognition of the end of a WebSocket frame is
requisite.
I would like to introduce MSG_DONE in HTTP2 so that intermediaries that
don't fully understand WS over HTTP2 can relay it.
Since there are not so much space in FLAGS field, Introducing FRAME_DONE
may decrease the extensiveness of the spec.
Do you have a strong preference to introduce FRAME_DONE?



On Thu, Jan 9, 2014 at 6:53 PM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> Hi Yutaka,
>
>
>
> 1) Yes, I understand. Thanks!
>
>
>
> 2) Ok, both WS frame and message boundaries are preserved.
>
>
>
> But: I cannot recognize the end of a WS frame until the next WS frame
> starts (for which I receive a HEADERS frame, which is the implicit signal
> that the previous WS frame ended), right?
>
>
>
> And if so, why is there no FRAME_DONE bit on the SPDY frame that
> corresponds to the end of a WS frame?
>
>
>
> 3) Ok, great! Hope this works, since SPDY provides header compression, but
> no sophisticated bidirectional payload compression as "WS
> permessage-deflate" does.
>
>
>
> Cheers,
>
> /Tobias
>
>
>
>
>
>
>
> *Von:* yhirano@google.com [mailto:yhirano@google.com] *Im Auftrag von *Yutaka
> Hirano
> *Gesendet:* Donnerstag, 9. Januar 2014 10:37
> *An:* Tobias Oberstein
> *Cc:* Takeshi Yoshino; hybi@ietf.org
> *Betreff:* Re: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
>
> Hi Tobias,
>
> Thank you for the comments.
>
>
>
> 1) Frame Mapping
> """
> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
> following plural SPDY data frames. The SPDY HEADERS frame must contain
> WebSocket frame fields and following SPDY data frames contain payload data.
> These data frames are free to be reframed.
> """
>
> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
> So reframing of WebSocket frames (the payload) into multiple SPDY data
> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
> is free to choose the boundaries within WS frame payload at which to
> reframe into different SPDY data frames as it likes?
>
> A WebSocket frame MAY be split to multiple SPDY DATA frames.
>
> Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA frame
> into multiple DATA frames.
>
> Is this answering your question?
>
>
>
> 2) End of WS Frame
> The end of a WS message is signaled with a SPDY data frame having MSG_DONE
> set.
> How is the end of a WS frame signaled, given that the WS frame may have
> been reframed into multiple SPDY data frames?
>
> When you receive a HEADERS frame, you can recognize that the last SPDY
> frame is that last frame in a WebSocket frame.
>
> When you receive a SPDY frame with MSG_DONE, you can recognize that it is
> the last SPDY frame in a WebSocket message and hence the last SPDY frame in
> a WebSocket frame.
>
>
>
> 3) WS extension
> Are WS extensions fully supported? In particular, is "permessage-deflate"
> supported when running WS/SPDY?
>
> Yes. We have reserved bits and we can emulate the opening handshake, I
> think.
>
>
>
> On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
>
> Hi Takeshi,
>
> it's a pity, but I can understand the motivation to shove everything into
> HTTP2.
>
> Anyway, some questions:
>
> 1) Frame Mapping
> """
> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
> following plural SPDY data frames. The SPDY HEADERS frame must contain
> WebSocket frame fields and following SPDY data frames contain payload data.
> These data frames are free to be reframed.
> """
>
> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
>
> So reframing of WebSocket frames (the payload) into multiple SPDY data
> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
> is free to choose the boundaries within WS frame payload at which to
> reframe into different SPDY data frames as it likes?
>
> 2) End of WS Frame
>
> The end of a WS message is signaled with a SPDY data frame having MSG_DONE
> set.
>
> How is the end of a WS frame signaled, given that the WS frame may have
> been reframed into multiple SPDY data frames?
>
> 3) WS extension
>
> Are WS extensions fully supported? In particular, is "permessage-deflate"
> supported when running WS/SPDY?
>
> Thanks!
> /Tobias
>
>
> Von: hybi [mailto:hybi-bounces@ietf.org] Im Auftrag von Takeshi Yoshino
> Gesendet: Dienstag, 7. Januar 2014 08:36
> An: hybi@ietf.org
> Betreff: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
> Hi all,
>
> We've been working on mux extension to make WebSocket scalable for long
> time, but as SPDY got widely supported and adopted quickly, now there seems
> to be no longer any good reason to continue this work which is WebSocket
> dedicated.
>
> Scalability is still important issue to address. From now, I'd like to
> focus on WS/SPDY effort which we've been working on in parallel with the
> mux extension.
>
> Please take a look at the latest proposal in this post.
> https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc
>
> If you really want the dedicated mux extension than one over SPDY, sorry
> but please take over the work.
>
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>
>
>

--047d7b6d99b0cb1f1904ef95afbb
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div>+Roberto as an HTTP2 expert.</div><div>=C2=A0</div><b=
lockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-le=
ft-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;pad=
ding-left:1ex">
<span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,7=
3,125)">2) Ok, both WS frame and message boundaries are preserved.<br></spa=
n><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31=
,73,125)">=C2=A0<br>
</span><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:r=
gb(31,73,125)">But: I cannot recognize the end of a WS frame until the next=
 WS frame starts (for which I receive a HEADERS frame, which is the implici=
t signal that the previous WS frame ended), right?<br>
</span><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:r=
gb(31,73,125)"></span></blockquote>You are right, you can&#39;t.<div>=C2=A0=
</div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;b=
order-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:s=
olid;padding-left:1ex">
<span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,7=
3,125)">=C2=A0<br></span><span style=3D"font-size:11pt;font-family:Calibri,=
sans-serif;color:rgb(31,73,125)">And if so, why is there no FRAME_DONE bit =
on the SPDY frame that corresponds to the end of a WS frame?</span></blockq=
uote>
Given that WebSocket frames &quot;have no semantic meaning&quot; as in RFC6=
455, I don&#39;t think the immediate recognition of the end of a WebSocket =
frame is requisite.<div>I would like to introduce MSG_DONE in HTTP2 so that=
 intermediaries that don&#39;t fully understand WS over HTTP2 can relay it.=
</div>
<div>Since there are not so much space in FLAGS field, Introducing FRAME_DO=
NE may decrease the extensiveness of the spec.</div><div>Do you have a stro=
ng preference to introduce FRAME_DONE?</div><div><br></div></div><div class=
=3D"gmail_extra">
<br><br><div class=3D"gmail_quote">On Thu, Jan 9, 2014 at 6:53 PM, Tobias O=
berstein <span dir=3D"ltr">&lt;<a href=3D"mailto:tobias.oberstein@tavendo.d=
e" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt;</span> wrote:<br><=
blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px=
 #ccc solid;padding-left:1ex">
<div lang=3D"DE" link=3D"blue" vlink=3D"purple"><div><p class=3D"MsoNormal"=
><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans=
-serif&quot;;color:#1f497d">Hi Yutaka,<u></u><u></u></span></p><p class=3D"=
MsoNormal">
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></span></p><p class=3D"MsoNo=
rmal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot=
;sans-serif&quot;;color:#1f497d">1) Yes, I understand. Thanks!<u></u><u></u=
></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">2) Ok, both WS fram=
e and message boundaries are preserved.<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">But: I cannot recog=
nize the end of a WS frame until the next WS frame starts (for which I rece=
ive a HEADERS frame, which is the implicit signal that the previous WS fram=
e ended), right?<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">And if so, why is t=
here no FRAME_DONE bit on the SPDY frame that corresponds to the end of a W=
S frame?<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">3) Ok, great! Hope =
this works, since SPDY provides header compression, but no sophisticated bi=
directional payload compression as &quot;WS permessage-deflate&quot; does.<=
u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Cheers,<u></u><u></=
u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">/Tobias<u></u><u></u></sp=
an></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&=
quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></=
u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u=
></span></p>
<div style=3D"border:none;border-left:solid blue 1.5pt;padding:0cm 0cm 0cm =
4.0pt"><div><div style=3D"border:none;border-top:solid #e1e1e1 1.0pt;paddin=
g:3.0pt 0cm 0cm 0cm"><p class=3D"MsoNormal"><b><span style=3D"font-size:11.=
0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;">Von:</span></b>=
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;"> <a href=3D"mailto:yhirano@google.com" target=3D"_blank">yhira=
no@google.com</a> [mailto:<a href=3D"mailto:yhirano@google.com" target=3D"_=
blank">yhirano@google.com</a>] <b>Im Auftrag von </b>Yutaka Hirano<br>
<b>Gesendet:</b> Donnerstag, 9. Januar 2014 10:37<br><b>An:</b> Tobias Ober=
stein<br><b>Cc:</b> Takeshi Yoshino; <a href=3D"mailto:hybi@ietf.org" targe=
t=3D"_blank">hybi@ietf.org</a><br><b>Betreff:</b> Re: [hybi] Discontinuatio=
n of mux standardizaton in favor of WS/HTTP/2.0<u></u><u></u></span></p>
</div></div><div><div class=3D"h5"><p class=3D"MsoNormal"><u></u>=C2=A0<u><=
/u></p><div><p class=3D"MsoNormal">Hi Tobias,<u></u><u></u></p><div><p clas=
s=3D"MsoNormal">Thank you for the comments.<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">
<u></u>=C2=A0<u></u></p><blockquote style=3D"border:none;border-left:solid =
#cccccc 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm"=
><p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;A=
rial&quot;,&quot;sans-serif&quot;">1) Frame Mapping<br>
&quot;&quot;&quot;<br>One WebSocket frame is mapped into one preceding SPDY=
 HEADERS frame and following plural SPDY data frames. The SPDY HEADERS fram=
e must contain WebSocket frame fields and following SPDY data frames contai=
n payload data. These data frames are free to be reframed.<br>
&quot;&quot;&quot;<br></span><a href=3D"https://docs.google.com/document/d/=
124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank"><span =
style=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&qu=
ot;">https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw=
9aWzSHijk/edit</span></a><span style=3D"font-size:10.0pt;font-family:&quot;=
Arial&quot;,&quot;sans-serif&quot;"><br>
So reframing of WebSocket frames (the payload) into multiple SPDY data fram=
es MAY occur, but does not need to occur? And a WS/SPDY implementation is f=
ree to choose the boundaries within WS frame payload at which to reframe in=
to different SPDY data frames as it likes?</span><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal">A WebSocket frame MAY be split to =
multiple SPDY DATA frames.<u></u><u></u></p></div><div><p class=3D"MsoNorma=
l">Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA fram=
e into multiple DATA frames.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">Is this answering your question?<u></u><u=
></u></p></div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><b=
lockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm =
0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm">
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ar=
ial&quot;,&quot;sans-serif&quot;">2) End of WS Frame<br>The end of a WS mes=
sage is signaled with a SPDY data frame having MSG_DONE set.<br>How is the =
end of a WS frame signaled, given that the WS frame may have been reframed =
into multiple SPDY data frames?</span><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal">When you receive a HEADERS frame, =
you can recognize that the last SPDY frame is that last frame in a WebSocke=
t frame.<u></u><u></u></p></div><div><p class=3D"MsoNormal">When you receiv=
e a SPDY frame with MSG_DONE, you can recognize that it is the last SPDY fr=
ame in a WebSocket message and hence the last SPDY frame in a WebSocket fra=
me.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><blockquote=
 style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm 6=
.0pt;margin-left:4.8pt;margin-right:0cm"><p class=3D"MsoNormal"><span style=
=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&quot;">=
3) WS extension<br>
Are WS extensions fully supported? In particular, is &quot;permessage-defla=
te&quot; supported when running WS/SPDY?</span><u></u><u></u></p></blockquo=
te><div><p class=3D"MsoNormal">Yes. We have reserved bits and we can emulat=
e the opening handshake, I think.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div><p cla=
ss=3D"MsoNormal">On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein &lt;<a hr=
ef=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstei=
n@tavendo.de</a>&gt; wrote:<u></u><u></u></p>
<blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0c=
m 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm"><p class=3D"MsoNormal">=
Hi Takeshi,<br><br>it&#39;s a pity, but I can understand the motivation to =
shove everything into HTTP2.<br>
<br>Anyway, some questions:<br><br>1) Frame Mapping<br>&quot;&quot;&quot;<b=
r>One WebSocket frame is mapped into one preceding SPDY HEADERS frame and f=
ollowing plural SPDY data frames. The SPDY HEADERS frame must contain WebSo=
cket frame fields and following SPDY data frames contain payload data. Thes=
e data frames are free to be reframed.<br>
&quot;&quot;&quot;<br><a href=3D"https://docs.google.com/document/d/124vB5N=
V1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank">https://docs.=
google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit</a>=
<br>
<br>So reframing of WebSocket frames (the payload) into multiple SPDY data =
frames MAY occur, but does not need to occur? And a WS/SPDY implementation =
is free to choose the boundaries within WS frame payload at which to refram=
e into different SPDY data frames as it likes?<br>
<br>2) End of WS Frame<br><br>The end of a WS message is signaled with a SP=
DY data frame having MSG_DONE set.<br><br>How is the end of a WS frame sign=
aled, given that the WS frame may have been reframed into multiple SPDY dat=
a frames?<br>
<br>3) WS extension<br><br>Are WS extensions fully supported? In particular=
, is &quot;permessage-deflate&quot; supported when running WS/SPDY?<br><br>=
Thanks!<br>/Tobias<br><br><br>Von: hybi [mailto:<a href=3D"mailto:hybi-boun=
ces@ietf.org" target=3D"_blank">hybi-bounces@ietf.org</a>] Im Auftrag von T=
akeshi Yoshino<br>
Gesendet: Dienstag, 7. Januar 2014 08:36<br>An: <a href=3D"mailto:hybi@ietf=
.org" target=3D"_blank">hybi@ietf.org</a><br>Betreff: [hybi] Discontinuatio=
n of mux standardizaton in favor of WS/HTTP/2.0<u></u><u></u></p><div><div>
<p class=3D"MsoNormal"><br>Hi all,<br><br>We&#39;ve been working on mux ext=
ension to make WebSocket scalable for long time, but as SPDY got widely sup=
ported and adopted quickly, now there seems to be no longer any good reason=
 to continue this work which is WebSocket dedicated.<br>
<br>Scalability is still important issue to address. From now, I&#39;d like=
 to focus on WS/SPDY effort which we&#39;ve been working on in parallel wit=
h the mux extension.<br><br>Please take a look at the latest proposal in th=
is post.<br>
<a href=3D"https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc" ta=
rget=3D"_blank">https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3gl=
c</a><br><br>If you really want the dedicated mux extension than one over S=
PDY, sorry but please take over the work.<u></u><u></u></p>
</div></div><div><div><p class=3D"MsoNormal">______________________________=
_________________<br>hybi mailing list<br><a href=3D"mailto:hybi@ietf.org" =
target=3D"_blank">hybi@ietf.org</a><br><a href=3D"https://www.ietf.org/mail=
man/listinfo/hybi" target=3D"_blank">https://www.ietf.org/mailman/listinfo/=
hybi</a><u></u><u></u></p>
</div></div></blockquote></div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u><=
/p></div></div></div></div></div></div></div></blockquote></div><br></div>

--047d7b6d99b0cb1f1904ef95afbb--

From tobias.oberstein@tavendo.de  Fri Jan 10 00:15:03 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id BD6A21ADFCE for <hybi@ietfa.amsl.com>; Fri, 10 Jan 2014 00:15:03 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id BpG_LnRya3Po for <hybi@ietfa.amsl.com>; Fri, 10 Jan 2014 00:15:01 -0800 (PST)
Received: from EXHUB020-3.exch020.serverdata.net (exhub020-3.exch020.serverdata.net [206.225.164.30]) by ietfa.amsl.com (Postfix) with ESMTP id A6D4D1ADF34 for <hybi@ietf.org>; Fri, 10 Jan 2014 00:15:01 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-3.exch020.serverdata.net ([206.225.164.30]) with mapi; Fri, 10 Jan 2014 00:14:51 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Yutaka Hirano <yhirano@chromium.org>, Roberto Peon <fenix@google.com>
Date: Fri, 10 Jan 2014 00:14:42 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8Nt7RHEkGX4O1+SqCaJHtGOzmMvAAIdQNQ
Message-ID: <634914A010D0B943A035D226786325D4446BE40A18@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com>
In-Reply-To: <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 10 Jan 2014 08:15:03 -0000
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From tobias.oberstein@tavendo.de  Fri Jan 10 00:38:00 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 00BD11AC7F3 for <hybi@ietfa.amsl.com>; Fri, 10 Jan 2014 00:38:00 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id PBLRcR6GvXXT for <hybi@ietfa.amsl.com>; Fri, 10 Jan 2014 00:37:58 -0800 (PST)
Received: from EXHUB020-1.exch020.serverdata.net (exhub020-1.exch020.serverdata.net [206.225.164.28]) by ietfa.amsl.com (Postfix) with ESMTP id A7AF61AD738 for <hybi@ietf.org>; Fri, 10 Jan 2014 00:37:58 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-1.exch020.serverdata.net ([206.225.164.28]) with mapi; Fri, 10 Jan 2014 00:37:48 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Takeshi Yoshino <tyoshino@google.com>
Date: Fri, 10 Jan 2014 00:37:43 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8NNs32KkI7AyDhQqCZPa7GhKSYOgApTvBg
Message-ID: <634914A010D0B943A035D226786325D4446BE40A19@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E0@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZfBMEL1zWx+Eh9+rTJpTYV=q3Be0ZeW-WQQ7NXszharA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E9@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJYyA+-5LPTNPH+uNLGGcHFq1XVBNbEqP=grHpBgTTxN3g@mail.gmail.com>
In-Reply-To: <CAH9hSJYyA+-5LPTNPH+uNLGGcHFq1XVBNbEqP=grHpBgTTxN3g@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 10 Jan 2014 08:38:00 -0000

Pj4xKSBXaGF0IEkgbWVhbnQgaXMga2VlcCB0aGUgV1MgQVBJIChJIHN1cHBvc2UgeW91IG1lYW4g
dGhlIG9uZSBpbiBIVE1MNSksIGJ1dCBsYXllciBXUy1saWtlIG1lc3NhZ2luZyBkaXJlY3RseSBv
biBTUERZLA0KPlNvcnJ5IGZvciBhc2tpbmcgZm9yIGNsYXJpZmljYXRpb24gcmVwZWF0ZWRseS4N
Cj5EbyB5b3UgbWVhbiBsYXllcmluZyBXUyBwcm90b2NvbCBsaWtlIG1lc3NhZ2luZyBvdmVyIEhU
VFAvMS4xIG92ZXIgU1BEWT8gSWYgc28sIGl0IGhhcyBiZWVuIGRpc2N1c3NlZCBhdCBzcGR5LWRl
dkAgYW5kIHdhcyBpbiBvbGRlciBzcGVjIGRyYWZ0Lg0KDQpObywgSSBtZWFuIGEgbW9yZSBlZmZp
Y2llbnQgbWFwcGluZyBvZiB0aGUgV1MgX0FQSV8gdG8gU1BEWSAuLiBzZWUgYmVsb3cgLi4NCsKg
DQo+PmRvaW5nIGF3YXkgd2l0aCB0aGUgcmVmcmFtaW5nL2RvdWJsZS1mcmFtaW5nIHdpdGggV1Mv
U1BEWQ0KPkhtbSwgY3VycmVudCBwcm9wb3NhbCBpcyBraW5kYSBkaXJlY3QgdXNlIG9mIFNQRFku
IFBsZWFzZSBlbGFib3JhdGUuDQoNClRoZSBjdXJyZW50IHByb3Bvc2FsIHNob2Vob3JucyBXUyBp
bnRvIFNQRFkgLi4gZWcuIGFsbCB0aGF0IGluZWZmaWNpZW50IG1hcHBpbmcgb2Ygc3RyaW5nLWJh
c2VkIGtleS12YWx1ZSBwYWlycyBvZiBXUyBmcmFtZSBoZWFkZXJzLg0KDQpXaHkgbm90IG1hcCBX
UyBtZXNzYWdlcyBkaXJlY3RseSBpbnRvIFNQRFkgZGF0YSBmcmFtZXMgbGlrZSBzbzoNCg0KSGF2
ZSAyIGJpdHMgaW4gU1BEWSBkYXRhIGZyYW1lczoNCg0KMHgwMyA9IE1TR19UWVBFIDogMCA9PSB0
ZXh0LCAxID09IGJpbmFyeQ0KMHgwNCA9IE1TR19ET05FDQoNCkEgV1MgdGV4dCBtZXNzYWdlIHdv
dWxkIHNpbXBseSBiZSBtYXBwZWQgaW50byBhIHNlcmllcyBvZiBTUERZIGRhdGEgZnJhbWVzOg0K
DQpNU0dfVFlQRSA9IDAsIE1TR19ET05FID0gMA0KTVNHX1RZUEUgPSAwLCBNU0dfRE9ORSA9IDAN
Ci4uLg0KTVNHX1RZUEUgPSAwLCBNU0dfRE9ORSA9IDENCg0KU2ltaWxhciBmb3IgYSBXUyBiaW5h
cnkgbWVzc2FnZS4NCsKgDQpEb2luZyBsaWtlIGFib3ZlIHdpbGwgYWxsb3cgdG8gcHJlc2VydmUg
dGhlIGN1cnJlbnQgV1MgQVBJIGluIEhUTUw1ICh3aGljaCBuZWVkcyBtZXNzYWdlcywgYW5kIG5l
ZWRzIDIgbWVzc2FnZSB0eXBlcykuDQoNCkl0IHdvbid0IGFsbG93IHRvIHByZXNlcnZlIFdTIHBy
b3RvY29sIGZlYXR1cmVzIGFzIGluIFJGQzY0NTUuIFBlcnNvbmFsbHksIEkgd291bGRuJ3QgY2Fy
ZSBzbyBtdWNoLg0KDQpUaGUgcG9pbnQgaXM6IFdTL1NQRFksIHRha2luZyB0b2dldGhlciB0aGUg
c3BlY3MgaW52b2x2ZWQsIHdvdWxkIHNlZW0gbGlrZSBhbiBpbmNyZWRpYmxlIGNvbXBsZXggZGVz
aWduIGZvciBhIHNpbXBsZSB0aGluZzogbWVzc2FnZXMgb2YgMiB0eXBlcyBvdmVyIGEgYnl0ZSBz
dHJlYW0uDQoNCj09DQoNCkkgd2lsbCBjb250aW51ZSB0aGUgb3RoZXIgZGlzY3Vzc2lvbiBpbiBh
IHNlcGFyYXRlIG1haWwgLi4NCg0KL1RvYmlhcw0K

From arman@noemax.com  Fri Jan 10 00:40:39 2014
Return-Path: <arman@noemax.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 47F841ADF71 for <hybi@ietfa.amsl.com>; Fri, 10 Jan 2014 00:40:39 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: 0.161
X-Spam-Level: 
X-Spam-Status: No, score=0.161 tagged_above=-999 required=5 tests=[BAYES_50=0.8, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id s67Td1P1j2AQ for <hybi@ietfa.amsl.com>; Fri, 10 Jan 2014 00:40:37 -0800 (PST)
Received: from mail.noemax.com (mail.noemax.com [74.208.113.37]) by ietfa.amsl.com (Postfix) with ESMTP id 20B881AD738 for <hybi@ietf.org>; Fri, 10 Jan 2014 00:40:37 -0800 (PST)
DKIM-Signature: a=rsa-sha1; t=1389343172; x=1389947972; s=m2048; d=noemax.com; c=relaxed/relaxed; v=1; bh=6FDV6PlkYd/ByeSGWK97JyaRG/s=; h=From:Subject:Date:Message-ID:To:MIME-Version:Content-Type:Content-Transfer-Encoding; b=noAsTowqr1F4bhpZcKlF0UO1l6tAtAtvFSjdsQQQ++6h+qMwCtHjksgM2O0vw+KCXadLoF6GP7WJEiZpTwrJM0pLzPPNRtj90EA8B/gVOQYExFUnuSoFXSlG0pC5X+wMIyOaPn1N/dYWdjr0vcqew1nt89+4F4PJ2z3Z5LOuVet6FlvQS7dJ5roTpvGXO0yUv0TOldX2N2M/dpd2HZlLviC+RbHFgqRU0fKWxZnqdfnW2A6AGGRD9e4SjhFCcJAVczBSIOLQChoE27SbLfDMn5WVIkP6DnR3aK6/P6j2jOeaqimkXrN0e9C4j8L9AM4twVyFjJ4oe7xZMTMhvyrigw==
Received: from mail.noemax.com by mail.noemax.com (Noemax Mail Server) with ASMTP (SSL) id 201401101039310847 for <hybi@ietf.org>; Fri, 10 Jan 2014 10:39:31 +0200
From: "Arman Djusupov" <arman@noemax.com>
To: <hybi@ietf.org>
Date: Fri, 10 Jan 2014 10:40:12 +0200
Message-ID: <003c01cf0ddf$985d4070$c917c150$@noemax.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable
X-Mailer: Microsoft Outlook 14.0
Thread-Index: Ac8N3zH9c8r1oFHcSmOWzMgUhTa7lg==
Content-Language: en-us
Subject: [hybi] Permessage-deflate final frame
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 10 Jan 2014 08:40:39 -0000

Oops, apologies for inadvertently hijacking the thread, I am resending =
under a new topic. Arman


Hello Takeshi,

In the current draft, the description of the compression process states =
:

   1.  Compress all the octets of the payload of the message using
       DEFLATE.

This doesn=E2=80=99t take into account outbound messages of arbitrary =
size that NEED to be fragmented, for which it is more favorable to use =
fragmentation and to compress & send fragments until the end of the =
message is reached.

In such cases the implementation buffers the input from its source and =
then flushes it out into a compressed fragment, repeating this process =
until the end of the source data is reached. At that point it flushes =
any remaining bytes buffered into a final frame, or flushes a final =
frame of 0 length.


When the implementation produces compressed fragments, it periodically =
produces frames with 0x00 0x00 0xFF 0xFF at the end of the frame due to =
flushing DEFLATE. But because of the following requirements:

   2. If the resulting data does not end with an empty DEFLATE block
       with no compression (the "BTYPE" bits is set to 00), append an
       empty DEFLATE block with no compression to the tail end.

   3.  Remove 4 octets (that are 0x00 0x00 0xff 0xff) from the tail end.
       After this step, the last octet of the compressed data contains
       (possibly part of) the DEFLATE header bits with the "BTYPE" bits
       set to 00.

the implementation must ensure that the message ends empty blocks =
without a trailing 0x0000FFFF, so the implementation must keep track on =
what type of block was at the end of the frame that was sent last. If =
the last frame sent ends with 0x0000FFFF then the implementation cannot =
remove those 4 bytes from the wire, but it must artificially produce an =
empty DEFLATE block to send as the final frame. This is only due to the =
requirement to remove 0x0000FFFF.

Wouldn=E2=80=99t it be easier to make the removal 0x00 0x00 0xFF 0xFF =
optional, while at the same time requiring that the receiving side =
appends a 0x00 0x00 0xFF 0xFF to the final frame in case when its =
missing?

It is much simpler to just check the final bytes of the message instead =
of having to remember the state of the previous frame.

With best regards,
Arman


From tobias.oberstein@tavendo.de  Fri Jan 10 02:29:53 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 0B2781AE004 for <hybi@ietfa.amsl.com>; Fri, 10 Jan 2014 02:29:53 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id OCBb-eULKx72 for <hybi@ietfa.amsl.com>; Fri, 10 Jan 2014 02:29:52 -0800 (PST)
Received: from EXHUB020-1.exch020.serverdata.net (exhub020-1.exch020.serverdata.net [206.225.164.28]) by ietfa.amsl.com (Postfix) with ESMTP id CAA2D1ADFEF for <hybi@ietf.org>; Fri, 10 Jan 2014 02:29:51 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-1.exch020.serverdata.net ([206.225.164.28]) with mapi; Fri, 10 Jan 2014 02:29:41 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Takeshi Yoshino <tyoshino@google.com>
Date: Fri, 10 Jan 2014 02:29:35 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8NNs32KkI7AyDhQqCZPa7GhKSYOgAtl3DQ
Message-ID: <634914A010D0B943A035D226786325D4446BE40A1B@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E0@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZfBMEL1zWx+Eh9+rTJpTYV=q3Be0ZeW-WQQ7NXszharA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E9@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJYyA+-5LPTNPH+uNLGGcHFq1XVBNbEqP=grHpBgTTxN3g@mail.gmail.com>
In-Reply-To: <CAH9hSJYyA+-5LPTNPH+uNLGGcHFq1XVBNbEqP=grHpBgTTxN3g@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 10 Jan 2014 10:29:53 -0000
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From tobias.oberstein@tavendo.de  Fri Jan 10 13:12:12 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 2BE051AE19D for <hybi@ietfa.amsl.com>; Fri, 10 Jan 2014 13:12:12 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ba_M-pLhSbIG for <hybi@ietfa.amsl.com>; Fri, 10 Jan 2014 13:12:10 -0800 (PST)
Received: from EXHUB020-5.exch020.serverdata.net (exhub020-5.exch020.serverdata.net [206.225.164.32]) by ietfa.amsl.com (Postfix) with ESMTP id 51C791AE192 for <hybi@ietf.org>; Fri, 10 Jan 2014 13:12:10 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-5.exch020.serverdata.net ([206.225.164.32]) with mapi; Fri, 10 Jan 2014 13:12:00 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Takeshi Yoshino <tyoshino@google.com>, "hybi@ietf.org" <hybi@ietf.org>
Date: Fri, 10 Jan 2014 13:11:50 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8LezfCoM4V94vlQsauRUb5fEzOMwCyPYpw
Message-ID: <634914A010D0B943A035D226786325D4446BED91E4@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com>
In-Reply-To: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 10 Jan 2014 21:12:12 -0000
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From tyoshino@google.com  Mon Jan 13 20:05:41 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C12D01AE1C9 for <hybi@ietfa.amsl.com>; Mon, 13 Jan 2014 20:05:41 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Wf-yiA8Kl_qs for <hybi@ietfa.amsl.com>; Mon, 13 Jan 2014 20:05:39 -0800 (PST)
Received: from mail-wg0-x232.google.com (mail-wg0-x232.google.com [IPv6:2a00:1450:400c:c00::232]) by ietfa.amsl.com (Postfix) with ESMTP id 26B9D1AE1BF for <hybi@ietf.org>; Mon, 13 Jan 2014 20:05:38 -0800 (PST)
Received: by mail-wg0-f50.google.com with SMTP id l18so6439850wgh.17 for <hybi@ietf.org>; Mon, 13 Jan 2014 20:05:27 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:from:date:message-id:subject:to:cc:content-type; bh=Tnbw7Y2UqrNnLB+N5ZC4fgPa3fBpN6nMBU/SIck+FZw=; b=UzWki2Z1f4CUuCOLmdRxSAt+BdlNL9wNHWjmxo04WuL7IwYwu32ae3DsMbF6/umx2v rHxNZEV00/gfY26ueBMYtrnqzH38DBthhJVQiIeV9PVIcbkCef32RC6vqOWVC2++sQXS oWkfQFEOxcuLAFi0iv0cJDozwuaBtPV8i5FJK0etZ0FpBBhYpVvxky3c4bUiwbGzv7bL gt7YIlbj7ikVS6Dm/+wSWtoSeAI3NDwAvfsbCLmz/1OtMqtB+LpEliFd7ZMHfRu8+7Uj 4HIWTTAsu2Ofsi/giLXeMhSFWwvf/cpK+55JwevPmg+NY/Uc6+xsxyWFMgV1WMJNLoz3 fjEg==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:from:date:message-id:subject:to:cc :content-type; bh=Tnbw7Y2UqrNnLB+N5ZC4fgPa3fBpN6nMBU/SIck+FZw=; b=JaOgAuEDVNPzWJRRpJ6U9M1VZ/ON/sYcELHPnZbTTS6AnhPW9StgY3cO9Un7heXejY 9bCXhNOljeCtXuwS55JwuBQlNJ9hDUsiO1vUYSePTBgRq68745Iu2G4JoQuszCTHXFbE mzGGoP6dzy2aKdgksoRZW6TFlFuF3A+EOnwiyE0xLNEpuDcrwybfR9NrMVJFCb56GnBp kqQE/oPZoxZDP/xvVXx/rUA+YN8Sx26A/K5qXCp7vie4SWEWyqF38wfXr8Pmj36pPUXM A3hZEwxxt2YV+TsJDSfOsEXBLbiPQpVGsHO540ahuGK8Pc40jqbT8NFx9iSfStB9+uTM rg3w==
X-Gm-Message-State: ALoCoQm7ZixlRNOyI6TnMR6yOZAjkU2daCykISDyUD+Jx9bffuMCJTtGbAPujU1ykQpEh4VNDvhzrummwYvCT1mj4NsAnxLqyPK7IgCTaEQ/Jw+lhhQW0LT2ZgDFN3hq+tancTc7hpeoGe4yGFufgIIO30UhVh/xy3zw3wodE1hfXTUZ4aZxgIB4HU96/d2JPtYH0Q8VI2WF
X-Received: by 10.180.107.136 with SMTP id hc8mr16437330wib.11.1389672327483;  Mon, 13 Jan 2014 20:05:27 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Mon, 13 Jan 2014 20:05:07 -0800 (PST)
From: Takeshi Yoshino <tyoshino@google.com>
Date: Tue, 14 Jan 2014 13:05:07 +0900
Message-ID: <CAH9hSJbeY5VOY_iuwrdBq-KYcoVkW_8KArPp70hP4tdZj6eQfg@mail.gmail.com>
To: Arman Djusupov <arman@noemax.com>
Content-Type: multipart/alternative; boundary=e89a8f3bad456f293504efe64deb
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: [hybi] [permessage-deflate] Compressing fragmented data
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 14 Jan 2014 04:05:42 -0000

--e89a8f3bad456f293504efe64deb
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Hi Arman,

Let me change the subject.

On Thu, Jan 9, 2014 at 11:46 PM, Arman Djusupov <arman@noemax.com> wrote:

> Hello Takeshi,
>
>
>
> In the current draft, the description of the compression process states :
>
>
>
>    1.  Compress all the octets of the payload of the message using
>
>        DEFLATE.
>
>
>
> This doesn=E2=80=99t take into account outbound messages of arbitrary siz=
e that
> NEED to be fragmented, for which it is more favorable to use
> fragmentation and to compress & send fragments until the end of the
> message is reached.
>

OK. I'll add some text to explain how this works for fragmented messages.


>
>
> In such cases the implementation buffers the input from its source and
> then flushes it out into a compressed fragment, repeating this processunt=
il the end of the source data is
> reached. At that point it flushes any remaining bytes buffered into a
> final frame, or flushes a final frame of 0 length.
>
>
>
>
>
> When the implementation produces compressed fragments, it periodically
> produces frames with 0x00 0x00 0xFF 0xFF at the end of the frame due to
> flushing DEFLATE. But because of the following requirements:
>
>
>
>    2. If the resulting data does not end with an empty DEFLATE block
>
>        with no compression (the "BTYPE" bits is set to 00), append an
>
>        empty DEFLATE block with no compression to the tail end.
>
>
>
>    3.  Remove 4 octets (that are 0x00 0x00 0xff 0xff) from the tail end.
>
>        After this step, the last octet of the compressed data contains
>
>        (possibly part of) the DEFLATE header bits with the "BTYPE" bits
>
>        set to 00.
>
>
>
> the implementation must ensure that the message ends empty blocks without=
 a
> trailing 0x0000FFFF, so the implementation must keep track on what type
> of block was at the end of the frame that was sent last. If the last fram=
e
> sent ends with 0x0000FFFF then the implementation cannot remove those 4
> bytes from the wire, but it must artificially produce an empty DEFLATE
> block to send as the final frame. This is only due to the requirement to
> remove 0x0000FFFF.
>

Yes. It's just two byte long. DEFLATE block header (3bit) + Fixed Huffman
code for end of block symbol (7bit) + padding (6bit).


>
>

>
> Wouldn=E2=80=99t it be easier to make the removal 0x00 0x00 0xFF 0xFF opt=
ional,
> while at the same time requiring that the receiving side appends a 0x00
> 0x00 0xFF 0xFF to the final frame in case when its missing?
>
>
>

I think it complicates decoders. Code which doesn't know DEFLATE
decompressor's state cannot check if 0x00 0x00 0xFF 0xFF at the end is
really an uncompressed block body or not. It could be something else, e.g.
body of compressed block.


>  It is much simpler to just check the final bytes of the message instead
> of having to remember the state of the previous frame.
>

--e89a8f3bad456f293504efe64deb
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">Hi A=
rman,</div><div class=3D"gmail_quote"><br></div><div class=3D"gmail_quote">=
Let me change the subject.</div><div class=3D"gmail_quote"><br></div><div c=
lass=3D"gmail_quote">

On Thu, Jan 9, 2014 at 11:46 PM, Arman Djusupov <span dir=3D"ltr">&lt;<a hr=
ef=3D"mailto:arman@noemax.com" target=3D"_blank">arman@noemax.com</a>&gt;</=
span> wrote:<br><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8e=
x;border-left:1px #ccc solid;padding-left:1ex">

<div lang=3D"EN-US" link=3D"blue" vlink=3D"purple"><div><p class=3D"MsoNorm=
al"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;s=
ans-serif&quot;;color:#1f497d">Hello Takeshi,<u></u><u></u></span></p><p cl=
ass=3D"MsoNormal">

<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></span></p><p class=3D"MsoNo=
rmal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot=
;sans-serif&quot;;color:#1f497d">In the current draft, t</span><span style=
=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;=
;color:#44546a">he description of </span><span style=3D"font-size:11.0pt;fo=
nt-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">the </s=
pan><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;s=
ans-serif&quot;;color:#44546a">compression process states :<u></u><u></u></=
span></p>

<pre><u></u>=C2=A0<u></u></pre><pre>=C2=A0=C2=A0 1.=C2=A0 Compress all the =
octets of the payload of the message using<u></u><u></u></pre><pre>=C2=A0=
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 DEFLATE.<u></u><u></u></pre><p class=3D"MsoN=
ormal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quo=
t;sans-serif&quot;;color:#44546a"><u></u>=C2=A0<u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">T</span><span style=3D"fo=
nt-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color=
:#44546a">his doesn=E2=80=99t</span><span style=3D"font-size:11.0pt;font-fa=
mily:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"> take</span>=
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#44546a"> into account </span><span style=3D"font-size:11=
.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">=
outbound </span><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&q=
uot;,&quot;sans-serif&quot;;color:#44546a">messages of arbitrary size that =
NEED to be fragmented, </span><span style=3D"font-size:11.0pt;font-family:&=
quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">for which it is</s=
pan><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;s=
ans-serif&quot;;color:#44546a"> more favorable to use fragmentation and </s=
pan><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;s=
ans-serif&quot;;color:#1f497d">to </span><span style=3D"font-size:11.0pt;fo=
nt-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a">compres=
s</span><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&qu=
ot;sans-serif&quot;;color:#1f497d"> &amp; send</span><span style=3D"font-si=
ze:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#445=
46a"> fragments until the </span><span style=3D"font-size:11.0pt;font-famil=
y:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">end of the mess=
age is reached</span><span style=3D"font-size:11.0pt;font-family:&quot;Cali=
bri&quot;,&quot;sans-serif&quot;;color:#44546a">.</span></p>

</div></div></blockquote><div><br></div><div>OK. I&#39;ll add some text to =
explain how this works for fragmented messages.</div><div>=C2=A0</div><bloc=
kquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px #cc=
c solid;padding-left:1ex">

<div lang=3D"EN-US" link=3D"blue" vlink=3D"purple"><div><p class=3D"MsoNorm=
al"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;s=
ans-serif&quot;;color:#44546a"><u></u><u></u></span></p><p class=3D"MsoNorm=
al"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;s=
ans-serif&quot;;color:#44546a"><u></u>=C2=A0<u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">In such cases the </span>=
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#44546a">implementation buffers </span><span style=3D"fon=
t-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:=
#1f497d">the in</span><span style=3D"font-size:11.0pt;font-family:&quot;Cal=
ibri&quot;,&quot;sans-serif&quot;;color:#44546a">put </span><span style=3D"=
font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;col=
or:#1f497d">from its source </span><span style=3D"font-size:11.0pt;font-fam=
ily:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a">and then flus=
hes it out into a compressed fragment</span><span style=3D"font-size:11.0pt=
;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">, re=
peating this process</span><span style=3D"font-size:11.0pt;font-family:&quo=
t;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a"> until the end of the=
 source data </span><span style=3D"font-size:11.0pt;font-family:&quot;Calib=
ri&quot;,&quot;sans-serif&quot;;color:#1f497d">is </span><span style=3D"fon=
t-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:=
#44546a">reached</span><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">. At that point </span><s=
pan style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-se=
rif&quot;;color:#44546a">it </span><span style=3D"font-size:11.0pt;font-fam=
ily:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">flushes any r=
emaining bytes buffered into a final frame, or </span><span style=3D"font-s=
ize:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44=
546a">flushes </span><span style=3D"font-size:11.0pt;font-family:&quot;Cali=
bri&quot;,&quot;sans-serif&quot;;color:#1f497d">a</span><span style=3D"font=
-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#=
44546a"> final frame of 0 length.<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a"><u></u>=C2=A0<u></u=
></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#44546a">When </span><span style=
=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;=
;color:#1f497d">the </span><span style=3D"font-size:11.0pt;font-family:&quo=
t;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a">implementation produc=
es compressed </span><span style=3D"font-size:11.0pt;font-family:&quot;Cali=
bri&quot;,&quot;sans-serif&quot;;color:#1f497d">fragments,</span><span styl=
e=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot=
;;color:#44546a"> it periodically produces frames with 0x00 0x00 0xFF 0xFF =
at the end of the frame</span><span style=3D"font-size:11.0pt;font-family:&=
quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"> due to flushing D=
EFLATE</span><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot=
;,&quot;sans-serif&quot;;color:#44546a">. But </span><span style=3D"font-si=
ze:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f4=
97d">because of the</span><span style=3D"font-size:11.0pt;font-family:&quot=
;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a"> following requirement=
s:<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#44546a"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&q=
uot;Courier New&quot;">=C2=A0=C2=A0 2. If the resulting data does not end w=
ith an empty DEFLATE block<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Co=
urier New&quot;">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 with no compression (=
the &quot;BTYPE&quot; bits is set to 00), append an<u></u><u></u></span></p=
><p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;C=
ourier New&quot;">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 empty DEFLATE block =
with no compression to the tail end.<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Co=
urier New&quot;"><u></u>=C2=A0<u></u></span></p><p class=3D"MsoNormal"><spa=
n style=3D"font-size:10.0pt;font-family:&quot;Courier New&quot;">=C2=A0=C2=
=A0 3.=C2=A0 Remove 4 octets (that are 0x00 0x00 0xff 0xff) from the tail e=
nd.<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Co=
urier New&quot;">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 After this step, the =
last octet of the compressed data contains<u></u><u></u></span></p><p class=
=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Courier Ne=
w&quot;">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 (possibly part of) the DEFLAT=
E header bits with the &quot;BTYPE&quot; bits<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Co=
urier New&quot;">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 set to 00.<u></u><u><=
/u></span></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-f=
amily:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a"><u></u>=C2=
=A0<u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">the </span><span style=3D=
"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;co=
lor:#44546a">implementation must </span><span style=3D"font-size:11.0pt;fon=
t-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">e</span>=
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#44546a">nsure that the message ends empty block</span><s=
pan style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-se=
rif&quot;;color:#1f497d">s</span><span style=3D"font-size:11.0pt;font-famil=
y:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a"> without </span=
><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans=
-serif&quot;;color:#1f497d">a trailing</span><span style=3D"font-size:11.0p=
t;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a">=C2=
=A00x0000FFFF, so </span><span style=3D"font-size:11.0pt;font-family:&quot;=
Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">the </span><span style=
=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;=
;color:#44546a">implementation must keep track on what type of block was at=
 the end of the frame that was sent last. If the last frame sent ends with =
0x0000FFFF</span><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&=
quot;,&quot;sans-serif&quot;;color:#1f497d"> then</span><span style=3D"font=
-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#=
44546a"> </span><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&q=
uot;,&quot;sans-serif&quot;;color:#1f497d">the </span><span style=3D"font-s=
ize:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44=
546a">implementation cannot remove those 4 bytes from the wire</span><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&=
quot;;color:#1f497d">, but </span><span style=3D"font-size:11.0pt;font-fami=
ly:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a">it must artifi=
cially produce an empty DEFLATE block to send </span><span style=3D"font-si=
ze:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f4=
97d">as the</span><span style=3D"font-size:11.0pt;font-family:&quot;Calibri=
&quot;,&quot;sans-serif&quot;;color:#44546a"> final frame. </span><span sty=
le=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quo=
t;;color:#1f497d">This is only due to the requirement to remove </span><spa=
n style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-seri=
f&quot;;color:#44546a">0x0000FFFF.</span></p>

</div></div></blockquote><div><br></div><div>Yes. It&#39;s just two byte lo=
ng. DEFLATE block header (3bit) + Fixed Huffman code for end of block symbo=
l (7bit) + padding (6bit).</div><div>=C2=A0</div><blockquote class=3D"gmail=
_quote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:=
1ex">

<div lang=3D"EN-US" link=3D"blue" vlink=3D"purple"><div><p class=3D"MsoNorm=
al">=C2=A0</p></div></div></blockquote><blockquote class=3D"gmail_quote" st=
yle=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div =
lang=3D"EN-US" link=3D"blue" vlink=3D"purple">

<div><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&qu=
ot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a"><u></u><u></u></span=
></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&qu=
ot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a"><u></u>=C2=A0<u></u>=
</span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#44546a">Wouldn=E2=80=99t it be ea=
sier to make </span><span style=3D"font-size:11.0pt;font-family:&quot;Calib=
ri&quot;,&quot;sans-serif&quot;;color:#1f497d">the </span><span style=3D"fo=
nt-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color=
:#44546a">remov</span><span style=3D"font-size:11.0pt;font-family:&quot;Cal=
ibri&quot;,&quot;sans-serif&quot;;color:#1f497d">al</span><span style=3D"fo=
nt-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color=
:#44546a"> 0x00 0x00 0xFF 0xFF optional</span><span style=3D"font-size:11.0=
pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">, =
while at the same time requiring that the </span><span style=3D"font-size:1=
1.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a"=
>receiving side a</span><span style=3D"font-size:11.0pt;font-family:&quot;C=
alibri&quot;,&quot;sans-serif&quot;;color:#1f497d">ppends a</span><span sty=
le=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quo=
t;;color:#44546a"> 0x00 0x00 0xFF 0xFF </span><span style=3D"font-size:11.0=
pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">to=
</span><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quo=
t;sans-serif&quot;;color:#44546a"> the final frame in case when </span><spa=
n style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-seri=
f&quot;;color:#1f497d">its</span><span style=3D"font-size:11.0pt;font-famil=
y:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a"> missing</span>=
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#1f497d">?</span><span style=3D"font-size:11.0pt;font-fam=
ily:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a"><u></u><u></u=
></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#44546a"><u></u>=C2=A0</span></p><=
/div></div></blockquote><div><br></div><div>I think it complicates decoders=
. Code which doesn&#39;t know DEFLATE decompressor&#39;s state cannot check=
 if 0x00 0x00 0xFF 0xFF at the end is really an uncompressed block body or =
not. It could be something else, e.g. body of compressed block.</div>

<div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8=
ex;border-left:1px #ccc solid;padding-left:1ex"><div lang=3D"EN-US" link=3D=
"blue" vlink=3D"purple"><p class=3D"MsoNormal"><span style=3D"font-size:11.=
0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a"><=
u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#44546a">It </span><span style=3D"=
font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;col=
or:#1f497d">is </span><span style=3D"font-size:11.0pt;font-family:&quot;Cal=
ibri&quot;,&quot;sans-serif&quot;;color:#44546a">much simpler to </span><sp=
an style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-ser=
if&quot;;color:#1f497d">just</span><span style=3D"font-size:11.0pt;font-fam=
ily:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a"> check </span=
><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans=
-serif&quot;;color:#1f497d">the </span><span style=3D"font-size:11.0pt;font=
-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a">final byt=
es of the message </span><span style=3D"font-size:11.0pt;font-family:&quot;=
Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">instead of</span><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&=
quot;;color:#44546a"> </span><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">having to </span><s=
pan style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-se=
rif&quot;;color:#44546a">remember the state of </span><span style=3D"font-s=
ize:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f=
497d">the </span><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&=
quot;,&quot;sans-serif&quot;;color:#44546a">previous frame.</span></p>

</div></blockquote></div></div></div>

--e89a8f3bad456f293504efe64deb--

From internet-drafts@ietf.org  Mon Jan 13 20:30:09 2014
Return-Path: <internet-drafts@ietf.org>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 533DF1AD8F9; Mon, 13 Jan 2014 20:30:09 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 8ncgtwQUi1al; Mon, 13 Jan 2014 20:30:07 -0800 (PST)
Received: from ietfa.amsl.com (localhost [IPv6:::1]) by ietfa.amsl.com (Postfix) with ESMTP id 9B24B1ADDD0; Mon, 13 Jan 2014 20:30:06 -0800 (PST)
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable
From: internet-drafts@ietf.org
To: i-d-announce@ietf.org
X-Test-IDTracker: no
X-IETF-IDTracker: 4.90.p2
Auto-Submitted: auto-generated
Precedence: bulk
Message-ID: <20140114043006.25016.82793.idtracker@ietfa.amsl.com>
Date: Mon, 13 Jan 2014 20:30:06 -0800
Cc: hybi@ietf.org
Subject: [hybi] I-D Action: draft-ietf-hybi-permessage-compression-16.txt
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 14 Jan 2014 04:30:09 -0000

A New Internet-Draft is available from the on-line Internet-Drafts director=
ies.
 This draft is a work item of the BiDirectional or Server-Initiated HTTP Wo=
rking Group of the IETF.

        Title           : Compression Extensions for WebSocket
        Author          : Takeshi Yoshino
	Filename        : draft-ietf-hybi-permessage-compression-16.txt
	Pages           : 32
	Date            : 2014-01-13

Abstract:
   This document specifies a framework for creating WebSocket extensions
   that add compression functionality to the WebSocket Protocol.  An
   extension based on this framework compresses the payload data portion
   of non-control WebSocket messages on a per-message basis using
   parameters negotiated during the opening handshake.  This framework
   provides a general method to apply a compression algorithm to the
   contents of WebSocket messages.  For each compression algorithm, an
   extension is defined by specifying parameter negotiation and
   compression algorithm in detail.  This document also specifies one
   specific compression extension using the DEFLATE algorithm.

   Please send feedback to the hybi@ietf.org mailing list.


The IETF datatracker status page for this draft is:
https://datatracker.ietf.org/doc/draft-ietf-hybi-permessage-compression/

There's also a htmlized version available at:
http://tools.ietf.org/html/draft-ietf-hybi-permessage-compression-16

A diff from the previous version is available at:
http://www.ietf.org/rfcdiff?url2=3Ddraft-ietf-hybi-permessage-compression-16


Please note that it may take a couple of minutes from the time of submission
until the htmlized version and diff are available at tools.ietf.org.

Internet-Drafts are also available by anonymous FTP at:
ftp://ftp.ietf.org/internet-drafts/


From tyoshino@google.com  Mon Jan 13 20:33:13 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C1F981AD8F9 for <hybi@ietfa.amsl.com>; Mon, 13 Jan 2014 20:33:13 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id UJAeiRLXDzze for <hybi@ietfa.amsl.com>; Mon, 13 Jan 2014 20:33:12 -0800 (PST)
Received: from mail-wi0-x234.google.com (mail-wi0-x234.google.com [IPv6:2a00:1450:400c:c05::234]) by ietfa.amsl.com (Postfix) with ESMTP id 23D481ACCF8 for <hybi@ietf.org>; Mon, 13 Jan 2014 20:33:11 -0800 (PST)
Received: by mail-wi0-f180.google.com with SMTP id d13so203046wiw.7 for <hybi@ietf.org>; Mon, 13 Jan 2014 20:33:00 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :content-type; bh=VKLjKvPiEltqCf3NhjEm7Eu6MVVduEU3rShxmuu3FdE=; b=EegRFAIkAXY3547XLpeH5R5xNkvOnK3wJQlfYuZbAA9B36I4uhEK30png6au4gVkM7 zJSvIx1RkqTkVo9CQiwZ2EkJjbOJeQJLMiy5n83I2WRRRvGaAqgaXwhwe0Jc72ZuYpDz 5oaejMIVuTZ50caVtl579R538iqjCWEvYYiIjERC6gYu+ljaTy1D7KTr7/rbG7bd2PVI Tv6Fn1qqvLiJSseWNOm84z2FXUTdVI/oydidmKIi9mJlqktWc90Jb3rmL4PSr1h2iy8B CzqB02T/q85G+sFVgpbkF6vNGUd19rYiF6ofrLagXzbOTKGfsaJtMPiKtrqvOjDc6uKJ PZRg==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:content-type; bh=VKLjKvPiEltqCf3NhjEm7Eu6MVVduEU3rShxmuu3FdE=; b=A0CJNDOcJkYlhEatU+k935OGxhrviHBl5HvUTDSoaqNVxA8ELNxHawK6xrolnZ2x3V mF4nDTz+BDK11lCJDG0/qm7fhlr3QakB58tXM2WtCoiufW5GOnH9P9zx5dUuzlTky7sq oPXIB86JGdP/Izuxz59PBpzYGXYeH5Na7i7oMGlcSufLrol0dNb2K2iBkkAvTDCnGkm9 qDy/Rt79VIYcnQQpXGKl4Kl/SzL8nifUq+jEiJCHzfl9YpnWmUfbtwdoOsJCFqqEaWpC FgJKqgWGsuIomk3UfEw5qSuusdmAPKkyi8dKCiPc9nnvK37Ho5TgylAgW7q/b8W5LL8H f4bQ==
X-Gm-Message-State: ALoCoQkuIn2v5CAOn5tMj1iQNekWTzbsGL9kw4VudN2YcHWy5q+L3m6hpTD2jpP3odArf1iqCWrhrX6lIxc64fmt60YlDnjenjCHlqkOlgkAejU7zovPNnrFJkyiFphdRUx553ZYmF4gcIc30TMKRwmasF+3heXc9OJL2f6AtMdJpCr95b0YXMATZGYg4PbvBcDY3IxdKqLr
X-Received: by 10.180.188.197 with SMTP id gc5mr723740wic.30.1389673980511; Mon, 13 Jan 2014 20:33:00 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Mon, 13 Jan 2014 20:32:40 -0800 (PST)
In-Reply-To: <20140114043006.25016.82793.idtracker@ietfa.amsl.com>
References: <20140114043006.25016.82793.idtracker@ietfa.amsl.com>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Tue, 14 Jan 2014 13:32:40 +0900
Message-ID: <CAH9hSJadk51G=-C5x02zn+uoJvvi9rxqoG8E6tNNu9_+t4PjNg@mail.gmail.com>
To: "hybi@ietf.org" <hybi@ietf.org>
Content-Type: multipart/alternative; boundary=001a11c3841af6667504efe6af70
Subject: Re: [hybi] I-D Action: draft-ietf-hybi-permessage-compression-16.txt
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 14 Jan 2014 04:33:14 -0000

--001a11c3841af6667504efe6af70
Content-Type: text/plain; charset=UTF-8

Added texts to explain how DEFLATE works with fragmentation.
Per Arman's request
http://www.ietf.org/mail-archive/web/hybi/current/msg10283.html

--001a11c3841af6667504efe6af70
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra">Added texts to explain how DEFL=
ATE works with fragmentation.</div><div class=3D"gmail_extra">Per Arman&#39=
;s request</div><div class=3D"gmail_extra"><a href=3D"http://www.ietf.org/m=
ail-archive/web/hybi/current/msg10283.html">http://www.ietf.org/mail-archiv=
e/web/hybi/current/msg10283.html</a><br>

</div></div>

--001a11c3841af6667504efe6af70--

From yhirano@google.com  Tue Jan 14 02:16:05 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id CD6111AE07B for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 02:16:05 -0800 (PST)
X-Quarantine-ID: <v3nMkOSKyuA9>
X-Virus-Scanned: amavisd-new at amsl.com
X-Amavis-Alert: BANNED, message contains text/plain,.exe
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id v3nMkOSKyuA9 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 02:16:03 -0800 (PST)
Received: from mail-qc0-x236.google.com (mail-qc0-x236.google.com [IPv6:2607:f8b0:400d:c01::236]) by ietfa.amsl.com (Postfix) with ESMTP id F0BC31AE070 for <hybi@ietf.org>; Tue, 14 Jan 2014 02:16:02 -0800 (PST)
Received: by mail-qc0-f182.google.com with SMTP id c9so6003852qcz.13 for <hybi@ietf.org>; Tue, 14 Jan 2014 02:15:51 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=+qd/o/9Kgkrdx2/eUiJ0MGXmp1qlQ78q1HcnVwo90Ow=; b=dE9Wx3FOqgtDr8egEsupRyrfkK0G9d2t42JP2ehafzZEEBud33VcNZvml6fn66rMoH Cay/rDLJvP3l11mVZsj+UxoLS9BXwpbLtsA1M5lJrJgEMUPg39TgbRk2nhN8p1RLD4ap YG0Cqh9ncI7Se4d2QPfuUUVWxSDqjIMMVTsVLKOMLpi0e77KZ9vvjEDuMIwDa3/jn+G3 KII2mJbU5+t9Ga6Ihh+6lmvnF5fzDoP1PuDbYzcZOVTujLEUekummUyKWIOooNux3AFq Ty+6TGgcF1sedWqWFoZaBSmy+gqtJaEctyh/rAuTZRC/xsgMarRGMpGjGy+tgGLTQ+rR mH9A==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=+qd/o/9Kgkrdx2/eUiJ0MGXmp1qlQ78q1HcnVwo90Ow=; b=JKy7pq1wjjXmSBMQGAClRDzje+wzJCBAFDipdSmkpRQ5SvGfLYbklQYbaBa+jZyXKS mCLZw865ZIWkhNlfjiRPDJg5Wors1z7XcJ5Nml3kKfnolRRH0CC4DlTXD3GqddZwIIMv rFiB+CuSJwBtUxiyREWbUT7FyRtpCW0v01R5k=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=+qd/o/9Kgkrdx2/eUiJ0MGXmp1qlQ78q1HcnVwo90Ow=; b=Vncp/tpPfd7QxNQDVfLp7oj8c3cbvm6KzkNQuvUNMFwmzXXlfUCfj+GCbQXt5hRTog tkyJLgTUOkogEFrAgOHE+PAMJr7Su/fXPcMNfgIVVYi/ymyK7jRgAUF4j39puROdJtAg bujsg0+PBdex83LIYBckdfw2QbIqs41wVy7pluIk4pqZRsW8dFkSdHuP9Ksn6JbbnH7J EY/SFNSPGumRi+80lsuCz84Y+Ywjam2OWbstyTsLckiuVffgYyEgE0lGxVMVarz7gh6W mM5eBoNhMUUYovtKJvosUyvoUPuJAB8ubWiAHfEyDCOo+7szV2D9g+Xg7ElxDJzXO4RP Sy9Q==
X-Gm-Message-State: ALoCoQnFyBbZ7fqddfiJaDtwck8XSVuj1AVtiB92g7ub7yo0cx9en6q68bAfOK1JyDixx8DxoFGm1R/sOoQ/DzZlLMUD3E4z79GuJXayVUpGieuryq6XVq4eHrw/VQfgrjlyJwbHKNoZ3L75DLLOxSDaQ2sGWv4OUyMJY0/Y9o2ukGI1r1idPmirQr0rQeDINR2f4vzfd4IM
MIME-Version: 1.0
X-Received: by 10.49.49.40 with SMTP id r8mr1111720qen.31.1389694551428; Tue, 14 Jan 2014 02:15:51 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.140.109.202 with HTTP; Tue, 14 Jan 2014 02:15:51 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BE40A18@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE40A18@EXVMBX020-12.exch020.serverdata.net>
Date: Tue, 14 Jan 2014 19:15:51 +0900
X-Google-Sender-Auth: m7gU24UGLnWdSU4dfTeV6L1pjaY
Message-ID: <CABihn6GEizG2+F-n=2HvHmbj3Z7ZiURv96upPokN4fuMJGCcyw@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=047d7b6783f815bc3c04efeb7a6c
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 14 Jan 2014 10:16:06 -0000

--047d7b6783f815bc3c04efeb7a6c
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

>
> The WS base protocol does not assign semantics to frame boundaries, but W=
S
> extensions might do.

Yes, you are right. Thank you for pointing it out.

So if WS/SPDY does not preserve WS frame boundaries, future WS extensions
> which might assign semantics to frame boundaries might be incompatible wi=
th
> WS/SPDY.
>
WS/SPDY preserves WS frame boundaries, though endpoints can't recognize the
end of a WS frame immediately.

There are several ways to address this issue:
1: Use MSG_DONE for the WS frame boundaries (not WS message boundaries) and
introduce ":fin" in the HEADERS frame.
2: Introduce FRAME_DONE FLAG in SPDY frame for the WS frame boundaries.
3: Introduce ":fin" and ":length".

Why remove "length" and "fin" from the headers?

To reduce the size of HEADERS frames.
Though ":fin" will be well-compressed by HPACK, having ":length" increases
the overhead in a measure (around 6 bytes).



On Fri, Jan 10, 2014 at 5:14 PM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> >>2) Ok, both WS frame and message boundaries are preserved.
>
> >But: I cannot recognize the end of a WS frame until the next WS frame
> starts (for which I receive a HEADERS frame, which is the implicit signal
> that the previous WS frame ended), right?
> >You are right, you can't.
>
> I have reread the proposed WS/SPDY
>
> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aW=
zSHijk/edit?usp=3Dsharing
>
> """
> The SPDY HEADERS frame must contain WebSocket frame fields and following
> SPDY data frames contain payload data
> """
>
> However:
>
> """
> WebSocket uses MSG_DONE flag as the message boundary mark (It is not yet
> introduced, we should ask for the introduction). Then, =E2=80=9C:length=
=E2=80=9D and =E2=80=9C:fin=E2=80=9D
> can be omitted from HEADERS frame. Frames are delimited by a HEADERS fram=
e.
> """
>
> So the "length" and "fin" fields from WS frame header are omitted?
>
> This is unfortunate, since if they would be present, the end of a WS fram=
e
> (and the end of a WS message) could be readily detected.
>
> More so: WebSocket implementations which (also) expose a frame-based API
> might want to signal the length of a new WS frame started right at the
> _beginning_. And this can't be done with the proposed WS/SPDY design also=
,
> since the frame length is only know _after_ buffering up all frame payloa=
d
> data, and after receiving the next SPDY HEADERS frame (implicit signal of
> WS frame end).
>
> Why remove "length" and "fin" from the headers?
>
> I think this complicates matters ..
>
> >And if so, why is there no FRAME_DONE bit on the SPDY frame that
> corresponds to the end of a WS frame?
> >Given that WebSocket frames "have no semantic meaning" as in RFC6455, I
> don't think the immediate recognition of the end of a WebSocket frame is
> requisite.
> >I would like to introduce MSG_DONE in HTTP2 so that intermediaries that
> don't fully understand WS over HTTP2 can relay it.
> >Since there are not so much space in FLAGS field, Introducing FRAME_DONE
> may decrease the extensiveness of the spec.
> >Do you have a strong preference to introduce FRAME_DONE?
>
> The WS base protocol does not assign semantics to frame boundaries, but W=
S
> extensions might do.
>
> This is what the spec says:
>
> """
> Unless specified otherwise by an extension, frames have no semantic
>    meaning.  An intermediary might coalesce and/or split frames, if no
>    extensions were negotiated by the client and the server or if some
>    extensions were negotiated, but the intermediary understood all the
>    extensions negotiated and knows how to coalesce and/or split frames
>    in the presence of these extensions.  One implication of this is that
>    in absence of extensions, senders and receivers must not depend on
>    the presence of specific frame boundaries.
> """
>
> So if WS/SPDY does not preserve WS frame boundaries, future WS extensions
> which might assign semantics to frame boundaries might be incompatible wi=
th
> WS/SPDY.
>
> Not that I think it's a particular good idea for WS extensions to assign
> semantics to frame boundaries - it's just that I think it should be a
> conscious decision ..
>
> /Tobias
>
>
>

--047d7b6783f815bc3c04efeb7a6c
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px =
0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-l=
eft-style:solid;padding-left:1ex"><span style=3D"font-family:arial,sans-ser=
if;font-size:13px">The WS base protocol does not assign semantics to frame =
boundaries, but WS ex</span><span style=3D"font-family:arial,sans-serif;fon=
t-size:13px">tensions might do.</span></blockquote>
<div>Yes, you are right. Thank you for pointing it out.</div><div><br></div=
><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border=
-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;=
padding-left:1ex">
<span style=3D"font-family:arial,sans-serif;font-size:13px">So if WS/SPDY d=
oes not preserve WS frame boundaries, future WS extensions which might assi=
gn semantics to frame boundaries might be incompatible with WS/SPDY.</span>=
<br>
</blockquote><div>WS/SPDY preserves WS frame boundaries, though endpoints c=
an&#39;t recognize the end of a WS frame immediately.</div><div><br></div><=
div>There are several ways to address this issue:</div><div>1: Use MSG_DONE=
=C2=A0for the WS frame boundaries (not WS message boundaries) and introduce=
 &quot;:fin&quot; in the HEADERS frame.</div>
<div>2: Introduce FRAME_DONE FLAG in SPDY frame for the WS frame boundaries=
.</div><div>3: Introduce &quot;:fin&quot; and &quot;:length&quot;.</div><di=
v><br></div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0=
.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-s=
tyle:solid;padding-left:1ex">
<span style=3D"font-family:arial,sans-serif;font-size:13px">Why remove &quo=
t;length&quot; and &quot;fin&quot; from the headers?</span></blockquote><di=
v>To reduce the size of HEADERS frames.</div><div>Though &quot;:fin&quot; w=
ill be well-compressed by HPACK, having &quot;:length&quot; increases the o=
verhead in a measure (around 6 bytes).</div>
<div><br></div></div><div class=3D"gmail_extra"><br><br><div class=3D"gmail=
_quote">On Fri, Jan 10, 2014 at 5:14 PM, Tobias Oberstein <span dir=3D"ltr"=
>&lt;<a href=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobia=
s.oberstein@tavendo.de</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div class=3D"im">&gt;&gt;2) Ok, both WS fra=
me and message boundaries are preserved.<br>
=C2=A0<br>
&gt;But: I cannot recognize the end of a WS frame until the next WS frame s=
tarts (for which I receive a HEADERS frame, which is the implicit signal th=
at the previous WS frame ended), right?<br>
&gt;You are right, you can&#39;t.<br>
=C2=A0<br>
</div>I have reread the proposed WS/SPDY<br>
<a href=3D"https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmC=
fc7ukw9aWzSHijk/edit?usp=3Dsharing" target=3D"_blank">https://docs.google.c=
om/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit?usp=3Dshari=
ng</a><br>

<div class=3D"im"><br>
&quot;&quot;&quot;<br>
The SPDY HEADERS frame must contain WebSocket frame fields and following SP=
DY data frames contain payload data<br>
&quot;&quot;&quot;<br>
<br>
</div>However:<br>
<br>
&quot;&quot;&quot;<br>
WebSocket uses MSG_DONE flag as the message boundary mark (It is not yet in=
troduced, we should ask for the introduction). Then, =E2=80=9C:length=E2=80=
=9D and =E2=80=9C:fin=E2=80=9D can be omitted from HEADERS frame. Frames ar=
e delimited by a HEADERS frame.<br>

&quot;&quot;&quot;<br>
<br>
So the &quot;length&quot; and &quot;fin&quot; fields from WS frame header a=
re omitted?<br>
<br>
This is unfortunate, since if they would be present, the end of a WS frame =
(and the end of a WS message) could be readily detected.<br>
<br>
More so: WebSocket implementations which (also) expose a frame-based API mi=
ght want to signal the length of a new WS frame started right at the _begin=
ning_. And this can&#39;t be done with the proposed WS/SPDY design also, si=
nce the frame length is only know _after_ buffering up all frame payload da=
ta, and after receiving the next SPDY HEADERS frame (implicit signal of WS =
frame end).<br>

<br>
Why remove &quot;length&quot; and &quot;fin&quot; from the headers?<br>
<br>
I think this complicates matters ..<br>
<div class=3D"im"><br>
&gt;And if so, why is there no FRAME_DONE bit on the SPDY frame that corres=
ponds to the end of a WS frame?<br>
&gt;Given that WebSocket frames &quot;have no semantic meaning&quot; as in =
RFC6455, I don&#39;t think the immediate recognition of the end of a WebSoc=
ket frame is requisite.<br>
&gt;I would like to introduce MSG_DONE in HTTP2 so that intermediaries that=
 don&#39;t fully understand WS over HTTP2 can relay it.<br>
&gt;Since there are not so much space in FLAGS field, Introducing FRAME_DON=
E may decrease the extensiveness of the spec.<br>
&gt;Do you have a strong preference to introduce FRAME_DONE?<br>
<br>
</div>The WS base protocol does not assign semantics to frame boundaries, b=
ut WS extensions might do.<br>
<br>
This is what the spec says:<br>
<br>
&quot;&quot;&quot;<br>
Unless specified otherwise by an extension, frames have no semantic<br>
=C2=A0 =C2=A0meaning. =C2=A0An intermediary might coalesce and/or split fra=
mes, if no<br>
=C2=A0 =C2=A0extensions were negotiated by the client and the server or if =
some<br>
=C2=A0 =C2=A0extensions were negotiated, but the intermediary understood al=
l the<br>
=C2=A0 =C2=A0extensions negotiated and knows how to coalesce and/or split f=
rames<br>
=C2=A0 =C2=A0in the presence of these extensions. =C2=A0One implication of =
this is that<br>
=C2=A0 =C2=A0in absence of extensions, senders and receivers must not depen=
d on<br>
=C2=A0 =C2=A0the presence of specific frame boundaries.<br>
&quot;&quot;&quot;<br>
<br>
So if WS/SPDY does not preserve WS frame boundaries, future WS extensions w=
hich might assign semantics to frame boundaries might be incompatible with =
WS/SPDY.<br>
<br>
Not that I think it&#39;s a particular good idea for WS extensions to assig=
n semantics to frame boundaries - it&#39;s just that I think it should be a=
 conscious decision ..<br>
<span class=3D"HOEnZb"><font color=3D"#888888"><br>
/Tobias<br>
<br>
<br>
</font></span></blockquote></div><br></div>

--047d7b6783f815bc3c04efeb7a6c--

From tobias.oberstein@tavendo.de  Tue Jan 14 02:47:06 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 597E71AE080 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 02:47:06 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ezfGiVNDRCP2 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 02:47:05 -0800 (PST)
Received: from EXHUB020-4.exch020.serverdata.net (exhub020-4.exch020.serverdata.net [206.225.164.31]) by ietfa.amsl.com (Postfix) with ESMTP id D86B81AE072 for <hybi@ietf.org>; Tue, 14 Jan 2014 02:47:04 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-4.exch020.serverdata.net ([206.225.164.31]) with mapi; Tue, 14 Jan 2014 02:46:53 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Yutaka Hirano <yhirano@chromium.org>
Date: Tue, 14 Jan 2014 02:46:48 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8REZsM76g70jjgQIadTNTrNg4bSwAA/Tzg
Message-ID: <634914A010D0B943A035D226786325D4446BED9952@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE40A18@EXVMBX020-12.exch020.serverdata.net> <CABihn6GEizG2+F-n=2HvHmbj3Z7ZiURv96upPokN4fuMJGCcyw@mail.gmail.com>
In-Reply-To: <CABihn6GEizG2+F-n=2HvHmbj3Z7ZiURv96upPokN4fuMJGCcyw@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 14 Jan 2014 10:47:06 -0000
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From yhirano@google.com  Tue Jan 14 03:24:54 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 77CD01AE097 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 03:24:54 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id QNYvJgWpx3aF for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 03:24:52 -0800 (PST)
Received: from mail-qe0-x234.google.com (mail-qe0-x234.google.com [IPv6:2607:f8b0:400d:c02::234]) by ietfa.amsl.com (Postfix) with ESMTP id A2E341AE089 for <hybi@ietf.org>; Tue, 14 Jan 2014 03:24:52 -0800 (PST)
Received: by mail-qe0-f52.google.com with SMTP id s1so4607494qeb.39 for <hybi@ietf.org>; Tue, 14 Jan 2014 03:24:41 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=RmKquKiigMvBEPTt7x3ck6lYodv2tm0ji8U34/e87c4=; b=C4mC9uhGuaUKgAQpUsc2sBvheRb9GlI4AFrdy6WYxZ0ohQehl2Ozr3Yom8XH2lkd2z mJVvn1OzkApurIDbDHG7cwaJr6S5SG1k/s3ChBjpjPne/NWQqE+HKiKk8DIIqtzzW0TX b7wg0RoMsBZXfqrhFl/SnvenjsYk2XwNCr4CzwHnYiATXi1dfmPMYJsmyhtEWtUPD8Sb vERsSjDUHwh0mL2lyWqo0mpdpKERMgh1+Oe8W2S002oimwlyp7QIf+nMYnMiY0kBhxCN r9QhtimuE8zpx4eABHJhsfXJKzLGUWHt2ebZnsQjBA6fkjGbWzqmiG8ZsxiEKW/NsTzN c5qQ==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=RmKquKiigMvBEPTt7x3ck6lYodv2tm0ji8U34/e87c4=; b=RMS7UNdPXoGGduJrBsVkFJAVoG86o/rQLypTxemMdqZuCu28ch/jfyvZaEcOGaUexm VGlQIsJ7sPKf2WN1zSRUMhBo4mF2YQ/cdNqts1KaRdTJujfx91W7nffhd5iu6YDk3G3V UHIqJmpIG9MiM9PNtXlUJnBVkgBzbPSXjO+7k=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=RmKquKiigMvBEPTt7x3ck6lYodv2tm0ji8U34/e87c4=; b=HjBnHzHy4NvqGifuUANjOJF/JWrj+4s0Z7IBs0xWuvko+dVv5cA4lBsqaQD2JAHO1s VcKrZeYt+tRGyI8yvRsay621bXDTuQz/4udYLoHUHFdK6HKlevEWjcG0HceNNc7oairp O920G6g8Joo4J+s+6Rniwb+V7eMKR2t/ZGd4HsZ0eytg+3EjoYqC4yCDCwvmdlp1NJau UF0N352BQ1cipIwT7rw/Zd48oTSI1A3ijgtHpO1bdln9UlmCL+kfmMPQJT23RGAwFGeR fRsblLDHokyu7f5sZl0b346TromnUW3ACbahVbYz+1dzpyPrzMkLLIdylSvbD8RfZ8X8 BozQ==
X-Gm-Message-State: ALoCoQlLtuyeppYCx9uW1/feFbVADIyn32ROV2kyIa+XRsn6Vwsv/FfgLTNNBVMPMxLwp2S78N7Ks7J28h+/271tAZ8/EpBU8soRxxQX4IvFSyIWYo/CeOj8kmASDHCjuFGkZ4tTbCkDU5317/jpx0dnD5XuvyZVrQSYSrCe4derRhy9rcT7S3OnD6g4yEr9WHMxbNaQVRQN
MIME-Version: 1.0
X-Received: by 10.49.56.68 with SMTP id y4mr1923839qep.8.1389698681149; Tue, 14 Jan 2014 03:24:41 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.140.109.202 with HTTP; Tue, 14 Jan 2014 03:24:41 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BED9952@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE40A18@EXVMBX020-12.exch020.serverdata.net> <CABihn6GEizG2+F-n=2HvHmbj3Z7ZiURv96upPokN4fuMJGCcyw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9952@EXVMBX020-12.exch020.serverdata.net>
Date: Tue, 14 Jan 2014 20:24:41 +0900
X-Google-Sender-Auth: lecg5apFyCMSZZ6otBv_hVvj-iI
Message-ID: <CABihn6HXpTpKhhuBbyGqodXBMJzcU6oAOhpqAGpS+2btgaW+uQ@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=001a113489643c2b2d04efec70e0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 14 Jan 2014 11:24:54 -0000

--001a113489643c2b2d04efec70e0
Content-Type: text/plain; charset=UTF-8

>
> What's the advantage of using SPDY HEADER frames in the context of
> WebSocket over SPDY?

There are two advantages I think.
- Endpoints do not have to implement WebSocket framing.
  Especially if the endpoint need only WS/SPDY, not native WebSocket.
- Intermediaries can easily recognize the semantics.




On Tue, Jan 14, 2014 at 7:46 PM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> >>So if WS/SPDY does not preserve WS frame boundaries, future WS
> extensions which might assign semantics to frame boundaries might be
> incompatible with WS/SPDY.
> >WS/SPDY preserves WS frame boundaries, though endpoints can't recognize
> the end of a WS frame immediately.
>
> >There are several ways to address this issue:
> >1: Use MSG_DONE for the WS frame boundaries (not WS message boundaries)
> and introduce ":fin" in the HEADERS frame.
> >2: Introduce FRAME_DONE FLAG in SPDY frame for the WS frame boundaries.
> >3: Introduce ":fin" and ":length".
>
> >>Why remove "length" and "fin" from the headers?
> >To reduce the size of HEADERS frames.
> >Though ":fin" will be well-compressed by HPACK, having ":length"
> increases the overhead in a measure (around 6 bytes).
>
> Why not simply pack WebSocket frames (_including_ full WS frame header) as
> _payload_ of SPDY data frames only (no SPDY header frames)?
>
> What's the advantage of using SPDY HEADER frames in the context of
> WebSocket over SPDY?
>
> /Tobias
>

--001a113489643c2b2d04efec70e0
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px =
0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-l=
eft-style:solid;padding-left:1ex">What&#39;s the advantage of using SPDY HE=
ADER frames in the context of WebSocket over SPDY?</blockquote>
<div>There are two advantages I think.=C2=A0</div><div>- Endpoints do not h=
ave to implement WebSocket framing.</div><div>=C2=A0 Especially if the endp=
oint need only WS/SPDY, not native WebSocket.</div>- Intermediaries can eas=
ily recognize the semantics.<br>
<br>=C2=A0</div><div class=3D"gmail_extra"><br><br><div class=3D"gmail_quot=
e">On Tue, Jan 14, 2014 at 7:46 PM, Tobias Oberstein <span dir=3D"ltr">&lt;=
<a href=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.obe=
rstein@tavendo.de</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div class=3D"im">&gt;&gt;So if WS/SPDY does=
 not preserve WS frame boundaries, future WS extensions which might assign =
semantics to frame boundaries might be incompatible with WS/SPDY.<br>

&gt;WS/SPDY preserves WS frame boundaries, though endpoints can&#39;t recog=
nize the end of a WS frame immediately.<br>
<br>
&gt;There are several ways to address this issue:<br>
&gt;1: Use MSG_DONE=C2=A0for the WS frame boundaries (not WS message bounda=
ries) and introduce &quot;:fin&quot; in the HEADERS frame.<br>
&gt;2: Introduce FRAME_DONE FLAG in SPDY frame for the WS frame boundaries.=
<br>
&gt;3: Introduce &quot;:fin&quot; and &quot;:length&quot;.<br>
<br>
&gt;&gt;Why remove &quot;length&quot; and &quot;fin&quot; from the headers?=
<br>
&gt;To reduce the size of HEADERS frames.<br>
&gt;Though &quot;:fin&quot; will be well-compressed by HPACK, having &quot;=
:length&quot; increases the overhead in a measure (around 6 bytes).<br>
<br>
</div><div class=3D"im">Why not simply pack WebSocket frames (_including_ f=
ull WS frame header) as _payload_ of SPDY data frames only (no SPDY header =
frames)?<br>
<br>
</div>What&#39;s the advantage of using SPDY HEADER frames in the context o=
f WebSocket over SPDY?<br>
<span class=3D"HOEnZb"><font color=3D"#888888"><br>
/Tobias<br>
</font></span></blockquote></div><br></div>

--001a113489643c2b2d04efec70e0--

From tobias.oberstein@tavendo.de  Tue Jan 14 03:57:36 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 082FC1AE056 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 03:57:36 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id aEj5Sl0bAh8e for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 03:57:34 -0800 (PST)
Received: from EXHUB020-5.exch020.serverdata.net (exhub020-5.exch020.serverdata.net [206.225.164.32]) by ietfa.amsl.com (Postfix) with ESMTP id 6D0E81AE04D for <hybi@ietf.org>; Tue, 14 Jan 2014 03:57:34 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-5.exch020.serverdata.net ([206.225.164.32]) with mapi; Tue, 14 Jan 2014 03:57:23 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Yutaka Hirano <yhirano@chromium.org>
Date: Tue, 14 Jan 2014 03:57:17 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8RGzf+sYqwbnFjRDOOZOZD75SRFQAA6W3A
Message-ID: <634914A010D0B943A035D226786325D4446BED9953@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE40A18@EXVMBX020-12.exch020.serverdata.net> <CABihn6GEizG2+F-n=2HvHmbj3Z7ZiURv96upPokN4fuMJGCcyw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9952@EXVMBX020-12.exch020.serverdata.net> <CABihn6HXpTpKhhuBbyGqodXBMJzcU6oAOhpqAGpS+2btgaW+uQ@mail.gmail.com>
In-Reply-To: <CABihn6HXpTpKhhuBbyGqodXBMJzcU6oAOhpqAGpS+2btgaW+uQ@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 14 Jan 2014 11:57:36 -0000
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From tyoshino@google.com  Tue Jan 14 04:41:31 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 7AC301AE0C9 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 04:41:31 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 5LrPpPd-tXeH for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 04:41:29 -0800 (PST)
Received: from mail-we0-x22c.google.com (mail-we0-x22c.google.com [IPv6:2a00:1450:400c:c03::22c]) by ietfa.amsl.com (Postfix) with ESMTP id 2204A1AE0C8 for <hybi@ietf.org>; Tue, 14 Jan 2014 04:41:28 -0800 (PST)
Received: by mail-we0-f172.google.com with SMTP id q58so331978wes.3 for <hybi@ietf.org>; Tue, 14 Jan 2014 04:41:17 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=DBoDK/mHO0P6LkbKnuGy3ylmXdl3pBqbLf51oLP2EyE=; b=FDRW+4S18MlBFmOl9x5ELAbXj9Hh6aIB0VvOeCN7T66av5otzS8T+VQWntozGEdpwf eRA+HaRhkQvSmuivwxbHaaTBMII9Kvj7/vIrJiC520LFHZXtHAXhuPzP+KFadQrU4hJL CP8VSSp8RNOJZG41arfYvdM76rTDLG55KGUGY9Pez+pGYWhsjC0WfyDozFI/CH5hhlHw Mppw+zSdMu4dTq4Y6LSaA7ztcWUtc5VPxpC/+RSQXCaa98eCNoFK6lmh3lNm+khkuyKh oljPXmbNI134AES9pt2R45ldpR7BuqD58nFRk3qem8kxU4YfARAmIuYyukxAm+4C20Hn mOZw==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=DBoDK/mHO0P6LkbKnuGy3ylmXdl3pBqbLf51oLP2EyE=; b=byiTJwZYamP7copN30nc4JAgaM6PZ0oel1Xb2zmNE832Rr8XnlxL3x4joJ6oo4+Coa YbzDOhBDgsSQ8scKEEu3uyC4XAu7DQ7ZfaOf/pDQj9qsiXJ5nlIRY2vsy4/5X/jsYIAu eyB75e9M7oSSTAnrHne6Ka5bj0te2UgPAjaUVkx4VgqPh8HjlRKRCexakxVjBfLqk04k OYhZ33eiwIywCL8RoUbv2ATAdrauIk0IBXTMPz8KyNx6b4lVPXGhsFk61/I87TMQxb0z toX88u+XXTeW46Bj65YUX5mBsSGZAVbv9K49QLHUmC+xsTtaptOVl/OI+Vv4Y2jNCeaw T4Ew==
X-Gm-Message-State: ALoCoQnImvs7UX9z0Ax2ynzety0tUNUzLVONQXWfu6MIuFMJmhFvJGtZbbwlecdKdFMDhY+arZxQb/3uT7yIB8TEx6U/5Yu1BbTW9gLrjxqo0V2TpPTPvhBR4qg+M0mnCaR/jdWYcM0i0rTudGP8bge6ZXVcmUZl97xTDUWrP2XwioiSlVYSehTfPibjuAERph+NG5WTqcfF
X-Received: by 10.180.37.193 with SMTP id a1mr1798416wik.52.1389703277379; Tue, 14 Jan 2014 04:41:17 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Tue, 14 Jan 2014 04:40:57 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BED9953@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE40A18@EXVMBX020-12.exch020.serverdata.net> <CABihn6GEizG2+F-n=2HvHmbj3Z7ZiURv96upPokN4fuMJGCcyw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9952@EXVMBX020-12.exch020.serverdata.net> <CABihn6HXpTpKhhuBbyGqodXBMJzcU6oAOhpqAGpS+2btgaW+uQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9953@EXVMBX020-12.exch020.serverdata.net>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Tue, 14 Jan 2014 21:40:57 +0900
Message-ID: <CAH9hSJaEWTrPRMaMYtb5t71Xhr==Uj0KW+SwSvKXzRVDHzuzXg@mail.gmail.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=e89a8f502fa03114ce04efed8287
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 14 Jan 2014 12:41:31 -0000

--e89a8f502fa03114ce04efed8287
Content-Type: text/plain; charset=UTF-8

Tobias, thanks for input. Yutaka is representing consensus we reached at
spdy-dev. While we respect conclusion at spdy-dev, your suggestions are
also really welcome.

Regarding cost of implementing frame parser for RFC6455 WebSocket protocol,
it's possible that a service developer chooses to accept only WebSocket
with multiplexing and backup with fallback to HTTP/1.1 COMET. In such a
case, they don't implement RFC6455 parser. So, it's not completely
unrealistic.

Takeshi


On Tue, Jan 14, 2014 at 8:57 PM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> >What's the advantage of using SPDY HEADER frames in the context of
> WebSocket over SPDY?
> >There are two advantages I think.
> >- Endpoints do not have to implement WebSocket framing.
> >  Especially if the endpoint need only WS/SPDY, not native WebSocket.
> >- Intermediaries can easily recognize the semantics.
>
> I think it's likely that both endpoints and intermediaries that support
> WebSocket will either support only
> native WebSocket or both native and WS/SPDY.
>
> And in this case, it's easier to support both with just WS frames as
> payload of SPDY data frames.
>
> The current WS/SPDY proposal just complicates matters and is less
> efficient in terms of wire-level overhead.
>
> On Tue, Jan 14, 2014 at 7:46 PM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
> >>So if WS/SPDY does not preserve WS frame boundaries, future WS
> extensions which might assign semantics to frame boundaries might be
> incompatible with WS/SPDY.
> >WS/SPDY preserves WS frame boundaries, though endpoints can't recognize
> the end of a WS frame immediately.
>
> >There are several ways to address this issue:
> >1: Use MSG_DONE for the WS frame boundaries (not WS message boundaries)
> and introduce ":fin" in the HEADERS frame.
> >2: Introduce FRAME_DONE FLAG in SPDY frame for the WS frame boundaries.
> >3: Introduce ":fin" and ":length".
>
> >>Why remove "length" and "fin" from the headers?
> >To reduce the size of HEADERS frames.
> >Though ":fin" will be well-compressed by HPACK, having ":length"
> increases the overhead in a measure (around 6 bytes).
> Why not simply pack WebSocket frames (_including_ full WS frame header) as
> _payload_ of SPDY data frames only (no SPDY header frames)?
> What's the advantage of using SPDY HEADER frames in the context of
> WebSocket over SPDY?
>
> /Tobias
>
>

--e89a8f502fa03114ce04efed8287
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Tobias, thanks for input. Yutaka is representing consensus=
 we reached at spdy-dev. While we respect conclusion at spdy-dev, your sugg=
estions are also really welcome.<div><br></div><div>Regarding cost of imple=
menting frame parser for RFC6455 WebSocket protocol, it&#39;s possible that=
 a service developer chooses to accept only WebSocket with multiplexing and=
 backup with fallback to HTTP/1.1 COMET. In such a case, they don&#39;t imp=
lement RFC6455 parser. So, it&#39;s not completely unrealistic.</div>

</div><div class=3D"gmail_extra"><br clear=3D"all"><div>Takeshi</div>
<br><br><div class=3D"gmail_quote">On Tue, Jan 14, 2014 at 8:57 PM, Tobias =
Oberstein <span dir=3D"ltr">&lt;<a href=3D"mailto:tobias.oberstein@tavendo.=
de" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt;</span> wrote:<br>=
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">

<div class=3D"im">&gt;What&#39;s the advantage of using SPDY HEADER frames =
in the context of WebSocket over SPDY?<br>
&gt;There are two advantages I think.=C2=A0<br>
&gt;- Endpoints do not have to implement WebSocket framing.<br>
&gt;=C2=A0 Especially if the endpoint need only WS/SPDY, not native WebSock=
et.<br>
&gt;- Intermediaries can easily recognize the semantics.<br>
<br>
</div>I think it&#39;s likely that both endpoints and intermediaries that s=
upport WebSocket will either support only<br>
native WebSocket or both native and WS/SPDY.<br>
<br>
And in this case, it&#39;s easier to support both with just WS frames as pa=
yload of SPDY data frames.<br>
<br>
The current WS/SPDY proposal just complicates matters and is less efficient=
 in terms of wire-level overhead.=C2=A0<br>
<div class=3D"HOEnZb"><div class=3D"h5"><br>
On Tue, Jan 14, 2014 at 7:46 PM, Tobias Oberstein &lt;<a href=3D"mailto:tob=
ias.oberstein@tavendo.de">tobias.oberstein@tavendo.de</a>&gt; wrote:<br>
&gt;&gt;So if WS/SPDY does not preserve WS frame boundaries, future WS exte=
nsions which might assign semantics to frame boundaries might be incompatib=
le with WS/SPDY.<br>
&gt;WS/SPDY preserves WS frame boundaries, though endpoints can&#39;t recog=
nize the end of a WS frame immediately.<br>
<br>
&gt;There are several ways to address this issue:<br>
&gt;1: Use MSG_DONE=C2=A0for the WS frame boundaries (not WS message bounda=
ries) and introduce &quot;:fin&quot; in the HEADERS frame.<br>
&gt;2: Introduce FRAME_DONE FLAG in SPDY frame for the WS frame boundaries.=
<br>
&gt;3: Introduce &quot;:fin&quot; and &quot;:length&quot;.<br>
<br>
&gt;&gt;Why remove &quot;length&quot; and &quot;fin&quot; from the headers?=
<br>
&gt;To reduce the size of HEADERS frames.<br>
&gt;Though &quot;:fin&quot; will be well-compressed by HPACK, having &quot;=
:length&quot; increases the overhead in a measure (around 6 bytes).<br>
Why not simply pack WebSocket frames (_including_ full WS frame header) as =
_payload_ of SPDY data frames only (no SPDY header frames)?<br>
What&#39;s the advantage of using SPDY HEADER frames in the context of WebS=
ocket over SPDY?<br>
<br>
/Tobias<br>
<br>
</div></div></blockquote></div><br></div>

--e89a8f502fa03114ce04efed8287--

From fenix@google.com  Tue Jan 14 10:30:36 2014
Return-Path: <fenix@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id BC6541AE153 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 10:30:36 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id XaM_J5HYVU6F for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 10:30:32 -0800 (PST)
Received: from mail-qe0-x22a.google.com (mail-qe0-x22a.google.com [IPv6:2607:f8b0:400d:c02::22a]) by ietfa.amsl.com (Postfix) with ESMTP id 2B84C1AE06C for <hybi@ietf.org>; Tue, 14 Jan 2014 10:30:32 -0800 (PST)
Received: by mail-qe0-f42.google.com with SMTP id b4so9216347qen.1 for <hybi@ietf.org>; Tue, 14 Jan 2014 10:30:20 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type; bh=7wYhjxs+SZ/oq7yFi4aQzf+UePCfQybE86hyi32kAUo=; b=O7B/OQFgibQXoagVA0RQnkCdr1K3zSbaHyOEmUrAojeV+2D8LT3BXuK9tXO3rabShQ t0i7471uU1oQtlcMg5rJdEg51+xx9bg0y8mFISoIJ1hW1RPtcnnwH2wDMDreBRZGYkSl pRvH48b38n5M96veMqn6mYCuhfkZY1ulMd0OQ9OOYzK9o9TRHyLBSxgdY0HiHqtj5GBg 9cyRrhZqsxR1hDmCiceGvQaxFmwACtIVn7vs6/Gnx8O+m7SYP8tUz8vvBt3MVgKT0q6W I3+dBi07YmACpHxUIe4gijTmqnLERqL6dIJA1TuTbT4C51BGVhhz3zQzAmYMupSBv65U Q6Zg==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=7wYhjxs+SZ/oq7yFi4aQzf+UePCfQybE86hyi32kAUo=; b=CH1hG0RGxRkoOAQyBAt1ok+torvMIvRiVZTlRMy23JZhMWgwCY8QMYJhCvdZkgQn2K 4JJI4xyf2d+dFIVpb8ZJUcv/6Re2binAqapzFycmEZkkNUV794yHDlULUVghOR1CCYHq picnO4NX2CInKfQ7gYo0jcU/ArUL4BIc5UWB+TMhU3Ojuye9xebIoUgFY2AfcAU+WyHo 9uIrzJMSoBBdsPbp1T9eFQ4Uvxdf4qEVI8x50htI3k7IN/uBMqb7Ot3LVB18Aq4h/+no BlrStaUaoVlkI9C8qYtKc52OxvRQWKZnsDno1lKQbPeSrB7e9ERvHUc9l3JxdD8mgXLm jClQ==
X-Gm-Message-State: ALoCoQmui5Y5RIsGynN+cyJeBFIljWTB8Ql8xYTtC9qM8IQnQn/dpul3vIJye1IctT9aiOMIN8qP5Y6AovNjQ23awNGbsSjdUPQKIHR3rzJNDXpFf6DGwxzqH44iZkMvolKUwqt/N7CvYVWXsYS4cPEfpVcGtP9suJRPlaK7E//N2PJApbEc6D/eVVxVmzHima9m2dgTU1zY
MIME-Version: 1.0
X-Received: by 10.49.98.35 with SMTP id ef3mr5586817qeb.30.1389724220593; Tue, 14 Jan 2014 10:30:20 -0800 (PST)
Received: by 10.229.13.7 with HTTP; Tue, 14 Jan 2014 10:30:20 -0800 (PST)
In-Reply-To: <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com>
Date: Tue, 14 Jan 2014 10:30:20 -0800
Message-ID: <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com>
From: Roberto Peon <fenix@google.com>
To: Yutaka Hirano <yhirano@chromium.org>
Content-Type: multipart/alternative; boundary=047d7bea3cb6812ac104eff262f9
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 14 Jan 2014 18:30:36 -0000

--047d7bea3cb6812ac104eff262f9
Content-Type: text/plain; charset=UTF-8

I've been lurking anyway :)

As I understand the proposal, (and it seems that Takeshi and Yukata are
covering this nicely, so what i say is probably redundant):
Message boundaries, having semantic meaning, are preserved.

Frame boundaries, having no semantic meaning, are not preserved. I don't
think it makes sense to worry about this greatly when the ability to extend
the protocol as implemented in HTTP2 is present. It doesn't make a lot of
sense to signal the frame boundaries through everything unless there is a
high-likelyhood and high-utility use-case for such. If such is found, then
HTTP2 can be revised to include it. This is the approach we've been taking
with most of the HTTP2 stuff, since there has been great effort placed on
allowing new versions to be deployed in the future.

The mapping of WS onto HTTP2 is fairly direct-- we reserved the MSG_DONE
bit specifically for this kind of use-case (i.e. bidirectional,
full-duplex, message-based).
The observation made is that there are essentially 3 types of data that
need to be preserved one way or another:
1) metadata: data about the data, e.g. text, binary, where it is supposed
to be going, etc.
2) payload
3) boundaries between these.

imho, metadata is handled with HEADERS (I'd have preferred it to be called
METADATA, but was voted down at the time :) ), payload is handled with DATA
frames, and the boundaries are delimited either through the use of HEADERS
frames, or via the MSG_DONE or STREAM_DONE bits.
In the presence of extensions, assuming that MSG_DONE is required to signal
the end of a message, then HEADERS can be used to signal any boundary
desired.


I totally agree that it would be nice to expose the priority of various
messages/channels with respect to each other. That is really an issue that
is outside the WG's purvue, though.



On Thu, Jan 9, 2014 at 7:54 PM, Yutaka Hirano <yhirano@chromium.org> wrote:

> +Roberto as an HTTP2 expert.
>
>
>> 2) Ok, both WS frame and message boundaries are preserved.
>>
>> But: I cannot recognize the end of a WS frame until the next WS frame
>> starts (for which I receive a HEADERS frame, which is the implicit signal
>> that the previous WS frame ended), right?
>>
> You are right, you can't.
>
>
>>
>> And if so, why is there no FRAME_DONE bit on the SPDY frame that
>> corresponds to the end of a WS frame?
>
> Given that WebSocket frames "have no semantic meaning" as in RFC6455, I
> don't think the immediate recognition of the end of a WebSocket frame is
> requisite.
> I would like to introduce MSG_DONE in HTTP2 so that intermediaries that
> don't fully understand WS over HTTP2 can relay it.
> Since there are not so much space in FLAGS field, Introducing FRAME_DONE
> may decrease the extensiveness of the spec.
> Do you have a strong preference to introduce FRAME_DONE?
>
>
>
> On Thu, Jan 9, 2014 at 6:53 PM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
>
>> Hi Yutaka,
>>
>>
>>
>> 1) Yes, I understand. Thanks!
>>
>>
>>
>> 2) Ok, both WS frame and message boundaries are preserved.
>>
>>
>>
>> But: I cannot recognize the end of a WS frame until the next WS frame
>> starts (for which I receive a HEADERS frame, which is the implicit signal
>> that the previous WS frame ended), right?
>>
>>
>>
>> And if so, why is there no FRAME_DONE bit on the SPDY frame that
>> corresponds to the end of a WS frame?
>>
>>
>>
>> 3) Ok, great! Hope this works, since SPDY provides header compression,
>> but no sophisticated bidirectional payload compression as "WS
>> permessage-deflate" does.
>>
>>
>>
>> Cheers,
>>
>> /Tobias
>>
>>
>>
>>
>>
>>
>>
>> *Von:* yhirano@google.com [mailto:yhirano@google.com] *Im Auftrag von *Yutaka
>> Hirano
>> *Gesendet:* Donnerstag, 9. Januar 2014 10:37
>> *An:* Tobias Oberstein
>> *Cc:* Takeshi Yoshino; hybi@ietf.org
>> *Betreff:* Re: [hybi] Discontinuation of mux standardizaton in favor of
>> WS/HTTP/2.0
>>
>>
>>
>> Hi Tobias,
>>
>> Thank you for the comments.
>>
>>
>>
>> 1) Frame Mapping
>> """
>> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
>> following plural SPDY data frames. The SPDY HEADERS frame must contain
>> WebSocket frame fields and following SPDY data frames contain payload data.
>> These data frames are free to be reframed.
>> """
>>
>> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
>> So reframing of WebSocket frames (the payload) into multiple SPDY data
>> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
>> is free to choose the boundaries within WS frame payload at which to
>> reframe into different SPDY data frames as it likes?
>>
>> A WebSocket frame MAY be split to multiple SPDY DATA frames.
>>
>> Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA frame
>> into multiple DATA frames.
>>
>> Is this answering your question?
>>
>>
>>
>> 2) End of WS Frame
>> The end of a WS message is signaled with a SPDY data frame having
>> MSG_DONE set.
>> How is the end of a WS frame signaled, given that the WS frame may have
>> been reframed into multiple SPDY data frames?
>>
>> When you receive a HEADERS frame, you can recognize that the last SPDY
>> frame is that last frame in a WebSocket frame.
>>
>> When you receive a SPDY frame with MSG_DONE, you can recognize that it is
>> the last SPDY frame in a WebSocket message and hence the last SPDY frame in
>> a WebSocket frame.
>>
>>
>>
>> 3) WS extension
>> Are WS extensions fully supported? In particular, is "permessage-deflate"
>> supported when running WS/SPDY?
>>
>> Yes. We have reserved bits and we can emulate the opening handshake, I
>> think.
>>
>>
>>
>> On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein <
>> tobias.oberstein@tavendo.de> wrote:
>>
>> Hi Takeshi,
>>
>> it's a pity, but I can understand the motivation to shove everything into
>> HTTP2.
>>
>> Anyway, some questions:
>>
>> 1) Frame Mapping
>> """
>> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
>> following plural SPDY data frames. The SPDY HEADERS frame must contain
>> WebSocket frame fields and following SPDY data frames contain payload data.
>> These data frames are free to be reframed.
>> """
>>
>> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
>>
>> So reframing of WebSocket frames (the payload) into multiple SPDY data
>> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
>> is free to choose the boundaries within WS frame payload at which to
>> reframe into different SPDY data frames as it likes?
>>
>> 2) End of WS Frame
>>
>> The end of a WS message is signaled with a SPDY data frame having
>> MSG_DONE set.
>>
>> How is the end of a WS frame signaled, given that the WS frame may have
>> been reframed into multiple SPDY data frames?
>>
>> 3) WS extension
>>
>> Are WS extensions fully supported? In particular, is "permessage-deflate"
>> supported when running WS/SPDY?
>>
>> Thanks!
>> /Tobias
>>
>>
>> Von: hybi [mailto:hybi-bounces@ietf.org] Im Auftrag von Takeshi Yoshino
>> Gesendet: Dienstag, 7. Januar 2014 08:36
>> An: hybi@ietf.org
>> Betreff: [hybi] Discontinuation of mux standardizaton in favor of
>> WS/HTTP/2.0
>>
>>
>> Hi all,
>>
>> We've been working on mux extension to make WebSocket scalable for long
>> time, but as SPDY got widely supported and adopted quickly, now there seems
>> to be no longer any good reason to continue this work which is WebSocket
>> dedicated.
>>
>> Scalability is still important issue to address. From now, I'd like to
>> focus on WS/SPDY effort which we've been working on in parallel with the
>> mux extension.
>>
>> Please take a look at the latest proposal in this post.
>> https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc
>>
>> If you really want the dedicated mux extension than one over SPDY, sorry
>> but please take over the work.
>>
>> _______________________________________________
>> hybi mailing list
>> hybi@ietf.org
>> https://www.ietf.org/mailman/listinfo/hybi
>>
>>
>>
>
>

--047d7bea3cb6812ac104eff262f9
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">I&#39;ve been lurking anyway :)<div><br></div><div>As I un=
derstand the proposal, (and it seems that Takeshi and Yukata are covering t=
his nicely, so what i say is probably redundant):</div><div>Message boundar=
ies, having semantic meaning, are preserved.</div>
<div><br></div><div>Frame boundaries, having no semantic meaning, are not p=
reserved. I don&#39;t think it makes sense to worry about this greatly when=
 the ability to extend the protocol as implemented in HTTP2 is present. It =
doesn&#39;t make a lot of sense to signal the frame boundaries through ever=
ything unless there is a high-likelyhood and high-utility use-case for such=
. If such is found, then HTTP2 can be revised to include it. This is the ap=
proach we&#39;ve been taking with most of the HTTP2 stuff, since there has =
been great effort placed on allowing new versions to be deployed in the fut=
ure.</div>
<div><br></div><div>The mapping of WS onto HTTP2 is fairly direct-- we rese=
rved the MSG_DONE bit specifically for this kind of use-case (i.e. bidirect=
ional, full-duplex, message-based).</div><div>The observation made is that =
there are essentially 3 types of data that need to be preserved one way or =
another:</div>
<div>1) metadata: data about the data, e.g. text, binary, where it is suppo=
sed to be going, etc.</div><div>2) payload</div><div>3) boundaries between =
these.</div><div><br></div><div>imho, metadata is handled with HEADERS (I&#=
39;d have preferred it to be called METADATA, but was voted down at the tim=
e :) ), payload is handled with DATA frames, and the boundaries are delimit=
ed either through the use of HEADERS frames, or via the MSG_DONE or STREAM_=
DONE bits.</div>
<div>In the presence of extensions, assuming that MSG_DONE is required to s=
ignal the end of a message, then HEADERS can be used to signal any boundary=
 desired.</div><div><br></div><div><br></div><div>I totally agree that it w=
ould be nice to expose the priority of various messages/channels with respe=
ct to each other. That is really an issue that is outside the WG&#39;s purv=
ue, though.</div>
<div><br></div></div><div class=3D"gmail_extra"><br><br><div class=3D"gmail=
_quote">On Thu, Jan 9, 2014 at 7:54 PM, Yutaka Hirano <span dir=3D"ltr">&lt=
;<a href=3D"mailto:yhirano@chromium.org" target=3D"_blank">yhirano@chromium=
.org</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div dir=3D"ltr"><div>+Roberto as an HTTP2 e=
xpert.</div><div class=3D"im"><div>=C2=A0</div><blockquote class=3D"gmail_q=
uote" style=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-c=
olor:rgb(204,204,204);border-left-style:solid;padding-left:1ex">

<span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,7=
3,125)">2) Ok, both WS frame and message boundaries are preserved.<br></spa=
n><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31=
,73,125)">=C2=A0<br>

</span><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:r=
gb(31,73,125)">But: I cannot recognize the end of a WS frame until the next=
 WS frame starts (for which I receive a HEADERS frame, which is the implici=
t signal that the previous WS frame ended), right?<br>

</span><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:r=
gb(31,73,125)"></span></blockquote></div>You are right, you can&#39;t.<div =
class=3D"im"><div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"ma=
rgin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,=
204);border-left-style:solid;padding-left:1ex">

<span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,7=
3,125)">=C2=A0<br></span><span style=3D"font-size:11pt;font-family:Calibri,=
sans-serif;color:rgb(31,73,125)">And if so, why is there no FRAME_DONE bit =
on the SPDY frame that corresponds to the end of a WS frame?</span></blockq=
uote>
</div>
Given that WebSocket frames &quot;have no semantic meaning&quot; as in RFC6=
455, I don&#39;t think the immediate recognition of the end of a WebSocket =
frame is requisite.<div>I would like to introduce MSG_DONE in HTTP2 so that=
 intermediaries that don&#39;t fully understand WS over HTTP2 can relay it.=
</div>

<div>Since there are not so much space in FLAGS field, Introducing FRAME_DO=
NE may decrease the extensiveness of the spec.</div><div>Do you have a stro=
ng preference to introduce FRAME_DONE?</div><div><br></div></div><div class=
=3D"HOEnZb">
<div class=3D"h5"><div class=3D"gmail_extra">
<br><br><div class=3D"gmail_quote">On Thu, Jan 9, 2014 at 6:53 PM, Tobias O=
berstein <span dir=3D"ltr">&lt;<a href=3D"mailto:tobias.oberstein@tavendo.d=
e" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt;</span> wrote:<br><=
blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px=
 #ccc solid;padding-left:1ex">

<div lang=3D"DE" link=3D"blue" vlink=3D"purple"><div><p class=3D"MsoNormal"=
><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans=
-serif&quot;;color:#1f497d">Hi Yutaka,<u></u><u></u></span></p><p class=3D"=
MsoNormal">

<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></span></p><p class=3D"MsoNo=
rmal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot=
;sans-serif&quot;;color:#1f497d">1) Yes, I understand. Thanks!<u></u><u></u=
></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">2) Ok, both WS fram=
e and message boundaries are preserved.<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">But: I cannot recog=
nize the end of a WS frame until the next WS frame starts (for which I rece=
ive a HEADERS frame, which is the implicit signal that the previous WS fram=
e ended), right?<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">And if so, why is t=
here no FRAME_DONE bit on the SPDY frame that corresponds to the end of a W=
S frame?<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">3) Ok, great! Hope =
this works, since SPDY provides header compression, but no sophisticated bi=
directional payload compression as &quot;WS permessage-deflate&quot; does.<=
u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Cheers,<u></u><u></=
u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">/Tobias<u></u><u></u></sp=
an></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&=
quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></=
u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u=
></span></p>

<div style=3D"border:none;border-left:solid blue 1.5pt;padding:0cm 0cm 0cm =
4.0pt"><div><div style=3D"border:none;border-top:solid #e1e1e1 1.0pt;paddin=
g:3.0pt 0cm 0cm 0cm"><p class=3D"MsoNormal"><b><span style=3D"font-size:11.=
0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;">Von:</span></b>=
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;"> <a href=3D"mailto:yhirano@google.com" target=3D"_blank">yhira=
no@google.com</a> [mailto:<a href=3D"mailto:yhirano@google.com" target=3D"_=
blank">yhirano@google.com</a>] <b>Im Auftrag von </b>Yutaka Hirano<br>

<b>Gesendet:</b> Donnerstag, 9. Januar 2014 10:37<br><b>An:</b> Tobias Ober=
stein<br><b>Cc:</b> Takeshi Yoshino; <a href=3D"mailto:hybi@ietf.org" targe=
t=3D"_blank">hybi@ietf.org</a><br><b>Betreff:</b> Re: [hybi] Discontinuatio=
n of mux standardizaton in favor of WS/HTTP/2.0<u></u><u></u></span></p>

</div></div><div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><div><=
p class=3D"MsoNormal">Hi Tobias,<u></u><u></u></p><div><p class=3D"MsoNorma=
l">Thank you for the comments.<u></u><u></u></p></div><div><p class=3D"MsoN=
ormal">
<u></u>=C2=A0<u></u></p><blockquote style=3D"border:none;border-left:solid =
#cccccc 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm"=
><p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;A=
rial&quot;,&quot;sans-serif&quot;">1) Frame Mapping<br>

&quot;&quot;&quot;<br>One WebSocket frame is mapped into one preceding SPDY=
 HEADERS frame and following plural SPDY data frames. The SPDY HEADERS fram=
e must contain WebSocket frame fields and following SPDY data frames contai=
n payload data. These data frames are free to be reframed.<br>

&quot;&quot;&quot;<br></span><a href=3D"https://docs.google.com/document/d/=
124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank"><span =
style=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&qu=
ot;">https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw=
9aWzSHijk/edit</span></a><span style=3D"font-size:10.0pt;font-family:&quot;=
Arial&quot;,&quot;sans-serif&quot;"><br>

So reframing of WebSocket frames (the payload) into multiple SPDY data fram=
es MAY occur, but does not need to occur? And a WS/SPDY implementation is f=
ree to choose the boundaries within WS frame payload at which to reframe in=
to different SPDY data frames as it likes?</span><u></u><u></u></p>

</blockquote><div><p class=3D"MsoNormal">A WebSocket frame MAY be split to =
multiple SPDY DATA frames.<u></u><u></u></p></div><div><p class=3D"MsoNorma=
l">Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA fram=
e into multiple DATA frames.<u></u><u></u></p>

</div><div><p class=3D"MsoNormal">Is this answering your question?<u></u><u=
></u></p></div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><b=
lockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm =
0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm">

<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ar=
ial&quot;,&quot;sans-serif&quot;">2) End of WS Frame<br>The end of a WS mes=
sage is signaled with a SPDY data frame having MSG_DONE set.<br>How is the =
end of a WS frame signaled, given that the WS frame may have been reframed =
into multiple SPDY data frames?</span><u></u><u></u></p>

</blockquote><div><p class=3D"MsoNormal">When you receive a HEADERS frame, =
you can recognize that the last SPDY frame is that last frame in a WebSocke=
t frame.<u></u><u></u></p></div><div><p class=3D"MsoNormal">When you receiv=
e a SPDY frame with MSG_DONE, you can recognize that it is the last SPDY fr=
ame in a WebSocket message and hence the last SPDY frame in a WebSocket fra=
me.<u></u><u></u></p>

</div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><blockquote=
 style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm 6=
.0pt;margin-left:4.8pt;margin-right:0cm"><p class=3D"MsoNormal"><span style=
=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&quot;">=
3) WS extension<br>

Are WS extensions fully supported? In particular, is &quot;permessage-defla=
te&quot; supported when running WS/SPDY?</span><u></u><u></u></p></blockquo=
te><div><p class=3D"MsoNormal">Yes. We have reserved bits and we can emulat=
e the opening handshake, I think.<u></u><u></u></p>

</div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div><p cla=
ss=3D"MsoNormal">On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein &lt;<a hr=
ef=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstei=
n@tavendo.de</a>&gt; wrote:<u></u><u></u></p>

<blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0c=
m 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm"><p class=3D"MsoNormal">=
Hi Takeshi,<br><br>it&#39;s a pity, but I can understand the motivation to =
shove everything into HTTP2.<br>

<br>Anyway, some questions:<br><br>1) Frame Mapping<br>&quot;&quot;&quot;<b=
r>One WebSocket frame is mapped into one preceding SPDY HEADERS frame and f=
ollowing plural SPDY data frames. The SPDY HEADERS frame must contain WebSo=
cket frame fields and following SPDY data frames contain payload data. Thes=
e data frames are free to be reframed.<br>

&quot;&quot;&quot;<br><a href=3D"https://docs.google.com/document/d/124vB5N=
V1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank">https://docs.=
google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit</a>=
<br>

<br>So reframing of WebSocket frames (the payload) into multiple SPDY data =
frames MAY occur, but does not need to occur? And a WS/SPDY implementation =
is free to choose the boundaries within WS frame payload at which to refram=
e into different SPDY data frames as it likes?<br>

<br>2) End of WS Frame<br><br>The end of a WS message is signaled with a SP=
DY data frame having MSG_DONE set.<br><br>How is the end of a WS frame sign=
aled, given that the WS frame may have been reframed into multiple SPDY dat=
a frames?<br>

<br>3) WS extension<br><br>Are WS extensions fully supported? In particular=
, is &quot;permessage-deflate&quot; supported when running WS/SPDY?<br><br>=
Thanks!<br>/Tobias<br><br><br>Von: hybi [mailto:<a href=3D"mailto:hybi-boun=
ces@ietf.org" target=3D"_blank">hybi-bounces@ietf.org</a>] Im Auftrag von T=
akeshi Yoshino<br>

Gesendet: Dienstag, 7. Januar 2014 08:36<br>An: <a href=3D"mailto:hybi@ietf=
.org" target=3D"_blank">hybi@ietf.org</a><br>Betreff: [hybi] Discontinuatio=
n of mux standardizaton in favor of WS/HTTP/2.0<u></u><u></u></p><div><div>

<p class=3D"MsoNormal"><br>Hi all,<br><br>We&#39;ve been working on mux ext=
ension to make WebSocket scalable for long time, but as SPDY got widely sup=
ported and adopted quickly, now there seems to be no longer any good reason=
 to continue this work which is WebSocket dedicated.<br>

<br>Scalability is still important issue to address. From now, I&#39;d like=
 to focus on WS/SPDY effort which we&#39;ve been working on in parallel wit=
h the mux extension.<br><br>Please take a look at the latest proposal in th=
is post.<br>

<a href=3D"https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc" ta=
rget=3D"_blank">https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3gl=
c</a><br><br>If you really want the dedicated mux extension than one over S=
PDY, sorry but please take over the work.<u></u><u></u></p>

</div></div><div><div><p class=3D"MsoNormal">______________________________=
_________________<br>hybi mailing list<br><a href=3D"mailto:hybi@ietf.org" =
target=3D"_blank">hybi@ietf.org</a><br><a href=3D"https://www.ietf.org/mail=
man/listinfo/hybi" target=3D"_blank">https://www.ietf.org/mailman/listinfo/=
hybi</a><u></u><u></u></p>

</div></div></blockquote></div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u><=
/p></div></div></div></div></div></div></div></blockquote></div><br></div>
</div></div></blockquote></div><br></div>

--047d7bea3cb6812ac104eff262f9--

From tobias.oberstein@tavendo.de  Tue Jan 14 11:09:48 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9E02A1AE1DC for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 11:09:48 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.899
X-Spam-Level: 
X-Spam-Status: No, score=-1.899 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id I_p3T3ijH59i for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 11:09:45 -0800 (PST)
Received: from EXHUB020-5.exch020.serverdata.net (exhub020-5.exch020.serverdata.net [206.225.164.32]) by ietfa.amsl.com (Postfix) with ESMTP id 9F57B1AE1A8 for <hybi@ietf.org>; Tue, 14 Jan 2014 11:09:44 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-5.exch020.serverdata.net ([206.225.164.32]) with mapi; Tue, 14 Jan 2014 11:09:32 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Roberto Peon <fenix@google.com>, Yutaka Hirano <yhirano@chromium.org>
Date: Tue, 14 Jan 2014 11:09:29 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8RVq9V/GcOc/NpQLORHWawv32D/AAAf2/Q
Message-ID: <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com>
In-Reply-To: <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: multipart/alternative; boundary="_000_634914A010D0B943A035D226786325D4446BED9CDEEXVMBX02012ex_"
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 14 Jan 2014 19:09:48 -0000

--_000_634914A010D0B943A035D226786325D4446BED9CDEEXVMBX02012ex_
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_634914A010D0B943A035D226786325D4446BED9CDEEXVMBX02012ex_
Content-Type: text/html; charset="utf-8"
Content-Transfer-Encoding: base64
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=

--_000_634914A010D0B943A035D226786325D4446BED9CDEEXVMBX02012ex_--

From fenix@google.com  Tue Jan 14 12:07:43 2014
Return-Path: <fenix@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id D39091AE232 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 12:07:43 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.716
X-Spam-Level: 
X-Spam-Status: No, score=-0.716 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, J_CHICKENPOX_54=0.6, J_CHICKENPOX_57=0.6, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id B9IY9ERtu4Xu for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 12:07:38 -0800 (PST)
Received: from mail-qa0-x22c.google.com (mail-qa0-x22c.google.com [IPv6:2607:f8b0:400d:c00::22c]) by ietfa.amsl.com (Postfix) with ESMTP id 5042C1AE210 for <hybi@ietf.org>; Tue, 14 Jan 2014 12:07:38 -0800 (PST)
Received: by mail-qa0-f44.google.com with SMTP id w5so96554qac.17 for <hybi@ietf.org>; Tue, 14 Jan 2014 12:07:26 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type; bh=nFNWsMJtR+rS/D0nkAoc00YHUmI59hxLHhHrRjiX+P4=; b=iStDrno0wiUgU6AkCwBr+HDBDK9fPE842ZjlfBYaEjEG3D2cCQwfI1nC4w3dGblhGd WtDUzrxcGf7cIu6ZVbUHz45Id+xR/FLednkG7tdmgVA4ZH5JCDL+Z7sXuVj/WGfwwG/M yLItGdUNbPqLKERrfbmfYYOIXeAsUO5Hp8WYCJcjDT6Q31kgm51+xaPE7UVHgojRq7ge bj73aeK2WemSh/XUvBAzEtjww8gunMxift5ng9/5HHo1dhESKBj42eqzb3OnW2AriClc cT/uOA26kxsnosoukyvkN2VOwfB87ezhI+S4AKxhUOB1lNxFpZ4nSkPjeW9CDrjY1dbn ZvUQ==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=nFNWsMJtR+rS/D0nkAoc00YHUmI59hxLHhHrRjiX+P4=; b=RJwpGNjcBjMHom4sKi5RUD6zR9F4Z4P1tNvcpEjsmu2g9Bv7e9glcgbdLaL11g5EYK ACIlpZ/vU++Jkn7Qn0tPwhlAz51yiulJeUu47ykc3imdBaplxsK9Ps/Ve1fK0Scb40Bq WMc+c2bu9MbU19HHiK1bYw15xmamXyD4hSOemGpiZ3KMGS8A2D8aeyXskmm2l6otjzh9 DPYQOeoJ5/fNhrXJUeF9cmaJnMqe2ky4IXazAXOOwUACe6ojfjIAnIyVvIZo4zrFqpyP wAcV3UOycRL/TaXfv40cAM3BbO3X9VE5+EUBnoX/WlFsaKtitPeToYgV1wMpiOcPMEd2 4YlA==
X-Gm-Message-State: ALoCoQnwC1Lrk/cEI6ocQAgSF0CauPBc/vp2qczb95i2kXGSG9JAURIunX2cx4aPivUvfQskUbyiiJGtlXHFqGH5FWGNPfERphFY+wtmA0sLpGM+9m2zcPEd3V638lTRzBIWJmC1xxDvqzZSNJwCvtebKtQtwMU6iUEOv2U7mgy58alD441n+bGsYID5wqTq5DR1TjiHlo70
MIME-Version: 1.0
X-Received: by 10.49.34.200 with SMTP id b8mr3492024qej.7.1389730046608; Tue, 14 Jan 2014 12:07:26 -0800 (PST)
Received: by 10.229.13.7 with HTTP; Tue, 14 Jan 2014 12:07:26 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net>
Date: Tue, 14 Jan 2014 12:07:26 -0800
Message-ID: <CAGzyod7HAjdDHt1B4XRHT6h43DXnoupqGmr4buP6D+bQs86m+w@mail.gmail.com>
From: Roberto Peon <fenix@google.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=047d7b33cba0c31fd904eff3bd41
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 14 Jan 2014 20:07:44 -0000

--047d7b33cba0c31fd904eff3bd41
Content-Type: text/plain; charset=UTF-8

On Tue, Jan 14, 2014 at 11:09 AM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> Hi Roberto,
>
>
>
> not sure if you agree, but FWIW, here is a summary as I see it (currently):
>
>
>
> 1) WebSocket frame boundaries _can_ have semantic meaning in WebSocket
> _extensions_
>

Agreed. And there are none deplyed that we know about where this is a
concern.


>
>
> 2) WebSocket message and frame boundaries are indeed preserved in the
> current WS/SPDY proposal,
>
> however, non-final WebSocket frame ends cannot be _immediately_ detected.
>

I'm not sure what you mean here?


>
>
> 3) The current WS/SPDY proposal
>
>   * is comfortable to implement for implementations that _only_ want to
> implement WS/SPDY, and not native WS at the same time
>
>  * makes things complicated for WS implementations that want to support
> both WS/SPDY _and_ native WS at the same time
>
>  * has significant wire-level overhead compared to native WS (and compared
> to the proposal below)
>
>
>

The complication here is relative. Using the approach proposed in terms of
using MSG_DONE, etc. would make deployment at large sites like Google
substantially easier as large sites typically want to see the semantic
boundaries easily, quickly, and consistently, as there needs to be robust
mechanisms for handling failover, etc. The more layers of interpretation
there are when doing this, the bigger the number of problems.

I'm honestly not worried about the potential increased overhead-- arguably
HTTP2 reduces overhead substantially simply be requiring fewer handshakes
(the connection setup is a non-trivial amount of bytes and latency. That
this was ignored was hugely frustrating to me during the initial
standardization process here in HyBi). Even discounting this, I haven't
actually heard of use-cases where a well-written application sends so many
messages that the bandwidth is saturated. I am for efficiency (had to work
on compression for HTTP2, for instance), but not when it increases
complexity in the more general case.

In terms of total amount of code:
In any layering scheme, we'd need to use a HEADER frame one way or another
because we need to at least announce that we're connecting to a WS endpoint.
In terms of decode, we're essentially talking the following state machine:

if frame.type == HEADERS:
    streams[frame.stream_id] = Stream(frame.headers)
elif frame.type == DATA:
    payload[frame.stream_id] += frame.payload
    if frame.MSG_DONE or frame.STREAM_DONE:
      passMessageToApplication(payload[frame.stream_id])
      del payload[frame.stream_id]
if frame.STREAM_DONE:
  del streams[frame.stream_id]

While the above is pseudocode and leaves out error handling, one of the
interesting things is that it is the same for both HTTP, WS, or any other
protocol layering over HTTP2.




> I'd like to suggest a different WS/SPDY:
>
>
>
> Pack WebSocket frames _including_ full WS frame header as _payload_ of
> SPDY data frames only - no SPDY header frames at all.
>
>
>
> In the absence of negotiated WebSocket extensions (or when all WS
> extensions in place do not assign semantics to WS frame boundaries),
>
> fragment WebSocket payload at the WS level such that WS frame
> header+payload fits into the payload of SPDY data frames.
>
>
>
> In the latter case, this essentially adds a small SPDY data frame header
> (of fixed length) to WS frames. Easy to implement and efficient on wire.
>
>
>
> IMHO, this has advantages compared to current proposal and considering 3)
> above.
>
>
>
> The only disadvantage is for implementations that _only_ want to implement
> WS/SPDY and not native WS at the same time.
>

One of the nice things about the current plan as compared to your proposal
is that it allows for generic infrastructure to be built to examine and
handle things for DoS and loadbalancing concerns across any protocol using
HTTP2 as a session layer since all boundaries are clearly demarcated. It
seems like a very nice thing indeed that HTTP, WS, various internal API
things, etc. can all end up using the same infrastructure.

Additionally, what happens if/when someone defines a mapping for HTTP2 onto
something like QUIC, SCTP, TCP-Minion, or perhaps the RTCWeb stack? With
the proposed design where the HTTP2 layer has full visibility into the
semantic boundaries, this mapping needs to be done only once for the HTTP2
session layer and all protocols layered atop it are now able to take
advantage of it (one would need signaling about HOL blocking properties of
streams to the HTTP2 layer in addition, of course). When we hide the
semantic boundaries, this requires work from every implementor, and impedes
adoption either of the new HTTP2 mapping onto the better transport, or
impedes adoption of WS.

I want to see the WS API succeed at scale and continue to do so into the
future, so, my thought is that the advantages in regularity/consistency
(today for DoS and loadbalancing, tomorrow on new transports) outweigh the
relatively small theoretical byte efficiency tradeoff.
As I mentioned earlier, this is the tradeoff that HTTP2 has made for HTTP
at the application layer as well, so even the approach is consistent :)


>
> ==
>
>
>
> Regarding "permessage-priority" with ordering guarantees: should I find
> time and if there is interest, I'll write something up.
>
> This could be a WebSocket extension. To be useful, the WS API in browsers
> would need to expose it though ..
>

Yup. The same thing is happening for HTTP (people finding that expressing
priority for resources is useful), btw, from what I'm led to understand.
I'm hopefuly for the day when APIs allow us to easily signal both priority,
and also other important properties like drop-dead-times (get this message
to the other side at most in X seconds from now, else leave it on the
floor), and ordering requirements (e.g. don't care about ordering, care
about ordering, but only to know when something has been dropped, etc.).
The interesting thing about priorities is that it is really only useful
when you're muxing over one connection/session.
-=R


>
> Just my 2cts,
>
> Tobias
>
>
>
>
>
>
>
>
>
> *Von:* Roberto Peon [mailto:fenix@google.com]
> *Gesendet:* Dienstag, 14. Januar 2014 19:30
> *An:* Yutaka Hirano
> *Cc:* Tobias Oberstein; Takeshi Yoshino; hybi@ietf.org
>
> *Betreff:* Re: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
>
> I've been lurking anyway :)
>
>
>
> As I understand the proposal, (and it seems that Takeshi and Yukata are
> covering this nicely, so what i say is probably redundant):
>
> Message boundaries, having semantic meaning, are preserved.
>
>
>
> Frame boundaries, having no semantic meaning, are not preserved. I don't
> think it makes sense to worry about this greatly when the ability to extend
> the protocol as implemented in HTTP2 is present. It doesn't make a lot of
> sense to signal the frame boundaries through everything unless there is a
> high-likelyhood and high-utility use-case for such. If such is found, then
> HTTP2 can be revised to include it. This is the approach we've been taking
> with most of the HTTP2 stuff, since there has been great effort placed on
> allowing new versions to be deployed in the future.
>
>
>
> The mapping of WS onto HTTP2 is fairly direct-- we reserved the MSG_DONE
> bit specifically for this kind of use-case (i.e. bidirectional,
> full-duplex, message-based).
>
> The observation made is that there are essentially 3 types of data that
> need to be preserved one way or another:
>
> 1) metadata: data about the data, e.g. text, binary, where it is supposed
> to be going, etc.
>
> 2) payload
>
> 3) boundaries between these.
>
>
>
> imho, metadata is handled with HEADERS (I'd have preferred it to be called
> METADATA, but was voted down at the time :) ), payload is handled with DATA
> frames, and the boundaries are delimited either through the use of HEADERS
> frames, or via the MSG_DONE or STREAM_DONE bits.
>
> In the presence of extensions, assuming that MSG_DONE is required to
> signal the end of a message, then HEADERS can be used to signal any
> boundary desired.
>
>
>
>
>
> I totally agree that it would be nice to expose the priority of various
> messages/channels with respect to each other. That is really an issue that
> is outside the WG's purvue, though.
>
>
>
>
>
> On Thu, Jan 9, 2014 at 7:54 PM, Yutaka Hirano <yhirano@chromium.org>
> wrote:
>
> +Roberto as an HTTP2 expert.
>
>
>
> 2) Ok, both WS frame and message boundaries are preserved.
>
> But: I cannot recognize the end of a WS frame until the next WS frame
> starts (for which I receive a HEADERS frame, which is the implicit signal
> that the previous WS frame ended), right?
>
> You are right, you can't.
>
>
>
>
> And if so, why is there no FRAME_DONE bit on the SPDY frame that
> corresponds to the end of a WS frame?
>
> Given that WebSocket frames "have no semantic meaning" as in RFC6455, I
> don't think the immediate recognition of the end of a WebSocket frame is
> requisite.
>
> I would like to introduce MSG_DONE in HTTP2 so that intermediaries that
> don't fully understand WS over HTTP2 can relay it.
>
> Since there are not so much space in FLAGS field, Introducing FRAME_DONE
> may decrease the extensiveness of the spec.
>
> Do you have a strong preference to introduce FRAME_DONE?
>
>
>
>
>
> On Thu, Jan 9, 2014 at 6:53 PM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
>
> Hi Yutaka,
>
>
>
> 1) Yes, I understand. Thanks!
>
>
>
> 2) Ok, both WS frame and message boundaries are preserved.
>
>
>
> But: I cannot recognize the end of a WS frame until the next WS frame
> starts (for which I receive a HEADERS frame, which is the implicit signal
> that the previous WS frame ended), right?
>
>
>
> And if so, why is there no FRAME_DONE bit on the SPDY frame that
> corresponds to the end of a WS frame?
>
>
>
> 3) Ok, great! Hope this works, since SPDY provides header compression, but
> no sophisticated bidirectional payload compression as "WS
> permessage-deflate" does.
>
>
>
> Cheers,
>
> /Tobias
>
>
>
>
>
>
>
> *Von:* yhirano@google.com [mailto:yhirano@google.com] *Im Auftrag von *Yutaka
> Hirano
> *Gesendet:* Donnerstag, 9. Januar 2014 10:37
> *An:* Tobias Oberstein
> *Cc:* Takeshi Yoshino; hybi@ietf.org
> *Betreff:* Re: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
>
> Hi Tobias,
>
> Thank you for the comments.
>
>
>
> 1) Frame Mapping
> """
> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
> following plural SPDY data frames. The SPDY HEADERS frame must contain
> WebSocket frame fields and following SPDY data frames contain payload data.
> These data frames are free to be reframed.
> """
>
> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
> So reframing of WebSocket frames (the payload) into multiple SPDY data
> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
> is free to choose the boundaries within WS frame payload at which to
> reframe into different SPDY data frames as it likes?
>
> A WebSocket frame MAY be split to multiple SPDY DATA frames.
>
> Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA frame
> into multiple DATA frames.
>
> Is this answering your question?
>
>
>
> 2) End of WS Frame
> The end of a WS message is signaled with a SPDY data frame having MSG_DONE
> set.
> How is the end of a WS frame signaled, given that the WS frame may have
> been reframed into multiple SPDY data frames?
>
> When you receive a HEADERS frame, you can recognize that the last SPDY
> frame is that last frame in a WebSocket frame.
>
> When you receive a SPDY frame with MSG_DONE, you can recognize that it is
> the last SPDY frame in a WebSocket message and hence the last SPDY frame in
> a WebSocket frame.
>
>
>
> 3) WS extension
> Are WS extensions fully supported? In particular, is "permessage-deflate"
> supported when running WS/SPDY?
>
> Yes. We have reserved bits and we can emulate the opening handshake, I
> think.
>
>
>
> On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
>
> Hi Takeshi,
>
> it's a pity, but I can understand the motivation to shove everything into
> HTTP2.
>
> Anyway, some questions:
>
> 1) Frame Mapping
> """
> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
> following plural SPDY data frames. The SPDY HEADERS frame must contain
> WebSocket frame fields and following SPDY data frames contain payload data.
> These data frames are free to be reframed.
> """
>
> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
>
> So reframing of WebSocket frames (the payload) into multiple SPDY data
> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
> is free to choose the boundaries within WS frame payload at which to
> reframe into different SPDY data frames as it likes?
>
> 2) End of WS Frame
>
> The end of a WS message is signaled with a SPDY data frame having MSG_DONE
> set.
>
> How is the end of a WS frame signaled, given that the WS frame may have
> been reframed into multiple SPDY data frames?
>
> 3) WS extension
>
> Are WS extensions fully supported? In particular, is "permessage-deflate"
> supported when running WS/SPDY?
>
> Thanks!
> /Tobias
>
>
> Von: hybi [mailto:hybi-bounces@ietf.org] Im Auftrag von Takeshi Yoshino
> Gesendet: Dienstag, 7. Januar 2014 08:36
> An: hybi@ietf.org
> Betreff: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
> Hi all,
>
> We've been working on mux extension to make WebSocket scalable for long
> time, but as SPDY got widely supported and adopted quickly, now there seems
> to be no longer any good reason to continue this work which is WebSocket
> dedicated.
>
> Scalability is still important issue to address. From now, I'd like to
> focus on WS/SPDY effort which we've been working on in parallel with the
> mux extension.
>
> Please take a look at the latest proposal in this post.
> https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc
>
> If you really want the dedicated mux extension than one over SPDY, sorry
> but please take over the work.
>
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>
>
>
>
>
>
>

--047d7b33cba0c31fd904eff3bd41
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><br><div class=3D"gmail_extra"><br><br><div class=3D"gmail=
_quote">On Tue, Jan 14, 2014 at 11:09 AM, Tobias Oberstein <span dir=3D"ltr=
">&lt;<a href=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobi=
as.oberstein@tavendo.de</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-=
left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;p=
adding-left:1ex"><div lang=3D"DE" link=3D"blue" vlink=3D"purple"><div><p cl=
ass=3D"MsoNormal">
<span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,7=
3,125)">Hi Roberto,<u></u><u></u></span></p><p class=3D"MsoNormal"><span st=
yle=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,73,125)">=
<u></u>=C2=A0<u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">not sure if you agree, but FWIW, here is a s=
ummary as I see it (currently):<u></u><u></u></span></p><p class=3D"MsoNorm=
al">
<span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,7=
3,125)"><u></u>=C2=A0<u></u></span></p><p class=3D"MsoNormal"><span style=
=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,73,125)">1) =
WebSocket frame boundaries _can_ have semantic meaning in WebSocket _extens=
ions_</span></p>
</div></div></blockquote><div><br></div><div>Agreed. And there are none dep=
lyed that we know about where this is a concern.</div><div>=C2=A0</div><blo=
ckquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left=
-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;paddi=
ng-left:1ex">
<div lang=3D"DE" link=3D"blue" vlink=3D"purple"><div><p class=3D"MsoNormal"=
><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,=
73,125)"><u></u><u></u></span></p><p class=3D"MsoNormal"><span style=3D"fon=
t-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,73,125)"><u></u>=C2=
=A0<u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">2) WebSocket message and frame boundaries ar=
e indeed preserved in the current WS/SPDY proposal,<u></u><u></u></span></p=
>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">however, non-final WebSocket frame ends cann=
ot be _immediately_ detected.</span></p></div></div></blockquote><div><br><=
/div>
<div>I&#39;m not sure what you mean here?</div><div>=C2=A0</div><blockquote=
 class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left-width:=
1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left=
:1ex">
<div lang=3D"DE" link=3D"blue" vlink=3D"purple"><div><p class=3D"MsoNormal"=
><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,=
73,125)"><u></u><u></u></span></p><p class=3D"MsoNormal"><span style=3D"fon=
t-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,73,125)"><u></u>=C2=
=A0<u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">3) The current WS/SPDY proposal<u></u><u></u=
></span></p><p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-famil=
y:Calibri,sans-serif;color:rgb(31,73,125)">=C2=A0 * is comfortable to imple=
ment for implementations that _only_ want to implement WS/SPDY, and not nat=
ive WS at the same time<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)"> =C2=A0* makes things complicated for WS imp=
lementations that want to support both WS/SPDY _and_ native WS at the same =
time<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)"> =C2=A0* has significant wire-level overhead=
 compared to native WS (and compared to the proposal below)<u></u><u></u></=
span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)"><u></u>=C2=A0</span></p></div></div></blockq=
uote><div><br></div><div>The complication here is relative. Using the appro=
ach proposed in terms of using MSG_DONE, etc. would make deployment at larg=
e sites like Google substantially easier as large sites typically want to s=
ee the semantic boundaries easily, quickly, and consistently, as there need=
s to be robust mechanisms for handling failover, etc. The more layers of in=
terpretation there are when doing this, the bigger the number of problems.=
=C2=A0</div>
<div><br></div><div>I&#39;m honestly not worried about the potential increa=
sed overhead-- arguably HTTP2 reduces overhead substantially simply be requ=
iring fewer handshakes (the connection setup is a non-trivial amount of byt=
es and latency. That this was ignored was hugely frustrating to me during t=
he initial standardization process here in HyBi). Even discounting this, I =
haven&#39;t actually heard of use-cases where a well-written application se=
nds so many messages that the bandwidth is saturated. I am for efficiency (=
had to work on compression for HTTP2, for instance), but not when it increa=
ses complexity in the more general case.</div>
<div><br></div><div>In terms of total amount of code:</div><div>In any laye=
ring scheme, we&#39;d need to use a HEADER frame one way or another because=
 we need to at least announce that we&#39;re connecting to a WS endpoint.</=
div>
<div>In terms of decode, we&#39;re essentially talking the following state =
machine:</div><div><br></div><div>if frame.type =3D=3D HEADERS:</div><div>=
=C2=A0 =C2=A0 streams[frame.stream_id] =3D Stream(frame.headers)<br></div><=
div>elif frame.type =3D=3D DATA:</div>
<div>=C2=A0 =C2=A0 payload[frame.stream_id] +=3D frame.payload<br></div><di=
v>=C2=A0 =C2=A0 if frame.MSG_DONE or frame.STREAM_DONE:</div><div>=C2=A0 =
=C2=A0 =C2=A0 passMessageToApplication(payload[frame.stream_id])</div><div>=
=C2=A0 =C2=A0 =C2=A0 del payload[frame.stream_id]</div>
<div>if frame.STREAM_DONE:</div><div>=C2=A0 del streams[frame.stream_id]</d=
iv><div><br></div><div>While the above is pseudocode and leaves out error h=
andling, one of the interesting things is that it is the same for both HTTP=
, WS, or any other protocol layering over HTTP2.</div>
<div><br></div><div><br></div><div>=C2=A0</div><blockquote class=3D"gmail_q=
uote" style=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-c=
olor:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><div lang=
=3D"DE" link=3D"blue" vlink=3D"purple">
<div><p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calib=
ri,sans-serif;color:rgb(31,73,125)"><u></u></span></p><p class=3D"MsoNormal=
"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31=
,73,125)">I&#39;d like to suggest a different WS/SPDY:<u></u><u></u></span>=
</p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)"><u></u>=C2=A0<u></u></span></p><p>Pack WebSo=
cket frames _including_ full WS frame header as _payload_ of SPDY data fram=
es only - no SPDY header frames at all.<u></u><u></u></p>
<p><u></u>=C2=A0<u></u></p><p>In the absence of negotiated WebSocket extens=
ions (or when all WS extensions in place do not assign semantics to WS fram=
e boundaries),<u></u><u></u></p><p>fragment WebSocket payload at the WS lev=
el such that WS frame header+payload fits into the payload of SPDY data fra=
mes.<u></u><u></u></p>
<p><u></u>=C2=A0<u></u></p><p>In the latter case, this essentially adds a s=
mall SPDY data frame header (of fixed length) to WS frames. Easy to impleme=
nt and efficient on wire.<u></u><u></u></p><p><u></u>=C2=A0<u></u></p><p>IM=
HO, this has advantages compared to current proposal and considering 3) abo=
ve.<u></u><u></u></p>
<p><u></u>=C2=A0<u></u></p><p>The only disadvantage is for implementations =
that _only_ want to implement WS/SPDY and not native WS at the same time.</=
p></div></div></blockquote><div><br></div><div>One of the nice things about=
 the current plan as compared to your proposal is that it allows for generi=
c infrastructure to be built to examine and handle things for DoS and loadb=
alancing concerns across any protocol using HTTP2 as a session layer since =
all boundaries are clearly demarcated. It seems like a very nice thing inde=
ed that HTTP, WS, various internal API things, etc. can all end up using th=
e same infrastructure.<br>
</div><div><br></div><div>Additionally, what happens if/when someone define=
s a mapping for HTTP2 onto something like QUIC, SCTP, TCP-Minion, or perhap=
s the RTCWeb stack? With the proposed design where the HTTP2 layer has full=
 visibility into the semantic boundaries, this mapping needs to be done onl=
y once for the HTTP2 session layer and all protocols layered atop it are no=
w able to take advantage of it (one would need signaling about HOL blocking=
 properties of streams to the HTTP2 layer in addition, of course). When we =
hide the semantic boundaries, this requires work from every implementor, an=
d impedes adoption either of the new HTTP2 mapping onto the better transpor=
t, or impedes adoption of WS.</div>
<div><br></div><div>I want to see the WS API succeed at scale and continue =
to do so into the future, so, my thought is that the advantages in regulari=
ty/consistency (today for DoS and loadbalancing, tomorrow on new transports=
) outweigh the relatively small theoretical byte efficiency tradeoff.</div>
<div>As I mentioned earlier, this is the tradeoff that HTTP2 has made for H=
TTP at the application layer as well, so even the approach is consistent :)=
</div><div><br></div><blockquote class=3D"gmail_quote" style=3D"margin:0px =
0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);bord=
er-left-style:solid;padding-left:1ex">
<div lang=3D"DE" link=3D"blue" vlink=3D"purple"><div><p>=C2=A0<u></u></p><p=
>=3D=3D<u></u><u></u></p><p><u></u>=C2=A0<u></u></p><p>Regarding &quot;perm=
essage-priority&quot; with ordering guarantees: should I find time and if t=
here is interest, I&#39;ll write something up.<u></u><u></u></p>
<p>This could be a WebSocket extension. To be useful, the WS API in browser=
s would need to expose it though ..</p></div></div></blockquote><div><br></=
div><div>Yup. The same thing is happening for HTTP (people finding that exp=
ressing priority for resources is useful), btw, from what I&#39;m led to un=
derstand.=C2=A0</div>
<div>I&#39;m hopefuly for the day when APIs allow us to easily signal both =
priority, and also other important properties like drop-dead-times (get thi=
s message to the other side at most in X seconds from now, else leave it on=
 the floor), and ordering requirements (e.g. don&#39;t care about ordering,=
 care about ordering, but only to know when something has been dropped, etc=
.).</div>
<div>The interesting thing about priorities is that it is really only usefu=
l when you&#39;re muxing over one connection/session.</div><div>-=3DR</div>=
<div><br></div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0p=
x 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-lef=
t-style:solid;padding-left:1ex">
<div lang=3D"DE" link=3D"blue" vlink=3D"purple"><div><p><u></u><u></u></p><=
p><u></u>=C2=A0<u></u></p><p>Just my 2cts,<u></u><u></u></p><p>Tobias<u></u=
><u></u></p><p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-famil=
y:Calibri,sans-serif;color:rgb(31,73,125)"><u></u>=C2=A0<u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)"><u></u>=C2=A0<u></u></span></p><p class=3D"M=
soNormal"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;colo=
r:rgb(31,73,125)"><u></u>=C2=A0<u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)"><u></u>=C2=A0<u></u></span></p><div style=3D=
"border-style:none none none solid;border-left-color:blue;border-left-width=
:1.5pt;padding:0cm 0cm 0cm 4pt">
<div><div style=3D"border-style:solid none none;border-top-color:rgb(225,22=
5,225);border-top-width:1pt;padding:3pt 0cm 0cm"><p class=3D"MsoNormal"><b>=
<span style=3D"font-size:11pt;font-family:Calibri,sans-serif">Von:</span></=
b><span style=3D"font-size:11pt;font-family:Calibri,sans-serif"> Roberto Pe=
on [mailto:<a href=3D"mailto:fenix@google.com" target=3D"_blank">fenix@goog=
le.com</a>] <br>
<b>Gesendet:</b> Dienstag, 14. Januar 2014 19:30<br><b>An:</b> Yutaka Hiran=
o<br><b>Cc:</b> Tobias Oberstein; Takeshi Yoshino; <a href=3D"mailto:hybi@i=
etf.org" target=3D"_blank">hybi@ietf.org</a></span></p><div><div class=3D"h=
5">
<br><b>Betreff:</b> Re: [hybi] Discontinuation of mux standardizaton in fav=
or of WS/HTTP/2.0<u></u><u></u></div></div><p></p></div></div><div><div cla=
ss=3D"h5"><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><div><p class=3D"M=
soNormal">
I&#39;ve been lurking anyway :)<u></u><u></u></p><div><p class=3D"MsoNormal=
"><u></u>=C2=A0<u></u></p></div><div><p class=3D"MsoNormal">As I understand=
 the proposal, (and it seems that Takeshi and Yukata are covering this nice=
ly, so what i say is probably redundant):<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">Message boundaries, having semantic meani=
ng, are preserved.<u></u><u></u></p></div><div><p class=3D"MsoNormal"><u></=
u>=C2=A0<u></u></p></div><div><p class=3D"MsoNormal">Frame boundaries, havi=
ng no semantic meaning, are not preserved. I don&#39;t think it makes sense=
 to worry about this greatly when the ability to extend the protocol as imp=
lemented in HTTP2 is present. It doesn&#39;t make a lot of sense to signal =
the frame boundaries through everything unless there is a high-likelyhood a=
nd high-utility use-case for such. If such is found, then HTTP2 can be revi=
sed to include it. This is the approach we&#39;ve been taking with most of =
the HTTP2 stuff, since there has been great effort placed on allowing new v=
ersions to be deployed in the future.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div><p cla=
ss=3D"MsoNormal">The mapping of WS onto HTTP2 is fairly direct-- we reserve=
d the MSG_DONE bit specifically for this kind of use-case (i.e. bidirection=
al, full-duplex, message-based).<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">The observation made is that there are es=
sentially 3 types of data that need to be preserved one way or another:<u><=
/u><u></u></p></div><div><p class=3D"MsoNormal">1) metadata: data about the=
 data, e.g. text, binary, where it is supposed to be going, etc.<u></u><u><=
/u></p>
</div><div><p class=3D"MsoNormal">2) payload<u></u><u></u></p></div><div><p=
 class=3D"MsoNormal">3) boundaries between these.<u></u><u></u></p></div><d=
iv><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div><p class=3D"Ms=
oNormal">
imho, metadata is handled with HEADERS (I&#39;d have preferred it to be cal=
led METADATA, but was voted down at the time :) ), payload is handled with =
DATA frames, and the boundaries are delimited either through the use of HEA=
DERS frames, or via the MSG_DONE or STREAM_DONE bits.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">In the presence of extensions, assuming t=
hat MSG_DONE is required to signal the end of a message, then HEADERS can b=
e used to signal any boundary desired.<u></u><u></u></p></div><div><p class=
=3D"MsoNormal">
<u></u>=C2=A0<u></u></p></div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u><=
/u></p></div><div><p class=3D"MsoNormal">I totally agree that it would be n=
ice to expose the priority of various messages/channels with respect to eac=
h other. That is really an issue that is outside the WG&#39;s purvue, thoug=
h.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div></div><div>=
<p class=3D"MsoNormal" style=3D"margin-bottom:12pt"><u></u>=C2=A0<u></u></p=
><div><p class=3D"MsoNormal">On Thu, Jan 9, 2014 at 7:54 PM, Yutaka Hirano =
&lt;<a href=3D"mailto:yhirano@chromium.org" target=3D"_blank">yhirano@chrom=
ium.org</a>&gt; wrote:<u></u><u></u></p>
<blockquote style=3D"border-style:none none none solid;border-left-color:rg=
b(204,204,204);border-left-width:1pt;padding:0cm 0cm 0cm 6pt;margin:5pt 0cm=
 5pt 4.8pt"><div><div><p class=3D"MsoNormal">+Roberto as an HTTP2 expert.<u=
></u><u></u></p>
</div><div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><block=
quote style=3D"border-style:none none none solid;border-left-color:rgb(204,=
204,204);border-left-width:1pt;padding:0cm 0cm 0cm 6pt;margin:5pt 0cm 5pt 4=
.8pt">
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">2) Ok, both WS frame and message boundaries =
are preserved.<br>=C2=A0<br>But: I cannot recognize the end of a WS frame u=
ntil the next WS frame starts (for which I receive a HEADERS frame, which i=
s the implicit signal that the previous WS frame ended), right?</span><u></=
u><u></u></p>
</blockquote></div><p class=3D"MsoNormal">You are right, you can&#39;t.<u><=
/u><u></u></p><div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></di=
v><blockquote style=3D"border-style:none none none solid;border-left-color:=
rgb(204,204,204);border-left-width:1pt;padding:0cm 0cm 0cm 6pt;margin:5pt 0=
cm 5pt 4.8pt">
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">=C2=A0<br>And if so, why is there no FRAME_D=
ONE bit on the SPDY frame that corresponds to the end of a WS frame?</span>=
<u></u><u></u></p>
</blockquote></div><p class=3D"MsoNormal">Given that WebSocket frames &quot=
;have no semantic meaning&quot; as in RFC6455, I don&#39;t think the immedi=
ate recognition of the end of a WebSocket frame is requisite.<u></u><u></u>=
</p>
<div><p class=3D"MsoNormal">I would like to introduce MSG_DONE in HTTP2 so =
that intermediaries that don&#39;t fully understand WS over HTTP2 can relay=
 it.<u></u><u></u></p></div><div><p class=3D"MsoNormal">Since there are not=
 so much space in FLAGS field, Introducing FRAME_DONE may decrease the exte=
nsiveness of the spec.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">Do you have a strong preference to introd=
uce FRAME_DONE?<u></u><u></u></p></div><div><p class=3D"MsoNormal"><u></u>=
=C2=A0<u></u></p></div></div><div><div><div><p class=3D"MsoNormal" style=3D=
"margin-bottom:12pt">
<u></u>=C2=A0<u></u></p><div><p class=3D"MsoNormal">On Thu, Jan 9, 2014 at =
6:53 PM, Tobias Oberstein &lt;<a href=3D"mailto:tobias.oberstein@tavendo.de=
" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt; wrote:<u></u><u></u=
></p><blockquote style=3D"border-style:none none none solid;border-left-col=
or:rgb(204,204,204);border-left-width:1pt;padding:0cm 0cm 0cm 6pt;margin:5p=
t 0cm 5pt 4.8pt">
<div><div><p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:=
Calibri,sans-serif;color:rgb(31,73,125)">Hi Yutaka,</span><u></u><u></u></p=
><p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,s=
ans-serif;color:rgb(31,73,125)">=C2=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">1) Yes, I understand. Thanks!</span><u></u><=
u></u></p><p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:=
Calibri,sans-serif;color:rgb(31,73,125)">=C2=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">2) Ok, both WS frame and message boundaries =
are preserved.</span><u></u><u></u></p><p class=3D"MsoNormal"><span style=
=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,73,125)">=C2=
=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">But: I cannot recognize the end of a WS fram=
e until the next WS frame starts (for which I receive a HEADERS frame, whic=
h is the implicit signal that the previous WS frame ended), right?</span><u=
></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">=C2=A0</span><u></u><u></u></p><p class=3D"M=
soNormal"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;colo=
r:rgb(31,73,125)">And if so, why is there no FRAME_DONE bit on the SPDY fra=
me that corresponds to the end of a WS frame?</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">=C2=A0</span><u></u><u></u></p><p class=3D"M=
soNormal"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;colo=
r:rgb(31,73,125)">3) Ok, great! Hope this works, since SPDY provides header=
 compression, but no sophisticated bidirectional payload compression as &qu=
ot;WS permessage-deflate&quot; does.</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">=C2=A0</span><u></u><u></u></p><p class=3D"M=
soNormal"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;colo=
r:rgb(31,73,125)">Cheers,</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">/Tobias</span><u></u><u></u></p><p class=3D"=
MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;col=
or:rgb(31,73,125)">=C2=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">=C2=A0</span><u></u><u></u></p><p class=3D"M=
soNormal"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;colo=
r:rgb(31,73,125)">=C2=A0</span><u></u><u></u></p>
<div style=3D"border-style:none none none solid;border-left-color:blue;bord=
er-left-width:1.5pt;padding:0cm 0cm 0cm 4pt"><div><div style=3D"border-styl=
e:solid none none;border-top-color:rgb(225,225,225);border-top-width:1pt;pa=
dding:3pt 0cm 0cm">
<p class=3D"MsoNormal"><b><span style=3D"font-size:11pt;font-family:Calibri=
,sans-serif">Von:</span></b><span style=3D"font-size:11pt;font-family:Calib=
ri,sans-serif"> </span><a href=3D"mailto:yhirano@google.com" target=3D"_bla=
nk"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif">yhirano@g=
oogle.com</span></a><span style=3D"font-size:11pt;font-family:Calibri,sans-=
serif"> [mailto:</span><a href=3D"mailto:yhirano@google.com" target=3D"_bla=
nk"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif">yhirano@g=
oogle.com</span></a><span style=3D"font-size:11pt;font-family:Calibri,sans-=
serif">] <b>Im Auftrag von </b>Yutaka Hirano<br>
<b>Gesendet:</b> Donnerstag, 9. Januar 2014 10:37<br><b>An:</b> Tobias Ober=
stein<br><b>Cc:</b> Takeshi Yoshino; </span><a href=3D"mailto:hybi@ietf.org=
" target=3D"_blank"><span style=3D"font-size:11pt;font-family:Calibri,sans-=
serif">hybi@ietf.org</span></a><span style=3D"font-size:11pt;font-family:Ca=
libri,sans-serif"><br>
<b>Betreff:</b> Re: [hybi] Discontinuation of mux standardizaton in favor o=
f WS/HTTP/2.0</span><u></u><u></u></p></div></div><div><div><p class=3D"Mso=
Normal">=C2=A0<u></u><u></u></p><div><p class=3D"MsoNormal">Hi Tobias,<u></=
u><u></u></p>
<div><p class=3D"MsoNormal">Thank you for the comments.<u></u><u></u></p></=
div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p><blockquote style=
=3D"border-style:none none none solid;border-left-color:rgb(204,204,204);bo=
rder-left-width:1pt;padding:0cm 0cm 0cm 6pt;margin:5pt 0cm 5pt 4.8pt">
<p class=3D"MsoNormal"><span style=3D"font-size:10pt;font-family:Arial,sans=
-serif">1) Frame Mapping<br>&quot;&quot;&quot;<br>One WebSocket frame is ma=
pped into one preceding SPDY HEADERS frame and following plural SPDY data f=
rames. The SPDY HEADERS frame must contain WebSocket frame fields and follo=
wing SPDY data frames contain payload data. These data frames are free to b=
e reframed.<br>
&quot;&quot;&quot;<br></span><a href=3D"https://docs.google.com/document/d/=
124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank"><span =
style=3D"font-size:10pt;font-family:Arial,sans-serif">https://docs.google.c=
om/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit</span></a><=
span style=3D"font-size:10pt;font-family:Arial,sans-serif"><br>
So reframing of WebSocket frames (the payload) into multiple SPDY data fram=
es MAY occur, but does not need to occur? And a WS/SPDY implementation is f=
ree to choose the boundaries within WS frame payload at which to reframe in=
to different SPDY data frames as it likes?</span><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal">A WebSocket frame MAY be split to =
multiple SPDY DATA frames.<u></u><u></u></p></div><div><p class=3D"MsoNorma=
l">Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA fram=
e into multiple DATA frames.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">Is this answering your question?<u></u><u=
></u></p></div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><b=
lockquote style=3D"border-style:none none none solid;border-left-color:rgb(=
204,204,204);border-left-width:1pt;padding:0cm 0cm 0cm 6pt;margin:5pt 0cm 5=
pt 4.8pt">
<p class=3D"MsoNormal"><span style=3D"font-size:10pt;font-family:Arial,sans=
-serif">2) End of WS Frame<br>The end of a WS message is signaled with a SP=
DY data frame having MSG_DONE set.<br>How is the end of a WS frame signaled=
, given that the WS frame may have been reframed into multiple SPDY data fr=
ames?</span><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal">When you receive a HEADERS frame, =
you can recognize that the last SPDY frame is that last frame in a WebSocke=
t frame.<u></u><u></u></p></div><div><p class=3D"MsoNormal">When you receiv=
e a SPDY frame with MSG_DONE, you can recognize that it is the last SPDY fr=
ame in a WebSocket message and hence the last SPDY frame in a WebSocket fra=
me.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><blockquote=
 style=3D"border-style:none none none solid;border-left-color:rgb(204,204,2=
04);border-left-width:1pt;padding:0cm 0cm 0cm 6pt;margin:5pt 0cm 5pt 4.8pt"=
><p class=3D"MsoNormal">
<span style=3D"font-size:10pt;font-family:Arial,sans-serif">3) WS extension=
<br>Are WS extensions fully supported? In particular, is &quot;permessage-d=
eflate&quot; supported when running WS/SPDY?</span><u></u><u></u></p></bloc=
kquote>
<div><p class=3D"MsoNormal">Yes. We have reserved bits and we can emulate t=
he opening handshake, I think.<u></u><u></u></p></div><div><p class=3D"MsoN=
ormal">=C2=A0<u></u><u></u></p></div><div><p class=3D"MsoNormal">On Thu, Ja=
n 9, 2014 at 6:11 PM, Tobias Oberstein &lt;<a href=3D"mailto:tobias.oberste=
in@tavendo.de" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt; wrote:=
<u></u><u></u></p>
<blockquote style=3D"border-style:none none none solid;border-left-color:rg=
b(204,204,204);border-left-width:1pt;padding:0cm 0cm 0cm 6pt;margin:5pt 0cm=
 5pt 4.8pt"><p class=3D"MsoNormal">Hi Takeshi,<br><br>it&#39;s a pity, but =
I can understand the motivation to shove everything into HTTP2.<br>
<br>Anyway, some questions:<br><br>1) Frame Mapping<br>&quot;&quot;&quot;<b=
r>One WebSocket frame is mapped into one preceding SPDY HEADERS frame and f=
ollowing plural SPDY data frames. The SPDY HEADERS frame must contain WebSo=
cket frame fields and following SPDY data frames contain payload data. Thes=
e data frames are free to be reframed.<br>
&quot;&quot;&quot;<br><a href=3D"https://docs.google.com/document/d/124vB5N=
V1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank">https://docs.=
google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit</a>=
<br>
<br>So reframing of WebSocket frames (the payload) into multiple SPDY data =
frames MAY occur, but does not need to occur? And a WS/SPDY implementation =
is free to choose the boundaries within WS frame payload at which to refram=
e into different SPDY data frames as it likes?<br>
<br>2) End of WS Frame<br><br>The end of a WS message is signaled with a SP=
DY data frame having MSG_DONE set.<br><br>How is the end of a WS frame sign=
aled, given that the WS frame may have been reframed into multiple SPDY dat=
a frames?<br>
<br>3) WS extension<br><br>Are WS extensions fully supported? In particular=
, is &quot;permessage-deflate&quot; supported when running WS/SPDY?<br><br>=
Thanks!<br>/Tobias<br><br><br>Von: hybi [mailto:<a href=3D"mailto:hybi-boun=
ces@ietf.org" target=3D"_blank">hybi-bounces@ietf.org</a>] Im Auftrag von T=
akeshi Yoshino<br>
Gesendet: Dienstag, 7. Januar 2014 08:36<br>An: <a href=3D"mailto:hybi@ietf=
.org" target=3D"_blank">hybi@ietf.org</a><br>Betreff: [hybi] Discontinuatio=
n of mux standardizaton in favor of WS/HTTP/2.0<u></u><u></u></p><div><div>
<p class=3D"MsoNormal"><br>Hi all,<br><br>We&#39;ve been working on mux ext=
ension to make WebSocket scalable for long time, but as SPDY got widely sup=
ported and adopted quickly, now there seems to be no longer any good reason=
 to continue this work which is WebSocket dedicated.<br>
<br>Scalability is still important issue to address. From now, I&#39;d like=
 to focus on WS/SPDY effort which we&#39;ve been working on in parallel wit=
h the mux extension.<br><br>Please take a look at the latest proposal in th=
is post.<br>
<a href=3D"https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc" ta=
rget=3D"_blank">https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3gl=
c</a><br><br>If you really want the dedicated mux extension than one over S=
PDY, sorry but please take over the work.<u></u><u></u></p>
</div></div><div><div><p class=3D"MsoNormal">______________________________=
_________________<br>hybi mailing list<br><a href=3D"mailto:hybi@ietf.org" =
target=3D"_blank">hybi@ietf.org</a><br><a href=3D"https://www.ietf.org/mail=
man/listinfo/hybi" target=3D"_blank">https://www.ietf.org/mailman/listinfo/=
hybi</a><u></u><u></u></p>
</div></div></blockquote></div><p class=3D"MsoNormal">=C2=A0<u></u><u></u><=
/p></div></div></div></div></div></div></div></blockquote></div><p class=3D=
"MsoNormal"><u></u>=C2=A0<u></u></p></div></div></div></blockquote></div><p=
 class=3D"MsoNormal">
<u></u>=C2=A0<u></u></p></div></div></div></div></div></div></blockquote></=
div><br></div></div>

--047d7b33cba0c31fd904eff3bd41--


From ricea@google.com  Tue Jan 14 20:42:10 2014
Return-Path: <ricea@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id A82971AE2B8 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 20:42:10 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id cWPeA0zxOm0Z for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 20:42:09 -0800 (PST)
Received: from mail-ob0-x234.google.com (mail-ob0-x234.google.com [IPv6:2607:f8b0:4003:c01::234]) by ietfa.amsl.com (Postfix) with ESMTP id 38C9E1AE2B3 for <hybi@ietf.org>; Tue, 14 Jan 2014 20:42:09 -0800 (PST)
Received: by mail-ob0-f180.google.com with SMTP id wm4so626870obc.39 for <hybi@ietf.org>; Tue, 14 Jan 2014 20:41:57 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type; bh=Hppy2RHxPZNu7Go5t9kKDWKxuSpqQrnFKUIIhjHqDxI=; b=HuWYbvdwvB+2CT3B5AADu6BHQJB9VWr+BnQS1IPxofyjWwd5y5VPByg3HLqd32m6YJ 9TSiKUDeLacO4BY/5lPm04BXqL1nFoW/jKHhOlrzwOGoJnIVD3ekW0v8xJB32O4xqwhK ZQScfOcGJowwhDaVelPXTYkFhVAJgUPoRN4Dt/t3Jacmd5U75CT177BbHkImbPjU09Za aGb6lXFQJa1KhxmAd1sosX8nonQu4p+rQjR1zSStMCzWhOUrrNhX7dh+nVT14neDh1CO y9Oxwbq4NgnkNdYt47wIn4l8ITMQBpoJ8xJHRHmuZzN3Lc+GlTtWIBzHUwMr4rtpUjAQ R5Nw==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=Hppy2RHxPZNu7Go5t9kKDWKxuSpqQrnFKUIIhjHqDxI=; b=k2v0xHF0BJ//lUAX2ygwceXPVgZKrqbFkN9Isaxe/L4J5QPLSkG9SfelsyTG0nyq7Q /2Fs+Svp5/4DYX2Gto1CU/iZGrgp+yLy/O29p1gYSFY9wJH60LRn5Erb60nayXcdlGcx B/k7rtmVkEZ0MloQ9py3xApkY2qY/oWujrKJemfxd7X1Mm3colAy3n5xyCy5wKVE2ocu oZnAFy8LyCiAj+TuZMf2jTbx2nfiySRMORTb49I/JEtV5fZZi+oZSYB8bMeMFoIq6G/r qKyZZcgYvAtqyJGl3yu5DL15Duiiwehcg4vtkMpBAYwaYaD17VotPs7THa+hoyv9wfZX KUSA==
X-Gm-Message-State: ALoCoQm//yeUfNTJTtnc1ddWO9w48Kbv0CYk0DIR4X+AM1nUGtr4cFxHLlD5Vwo787xUe3mqT9z2Zt6ad1HPEChU86tN0tp6iaUHR5FI5kksleeYINKcYakeSvhMbJx2vN6U2+fObIygJL2AuvSWVQh1mF399QIPlFo+E2PaXFtWDxO5S7D5noUhDliR6XArmpypW63j43rZ
MIME-Version: 1.0
X-Received: by 10.60.165.36 with SMTP id yv4mr4226329oeb.55.1389760917354; Tue, 14 Jan 2014 20:41:57 -0800 (PST)
Received: by 10.182.118.41 with HTTP; Tue, 14 Jan 2014 20:41:57 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net>
Date: Wed, 15 Jan 2014 13:41:57 +0900
Message-ID: <CAHixhFpVywsZPCx1r_HYXEUQ10WN31NnVYM5XT6-n78uAcrjig@mail.gmail.com>
From: Adam Rice <ricea@google.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=047d7b3a8674cd528604effaedd8
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 15 Jan 2014 04:42:10 -0000

--047d7b3a8674cd528604effaedd8
Content-Type: text/plain; charset=UTF-8

On 15 January 2014 04:09, Tobias Oberstein <tobias.oberstein@tavendo.de>wrote:

>
>  * has significant wire-level overhead compared to native WS (and compared
> to the proposal below)
>
>
If you accept the argument that unencrypted WebSockets are unusable on
heterogeneous networks due to interference from middleboxes, then the byte
overhead is a red-herring. Everything has to be encrypted, and so will get
padded out to the cipher block size anyway.

What is significant is the CPU overhead, which means battery life on mobile
devices and latency everywhere. The WebSocket frame header is trivial to
construct. SPDY HEADERS, not so much.

Adam Rice

--047d7b3a8674cd528604effaedd8
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On 1=
5 January 2014 04:09, Tobias Oberstein <span dir=3D"ltr">&lt;<a href=3D"mai=
lto:tobias.oberstein@tavendo.de" target=3D"_blank" class=3D"cremed">tobias.=
oberstein@tavendo.de</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div lang=3D"DE" link=3D"blue" vlink=3D"purp=
le"><p class=3D"MsoNormal"><br></p><p class=3D"MsoNormal"><span style=3D"fo=
nt-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color=
:#1f497d"> =C2=A0* has significant wire-level overhead compared to native W=
S (and compared to the proposal below)<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u></span></p></div><=
/blockquote><div><br></div><div>If you accept the argument that unencrypted=
 WebSockets are unusable on heterogeneous networks due to interference from=
 middleboxes, then the byte overhead is a red-herring. Everything has to be=
 encrypted, and so will get padded out to the cipher block size anyway.</di=
v>
<div><br></div><div>What is significant is the CPU overhead, which means ba=
ttery life on mobile devices and latency everywhere. The WebSocket frame he=
ader is trivial to construct. SPDY HEADERS, not so much.</div><div><br>
</div><div>Adam Rice</div><div><br></div></div></div></div>

--047d7b3a8674cd528604effaedd8--

From tyoshino@google.com  Tue Jan 14 21:43:12 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 7B9511AE2C6 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 21:43:12 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id EQ_bNJkWOnPG for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 21:43:11 -0800 (PST)
Received: from mail-we0-x22c.google.com (mail-we0-x22c.google.com [IPv6:2a00:1450:400c:c03::22c]) by ietfa.amsl.com (Postfix) with ESMTP id F18DA1AE1BF for <hybi@ietf.org>; Tue, 14 Jan 2014 21:43:10 -0800 (PST)
Received: by mail-we0-f172.google.com with SMTP id q58so1331626wes.31 for <hybi@ietf.org>; Tue, 14 Jan 2014 21:42:58 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=qQCprfB0CUvu88+3F624swTjHyLvzi2p6LAtjOu2ZVk=; b=fq0S8gU8R/FQvgVXmyeAkkgaLIZAVqFnnmUbiJ7z8tfJpoP7hpC55yfxhvgvHpkwLP JcpPaKbmQ2FbInhJpsqANk9ztxGvTnMNzNKhOZvgNXAj1vIdJhyKg0/+3VpncO9b/jpe e2Alj5WdkmhSKzULCiCRfwii1ITfr3xzYO+HLRGO95l10G4ELpIYs3ih2yEcDuG3amNm 79fLmGPSovhBg9crs2+qHsP0trMwD24EpT4qGg8359PnfHL8uQiAcmAmDAFLOGfbHiil rQ70FYkkMRysbW6MnVAR8svNfdhshQaxJiTWEo8Mq2l7UhH7dXb7U6MHVsU/um2sFUx0 rZtw==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=qQCprfB0CUvu88+3F624swTjHyLvzi2p6LAtjOu2ZVk=; b=dyN4knE4oXoBRcOGVE2JKsnWmQgNUNgZW+tK4wPJydmdIdj0q+63mckQRBUvRITrd0 OR1Odjxw7H9Fhi8KKR2b1lpBZKM8+VfUyviw2YlQXIeXkYNs5EiwbeYpPVQMIcd7/ZiL PSxr/dz4nKTyElJldAlP8vD6wIA8SUDQ5l5ThGXIWG2OKSgY/2CzG1Qu1rsLw+G7f1Na vMhj6Y6RtWlhTixjvRvDfguo2rUvul/IULzXY4xm941jxTs87WfhTajdLBB2JflagS51 9fmTJSr+/31Iu4bTwGDl7SHutfSQpzoo+mQhG/91mmc+VoYIPGEQPh5ze1zRVWtbPFF2 0lTg==
X-Gm-Message-State: ALoCoQlgLn5fI3YKfDCbMQEZXS6o9hoFWElMnRO98PSBOBBGAX8aWUmxBTd2Cq4gimQx67ghOSd2YALBY45EIP7lrJNs2chKPeLnhXL9WVNHxzWZv5G9Yn5CgahHeOB/VwCXUAT5PzHe/jap4rma1Lmeug8LGmCcsz8C/GJjJPm2AzwursVE1HeLrl7Wipoq9rZx+/AfFB8F
X-Received: by 10.194.173.163 with SMTP id bl3mr287086wjc.10.1389764578703; Tue, 14 Jan 2014 21:42:58 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Tue, 14 Jan 2014 21:42:37 -0800 (PST)
In-Reply-To: <CAGzyod7HAjdDHt1B4XRHT6h43DXnoupqGmr4buP6D+bQs86m+w@mail.gmail.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CAGzyod7HAjdDHt1B4XRHT6h43DXnoupqGmr4buP6D+bQs86m+w@mail.gmail.com>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Wed, 15 Jan 2014 14:42:37 +0900
Message-ID: <CAH9hSJY=T3mALqsTJWXW5=oq0=knzpWUwqwpUatZ==S2fDPSKg@mail.gmail.com>
To: Roberto Peon <fenix@google.com>
Content-Type: multipart/alternative; boundary=089e013c622009203004effbc825
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 15 Jan 2014 05:43:12 -0000

--089e013c622009203004effbc825
Content-Type: text/plain; charset=UTF-8

On Wed, Jan 15, 2014 at 5:07 AM, Roberto Peon <fenix@google.com> wrote:

>
>>
>> 2) WebSocket message and frame boundaries are indeed preserved in the
>> current WS/SPDY proposal,
>>
>> however, non-final WebSocket frame ends cannot be _immediately_ detected.
>>
>
>  I'm not sure what you mean here?
>

In current proposal, we mark only the end of each message (or the end of
control frame) using MSG_DONE. So, until an endpoint sees the next HEADERS
frame, it cannot determine the end of the frame is reached or not. I think
that's what Tobias meant.

--089e013c622009203004effbc825
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On W=
ed, Jan 15, 2014 at 5:07 AM, Roberto Peon <span dir=3D"ltr">&lt;<a href=3D"=
mailto:fenix@google.com" target=3D"_blank">fenix@google.com</a>&gt;</span> =
wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div dir=3D"ltr"><div class=3D"gmail_extra">=
<div class=3D"gmail_quote"><div class=3D"im"><blockquote class=3D"gmail_quo=
te" style=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-col=
or:rgb(204,204,204);border-left-style:solid;padding-left:1ex">

<div lang=3D"DE" link=3D"blue" vlink=3D"purple"><div><p class=3D"MsoNormal"=
><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,=
73,125)">=C2=A0<u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">2) WebSocket message and frame boundaries ar=
e indeed preserved in the current WS/SPDY proposal,<u></u><u></u></span></p=
>


<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">however, non-final WebSocket frame ends cann=
ot be _immediately_ detected.</span></p></div></div></blockquote><div><br>

</div>
</div><div>I&#39;m not sure what you mean here?</div></div></div></div></bl=
ockquote><div><br></div><div>In current proposal, we mark only the end of e=
ach message (or the end of control frame) using MSG_DONE. So, until an endp=
oint sees the next HEADERS frame, it cannot determine the end of the frame =
is reached or not. I think that&#39;s what Tobias meant.</div>

</div></div></div>

--089e013c622009203004effbc825--

From yhirano@google.com  Tue Jan 14 21:52:01 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 16EE71ADEA0 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 21:52:01 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.716
X-Spam-Level: 
X-Spam-Status: No, score=-0.716 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, J_CHICKENPOX_54=0.6, J_CHICKENPOX_57=0.6, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id zgvGLIgy3tYZ for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 21:51:56 -0800 (PST)
Received: from mail-qe0-x22f.google.com (mail-qe0-x22f.google.com [IPv6:2607:f8b0:400d:c02::22f]) by ietfa.amsl.com (Postfix) with ESMTP id EC03D1ADE7C for <hybi@ietf.org>; Tue, 14 Jan 2014 21:51:55 -0800 (PST)
Received: by mail-qe0-f47.google.com with SMTP id 5so637841qeb.20 for <hybi@ietf.org>; Tue, 14 Jan 2014 21:51:44 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=V7KYneyLhG2imU94Lhyn+SK5tKQeNvhiuKsaxjfmAig=; b=W78ptlyVbi6zbGOtqTMloxcsaO+OviyHVs1XEKCUUuXVn4K4HYWiTA74VSpvOJ4CST CDz2xTbDprSA3cnOUnFD2PsnFlkd/vT59pNCgCl13drAF3LlhpMMbBs0WZv7xehC/KoJ zeJ4zfJXXPXnx3yAjwIdzMs1WO/Bnsl/6V3BOrzPvr8vcN1R69P4ivzORb+kXtypOmZj 97ZwEHCXHFYm4ncZ9gU7G6qvPgeZW+mG05oLYvJdV9JwJEPF8C2EVRaemaKeIs46wPS/ zpsyYIZ5blR60Z0SiK3Qrf5AyXJ/EsbKOojglb8qU4Pq9tguLHtKan1RHDZ5YkSVJxC/ +mgA==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=V7KYneyLhG2imU94Lhyn+SK5tKQeNvhiuKsaxjfmAig=; b=HrenrQN5C5kqPb+f6r0qSG51WCZYvNhBAu5ZROBxccey5jueyre/r2BNU6CNbO4hn+ A7MFpRLezdXaiURDDqn/5acZzufv4KOmpX6njWXUrXUcQ1zvxeJqfagR0EJ6DwH1sLfc wg6mM/2N/iNVlAlm7zK+dFDIsTn19n8dt22l0=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=V7KYneyLhG2imU94Lhyn+SK5tKQeNvhiuKsaxjfmAig=; b=S3qBPdOotrV/N4XsgVxS8jdkRguefCgwfiRcyVtGb2qbkp08Zmkiafq1RfGu9ngQsn Ku4xkAzHjUpVQI2PzJiJrj5QCvHGesePGsYY5tAsU9BKyUPkKEdVaziwN3Ai9Szzawpq 8qhUEINwTbTWvXxNfhsSicDtez6wtHNowxWQ6+vSwTUH7DMWYYhEWMPuYmryTZsHLXnu qZFmns0MBc2FEHkEDTl0c+diBLg7jPnPqRkSMhE8aUN3M6DLxY1BARj5IWXveDTlj2xk J20Co+8ON+w6Lv9LlG8L4yvGKokLd5n+fZIi/Gm70mxjZ1+IcpaCr3HWJb7Ix+tIbKCP ZQAA==
X-Gm-Message-State: ALoCoQmHrpVFrT0VCe4XRwHQqQiIm0RYbmh8kz4wSey9TCmB3FB5BuB8oijNLC3B83kjE/pYZ7KZUvlRTdOmKwZquAsyP9EHckJidYfdQt4LU4lbYYCMnp77zqw+tO8pFUoX3LKHLytJmG6w5Ti1mnqEx7ozRAW1MTnd7MqyYFHA+nEx6Nk7APCjRUgzyIzYYYHUF1+wFQRG
MIME-Version: 1.0
X-Received: by 10.224.132.133 with SMTP id b5mr484328qat.61.1389765104070; Tue, 14 Jan 2014 21:51:44 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.140.109.202 with HTTP; Tue, 14 Jan 2014 21:51:43 -0800 (PST)
In-Reply-To: <CAGzyod7HAjdDHt1B4XRHT6h43DXnoupqGmr4buP6D+bQs86m+w@mail.gmail.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CAGzyod7HAjdDHt1B4XRHT6h43DXnoupqGmr4buP6D+bQs86m+w@mail.gmail.com>
Date: Wed, 15 Jan 2014 14:51:43 +0900
X-Google-Sender-Auth: MeOfpj7BjA5Qwl8Wy_tXEzo2ADM
Message-ID: <CABihn6EFmB+O9Lubuse5K0geM6FnDCkqTe3APz+UQhA4UErWZg@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: Roberto Peon <fenix@google.com>
Content-Type: multipart/alternative; boundary=001a11c2ef125986ab04effbe70f
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 15 Jan 2014 05:52:01 -0000

--001a11c2ef125986ab04effbe70f
Content-Type: text/plain; charset=UTF-8

Roberto, let me confirm:

> In terms of total amount of code:
> In any layering scheme, we'd need to use a HEADER frame one way or another
> because we need to at least announce that we're connecting to a WS endpoint.
> In terms of decode, we're essentially talking the following state machine:
> if frame.type == HEADERS:
>     streams[frame.stream_id] = Stream(frame.headers)
> elif frame.type == DATA:
>     payload[frame.stream_id] += frame.payload
>     if frame.MSG_DONE or frame.STREAM_DONE:
>       passMessageToApplication(payload[frame.stream_id])
>       del payload[frame.stream_id]
> if frame.STREAM_DONE:
>   del streams[frame.stream_id]


In the proposal in the doc, there are two kind of HEADERS, for handshake
and for WS frame headers.
Then I think the pseudocode will be like this:

# TODO: message injection is not addressed in this code.
if frame.type == HEADERS:
  if frame.stream_id not in streams:
    # stream initialization
    streams[frame.stream_id] = Stream(frame.headers)
  else:
    # WS frame header
    stream = streams[frame.stream_id]
    if stream.current_processing_message is None:
       stream.current_processing_message = Message()
    stream.current_processing_frame = Frame(frame.headers)
elif frame.type == DATA:
  stream = streams[frame.stream_id]
  # How to add payload depends on the WS frame header
  stream.current_processing_message.addPayload(frame.payload,
stream.current_processing_frame)
if frame.MSG_DONE:
  stream = streams[frame.stream_id]
  passMessageToApplication(stream.current_processing_message)
  stream.current_processing_message = None
if frame.STREAM_DONE:
  del stream[frame.stream_id]

Am I missing something?




On Wed, Jan 15, 2014 at 5:07 AM, Roberto Peon <fenix@google.com> wrote:

>
>
>
> On Tue, Jan 14, 2014 at 11:09 AM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
>
>> Hi Roberto,
>>
>>
>>
>> not sure if you agree, but FWIW, here is a summary as I see it
>> (currently):
>>
>>
>>
>> 1) WebSocket frame boundaries _can_ have semantic meaning in WebSocket
>> _extensions_
>>
>
> Agreed. And there are none deplyed that we know about where this is a
> concern.
>
>
>>
>>
>> 2) WebSocket message and frame boundaries are indeed preserved in the
>> current WS/SPDY proposal,
>>
>> however, non-final WebSocket frame ends cannot be _immediately_ detected.
>>
>
> I'm not sure what you mean here?
>
>
>>
>>
>> 3) The current WS/SPDY proposal
>>
>>   * is comfortable to implement for implementations that _only_ want to
>> implement WS/SPDY, and not native WS at the same time
>>
>>  * makes things complicated for WS implementations that want to support
>> both WS/SPDY _and_ native WS at the same time
>>
>>  * has significant wire-level overhead compared to native WS (and
>> compared to the proposal below)
>>
>>
>>
>
> The complication here is relative. Using the approach proposed in terms of
> using MSG_DONE, etc. would make deployment at large sites like Google
> substantially easier as large sites typically want to see the semantic
> boundaries easily, quickly, and consistently, as there needs to be robust
> mechanisms for handling failover, etc. The more layers of interpretation
> there are when doing this, the bigger the number of problems.
>
> I'm honestly not worried about the potential increased overhead-- arguably
> HTTP2 reduces overhead substantially simply be requiring fewer handshakes
> (the connection setup is a non-trivial amount of bytes and latency. That
> this was ignored was hugely frustrating to me during the initial
> standardization process here in HyBi). Even discounting this, I haven't
> actually heard of use-cases where a well-written application sends so many
> messages that the bandwidth is saturated. I am for efficiency (had to work
> on compression for HTTP2, for instance), but not when it increases
> complexity in the more general case.
>
> In terms of total amount of code:
> In any layering scheme, we'd need to use a HEADER frame one way or another
> because we need to at least announce that we're connecting to a WS endpoint.
> In terms of decode, we're essentially talking the following state machine:
>
> if frame.type == HEADERS:
>     streams[frame.stream_id] = Stream(frame.headers)
> elif frame.type == DATA:
>     payload[frame.stream_id] += frame.payload
>     if frame.MSG_DONE or frame.STREAM_DONE:
>       passMessageToApplication(payload[frame.stream_id])
>       del payload[frame.stream_id]
> if frame.STREAM_DONE:
>   del streams[frame.stream_id]
>
> While the above is pseudocode and leaves out error handling, one of the
> interesting things is that it is the same for both HTTP, WS, or any other
> protocol layering over HTTP2.
>
>
>
>
>> I'd like to suggest a different WS/SPDY:
>>
>>
>>
>> Pack WebSocket frames _including_ full WS frame header as _payload_ of
>> SPDY data frames only - no SPDY header frames at all.
>>
>>
>>
>> In the absence of negotiated WebSocket extensions (or when all WS
>> extensions in place do not assign semantics to WS frame boundaries),
>>
>> fragment WebSocket payload at the WS level such that WS frame
>> header+payload fits into the payload of SPDY data frames.
>>
>>
>>
>> In the latter case, this essentially adds a small SPDY data frame header
>> (of fixed length) to WS frames. Easy to implement and efficient on wire.
>>
>>
>>
>> IMHO, this has advantages compared to current proposal and considering 3)
>> above.
>>
>>
>>
>> The only disadvantage is for implementations that _only_ want to
>> implement WS/SPDY and not native WS at the same time.
>>
>
> One of the nice things about the current plan as compared to your proposal
> is that it allows for generic infrastructure to be built to examine and
> handle things for DoS and loadbalancing concerns across any protocol using
> HTTP2 as a session layer since all boundaries are clearly demarcated. It
> seems like a very nice thing indeed that HTTP, WS, various internal API
> things, etc. can all end up using the same infrastructure.
>
> Additionally, what happens if/when someone defines a mapping for HTTP2
> onto something like QUIC, SCTP, TCP-Minion, or perhaps the RTCWeb stack?
> With the proposed design where the HTTP2 layer has full visibility into the
> semantic boundaries, this mapping needs to be done only once for the HTTP2
> session layer and all protocols layered atop it are now able to take
> advantage of it (one would need signaling about HOL blocking properties of
> streams to the HTTP2 layer in addition, of course). When we hide the
> semantic boundaries, this requires work from every implementor, and impedes
> adoption either of the new HTTP2 mapping onto the better transport, or
> impedes adoption of WS.
>
> I want to see the WS API succeed at scale and continue to do so into the
> future, so, my thought is that the advantages in regularity/consistency
> (today for DoS and loadbalancing, tomorrow on new transports) outweigh the
> relatively small theoretical byte efficiency tradeoff.
> As I mentioned earlier, this is the tradeoff that HTTP2 has made for HTTP
> at the application layer as well, so even the approach is consistent :)
>
>
>>
>> ==
>>
>>
>>
>> Regarding "permessage-priority" with ordering guarantees: should I find
>> time and if there is interest, I'll write something up.
>>
>> This could be a WebSocket extension. To be useful, the WS API in browsers
>> would need to expose it though ..
>>
>
> Yup. The same thing is happening for HTTP (people finding that expressing
> priority for resources is useful), btw, from what I'm led to understand.
> I'm hopefuly for the day when APIs allow us to easily signal both
> priority, and also other important properties like drop-dead-times (get
> this message to the other side at most in X seconds from now, else leave it
> on the floor), and ordering requirements (e.g. don't care about ordering,
> care about ordering, but only to know when something has been dropped,
> etc.).
> The interesting thing about priorities is that it is really only useful
> when you're muxing over one connection/session.
> -=R
>
>
>>
>> Just my 2cts,
>>
>> Tobias
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> *Von:* Roberto Peon [mailto:fenix@google.com]
>> *Gesendet:* Dienstag, 14. Januar 2014 19:30
>> *An:* Yutaka Hirano
>> *Cc:* Tobias Oberstein; Takeshi Yoshino; hybi@ietf.org
>>
>> *Betreff:* Re: [hybi] Discontinuation of mux standardizaton in favor of
>> WS/HTTP/2.0
>>
>>
>>
>> I've been lurking anyway :)
>>
>>
>>
>> As I understand the proposal, (and it seems that Takeshi and Yukata are
>> covering this nicely, so what i say is probably redundant):
>>
>> Message boundaries, having semantic meaning, are preserved.
>>
>>
>>
>> Frame boundaries, having no semantic meaning, are not preserved. I don't
>> think it makes sense to worry about this greatly when the ability to extend
>> the protocol as implemented in HTTP2 is present. It doesn't make a lot of
>> sense to signal the frame boundaries through everything unless there is a
>> high-likelyhood and high-utility use-case for such. If such is found, then
>> HTTP2 can be revised to include it. This is the approach we've been taking
>> with most of the HTTP2 stuff, since there has been great effort placed on
>> allowing new versions to be deployed in the future.
>>
>>
>>
>> The mapping of WS onto HTTP2 is fairly direct-- we reserved the MSG_DONE
>> bit specifically for this kind of use-case (i.e. bidirectional,
>> full-duplex, message-based).
>>
>> The observation made is that there are essentially 3 types of data that
>> need to be preserved one way or another:
>>
>> 1) metadata: data about the data, e.g. text, binary, where it is supposed
>> to be going, etc.
>>
>> 2) payload
>>
>> 3) boundaries between these.
>>
>>
>>
>> imho, metadata is handled with HEADERS (I'd have preferred it to be
>> called METADATA, but was voted down at the time :) ), payload is handled
>> with DATA frames, and the boundaries are delimited either through the use
>> of HEADERS frames, or via the MSG_DONE or STREAM_DONE bits.
>>
>> In the presence of extensions, assuming that MSG_DONE is required to
>> signal the end of a message, then HEADERS can be used to signal any
>> boundary desired.
>>
>>
>>
>>
>>
>> I totally agree that it would be nice to expose the priority of various
>> messages/channels with respect to each other. That is really an issue that
>> is outside the WG's purvue, though.
>>
>>
>>
>>
>>
>> On Thu, Jan 9, 2014 at 7:54 PM, Yutaka Hirano <yhirano@chromium.org>
>> wrote:
>>
>> +Roberto as an HTTP2 expert.
>>
>>
>>
>> 2) Ok, both WS frame and message boundaries are preserved.
>>
>> But: I cannot recognize the end of a WS frame until the next WS frame
>> starts (for which I receive a HEADERS frame, which is the implicit signal
>> that the previous WS frame ended), right?
>>
>> You are right, you can't.
>>
>>
>>
>>
>> And if so, why is there no FRAME_DONE bit on the SPDY frame that
>> corresponds to the end of a WS frame?
>>
>> Given that WebSocket frames "have no semantic meaning" as in RFC6455, I
>> don't think the immediate recognition of the end of a WebSocket frame is
>> requisite.
>>
>> I would like to introduce MSG_DONE in HTTP2 so that intermediaries that
>> don't fully understand WS over HTTP2 can relay it.
>>
>> Since there are not so much space in FLAGS field, Introducing FRAME_DONE
>> may decrease the extensiveness of the spec.
>>
>> Do you have a strong preference to introduce FRAME_DONE?
>>
>>
>>
>>
>>
>> On Thu, Jan 9, 2014 at 6:53 PM, Tobias Oberstein <
>> tobias.oberstein@tavendo.de> wrote:
>>
>> Hi Yutaka,
>>
>>
>>
>> 1) Yes, I understand. Thanks!
>>
>>
>>
>> 2) Ok, both WS frame and message boundaries are preserved.
>>
>>
>>
>> But: I cannot recognize the end of a WS frame until the next WS frame
>> starts (for which I receive a HEADERS frame, which is the implicit signal
>> that the previous WS frame ended), right?
>>
>>
>>
>> And if so, why is there no FRAME_DONE bit on the SPDY frame that
>> corresponds to the end of a WS frame?
>>
>>
>>
>> 3) Ok, great! Hope this works, since SPDY provides header compression,
>> but no sophisticated bidirectional payload compression as "WS
>> permessage-deflate" does.
>>
>>
>>
>> Cheers,
>>
>> /Tobias
>>
>>
>>
>>
>>
>>
>>
>> *Von:* yhirano@google.com [mailto:yhirano@google.com] *Im Auftrag von *Yutaka
>> Hirano
>> *Gesendet:* Donnerstag, 9. Januar 2014 10:37
>> *An:* Tobias Oberstein
>> *Cc:* Takeshi Yoshino; hybi@ietf.org
>> *Betreff:* Re: [hybi] Discontinuation of mux standardizaton in favor of
>> WS/HTTP/2.0
>>
>>
>>
>> Hi Tobias,
>>
>> Thank you for the comments.
>>
>>
>>
>> 1) Frame Mapping
>> """
>> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
>> following plural SPDY data frames. The SPDY HEADERS frame must contain
>> WebSocket frame fields and following SPDY data frames contain payload data.
>> These data frames are free to be reframed.
>> """
>>
>> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
>> So reframing of WebSocket frames (the payload) into multiple SPDY data
>> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
>> is free to choose the boundaries within WS frame payload at which to
>> reframe into different SPDY data frames as it likes?
>>
>> A WebSocket frame MAY be split to multiple SPDY DATA frames.
>>
>> Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA frame
>> into multiple DATA frames.
>>
>> Is this answering your question?
>>
>>
>>
>> 2) End of WS Frame
>> The end of a WS message is signaled with a SPDY data frame having
>> MSG_DONE set.
>> How is the end of a WS frame signaled, given that the WS frame may have
>> been reframed into multiple SPDY data frames?
>>
>> When you receive a HEADERS frame, you can recognize that the last SPDY
>> frame is that last frame in a WebSocket frame.
>>
>> When you receive a SPDY frame with MSG_DONE, you can recognize that it is
>> the last SPDY frame in a WebSocket message and hence the last SPDY frame in
>> a WebSocket frame.
>>
>>
>>
>> 3) WS extension
>> Are WS extensions fully supported? In particular, is "permessage-deflate"
>> supported when running WS/SPDY?
>>
>> Yes. We have reserved bits and we can emulate the opening handshake, I
>> think.
>>
>>
>>
>> On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein <
>> tobias.oberstein@tavendo.de> wrote:
>>
>> Hi Takeshi,
>>
>> it's a pity, but I can understand the motivation to shove everything into
>> HTTP2.
>>
>> Anyway, some questions:
>>
>> 1) Frame Mapping
>> """
>> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
>> following plural SPDY data frames. The SPDY HEADERS frame must contain
>> WebSocket frame fields and following SPDY data frames contain payload data.
>> These data frames are free to be reframed.
>> """
>>
>> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
>>
>> So reframing of WebSocket frames (the payload) into multiple SPDY data
>> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
>> is free to choose the boundaries within WS frame payload at which to
>> reframe into different SPDY data frames as it likes?
>>
>> 2) End of WS Frame
>>
>> The end of a WS message is signaled with a SPDY data frame having
>> MSG_DONE set.
>>
>> How is the end of a WS frame signaled, given that the WS frame may have
>> been reframed into multiple SPDY data frames?
>>
>> 3) WS extension
>>
>> Are WS extensions fully supported? In particular, is "permessage-deflate"
>> supported when running WS/SPDY?
>>
>> Thanks!
>> /Tobias
>>
>>
>> Von: hybi [mailto:hybi-bounces@ietf.org] Im Auftrag von Takeshi Yoshino
>> Gesendet: Dienstag, 7. Januar 2014 08:36
>> An: hybi@ietf.org
>> Betreff: [hybi] Discontinuation of mux standardizaton in favor of
>> WS/HTTP/2.0
>>
>>
>> Hi all,
>>
>> We've been working on mux extension to make WebSocket scalable for long
>> time, but as SPDY got widely supported and adopted quickly, now there seems
>> to be no longer any good reason to continue this work which is WebSocket
>> dedicated.
>>
>> Scalability is still important issue to address. From now, I'd like to
>> focus on WS/SPDY effort which we've been working on in parallel with the
>> mux extension.
>>
>> Please take a look at the latest proposal in this post.
>> https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc
>>
>> If you really want the dedicated mux extension than one over SPDY, sorry
>> but please take over the work.
>>
>> _______________________________________________
>> hybi mailing list
>> hybi@ietf.org
>> https://www.ietf.org/mailman/listinfo/hybi
>>
>>
>>
>>
>>
>>
>>
>
>

--001a11c2ef125986ab04effbe70f
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div>Roberto, let me confirm:=C2=A0</div><blockquote class=
=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;bo=
rder-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">=
In terms of total amount of code:<br>
In any layering scheme, we&#39;d need to use a HEADER frame one way or anot=
her because we need to at least announce that we&#39;re connecting to a WS =
endpoint.<br>In terms of decode, we&#39;re essentially talking the followin=
g state machine:<br>
if frame.type =3D=3D HEADERS:<br>=C2=A0 =C2=A0 streams[frame.stream_id] =3D=
 Stream(frame.headers)<br>elif frame.type =3D=3D DATA:<br>=C2=A0 =C2=A0 pay=
load[frame.stream_id] +=3D frame.payload<br>=C2=A0 =C2=A0 if frame.MSG_DONE=
 or frame.STREAM_DONE:<br>=C2=A0 =C2=A0 =C2=A0 passMessageToApplication(pay=
load[frame.stream_id])<br>
=C2=A0 =C2=A0 =C2=A0 del payload[frame.stream_id]<br>if frame.STREAM_DONE:<=
br>=C2=A0 del streams[frame.stream_id]</blockquote><div><br></div><div>In t=
he proposal in the doc, there are two kind of HEADERS, for handshake and fo=
r WS frame headers.</div>
<div>Then I think the=C2=A0<span style=3D"font-family:arial,sans-serif;font=
-size:13px">pseudocode</span>=C2=A0will be like this:</div><div><br></div><=
div># TODO: message injection is not addressed in this code.</div><div>if f=
rame.type =3D=3D HEADERS:</div>
<div>=C2=A0 if frame.stream_id not in streams:</div><div>=C2=A0 =C2=A0 # st=
ream initialization</div><div>=C2=A0 =C2=A0 streams[frame.stream_id] =3D St=
ream(frame.headers)</div><div>=C2=A0 else:</div><div>=C2=A0 =C2=A0 # WS fra=
me header</div><div>=C2=A0 =C2=A0 stream =3D streams[frame.stream_id]<br>
</div><div>=C2=A0 =C2=A0 if stream.current_processing_message is None:<br><=
/div><div>=C2=A0 =C2=A0 =C2=A0 =C2=A0stream.current_processing_message =3D =
Message()</div><div>=C2=A0 =C2=A0 stream.current_processing_frame =3D Frame=
(frame.headers)</div><div>elif frame.type =3D=3D DATA:</div>
<div>=C2=A0 stream =3D streams[frame.stream_id]</div><div>=C2=A0 # How to a=
dd payload depends on the WS frame header</div><div>=C2=A0 stream.current_p=
rocessing_message.addPayload(frame.payload, stream.current_processing_frame=
)</div><div>if frame.MSG_DONE:</div>
<div>=C2=A0 stream =3D streams[frame.stream_id]<br></div><div>=C2=A0 passMe=
ssageToApplication(stream.current_processing_message)</div><div>=C2=A0 stre=
am.current_processing_message =3D None</div><div>if frame.STREAM_DONE:</div=
><div>=C2=A0 del stream[frame.stream_id]</div>
<div><br></div><div>Am I missing something?</div><div><br></div><div>=C2=A0=
</div></div><div class=3D"gmail_extra"><br><br><div class=3D"gmail_quote">O=
n Wed, Jan 15, 2014 at 5:07 AM, Roberto Peon <span dir=3D"ltr">&lt;<a href=
=3D"mailto:fenix@google.com" target=3D"_blank">fenix@google.com</a>&gt;</sp=
an> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div dir=3D"ltr"><br><div class=3D"gmail_ext=
ra"><br><br><div class=3D"gmail_quote"><div class=3D"im">On Tue, Jan 14, 20=
14 at 11:09 AM, Tobias Oberstein <span dir=3D"ltr">&lt;<a href=3D"mailto:to=
bias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstein@tavendo.de</a=
>&gt;</span> wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-=
left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;p=
adding-left:1ex"><div lang=3D"DE" link=3D"blue" vlink=3D"purple"><div><p cl=
ass=3D"MsoNormal">

<span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,7=
3,125)">Hi Roberto,<u></u><u></u></span></p><p class=3D"MsoNormal"><span st=
yle=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,73,125)">=
<u></u>=C2=A0<u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">not sure if you agree, but FWIW, here is a s=
ummary as I see it (currently):<u></u><u></u></span></p><p class=3D"MsoNorm=
al">

<span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,7=
3,125)"><u></u>=C2=A0<u></u></span></p><p class=3D"MsoNormal"><span style=
=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,73,125)">1) =
WebSocket frame boundaries _can_ have semantic meaning in WebSocket _extens=
ions_</span></p>

</div></div></blockquote><div><br></div></div><div>Agreed. And there are no=
ne deplyed that we know about where this is a concern.</div><div class=3D"i=
m"><div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0=
px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);borde=
r-left-style:solid;padding-left:1ex">

<div lang=3D"DE" link=3D"blue" vlink=3D"purple"><div><p class=3D"MsoNormal"=
><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,=
73,125)"><u></u><u></u></span></p><p class=3D"MsoNormal"><span style=3D"fon=
t-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,73,125)"><u></u>=C2=
=A0<u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">2) WebSocket message and frame boundaries ar=
e indeed preserved in the current WS/SPDY proposal,<u></u><u></u></span></p=
>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">however, non-final WebSocket frame ends cann=
ot be _immediately_ detected.</span></p></div></div></blockquote><div><br>
</div>
</div><div>I&#39;m not sure what you mean here?</div><div class=3D"im"><div=
>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px =
0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-=
style:solid;padding-left:1ex">

<div lang=3D"DE" link=3D"blue" vlink=3D"purple"><div><p class=3D"MsoNormal"=
><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,=
73,125)"><u></u><u></u></span></p><p class=3D"MsoNormal"><span style=3D"fon=
t-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,73,125)"><u></u>=C2=
=A0<u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">3) The current WS/SPDY proposal<u></u><u></u=
></span></p><p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-famil=
y:Calibri,sans-serif;color:rgb(31,73,125)">=C2=A0 * is comfortable to imple=
ment for implementations that _only_ want to implement WS/SPDY, and not nat=
ive WS at the same time<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)"> =C2=A0* makes things complicated for WS imp=
lementations that want to support both WS/SPDY _and_ native WS at the same =
time<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)"> =C2=A0* has significant wire-level overhead=
 compared to native WS (and compared to the proposal below)<u></u><u></u></=
span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)"><u></u>=C2=A0</span></p></div></div></blockq=
uote><div><br></div></div><div>The complication here is relative. Using the=
 approach proposed in terms of using MSG_DONE, etc. would make deployment a=
t large sites like Google substantially easier as large sites typically wan=
t to see the semantic boundaries easily, quickly, and consistently, as ther=
e needs to be robust mechanisms for handling failover, etc. The more layers=
 of interpretation there are when doing this, the bigger the number of prob=
lems.=C2=A0</div>

<div><br></div><div>I&#39;m honestly not worried about the potential increa=
sed overhead-- arguably HTTP2 reduces overhead substantially simply be requ=
iring fewer handshakes (the connection setup is a non-trivial amount of byt=
es and latency. That this was ignored was hugely frustrating to me during t=
he initial standardization process here in HyBi). Even discounting this, I =
haven&#39;t actually heard of use-cases where a well-written application se=
nds so many messages that the bandwidth is saturated. I am for efficiency (=
had to work on compression for HTTP2, for instance), but not when it increa=
ses complexity in the more general case.</div>

<div><br></div><div>In terms of total amount of code:</div><div>In any laye=
ring scheme, we&#39;d need to use a HEADER frame one way or another because=
 we need to at least announce that we&#39;re connecting to a WS endpoint.</=
div>

<div>In terms of decode, we&#39;re essentially talking the following state =
machine:</div><div><br></div><div>if frame.type =3D=3D HEADERS:</div><div>=
=C2=A0 =C2=A0 streams[frame.stream_id] =3D Stream(frame.headers)<br></div><=
div>elif frame.type =3D=3D DATA:</div>

<div>=C2=A0 =C2=A0 payload[frame.stream_id] +=3D frame.payload<br></div><di=
v>=C2=A0 =C2=A0 if frame.MSG_DONE or frame.STREAM_DONE:</div><div>=C2=A0 =
=C2=A0 =C2=A0 passMessageToApplication(payload[frame.stream_id])</div><div>=
=C2=A0 =C2=A0 =C2=A0 del payload[frame.stream_id]</div>

<div>if frame.STREAM_DONE:</div><div>=C2=A0 del streams[frame.stream_id]</d=
iv><div><br></div><div>While the above is pseudocode and leaves out error h=
andling, one of the interesting things is that it is the same for both HTTP=
, WS, or any other protocol layering over HTTP2.</div>
<div class=3D"im">
<div><br></div><div><br></div><div>=C2=A0</div><blockquote class=3D"gmail_q=
uote" style=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-c=
olor:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><div lang=
=3D"DE" link=3D"blue" vlink=3D"purple">

<div><p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calib=
ri,sans-serif;color:rgb(31,73,125)"><u></u></span></p><p class=3D"MsoNormal=
"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31=
,73,125)">I&#39;d like to suggest a different WS/SPDY:<u></u><u></u></span>=
</p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)"><u></u>=C2=A0<u></u></span></p><p>Pack WebSo=
cket frames _including_ full WS frame header as _payload_ of SPDY data fram=
es only - no SPDY header frames at all.<u></u><u></u></p>

<p><u></u>=C2=A0<u></u></p><p>In the absence of negotiated WebSocket extens=
ions (or when all WS extensions in place do not assign semantics to WS fram=
e boundaries),<u></u><u></u></p><p>fragment WebSocket payload at the WS lev=
el such that WS frame header+payload fits into the payload of SPDY data fra=
mes.<u></u><u></u></p>

<p><u></u>=C2=A0<u></u></p><p>In the latter case, this essentially adds a s=
mall SPDY data frame header (of fixed length) to WS frames. Easy to impleme=
nt and efficient on wire.<u></u><u></u></p><p><u></u>=C2=A0<u></u></p><p>IM=
HO, this has advantages compared to current proposal and considering 3) abo=
ve.<u></u><u></u></p>

<p><u></u>=C2=A0<u></u></p><p>The only disadvantage is for implementations =
that _only_ want to implement WS/SPDY and not native WS at the same time.</=
p></div></div></blockquote><div><br></div></div><div>One of the nice things=
 about the current plan as compared to your proposal is that it allows for =
generic infrastructure to be built to examine and handle things for DoS and=
 loadbalancing concerns across any protocol using HTTP2 as a session layer =
since all boundaries are clearly demarcated. It seems like a very nice thin=
g indeed that HTTP, WS, various internal API things, etc. can all end up us=
ing the same infrastructure.<br>

</div><div><br></div><div>Additionally, what happens if/when someone define=
s a mapping for HTTP2 onto something like QUIC, SCTP, TCP-Minion, or perhap=
s the RTCWeb stack? With the proposed design where the HTTP2 layer has full=
 visibility into the semantic boundaries, this mapping needs to be done onl=
y once for the HTTP2 session layer and all protocols layered atop it are no=
w able to take advantage of it (one would need signaling about HOL blocking=
 properties of streams to the HTTP2 layer in addition, of course). When we =
hide the semantic boundaries, this requires work from every implementor, an=
d impedes adoption either of the new HTTP2 mapping onto the better transpor=
t, or impedes adoption of WS.</div>

<div><br></div><div>I want to see the WS API succeed at scale and continue =
to do so into the future, so, my thought is that the advantages in regulari=
ty/consistency (today for DoS and loadbalancing, tomorrow on new transports=
) outweigh the relatively small theoretical byte efficiency tradeoff.</div>

<div>As I mentioned earlier, this is the tradeoff that HTTP2 has made for H=
TTP at the application layer as well, so even the approach is consistent :)=
</div><div class=3D"im"><div><br></div><blockquote class=3D"gmail_quote" st=
yle=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb=
(204,204,204);border-left-style:solid;padding-left:1ex">

<div lang=3D"DE" link=3D"blue" vlink=3D"purple"><div><p>=C2=A0<u></u></p><p=
>=3D=3D<u></u><u></u></p><p><u></u>=C2=A0<u></u></p><p>Regarding &quot;perm=
essage-priority&quot; with ordering guarantees: should I find time and if t=
here is interest, I&#39;ll write something up.<u></u><u></u></p>

<p>This could be a WebSocket extension. To be useful, the WS API in browser=
s would need to expose it though ..</p></div></div></blockquote><div><br></=
div></div><div>Yup. The same thing is happening for HTTP (people finding th=
at expressing priority for resources is useful), btw, from what I&#39;m led=
 to understand.=C2=A0</div>

<div>I&#39;m hopefuly for the day when APIs allow us to easily signal both =
priority, and also other important properties like drop-dead-times (get thi=
s message to the other side at most in X seconds from now, else leave it on=
 the floor), and ordering requirements (e.g. don&#39;t care about ordering,=
 care about ordering, but only to know when something has been dropped, etc=
.).</div>

<div>The interesting thing about priorities is that it is really only usefu=
l when you&#39;re muxing over one connection/session.</div><span class=3D"H=
OEnZb"><font color=3D"#888888"><div>-=3DR</div></font></span><div><div clas=
s=3D"h5">
<div><br></div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0p=
x 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-lef=
t-style:solid;padding-left:1ex">
<div lang=3D"DE" link=3D"blue" vlink=3D"purple"><div><p><u></u><u></u></p><=
p><u></u>=C2=A0<u></u></p><p>Just my 2cts,<u></u><u></u></p><p>Tobias<u></u=
><u></u></p><p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-famil=
y:Calibri,sans-serif;color:rgb(31,73,125)"><u></u>=C2=A0<u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)"><u></u>=C2=A0<u></u></span></p><p class=3D"M=
soNormal"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;colo=
r:rgb(31,73,125)"><u></u>=C2=A0<u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)"><u></u>=C2=A0<u></u></span></p><div style=3D=
"border-style:none none none solid;border-left-color:blue;border-left-width=
:1.5pt;padding:0cm 0cm 0cm 4pt">

<div><div style=3D"border-style:solid none none;border-top-color:rgb(225,22=
5,225);border-top-width:1pt;padding:3pt 0cm 0cm"><p class=3D"MsoNormal"><b>=
<span style=3D"font-size:11pt;font-family:Calibri,sans-serif">Von:</span></=
b><span style=3D"font-size:11pt;font-family:Calibri,sans-serif"> Roberto Pe=
on [mailto:<a href=3D"mailto:fenix@google.com" target=3D"_blank">fenix@goog=
le.com</a>] <br>

<b>Gesendet:</b> Dienstag, 14. Januar 2014 19:30<br><b>An:</b> Yutaka Hiran=
o<br><b>Cc:</b> Tobias Oberstein; Takeshi Yoshino; <a href=3D"mailto:hybi@i=
etf.org" target=3D"_blank">hybi@ietf.org</a></span></p><div><div>
<br><b>Betreff:</b> Re: [hybi] Discontinuation of mux standardizaton in fav=
or of WS/HTTP/2.0<u></u><u></u></div></div><p></p></div></div><div><div><p =
class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><div><p class=3D"MsoNormal">
I&#39;ve been lurking anyway :)<u></u><u></u></p><div><p class=3D"MsoNormal=
"><u></u>=C2=A0<u></u></p></div><div><p class=3D"MsoNormal">As I understand=
 the proposal, (and it seems that Takeshi and Yukata are covering this nice=
ly, so what i say is probably redundant):<u></u><u></u></p>

</div><div><p class=3D"MsoNormal">Message boundaries, having semantic meani=
ng, are preserved.<u></u><u></u></p></div><div><p class=3D"MsoNormal"><u></=
u>=C2=A0<u></u></p></div><div><p class=3D"MsoNormal">Frame boundaries, havi=
ng no semantic meaning, are not preserved. I don&#39;t think it makes sense=
 to worry about this greatly when the ability to extend the protocol as imp=
lemented in HTTP2 is present. It doesn&#39;t make a lot of sense to signal =
the frame boundaries through everything unless there is a high-likelyhood a=
nd high-utility use-case for such. If such is found, then HTTP2 can be revi=
sed to include it. This is the approach we&#39;ve been taking with most of =
the HTTP2 stuff, since there has been great effort placed on allowing new v=
ersions to be deployed in the future.<u></u><u></u></p>

</div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div><p cla=
ss=3D"MsoNormal">The mapping of WS onto HTTP2 is fairly direct-- we reserve=
d the MSG_DONE bit specifically for this kind of use-case (i.e. bidirection=
al, full-duplex, message-based).<u></u><u></u></p>

</div><div><p class=3D"MsoNormal">The observation made is that there are es=
sentially 3 types of data that need to be preserved one way or another:<u><=
/u><u></u></p></div><div><p class=3D"MsoNormal">1) metadata: data about the=
 data, e.g. text, binary, where it is supposed to be going, etc.<u></u><u><=
/u></p>

</div><div><p class=3D"MsoNormal">2) payload<u></u><u></u></p></div><div><p=
 class=3D"MsoNormal">3) boundaries between these.<u></u><u></u></p></div><d=
iv><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div><p class=3D"Ms=
oNormal">

imho, metadata is handled with HEADERS (I&#39;d have preferred it to be cal=
led METADATA, but was voted down at the time :) ), payload is handled with =
DATA frames, and the boundaries are delimited either through the use of HEA=
DERS frames, or via the MSG_DONE or STREAM_DONE bits.<u></u><u></u></p>

</div><div><p class=3D"MsoNormal">In the presence of extensions, assuming t=
hat MSG_DONE is required to signal the end of a message, then HEADERS can b=
e used to signal any boundary desired.<u></u><u></u></p></div><div><p class=
=3D"MsoNormal">

<u></u>=C2=A0<u></u></p></div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u><=
/u></p></div><div><p class=3D"MsoNormal">I totally agree that it would be n=
ice to expose the priority of various messages/channels with respect to eac=
h other. That is really an issue that is outside the WG&#39;s purvue, thoug=
h.<u></u><u></u></p>

</div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div></div><div>=
<p class=3D"MsoNormal" style=3D"margin-bottom:12pt"><u></u>=C2=A0<u></u></p=
><div><p class=3D"MsoNormal">On Thu, Jan 9, 2014 at 7:54 PM, Yutaka Hirano =
&lt;<a href=3D"mailto:yhirano@chromium.org" target=3D"_blank">yhirano@chrom=
ium.org</a>&gt; wrote:<u></u><u></u></p>

<blockquote style=3D"border-style:none none none solid;border-left-color:rg=
b(204,204,204);border-left-width:1pt;padding:0cm 0cm 0cm 6pt;margin:5pt 0cm=
 5pt 4.8pt"><div><div><p class=3D"MsoNormal">+Roberto as an HTTP2 expert.<u=
></u><u></u></p>

</div><div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><block=
quote style=3D"border-style:none none none solid;border-left-color:rgb(204,=
204,204);border-left-width:1pt;padding:0cm 0cm 0cm 6pt;margin:5pt 0cm 5pt 4=
.8pt">

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">2) Ok, both WS frame and message boundaries =
are preserved.<br>=C2=A0<br>But: I cannot recognize the end of a WS frame u=
ntil the next WS frame starts (for which I receive a HEADERS frame, which i=
s the implicit signal that the previous WS frame ended), right?</span><u></=
u><u></u></p>

</blockquote></div><p class=3D"MsoNormal">You are right, you can&#39;t.<u><=
/u><u></u></p><div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></di=
v><blockquote style=3D"border-style:none none none solid;border-left-color:=
rgb(204,204,204);border-left-width:1pt;padding:0cm 0cm 0cm 6pt;margin:5pt 0=
cm 5pt 4.8pt">

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">=C2=A0<br>And if so, why is there no FRAME_D=
ONE bit on the SPDY frame that corresponds to the end of a WS frame?</span>=
<u></u><u></u></p>

</blockquote></div><p class=3D"MsoNormal">Given that WebSocket frames &quot=
;have no semantic meaning&quot; as in RFC6455, I don&#39;t think the immedi=
ate recognition of the end of a WebSocket frame is requisite.<u></u><u></u>=
</p>

<div><p class=3D"MsoNormal">I would like to introduce MSG_DONE in HTTP2 so =
that intermediaries that don&#39;t fully understand WS over HTTP2 can relay=
 it.<u></u><u></u></p></div><div><p class=3D"MsoNormal">Since there are not=
 so much space in FLAGS field, Introducing FRAME_DONE may decrease the exte=
nsiveness of the spec.<u></u><u></u></p>

</div><div><p class=3D"MsoNormal">Do you have a strong preference to introd=
uce FRAME_DONE?<u></u><u></u></p></div><div><p class=3D"MsoNormal"><u></u>=
=C2=A0<u></u></p></div></div><div><div><div><p class=3D"MsoNormal" style=3D=
"margin-bottom:12pt">

<u></u>=C2=A0<u></u></p><div><p class=3D"MsoNormal">On Thu, Jan 9, 2014 at =
6:53 PM, Tobias Oberstein &lt;<a href=3D"mailto:tobias.oberstein@tavendo.de=
" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt; wrote:<u></u><u></u=
></p><blockquote style=3D"border-style:none none none solid;border-left-col=
or:rgb(204,204,204);border-left-width:1pt;padding:0cm 0cm 0cm 6pt;margin:5p=
t 0cm 5pt 4.8pt">

<div><div><p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:=
Calibri,sans-serif;color:rgb(31,73,125)">Hi Yutaka,</span><u></u><u></u></p=
><p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,s=
ans-serif;color:rgb(31,73,125)">=C2=A0</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">1) Yes, I understand. Thanks!</span><u></u><=
u></u></p><p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:=
Calibri,sans-serif;color:rgb(31,73,125)">=C2=A0</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">2) Ok, both WS frame and message boundaries =
are preserved.</span><u></u><u></u></p><p class=3D"MsoNormal"><span style=
=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,73,125)">=C2=
=A0</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">But: I cannot recognize the end of a WS fram=
e until the next WS frame starts (for which I receive a HEADERS frame, whic=
h is the implicit signal that the previous WS frame ended), right?</span><u=
></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">=C2=A0</span><u></u><u></u></p><p class=3D"M=
soNormal"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;colo=
r:rgb(31,73,125)">And if so, why is there no FRAME_DONE bit on the SPDY fra=
me that corresponds to the end of a WS frame?</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">=C2=A0</span><u></u><u></u></p><p class=3D"M=
soNormal"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;colo=
r:rgb(31,73,125)">3) Ok, great! Hope this works, since SPDY provides header=
 compression, but no sophisticated bidirectional payload compression as &qu=
ot;WS permessage-deflate&quot; does.</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">=C2=A0</span><u></u><u></u></p><p class=3D"M=
soNormal"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;colo=
r:rgb(31,73,125)">Cheers,</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">/Tobias</span><u></u><u></u></p><p class=3D"=
MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;col=
or:rgb(31,73,125)">=C2=A0</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11pt;font-family:Calibri,sa=
ns-serif;color:rgb(31,73,125)">=C2=A0</span><u></u><u></u></p><p class=3D"M=
soNormal"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;colo=
r:rgb(31,73,125)">=C2=A0</span><u></u><u></u></p>

<div style=3D"border-style:none none none solid;border-left-color:blue;bord=
er-left-width:1.5pt;padding:0cm 0cm 0cm 4pt"><div><div style=3D"border-styl=
e:solid none none;border-top-color:rgb(225,225,225);border-top-width:1pt;pa=
dding:3pt 0cm 0cm">

<p class=3D"MsoNormal"><b><span style=3D"font-size:11pt;font-family:Calibri=
,sans-serif">Von:</span></b><span style=3D"font-size:11pt;font-family:Calib=
ri,sans-serif"> </span><a href=3D"mailto:yhirano@google.com" target=3D"_bla=
nk"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif">yhirano@g=
oogle.com</span></a><span style=3D"font-size:11pt;font-family:Calibri,sans-=
serif"> [mailto:</span><a href=3D"mailto:yhirano@google.com" target=3D"_bla=
nk"><span style=3D"font-size:11pt;font-family:Calibri,sans-serif">yhirano@g=
oogle.com</span></a><span style=3D"font-size:11pt;font-family:Calibri,sans-=
serif">] <b>Im Auftrag von </b>Yutaka Hirano<br>

<b>Gesendet:</b> Donnerstag, 9. Januar 2014 10:37<br><b>An:</b> Tobias Ober=
stein<br><b>Cc:</b> Takeshi Yoshino; </span><a href=3D"mailto:hybi@ietf.org=
" target=3D"_blank"><span style=3D"font-size:11pt;font-family:Calibri,sans-=
serif">hybi@ietf.org</span></a><span style=3D"font-size:11pt;font-family:Ca=
libri,sans-serif"><br>

<b>Betreff:</b> Re: [hybi] Discontinuation of mux standardizaton in favor o=
f WS/HTTP/2.0</span><u></u><u></u></p></div></div><div><div><p class=3D"Mso=
Normal">=C2=A0<u></u><u></u></p><div><p class=3D"MsoNormal">Hi Tobias,<u></=
u><u></u></p>

<div><p class=3D"MsoNormal">Thank you for the comments.<u></u><u></u></p></=
div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p><blockquote style=
=3D"border-style:none none none solid;border-left-color:rgb(204,204,204);bo=
rder-left-width:1pt;padding:0cm 0cm 0cm 6pt;margin:5pt 0cm 5pt 4.8pt">

<p class=3D"MsoNormal"><span style=3D"font-size:10pt;font-family:Arial,sans=
-serif">1) Frame Mapping<br>&quot;&quot;&quot;<br>One WebSocket frame is ma=
pped into one preceding SPDY HEADERS frame and following plural SPDY data f=
rames. The SPDY HEADERS frame must contain WebSocket frame fields and follo=
wing SPDY data frames contain payload data. These data frames are free to b=
e reframed.<br>

&quot;&quot;&quot;<br></span><a href=3D"https://docs.google.com/document/d/=
124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank"><span =
style=3D"font-size:10pt;font-family:Arial,sans-serif">https://docs.google.c=
om/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit</span></a><=
span style=3D"font-size:10pt;font-family:Arial,sans-serif"><br>

So reframing of WebSocket frames (the payload) into multiple SPDY data fram=
es MAY occur, but does not need to occur? And a WS/SPDY implementation is f=
ree to choose the boundaries within WS frame payload at which to reframe in=
to different SPDY data frames as it likes?</span><u></u><u></u></p>

</blockquote><div><p class=3D"MsoNormal">A WebSocket frame MAY be split to =
multiple SPDY DATA frames.<u></u><u></u></p></div><div><p class=3D"MsoNorma=
l">Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA fram=
e into multiple DATA frames.<u></u><u></u></p>

</div><div><p class=3D"MsoNormal">Is this answering your question?<u></u><u=
></u></p></div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><b=
lockquote style=3D"border-style:none none none solid;border-left-color:rgb(=
204,204,204);border-left-width:1pt;padding:0cm 0cm 0cm 6pt;margin:5pt 0cm 5=
pt 4.8pt">

<p class=3D"MsoNormal"><span style=3D"font-size:10pt;font-family:Arial,sans=
-serif">2) End of WS Frame<br>The end of a WS message is signaled with a SP=
DY data frame having MSG_DONE set.<br>How is the end of a WS frame signaled=
, given that the WS frame may have been reframed into multiple SPDY data fr=
ames?</span><u></u><u></u></p>

</blockquote><div><p class=3D"MsoNormal">When you receive a HEADERS frame, =
you can recognize that the last SPDY frame is that last frame in a WebSocke=
t frame.<u></u><u></u></p></div><div><p class=3D"MsoNormal">When you receiv=
e a SPDY frame with MSG_DONE, you can recognize that it is the last SPDY fr=
ame in a WebSocket message and hence the last SPDY frame in a WebSocket fra=
me.<u></u><u></u></p>

</div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><blockquote=
 style=3D"border-style:none none none solid;border-left-color:rgb(204,204,2=
04);border-left-width:1pt;padding:0cm 0cm 0cm 6pt;margin:5pt 0cm 5pt 4.8pt"=
><p class=3D"MsoNormal">

<span style=3D"font-size:10pt;font-family:Arial,sans-serif">3) WS extension=
<br>Are WS extensions fully supported? In particular, is &quot;permessage-d=
eflate&quot; supported when running WS/SPDY?</span><u></u><u></u></p></bloc=
kquote>

<div><p class=3D"MsoNormal">Yes. We have reserved bits and we can emulate t=
he opening handshake, I think.<u></u><u></u></p></div><div><p class=3D"MsoN=
ormal">=C2=A0<u></u><u></u></p></div><div><p class=3D"MsoNormal">On Thu, Ja=
n 9, 2014 at 6:11 PM, Tobias Oberstein &lt;<a href=3D"mailto:tobias.oberste=
in@tavendo.de" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt; wrote:=
<u></u><u></u></p>

<blockquote style=3D"border-style:none none none solid;border-left-color:rg=
b(204,204,204);border-left-width:1pt;padding:0cm 0cm 0cm 6pt;margin:5pt 0cm=
 5pt 4.8pt"><p class=3D"MsoNormal">Hi Takeshi,<br><br>it&#39;s a pity, but =
I can understand the motivation to shove everything into HTTP2.<br>

<br>Anyway, some questions:<br><br>1) Frame Mapping<br>&quot;&quot;&quot;<b=
r>One WebSocket frame is mapped into one preceding SPDY HEADERS frame and f=
ollowing plural SPDY data frames. The SPDY HEADERS frame must contain WebSo=
cket frame fields and following SPDY data frames contain payload data. Thes=
e data frames are free to be reframed.<br>

&quot;&quot;&quot;<br><a href=3D"https://docs.google.com/document/d/124vB5N=
V1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank">https://docs.=
google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit</a>=
<br>

<br>So reframing of WebSocket frames (the payload) into multiple SPDY data =
frames MAY occur, but does not need to occur? And a WS/SPDY implementation =
is free to choose the boundaries within WS frame payload at which to refram=
e into different SPDY data frames as it likes?<br>

<br>2) End of WS Frame<br><br>The end of a WS message is signaled with a SP=
DY data frame having MSG_DONE set.<br><br>How is the end of a WS frame sign=
aled, given that the WS frame may have been reframed into multiple SPDY dat=
a frames?<br>

<br>3) WS extension<br><br>Are WS extensions fully supported? In particular=
, is &quot;permessage-deflate&quot; supported when running WS/SPDY?<br><br>=
Thanks!<br>/Tobias<br><br><br>Von: hybi [mailto:<a href=3D"mailto:hybi-boun=
ces@ietf.org" target=3D"_blank">hybi-bounces@ietf.org</a>] Im Auftrag von T=
akeshi Yoshino<br>

Gesendet: Dienstag, 7. Januar 2014 08:36<br>An: <a href=3D"mailto:hybi@ietf=
.org" target=3D"_blank">hybi@ietf.org</a><br>Betreff: [hybi] Discontinuatio=
n of mux standardizaton in favor of WS/HTTP/2.0<u></u><u></u></p><div><div>

<p class=3D"MsoNormal"><br>Hi all,<br><br>We&#39;ve been working on mux ext=
ension to make WebSocket scalable for long time, but as SPDY got widely sup=
ported and adopted quickly, now there seems to be no longer any good reason=
 to continue this work which is WebSocket dedicated.<br>

<br>Scalability is still important issue to address. From now, I&#39;d like=
 to focus on WS/SPDY effort which we&#39;ve been working on in parallel wit=
h the mux extension.<br><br>Please take a look at the latest proposal in th=
is post.<br>

<a href=3D"https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc" ta=
rget=3D"_blank">https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3gl=
c</a><br><br>If you really want the dedicated mux extension than one over S=
PDY, sorry but please take over the work.<u></u><u></u></p>

</div></div><div><div><p class=3D"MsoNormal">______________________________=
_________________<br>hybi mailing list<br><a href=3D"mailto:hybi@ietf.org" =
target=3D"_blank">hybi@ietf.org</a><br><a href=3D"https://www.ietf.org/mail=
man/listinfo/hybi" target=3D"_blank">https://www.ietf.org/mailman/listinfo/=
hybi</a><u></u><u></u></p>

</div></div></blockquote></div><p class=3D"MsoNormal">=C2=A0<u></u><u></u><=
/p></div></div></div></div></div></div></div></blockquote></div><p class=3D=
"MsoNormal"><u></u>=C2=A0<u></u></p></div></div></div></blockquote></div><p=
 class=3D"MsoNormal">

<u></u>=C2=A0<u></u></p></div></div></div></div></div></div></blockquote></=
div></div></div><br></div></div>
</blockquote></div><br></div>

--001a11c2ef125986ab04effbe70f--

From yhirano@google.com  Tue Jan 14 22:30:11 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 13A501AE2E8 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 22:30:11 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 3yoXU9odmFcd for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 22:30:06 -0800 (PST)
Received: from mail-qa0-x22c.google.com (mail-qa0-x22c.google.com [IPv6:2607:f8b0:400d:c00::22c]) by ietfa.amsl.com (Postfix) with ESMTP id C85841AE2E2 for <hybi@ietf.org>; Tue, 14 Jan 2014 22:30:05 -0800 (PST)
Received: by mail-qa0-f44.google.com with SMTP id w5so563918qac.17 for <hybi@ietf.org>; Tue, 14 Jan 2014 22:29:54 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=7wt0Tu2/HUiWJ9wpK9uoHkvyIy1WaAEon0q5K7wW8Qo=; b=bWqZ5Aph7mmAH5LifFG3yoWNaSBbFsIJnokR6/OJgGGeFmYatwh7GDB7keUI3ouHCO I/veQfuQklEvI9Zi7x27VU0aba6yW//AkqLHiblbAFbMX7nQdIjLaBI+HmPOVmQuRgFh 9MlMrUDHwn8ScUHfPxo7RZl9gv886eap0YXc1OzkGK/Y/g3FIoVaBL0ix2gDrbJZ19nR B4uHqEHbG5UTqIRzBqTsn6KNGNbzAKv1XYHTOa6GM1rw9X9bC+JWb9hUzRCiaT843sUL Xp2oCwiwhWGF9wnxs4QFqszbB9sNx125a6Kegdcckos2tV4AUx9i3yYE6KvMqCZDLYqD PRxw==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=7wt0Tu2/HUiWJ9wpK9uoHkvyIy1WaAEon0q5K7wW8Qo=; b=m06H0Sjimdgv5l0cRQ70MMHesspIYIPi78tKMYPNy2G17anEok98FXiyqEZXPU0rsw Vx9BICWGBiVIB6wCUaj3liMituVGUpJ/D8LiSx78hMulOGYgjF+UqJmsfSBTVOHsHp/a oPaYR/tUX7T/IIc63VdjDrZO/Ia8fL04gZ1FU=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=7wt0Tu2/HUiWJ9wpK9uoHkvyIy1WaAEon0q5K7wW8Qo=; b=jYZMNSIHfpeZ3CBk5e0gkoc0ACcXlo56SwaCJC1/7et+/Hsg7CVvuHvtZgJxe9beAL HPLfAXQODkBTr5B67w2S5oMJ/e3S6uOXCdHtTLCnsUp0AfsSPviWfauD+rl/sXyO6oZ1 iSEC0W7tbbUaX52RIYhsHgtya3F0dyV/rNpMxOaNw+4ZiQApHXUUUwwSjF4dNuLRrnVD AJh8J3RuPj7NCHwLQunYHskqPJ22W4L1Sv0bYad2qrcUEOY1QY9pyKl8DGYP2PmC+nfH sRQ2tkwtkYUt0VGQEayO+Yq0al/2hBFGL7gNJP51kzOp07IJN8cnYgfgfGC28FVVL49q WUdA==
X-Gm-Message-State: ALoCoQnhDFUD4X7hZEQ6CuN8geKOUSAxqUB/IIEX3xNbpdNM5hHkm85Z5TJ8nY4irE9E46b+O7fX2xghe0b0XgXHkuFZQlh3eNoWaIqwQpf95EmG0OI19xpf3DHxD94v4HswAPx2XKSfCg1p9i9D+voQhwkEDsMCsoCMRsgsVGUJx1jBXmuwB8SGRGBAJl0IPpQC3Cvbno53
MIME-Version: 1.0
X-Received: by 10.229.90.199 with SMTP id j7mr1091506qcm.14.1389767393881; Tue, 14 Jan 2014 22:29:53 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.140.109.202 with HTTP; Tue, 14 Jan 2014 22:29:53 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net>
Date: Wed, 15 Jan 2014 15:29:53 +0900
X-Google-Sender-Auth: c8o0pPiWuBn8hxfH4JzSAmpQHVg
Message-ID: <CABihn6GJLk-mbLGvLgG4dsTGuAtnp2JKx-qtOdeaB1CiWeEmdg@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=001a11343cdcd543a604effc6f12
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 15 Jan 2014 06:30:11 -0000

--001a11343cdcd543a604effc6f12
Content-Type: text/plain; charset=UTF-8

FYI: I added the framing overhead measurement result at the last of the
document (Memo - Framing overhead section).

Thanks,



On Wed, Jan 15, 2014 at 4:09 AM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> Hi Roberto,
>
>
>
> not sure if you agree, but FWIW, here is a summary as I see it (currently):
>
>
>
> 1) WebSocket frame boundaries _can_ have semantic meaning in WebSocket
> _extensions_
>
>
>
> 2) WebSocket message and frame boundaries are indeed preserved in the
> current WS/SPDY proposal,
>
> however, non-final WebSocket frame ends cannot be _immediately_ detected.
>
>
>
> 3) The current WS/SPDY proposal
>
>   * is comfortable to implement for implementations that _only_ want to
> implement WS/SPDY, and not native WS at the same time
>
>  * makes things complicated for WS implementations that want to support
> both WS/SPDY _and_ native WS at the same time
>
>  * has significant wire-level overhead compared to native WS (and compared
> to the proposal below)
>
>
>
> I'd like to suggest a different WS/SPDY:
>
>
>
> Pack WebSocket frames _including_ full WS frame header as _payload_ of
> SPDY data frames only - no SPDY header frames at all.
>
>
>
> In the absence of negotiated WebSocket extensions (or when all WS
> extensions in place do not assign semantics to WS frame boundaries),
>
> fragment WebSocket payload at the WS level such that WS frame
> header+payload fits into the payload of SPDY data frames.
>
>
>
> In the latter case, this essentially adds a small SPDY data frame header
> (of fixed length) to WS frames. Easy to implement and efficient on wire.
>
>
>
> IMHO, this has advantages compared to current proposal and considering 3)
> above.
>
>
>
> The only disadvantage is for implementations that _only_ want to implement
> WS/SPDY and not native WS at the same time.
>
>
>
> ==
>
>
>
> Regarding "permessage-priority" with ordering guarantees: should I find
> time and if there is interest, I'll write something up.
>
> This could be a WebSocket extension. To be useful, the WS API in browsers
> would need to expose it though ..
>
>
>
> Just my 2cts,
>
> Tobias
>
>
>
>
>
>
>
>
>
> *Von:* Roberto Peon [mailto:fenix@google.com]
> *Gesendet:* Dienstag, 14. Januar 2014 19:30
> *An:* Yutaka Hirano
> *Cc:* Tobias Oberstein; Takeshi Yoshino; hybi@ietf.org
>
> *Betreff:* Re: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
>
> I've been lurking anyway :)
>
>
>
> As I understand the proposal, (and it seems that Takeshi and Yukata are
> covering this nicely, so what i say is probably redundant):
>
> Message boundaries, having semantic meaning, are preserved.
>
>
>
> Frame boundaries, having no semantic meaning, are not preserved. I don't
> think it makes sense to worry about this greatly when the ability to extend
> the protocol as implemented in HTTP2 is present. It doesn't make a lot of
> sense to signal the frame boundaries through everything unless there is a
> high-likelyhood and high-utility use-case for such. If such is found, then
> HTTP2 can be revised to include it. This is the approach we've been taking
> with most of the HTTP2 stuff, since there has been great effort placed on
> allowing new versions to be deployed in the future.
>
>
>
> The mapping of WS onto HTTP2 is fairly direct-- we reserved the MSG_DONE
> bit specifically for this kind of use-case (i.e. bidirectional,
> full-duplex, message-based).
>
> The observation made is that there are essentially 3 types of data that
> need to be preserved one way or another:
>
> 1) metadata: data about the data, e.g. text, binary, where it is supposed
> to be going, etc.
>
> 2) payload
>
> 3) boundaries between these.
>
>
>
> imho, metadata is handled with HEADERS (I'd have preferred it to be called
> METADATA, but was voted down at the time :) ), payload is handled with DATA
> frames, and the boundaries are delimited either through the use of HEADERS
> frames, or via the MSG_DONE or STREAM_DONE bits.
>
> In the presence of extensions, assuming that MSG_DONE is required to
> signal the end of a message, then HEADERS can be used to signal any
> boundary desired.
>
>
>
>
>
> I totally agree that it would be nice to expose the priority of various
> messages/channels with respect to each other. That is really an issue that
> is outside the WG's purvue, though.
>
>
>
>
>
> On Thu, Jan 9, 2014 at 7:54 PM, Yutaka Hirano <yhirano@chromium.org>
> wrote:
>
> +Roberto as an HTTP2 expert.
>
>
>
> 2) Ok, both WS frame and message boundaries are preserved.
>
> But: I cannot recognize the end of a WS frame until the next WS frame
> starts (for which I receive a HEADERS frame, which is the implicit signal
> that the previous WS frame ended), right?
>
> You are right, you can't.
>
>
>
>
> And if so, why is there no FRAME_DONE bit on the SPDY frame that
> corresponds to the end of a WS frame?
>
> Given that WebSocket frames "have no semantic meaning" as in RFC6455, I
> don't think the immediate recognition of the end of a WebSocket frame is
> requisite.
>
> I would like to introduce MSG_DONE in HTTP2 so that intermediaries that
> don't fully understand WS over HTTP2 can relay it.
>
> Since there are not so much space in FLAGS field, Introducing FRAME_DONE
> may decrease the extensiveness of the spec.
>
> Do you have a strong preference to introduce FRAME_DONE?
>
>
>
>
>
> On Thu, Jan 9, 2014 at 6:53 PM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
>
> Hi Yutaka,
>
>
>
> 1) Yes, I understand. Thanks!
>
>
>
> 2) Ok, both WS frame and message boundaries are preserved.
>
>
>
> But: I cannot recognize the end of a WS frame until the next WS frame
> starts (for which I receive a HEADERS frame, which is the implicit signal
> that the previous WS frame ended), right?
>
>
>
> And if so, why is there no FRAME_DONE bit on the SPDY frame that
> corresponds to the end of a WS frame?
>
>
>
> 3) Ok, great! Hope this works, since SPDY provides header compression, but
> no sophisticated bidirectional payload compression as "WS
> permessage-deflate" does.
>
>
>
> Cheers,
>
> /Tobias
>
>
>
>
>
>
>
> *Von:* yhirano@google.com [mailto:yhirano@google.com] *Im Auftrag von *Yutaka
> Hirano
> *Gesendet:* Donnerstag, 9. Januar 2014 10:37
> *An:* Tobias Oberstein
> *Cc:* Takeshi Yoshino; hybi@ietf.org
> *Betreff:* Re: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
>
> Hi Tobias,
>
> Thank you for the comments.
>
>
>
> 1) Frame Mapping
> """
> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
> following plural SPDY data frames. The SPDY HEADERS frame must contain
> WebSocket frame fields and following SPDY data frames contain payload data.
> These data frames are free to be reframed.
> """
>
> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
> So reframing of WebSocket frames (the payload) into multiple SPDY data
> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
> is free to choose the boundaries within WS frame payload at which to
> reframe into different SPDY data frames as it likes?
>
> A WebSocket frame MAY be split to multiple SPDY DATA frames.
>
> Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA frame
> into multiple DATA frames.
>
> Is this answering your question?
>
>
>
> 2) End of WS Frame
> The end of a WS message is signaled with a SPDY data frame having MSG_DONE
> set.
> How is the end of a WS frame signaled, given that the WS frame may have
> been reframed into multiple SPDY data frames?
>
> When you receive a HEADERS frame, you can recognize that the last SPDY
> frame is that last frame in a WebSocket frame.
>
> When you receive a SPDY frame with MSG_DONE, you can recognize that it is
> the last SPDY frame in a WebSocket message and hence the last SPDY frame in
> a WebSocket frame.
>
>
>
> 3) WS extension
> Are WS extensions fully supported? In particular, is "permessage-deflate"
> supported when running WS/SPDY?
>
> Yes. We have reserved bits and we can emulate the opening handshake, I
> think.
>
>
>
> On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
>
> Hi Takeshi,
>
> it's a pity, but I can understand the motivation to shove everything into
> HTTP2.
>
> Anyway, some questions:
>
> 1) Frame Mapping
> """
> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
> following plural SPDY data frames. The SPDY HEADERS frame must contain
> WebSocket frame fields and following SPDY data frames contain payload data.
> These data frames are free to be reframed.
> """
>
> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
>
> So reframing of WebSocket frames (the payload) into multiple SPDY data
> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
> is free to choose the boundaries within WS frame payload at which to
> reframe into different SPDY data frames as it likes?
>
> 2) End of WS Frame
>
> The end of a WS message is signaled with a SPDY data frame having MSG_DONE
> set.
>
> How is the end of a WS frame signaled, given that the WS frame may have
> been reframed into multiple SPDY data frames?
>
> 3) WS extension
>
> Are WS extensions fully supported? In particular, is "permessage-deflate"
> supported when running WS/SPDY?
>
> Thanks!
> /Tobias
>
>
> Von: hybi [mailto:hybi-bounces@ietf.org] Im Auftrag von Takeshi Yoshino
> Gesendet: Dienstag, 7. Januar 2014 08:36
> An: hybi@ietf.org
> Betreff: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
> Hi all,
>
> We've been working on mux extension to make WebSocket scalable for long
> time, but as SPDY got widely supported and adopted quickly, now there seems
> to be no longer any good reason to continue this work which is WebSocket
> dedicated.
>
> Scalability is still important issue to address. From now, I'd like to
> focus on WS/SPDY effort which we've been working on in parallel with the
> mux extension.
>
> Please take a look at the latest proposal in this post.
> https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc
>
> If you really want the dedicated mux extension than one over SPDY, sorry
> but please take over the work.
>
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>
>
>
>
>
>
>

--001a11343cdcd543a604effc6f12
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">FYI: I added the framing overhead measurement result at th=
e last of the document (Memo - Framing overhead section).<div><br></div><di=
v>Thanks,<br></div><div><div><br></div></div></div><div class=3D"gmail_extr=
a">
<br><br><div class=3D"gmail_quote">On Wed, Jan 15, 2014 at 4:09 AM, Tobias =
Oberstein <span dir=3D"ltr">&lt;<a href=3D"mailto:tobias.oberstein@tavendo.=
de" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt;</span> wrote:<br>=
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">
<div lang=3D"DE" link=3D"blue" vlink=3D"purple"><div><p class=3D"MsoNormal"=
><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans=
-serif&quot;;color:#1f497d">Hi Roberto,<u></u><u></u></span></p><p class=3D=
"MsoNormal">
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></span></p><p class=3D"MsoNo=
rmal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot=
;sans-serif&quot;;color:#1f497d">not sure if you agree, but FWIW, here is a=
 summary as I see it (currently):<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">1) WebSocket frame =
boundaries _can_ have semantic meaning in WebSocket _extensions_<u></u><u><=
/u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">2) WebSocket messag=
e and frame boundaries are indeed preserved in the current WS/SPDY proposal=
,<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">however, non-final WebSoc=
ket frame ends cannot be _immediately_ detected.<u></u><u></u></span></p><p=
 class=3D"MsoNormal">
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></span></p><p class=3D"MsoNo=
rmal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot=
;sans-serif&quot;;color:#1f497d">3) The current WS/SPDY proposal<u></u><u><=
/u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0 * is comfortable t=
o implement for implementations that _only_ want to implement WS/SPDY, and =
not native WS at the same time<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"> =C2=A0* makes things com=
plicated for WS implementations that want to support both WS/SPDY _and_ nat=
ive WS at the same time<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"> =C2=A0* has significant =
wire-level overhead compared to native WS (and compared to the proposal bel=
ow)<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">I&#39;d like to sug=
gest a different WS/SPDY:<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p>Pack WebSocket frames _including_ full WS frame header as _payload=
_ of SPDY data frames only - no SPDY header frames at all.<u></u><u></u></p=
>
<p><u></u>=C2=A0<u></u></p><p>In the absence of negotiated WebSocket extens=
ions (or when all WS extensions in place do not assign semantics to WS fram=
e boundaries),<u></u><u></u></p><p>fragment WebSocket payload at the WS lev=
el such that WS frame header+payload fits into the payload of SPDY data fra=
mes.<u></u><u></u></p>
<p><u></u>=C2=A0<u></u></p><p>In the latter case, this essentially adds a s=
mall SPDY data frame header (of fixed length) to WS frames. Easy to impleme=
nt and efficient on wire.<u></u><u></u></p><p><u></u>=C2=A0<u></u></p><p>IM=
HO, this has advantages compared to current proposal and considering 3) abo=
ve.<u></u><u></u></p>
<p><u></u>=C2=A0<u></u></p><p>The only disadvantage is for implementations =
that _only_ want to implement WS/SPDY and not native WS at the same time.<u=
></u><u></u></p><p><u></u>=C2=A0<u></u></p><p>=3D=3D<u></u><u></u></p><p><u=
></u>=C2=A0<u></u></p>
<p>Regarding &quot;permessage-priority&quot; with ordering guarantees: shou=
ld I find time and if there is interest, I&#39;ll write something up.<u></u=
><u></u></p><p>This could be a WebSocket extension. To be useful, the WS AP=
I in browsers would need to expose it though ..<u></u><u></u></p>
<p><u></u>=C2=A0<u></u></p><p>Just my 2cts,<u></u><u></u></p><p>Tobias<u></=
u><u></u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-fa=
mily:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=
=A0<u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u=
></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><div style=3D"border:none;border-left:solid blue 1.5pt;padding:0cm 0c=
m 0cm 4.0pt">
<div><div style=3D"border:none;border-top:solid #e1e1e1 1.0pt;padding:3.0pt=
 0cm 0cm 0cm"><p class=3D"MsoNormal"><b><span style=3D"font-size:11.0pt;fon=
t-family:&quot;Calibri&quot;,&quot;sans-serif&quot;">Von:</span></b><span s=
tyle=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&q=
uot;"> Roberto Peon [mailto:<a href=3D"mailto:fenix@google.com" target=3D"_=
blank">fenix@google.com</a>] <br>
<b>Gesendet:</b> Dienstag, 14. Januar 2014 19:30<br><b>An:</b> Yutaka Hiran=
o<br><b>Cc:</b> Tobias Oberstein; Takeshi Yoshino; <a href=3D"mailto:hybi@i=
etf.org" target=3D"_blank">hybi@ietf.org</a></span></p><div><div class=3D"h=
5">
<br><b>Betreff:</b> Re: [hybi] Discontinuation of mux standardizaton in fav=
or of WS/HTTP/2.0<u></u><u></u></div></div><p></p></div></div><div><div cla=
ss=3D"h5"><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><div><p class=3D"M=
soNormal">
I&#39;ve been lurking anyway :)<u></u><u></u></p><div><p class=3D"MsoNormal=
"><u></u>=C2=A0<u></u></p></div><div><p class=3D"MsoNormal">As I understand=
 the proposal, (and it seems that Takeshi and Yukata are covering this nice=
ly, so what i say is probably redundant):<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">Message boundaries, having semantic meani=
ng, are preserved.<u></u><u></u></p></div><div><p class=3D"MsoNormal"><u></=
u>=C2=A0<u></u></p></div><div><p class=3D"MsoNormal">Frame boundaries, havi=
ng no semantic meaning, are not preserved. I don&#39;t think it makes sense=
 to worry about this greatly when the ability to extend the protocol as imp=
lemented in HTTP2 is present. It doesn&#39;t make a lot of sense to signal =
the frame boundaries through everything unless there is a high-likelyhood a=
nd high-utility use-case for such. If such is found, then HTTP2 can be revi=
sed to include it. This is the approach we&#39;ve been taking with most of =
the HTTP2 stuff, since there has been great effort placed on allowing new v=
ersions to be deployed in the future.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div><p cla=
ss=3D"MsoNormal">The mapping of WS onto HTTP2 is fairly direct-- we reserve=
d the MSG_DONE bit specifically for this kind of use-case (i.e. bidirection=
al, full-duplex, message-based).<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">The observation made is that there are es=
sentially 3 types of data that need to be preserved one way or another:<u><=
/u><u></u></p></div><div><p class=3D"MsoNormal">1) metadata: data about the=
 data, e.g. text, binary, where it is supposed to be going, etc.<u></u><u><=
/u></p>
</div><div><p class=3D"MsoNormal">2) payload<u></u><u></u></p></div><div><p=
 class=3D"MsoNormal">3) boundaries between these.<u></u><u></u></p></div><d=
iv><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div><p class=3D"Ms=
oNormal">
imho, metadata is handled with HEADERS (I&#39;d have preferred it to be cal=
led METADATA, but was voted down at the time :) ), payload is handled with =
DATA frames, and the boundaries are delimited either through the use of HEA=
DERS frames, or via the MSG_DONE or STREAM_DONE bits.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">In the presence of extensions, assuming t=
hat MSG_DONE is required to signal the end of a message, then HEADERS can b=
e used to signal any boundary desired.<u></u><u></u></p></div><div><p class=
=3D"MsoNormal">
<u></u>=C2=A0<u></u></p></div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u><=
/u></p></div><div><p class=3D"MsoNormal">I totally agree that it would be n=
ice to expose the priority of various messages/channels with respect to eac=
h other. That is really an issue that is outside the WG&#39;s purvue, thoug=
h.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div></div><div>=
<p class=3D"MsoNormal" style=3D"margin-bottom:12.0pt"><u></u>=C2=A0<u></u><=
/p><div><p class=3D"MsoNormal">On Thu, Jan 9, 2014 at 7:54 PM, Yutaka Hiran=
o &lt;<a href=3D"mailto:yhirano@chromium.org" target=3D"_blank">yhirano@chr=
omium.org</a>&gt; wrote:<u></u><u></u></p>
<blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0c=
m 0cm 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-=
bottom:5.0pt"><div><div><p class=3D"MsoNormal">+Roberto as an HTTP2 expert.=
<u></u><u></u></p>
</div><div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><block=
quote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm =
0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom=
:5.0pt">
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">2) Ok, both WS frame and =
message boundaries are preserved.<br>=C2=A0<br>But: I cannot recognize the =
end of a WS frame until the next WS frame starts (for which I receive a HEA=
DERS frame, which is the implicit signal that the previous WS frame ended),=
 right?</span><u></u><u></u></p>
</blockquote></div><p class=3D"MsoNormal">You are right, you can&#39;t.<u><=
/u><u></u></p><div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></di=
v><blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:=
0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margi=
n-bottom:5.0pt">
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0<br>And if so, why =
is there no FRAME_DONE bit on the SPDY frame that corresponds to the end of=
 a WS frame?</span><u></u><u></u></p>
</blockquote></div><p class=3D"MsoNormal">Given that WebSocket frames &quot=
;have no semantic meaning&quot; as in RFC6455, I don&#39;t think the immedi=
ate recognition of the end of a WebSocket frame is requisite.<u></u><u></u>=
</p>
<div><p class=3D"MsoNormal">I would like to introduce MSG_DONE in HTTP2 so =
that intermediaries that don&#39;t fully understand WS over HTTP2 can relay=
 it.<u></u><u></u></p></div><div><p class=3D"MsoNormal">Since there are not=
 so much space in FLAGS field, Introducing FRAME_DONE may decrease the exte=
nsiveness of the spec.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">Do you have a strong preference to introd=
uce FRAME_DONE?<u></u><u></u></p></div><div><p class=3D"MsoNormal"><u></u>=
=C2=A0<u></u></p></div></div><div><div><div><p class=3D"MsoNormal" style=3D=
"margin-bottom:12.0pt">
<u></u>=C2=A0<u></u></p><div><p class=3D"MsoNormal">On Thu, Jan 9, 2014 at =
6:53 PM, Tobias Oberstein &lt;<a href=3D"mailto:tobias.oberstein@tavendo.de=
" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt; wrote:<u></u><u></u=
></p><blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;paddi=
ng:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;ma=
rgin-bottom:5.0pt">
<div><div><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-famil=
y:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Hi Yutaka,</spa=
n><u></u><u></u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;=
font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=
=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">1) Yes, I understand. Tha=
nks!</span><u></u><u></u></p><p class=3D"MsoNormal"><span style=3D"font-siz=
e:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f49=
7d">=C2=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">2) Ok, both WS frame and =
message boundaries are preserved.</span><u></u><u></u></p><p class=3D"MsoNo=
rmal">
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></u></p><p class=3D"MsoNo=
rmal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot=
;sans-serif&quot;;color:#1f497d">But: I cannot recognize the end of a WS fr=
ame until the next WS frame starts (for which I receive a HEADERS frame, wh=
ich is the implicit signal that the previous WS frame ended), right?</span>=
<u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">And if so, why is t=
here no FRAME_DONE bit on the SPDY frame that corresponds to the end of a W=
S frame?</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">3) Ok, great! Hope =
this works, since SPDY provides header compression, but no sophisticated bi=
directional payload compression as &quot;WS permessage-deflate&quot; does.<=
/span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Cheers,</span><u></=
u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">/Tobias</span><u></u><u><=
/u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&=
quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></=
u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u=
><u></u></p>
<div style=3D"border:none;border-left:solid blue 1.5pt;padding:0cm 0cm 0cm =
4.0pt"><div><div style=3D"border:none;border-top:solid #e1e1e1 1.0pt;paddin=
g:3.0pt 0cm 0cm 0cm"><p class=3D"MsoNormal"><b><span style=3D"font-size:11.=
0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;">Von:</span></b>=
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;"> </span><a href=3D"mailto:yhirano@google.com" target=3D"_blank=
"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;san=
s-serif&quot;">yhirano@google.com</span></a><span style=3D"font-size:11.0pt=
;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;"> [mailto:</span><a=
 href=3D"mailto:yhirano@google.com" target=3D"_blank"><span style=3D"font-s=
ize:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;">yhirano@=
google.com</span></a><span style=3D"font-size:11.0pt;font-family:&quot;Cali=
bri&quot;,&quot;sans-serif&quot;">] <b>Im Auftrag von </b>Yutaka Hirano<br>
<b>Gesendet:</b> Donnerstag, 9. Januar 2014 10:37<br><b>An:</b> Tobias Ober=
stein<br><b>Cc:</b> Takeshi Yoshino; </span><a href=3D"mailto:hybi@ietf.org=
" target=3D"_blank"><span style=3D"font-size:11.0pt;font-family:&quot;Calib=
ri&quot;,&quot;sans-serif&quot;">hybi@ietf.org</span></a><span style=3D"fon=
t-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;"><br>
<b>Betreff:</b> Re: [hybi] Discontinuation of mux standardizaton in favor o=
f WS/HTTP/2.0</span><u></u><u></u></p></div></div><div><div><p class=3D"Mso=
Normal">=C2=A0<u></u><u></u></p><div><p class=3D"MsoNormal">Hi Tobias,<u></=
u><u></u></p>
<div><p class=3D"MsoNormal">Thank you for the comments.<u></u><u></u></p></=
div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p><blockquote style=
=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm 6.0pt;m=
argin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom:5.0pt">
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ar=
ial&quot;,&quot;sans-serif&quot;">1) Frame Mapping<br>&quot;&quot;&quot;<br=
>One WebSocket frame is mapped into one preceding SPDY HEADERS frame and fo=
llowing plural SPDY data frames. The SPDY HEADERS frame must contain WebSoc=
ket frame fields and following SPDY data frames contain payload data. These=
 data frames are free to be reframed.<br>
&quot;&quot;&quot;<br></span><a href=3D"https://docs.google.com/document/d/=
124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank"><span =
style=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&qu=
ot;">https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw=
9aWzSHijk/edit</span></a><span style=3D"font-size:10.0pt;font-family:&quot;=
Arial&quot;,&quot;sans-serif&quot;"><br>
So reframing of WebSocket frames (the payload) into multiple SPDY data fram=
es MAY occur, but does not need to occur? And a WS/SPDY implementation is f=
ree to choose the boundaries within WS frame payload at which to reframe in=
to different SPDY data frames as it likes?</span><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal">A WebSocket frame MAY be split to =
multiple SPDY DATA frames.<u></u><u></u></p></div><div><p class=3D"MsoNorma=
l">Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA fram=
e into multiple DATA frames.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">Is this answering your question?<u></u><u=
></u></p></div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><b=
lockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm =
0cm 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bo=
ttom:5.0pt">
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ar=
ial&quot;,&quot;sans-serif&quot;">2) End of WS Frame<br>The end of a WS mes=
sage is signaled with a SPDY data frame having MSG_DONE set.<br>How is the =
end of a WS frame signaled, given that the WS frame may have been reframed =
into multiple SPDY data frames?</span><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal">When you receive a HEADERS frame, =
you can recognize that the last SPDY frame is that last frame in a WebSocke=
t frame.<u></u><u></u></p></div><div><p class=3D"MsoNormal">When you receiv=
e a SPDY frame with MSG_DONE, you can recognize that it is the last SPDY fr=
ame in a WebSocket message and hence the last SPDY frame in a WebSocket fra=
me.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><blockquote=
 style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm 6=
.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom:5.0p=
t"><p class=3D"MsoNormal">
<span style=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-se=
rif&quot;">3) WS extension<br>Are WS extensions fully supported? In particu=
lar, is &quot;permessage-deflate&quot; supported when running WS/SPDY?</spa=
n><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal">Yes. We have reserved bits and we =
can emulate the opening handshake, I think.<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><div><p class=3D"MsoNorma=
l">On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein &lt;<a href=3D"mailto:t=
obias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstein@tavendo.de</=
a>&gt; wrote:<u></u><u></u></p>
<blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0c=
m 0cm 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-=
bottom:5.0pt"><p class=3D"MsoNormal">Hi Takeshi,<br><br>it&#39;s a pity, bu=
t I can understand the motivation to shove everything into HTTP2.<br>
<br>Anyway, some questions:<br><br>1) Frame Mapping<br>&quot;&quot;&quot;<b=
r>One WebSocket frame is mapped into one preceding SPDY HEADERS frame and f=
ollowing plural SPDY data frames. The SPDY HEADERS frame must contain WebSo=
cket frame fields and following SPDY data frames contain payload data. Thes=
e data frames are free to be reframed.<br>
&quot;&quot;&quot;<br><a href=3D"https://docs.google.com/document/d/124vB5N=
V1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank">https://docs.=
google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit</a>=
<br>
<br>So reframing of WebSocket frames (the payload) into multiple SPDY data =
frames MAY occur, but does not need to occur? And a WS/SPDY implementation =
is free to choose the boundaries within WS frame payload at which to refram=
e into different SPDY data frames as it likes?<br>
<br>2) End of WS Frame<br><br>The end of a WS message is signaled with a SP=
DY data frame having MSG_DONE set.<br><br>How is the end of a WS frame sign=
aled, given that the WS frame may have been reframed into multiple SPDY dat=
a frames?<br>
<br>3) WS extension<br><br>Are WS extensions fully supported? In particular=
, is &quot;permessage-deflate&quot; supported when running WS/SPDY?<br><br>=
Thanks!<br>/Tobias<br><br><br>Von: hybi [mailto:<a href=3D"mailto:hybi-boun=
ces@ietf.org" target=3D"_blank">hybi-bounces@ietf.org</a>] Im Auftrag von T=
akeshi Yoshino<br>
Gesendet: Dienstag, 7. Januar 2014 08:36<br>An: <a href=3D"mailto:hybi@ietf=
.org" target=3D"_blank">hybi@ietf.org</a><br>Betreff: [hybi] Discontinuatio=
n of mux standardizaton in favor of WS/HTTP/2.0<u></u><u></u></p><div><div>
<p class=3D"MsoNormal"><br>Hi all,<br><br>We&#39;ve been working on mux ext=
ension to make WebSocket scalable for long time, but as SPDY got widely sup=
ported and adopted quickly, now there seems to be no longer any good reason=
 to continue this work which is WebSocket dedicated.<br>
<br>Scalability is still important issue to address. From now, I&#39;d like=
 to focus on WS/SPDY effort which we&#39;ve been working on in parallel wit=
h the mux extension.<br><br>Please take a look at the latest proposal in th=
is post.<br>
<a href=3D"https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc" ta=
rget=3D"_blank">https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3gl=
c</a><br><br>If you really want the dedicated mux extension than one over S=
PDY, sorry but please take over the work.<u></u><u></u></p>
</div></div><div><div><p class=3D"MsoNormal">______________________________=
_________________<br>hybi mailing list<br><a href=3D"mailto:hybi@ietf.org" =
target=3D"_blank">hybi@ietf.org</a><br><a href=3D"https://www.ietf.org/mail=
man/listinfo/hybi" target=3D"_blank">https://www.ietf.org/mailman/listinfo/=
hybi</a><u></u><u></u></p>
</div></div></blockquote></div><p class=3D"MsoNormal">=C2=A0<u></u><u></u><=
/p></div></div></div></div></div></div></div></blockquote></div><p class=3D=
"MsoNormal"><u></u>=C2=A0<u></u></p></div></div></div></blockquote></div><p=
 class=3D"MsoNormal">
<u></u>=C2=A0<u></u></p></div></div></div></div></div></div></blockquote></=
div><br></div>

--001a11343cdcd543a604effc6f12--


From tobias.oberstein@tavendo.de  Tue Jan 14 22:34:40 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 37F371AE2DC for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 22:34:40 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id puUznp5T249W for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 22:34:36 -0800 (PST)
Received: from EXHUB020-1.exch020.serverdata.net (exhub020-1.exch020.serverdata.net [206.225.164.28]) by ietfa.amsl.com (Postfix) with ESMTP id E2A6E1AE1EA for <hybi@ietf.org>; Tue, 14 Jan 2014 22:34:35 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-1.exch020.serverdata.net ([206.225.164.28]) with mapi; Tue, 14 Jan 2014 22:34:24 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: 'Roberto Peon' <fenix@google.com>
Date: Tue, 14 Jan 2014 22:34:20 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8RZElWEhH1XuqBTBG21iDvEEz6MwAAu/jA
Message-ID: <634914A010D0B943A035D226786325D4446BF98B9D@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CAGzyod7HAjdDHt1B4XRHT6h43DXnoupqGmr4buP6D+bQs86m+w@mail.gmail.com>
In-Reply-To: <CAGzyod7HAjdDHt1B4XRHT6h43DXnoupqGmr4buP6D+bQs86m+w@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 15 Jan 2014 06:34:40 -0000
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From tyoshino@google.com  Tue Jan 14 23:00:26 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 11CCA1AE305 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 23:00:26 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id I_pbe3pvW-y1 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 23:00:24 -0800 (PST)
Received: from mail-we0-x233.google.com (mail-we0-x233.google.com [IPv6:2a00:1450:400c:c03::233]) by ietfa.amsl.com (Postfix) with ESMTP id B8FB01AE2ED for <hybi@ietf.org>; Tue, 14 Jan 2014 23:00:23 -0800 (PST)
Received: by mail-we0-f179.google.com with SMTP id w62so1343038wes.24 for <hybi@ietf.org>; Tue, 14 Jan 2014 23:00:11 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=49V3NIVW6705pJpnZ+9OR0WMuuPTkGSVq4xokQJmj8k=; b=dFLaeNb/+gz/LWiPQ2tNfp1/U7A6Q+OjcZLD0Sctfzm4rxbWU5nFsDC20zOD1aYPo7 Mn6pPApFsLf5FfAl4bwcZWewJtRn/Z7KRY8WqAIxGDIoNWaj8WO2DDBhUl3h3qAQ6CLu pvMLL0S1AQf0SbTwYWCojzlimQM04BsxYLUMBxOA5U+hqfT1bO6RicuOFumv4XM3SJAM 2wi+f8mRE2IRUOPonq/3GK9gRwNUsRXjSxgoxEk5bv8pc9i9pWzvkPZnsDd8OM9iS3UJ Txif5dlDbPvFZKp2ZM/y/FQfNfPi2QocTJ9+VGhFm/LaAHcIK+0p1WyxAD0o2FR+gAb3 ly4A==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=49V3NIVW6705pJpnZ+9OR0WMuuPTkGSVq4xokQJmj8k=; b=JISHvxc4i2Mpasr1kyRQ9RnBh+Tb7D4hfeMMp/7BCriMNkrbgAb0+25A+p7Hlbln9c TQf/bFaPcm1RFzsFrKF7O8pf42OzWvo7Qc0F8NLETBO0WCzakEFkW9t/BHqUhKU7L8um pUMVvhi1+GuhvJR0dfp/y27e4FhSDmo5kugSoCOJ8cjSjEN85zhrKwB+bo/eEjJJtNfT vr9lLweOTeCWRX3V128s5CnuQWWcX4w003nNvNwRuFN8OUapkmqd60gNtpnEtdF0Lqnd v1yM/R9tnoOC9zObdpzT32VFIns+pOj+21oPrUpDrACC575oWDeaVJ8kZlpyvSglO8jT n2Fw==
X-Gm-Message-State: ALoCoQk9l5OIcuYgVYRNiz6/YbNy8cNhrcTc7QMLLC9C8GfPqQTt7mcvEZEHuff7f4K8NYcdhPcPIWbGexveA3xFgrAB36VranlZjjyAE2KlfNCiFU6dV1+Iyyubyyg0xxAc56oT2EMbReB8XESV6TsfMM51ZZ+2ZxhJXwD4G+ku9kxyVajh7psrM8pODZCcu0UbTrN0frcF
X-Received: by 10.180.37.193 with SMTP id a1mr555738wik.52.1389769211125; Tue, 14 Jan 2014 23:00:11 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Tue, 14 Jan 2014 22:59:51 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BF98B9D@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CAGzyod7HAjdDHt1B4XRHT6h43DXnoupqGmr4buP6D+bQs86m+w@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF98B9D@EXVMBX020-12.exch020.serverdata.net>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Wed, 15 Jan 2014 15:59:51 +0900
Message-ID: <CAH9hSJZVfNATVm2P6qGh8JAWEWq05Gn+7p81W7kHwqz85brRTA@mail.gmail.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=e89a8f502fa026438804effcdcdc
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 15 Jan 2014 07:00:26 -0000

--e89a8f502fa026438804effcdcdc
Content-Type: text/plain; charset=UTF-8

On Wed, Jan 15, 2014 at 3:34 PM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> >>In any layering scheme, we'd need to use a HEADER frame one way or
> another because we need to at least announce that we're connecting to a WS
> endpoint.
>
> No, the alternative WS/SPDY proposal uses SPDY HEADER only for the initial
> WS opening handshake, but not for the subsequent WS messaging traffic.


I think Roberto meant that we need to have code to parse HEADERS frame
regardless of how layering scheme is. As you say, we'll exchange opening
handshake using HEADERS. I.e. a server implementing your proposal will have
HEADERS parser. So, use of HEADERS itself doesn't add cost.

--e89a8f502fa026438804effcdcdc
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On W=
ed, Jan 15, 2014 at 3:34 PM, Tobias Oberstein <span dir=3D"ltr">&lt;<a href=
=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstein@=
tavendo.de</a>&gt;</span> wrote:<br>


<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div>&gt;&gt;In any layering scheme, we&#39;=
d need to use a HEADER frame one way or another because we need to at least=
 announce that we&#39;re connecting to a WS endpoint.<br>


</div><div>
<br>
</div>No, the alternative WS/SPDY proposal uses SPDY HEADER only for the in=
itial WS opening handshake, but not for the subsequent WS messaging traffic=
.</blockquote><div><br></div><div>I think Roberto meant that we need to hav=
e code to parse HEADERS frame regardless of how layering scheme is. As you =
say, we&#39;ll exchange opening handshake using HEADERS. I.e. a server impl=
ementing your proposal will have HEADERS parser. So, use of HEADERS itself =
doesn&#39;t add cost.<br>


</div></div></div></div>

--e89a8f502fa026438804effcdcdc--

From tobias.oberstein@tavendo.de  Tue Jan 14 23:01:11 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 4E8C21AE305 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 23:01:11 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id kPDN13kOem_t for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 23:01:10 -0800 (PST)
Received: from EXHUB020-1.exch020.serverdata.net (exhub020-1.exch020.serverdata.net [206.225.164.28]) by ietfa.amsl.com (Postfix) with ESMTP id E1DDA1AE2ED for <hybi@ietf.org>; Tue, 14 Jan 2014 23:01:09 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-1.exch020.serverdata.net ([206.225.164.28]) with mapi; Tue, 14 Jan 2014 23:00:58 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Adam Rice <ricea@google.com>
Date: Tue, 14 Jan 2014 23:00:56 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8RrB/pZv3Oy9DrS5GwR4m7iu07IgAEk4Og
Message-ID: <634914A010D0B943A035D226786325D4446BF98BA0@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CAHixhFpVywsZPCx1r_HYXEUQ10WN31NnVYM5XT6-n78uAcrjig@mail.gmail.com>
In-Reply-To: <CAHixhFpVywsZPCx1r_HYXEUQ10WN31NnVYM5XT6-n78uAcrjig@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 15 Jan 2014 07:01:11 -0000
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From tobias.oberstein@tavendo.de  Tue Jan 14 23:07:17 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 8EC471AE307 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 23:07:17 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id DllOoTzXClCa for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 23:07:15 -0800 (PST)
Received: from EXHUB020-3.exch020.serverdata.net (exhub020-3.exch020.serverdata.net [206.225.164.30]) by ietfa.amsl.com (Postfix) with ESMTP id 10DC91AE2F9 for <hybi@ietf.org>; Tue, 14 Jan 2014 23:07:15 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-3.exch020.serverdata.net ([206.225.164.30]) with mapi; Tue, 14 Jan 2014 23:07:03 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Takeshi Yoshino <tyoshino@google.com>
Date: Tue, 14 Jan 2014 23:07:01 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8Rv3ABFUm/UqA+R0SlpL+XDngg8AAADEFw
Message-ID: <634914A010D0B943A035D226786325D4446BF98BA1@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CAGzyod7HAjdDHt1B4XRHT6h43DXnoupqGmr4buP6D+bQs86m+w@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF98B9D@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZVfNATVm2P6qGh8JAWEWq05Gn+7p81W7kHwqz85brRTA@mail.gmail.com>
In-Reply-To: <CAH9hSJZVfNATVm2P6qGh8JAWEWq05Gn+7p81W7kHwqz85brRTA@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 15 Jan 2014 07:07:17 -0000
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From tyoshino@google.com  Tue Jan 14 23:15:25 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 690AF1AE049 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 23:15:25 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.916
X-Spam-Level: 
X-Spam-Status: No, score=-1.916 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id bfXmrRzkPzXY for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 23:15:23 -0800 (PST)
Received: from mail-we0-x22d.google.com (mail-we0-x22d.google.com [IPv6:2a00:1450:400c:c03::22d]) by ietfa.amsl.com (Postfix) with ESMTP id 7B8D51AE2DA for <hybi@ietf.org>; Tue, 14 Jan 2014 23:15:23 -0800 (PST)
Received: by mail-we0-f173.google.com with SMTP id t60so1359056wes.4 for <hybi@ietf.org>; Tue, 14 Jan 2014 23:15:11 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=BHnjg9IrVv3F0JOUXmPFtFv5d6365lXhM2pvyY53rlo=; b=Idp4ZOEhWpR9tHy+4+1Yqp1cExygtNQ5bO836WTUv43ZZLqtmxpl79tmYjRc1h0vrB I9N4WVKkyTjEl+jL6Fh2qj4b1GdAUvaD+LtuNFeO1WO4BINGAd10uP/gCPlOldRnFByK HEoefByv14soUS0PDQNehXc7aMiPR06aAMlCcWsF/z5uWpjagFgk4//ONpbe9oUnQ7DF xIa3OkkwgyaTDsDUbj+S+ddq1EoLzUm0+P2cU0Yq6Bmsc2IjYVsrxWwBWsPX41uWOfsR dKaQ8JALLDcebdxmUb62f3+1sT0vsp8bhyJe1QhwgwTBjiYCSBBoufcbgDaFiLbmPVRW k8Ow==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=BHnjg9IrVv3F0JOUXmPFtFv5d6365lXhM2pvyY53rlo=; b=R4ijv/rAlbYRL+wTSt/zU38obbK+V2ao9NNxRpZj7/ICERCdp4kn6271+8NRQPWNKQ J12a5CsNA36nuca5QEOvFM5Rz6LgPFgihgLYt2P6lpBtyXqnNsIN+x+BeusMaLLQG3qR iYnWDu/5mvZBf7Zr/HoJfaAXQJR/VZhaDsfMOOpln9tzJ0n2VnFQBK7wMLBtc/eHBSGd CCENi4nA00ObUSFMBSnYLHjikQOMK7z6pkZnb9BgNF8lOll9eJVPaIvmEEMft2m584bQ 8gzlPRukFH7iBAJKbcQwrGriWxhbgP80nfiOEEaYhpgc6u5UsF/mdNctp77wt1yXTJmK zTCQ==
X-Gm-Message-State: ALoCoQnGNh1M+M6YxQmZgaMK1L+vpcnXl0EHL48GFW8sVcQzBN8eoqXTT8DELNAmz1w9Jp4bmLD3CsEnfomcwDt+TI4M4KKoe4mwyO2H1WNP7LOU6vsUlbXVlRoxvZRdQgxFeHqhLBehLRUdsdVc1KZtuIoB9XXTmUJj322PU59ONw+d9E/NdYelGNcgBBT7Ru9i6ll4q6My
X-Received: by 10.181.13.11 with SMTP id eu11mr706834wid.30.1389770111417; Tue, 14 Jan 2014 23:15:11 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Tue, 14 Jan 2014 23:14:51 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BF98B9D@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CAGzyod7HAjdDHt1B4XRHT6h43DXnoupqGmr4buP6D+bQs86m+w@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF98B9D@EXVMBX020-12.exch020.serverdata.net>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Wed, 15 Jan 2014 16:14:51 +0900
Message-ID: <CAH9hSJafjpCfv15gv_NrAnFEHA__oL-nzT=yFTczovDsnxZDag@mail.gmail.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=f46d043c7de6cfab5804effd11c7
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 15 Jan 2014 07:15:25 -0000

--f46d043c7de6cfab5804effd11c7
Content-Type: text/plain; charset=UTF-8

On Wed, Jan 15, 2014 at 3:34 PM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> >>The complication here is relative. Using the approach proposed in terms
> of using MSG_DONE, etc. would make deployment at large sites like Google
> substantially easier as large sites typically want to see the semantic
> boundaries easily, quickly, and consistently, as >>there needs to be robust
> mechanisms for handling failover, etc. The more layers of interpretation
> there are when doing this, the bigger the number of problems.
>
> Mmh. The semantic boundary for a WebSocket connection isn't the individual
> WebSocket message, but the whole connection, since message order might have
> significance at application level.
>
> Failing over a WebSocket connection in the middle (being it at WS message
> boundary or not) might break things at app level.
>
> This is different from the HTTP request/response model, where each request
> and response basically stands on it's own.


It depends on how the app corresponds transactions to messages. We're
thinking of deploying L7 DoS protectors for WebSocket that are separate
from backends. In such a case, DoS protectors doesn't fully understand
what's going on at app level but we'll teach them how much cost each
message has and let them drop/throttle connection when there's flood based
on that metric.

Complexity of such infrastructure would be reduced if less framing is used,
comparatively.

--f46d043c7de6cfab5804effd11c7
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On W=
ed, Jan 15, 2014 at 3:34 PM, Tobias Oberstein <span dir=3D"ltr">&lt;<a href=
=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstein@=
tavendo.de</a>&gt;</span> wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-=
left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;p=
adding-left:1ex"><div class=3D"im">&gt;&gt;The complication here is relativ=
e. Using the approach proposed in terms of using MSG_DONE, etc. would make =
deployment at large sites like Google substantially easier as large sites t=
ypically want to see the semantic boundaries easily, quickly, and consisten=
tly, as &gt;&gt;there needs to be robust mechanisms for handling failover, =
etc. The more layers of interpretation there are when doing this, the bigge=
r the number of problems.=C2=A0<br>

</div><div class=3D"im">
<br>
</div>Mmh. The semantic boundary for a WebSocket connection isn&#39;t the i=
ndividual WebSocket message, but the whole connection, since message order =
might have significance at application level.<br>
<br>
Failing over a WebSocket connection in the middle (being it at WS message b=
oundary or not) might break things at app level.<br>
<br>
This is different from the HTTP request/response model, where each request =
and response basically stands on it&#39;s own.</blockquote><div><br></div><=
div>It depends on how the app corresponds transactions to messages. We&#39;=
re thinking of deploying L7 DoS protectors for WebSocket that are separate =
from backends. In such a case, DoS protectors doesn&#39;t fully understand =
what&#39;s going on at app level but we&#39;ll teach them how much cost eac=
h message has and let them drop/throttle connection when there&#39;s flood =
based on that metric.</div>

<div><br></div><div>Complexity of such infrastructure would be reduced if l=
ess framing is used, comparatively.</div></div></div></div>

--f46d043c7de6cfab5804effd11c7--

From tobias.oberstein@tavendo.de  Tue Jan 14 23:30:17 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 29E8D1AE318 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 23:30:17 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id pUPMgFObImik for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 23:30:15 -0800 (PST)
Received: from EXHUB020-1.exch020.serverdata.net (exhub020-1.exch020.serverdata.net [206.225.164.28]) by ietfa.amsl.com (Postfix) with ESMTP id 442FA1AE314 for <hybi@ietf.org>; Tue, 14 Jan 2014 23:30:15 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-1.exch020.serverdata.net ([206.225.164.28]) with mapi; Tue, 14 Jan 2014 23:29:57 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Takeshi Yoshino <tyoshino@google.com>
Date: Tue, 14 Jan 2014 23:29:55 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8RwYibHAUYi62sQi63z3k3lRn81wAAN8Yg
Message-ID: <634914A010D0B943A035D226786325D4446BF98BA3@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CAGzyod7HAjdDHt1B4XRHT6h43DXnoupqGmr4buP6D+bQs86m+w@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF98B9D@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJafjpCfv15gv_NrAnFEHA__oL-nzT=yFTczovDsnxZDag@mail.gmail.com>
In-Reply-To: <CAH9hSJafjpCfv15gv_NrAnFEHA__oL-nzT=yFTczovDsnxZDag@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 15 Jan 2014 07:30:17 -0000
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From tyoshino@google.com  Tue Jan 14 23:42:20 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 4D0511AE325 for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 23:42:20 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.316
X-Spam-Level: 
X-Spam-Status: No, score=-1.316 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, J_CHICKENPOX_52=0.6, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 2zlxP8dCzwCi for <hybi@ietfa.amsl.com>; Tue, 14 Jan 2014 23:42:19 -0800 (PST)
Received: from mail-wg0-x230.google.com (mail-wg0-x230.google.com [IPv6:2a00:1450:400c:c00::230]) by ietfa.amsl.com (Postfix) with ESMTP id DA0F21AE323 for <hybi@ietf.org>; Tue, 14 Jan 2014 23:42:18 -0800 (PST)
Received: by mail-wg0-f48.google.com with SMTP id x13so1396767wgg.27 for <hybi@ietf.org>; Tue, 14 Jan 2014 23:42:06 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=pH8SNGX5FaH42gWYKYHCVMADQQe1+HELUf6dDOqArhc=; b=R+0/q28P/RTyIShVK/d5e5xQABookAe89xnLpKs/lG6gVUKTeEiJWd6SRhFsDB5Heg euNcewbNP7HNac4QWbtnvqsc1jYMvELygWDBfD15ErWHQahtHS+oK/Im/og+G2iIwAjA hAEz0e7VWxvQgwyJXCjgYTzTeWGsm170/lv3pO7+P/OTLVGHwbtCQmuyrpuYDWkydXW1 NtGvbxUJGaXT+qvYJLyC4cW29dnOpwgNnHx7GXtThHlIjXAzUfBloGi4TflUgze71UuQ 8mQLP3N7VYIvN1mSODySs96PflS0dYSdA8uYHqZ926gzsn5RoHOjJKBnYEPTzzN2ijVs goMA==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=pH8SNGX5FaH42gWYKYHCVMADQQe1+HELUf6dDOqArhc=; b=hQDcl9KVQsSsMh010sApN09gZE1UpP4t4xblMvNI+Emia+Gd4I1Ox1aD+oC8shrzB1 95UtIYR/y6tuGH9qFEFOyMHyyXJFLKNOlIBOWdH4TPo+R7gupi2xLEYpBylr9RhMRubc LFBKAe8DDz6FicujDG30dRUiHv9w07BdRA5tKA+8X9IWM5AWIF+FWK3+NsQZFLnD+CQU mtC9jBTB8wexTA1TZQLv+FYgNag57VPe1pYHXZIfyXrgw40fzxlat/BwsspXn+DrS+jj FlXbci5auCRZLPwSdVJ8r7D4xCTG+nL5bAfY9xXV218jB4BKj/fyvEkLdqpCdJZpNJ7z lwCA==
X-Gm-Message-State: ALoCoQm1Nwb6bCFz/vcTE1CFAyGyvmQPbJD9TaHQrLn3LJ5fzkcZKDWw5x9EqcYxnAXBRPwHTo0G6XYvml9EVRPUA9izLDjpYGgYcjpbFdul5oLNtPw6by9GwkbcyzHLyLwfPFwOBRRLSQ50N0dm4MS29VDTI4Nxde72rIHy1SKL21RDJSeGL3D8k7HHaNDG0zolrCPJmVpP
X-Received: by 10.194.108.100 with SMTP id hj4mr143921wjb.83.1389771726635; Tue, 14 Jan 2014 23:42:06 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Tue, 14 Jan 2014 23:41:46 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BF98BA1@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CAGzyod7HAjdDHt1B4XRHT6h43DXnoupqGmr4buP6D+bQs86m+w@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF98B9D@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZVfNATVm2P6qGh8JAWEWq05Gn+7p81W7kHwqz85brRTA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF98BA1@EXVMBX020-12.exch020.serverdata.net>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Wed, 15 Jan 2014 16:41:46 +0900
Message-ID: <CAH9hSJapTnYmrs8Q=ojhK9X63Ceo3cUM4gonvEvwJgK-yTf-tw@mail.gmail.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=047d7bf109fe15e3fb04effd726f
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 15 Jan 2014 07:42:20 -0000

--047d7bf109fe15e3fb04effd726f
Content-Type: text/plain; charset=UTF-8

On Wed, Jan 15, 2014 at 4:07 PM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> >>>In any layering scheme, we'd need to use a HEADER frame one way or
> another because we need to at least announce that we're connecting to a WS
> endpoint.
>
> >>No, the alternative WS/SPDY proposal uses SPDY HEADER only for the
> initial WS opening handshake, but not for the subsequent WS messaging
> traffic.
>
> >I think Roberto meant that we need to have code to parse HEADERS frame
> regardless of how layering scheme is. As you say, we'll exchange opening
> handshake using HEADERS. I.e. a server implementing your proposal will have
> HEADERS parser. So, use of HEADERS >itself doesn't add cost.
>
> Yes, SPDY HEADER parsing code needs to be there in both, but:
>
> The implementation cost for the current WS/SPDY proposal is for WS
> implementations that want to both support native WS and WS/SPDY:
>
> the WS message data processing code needs to be intermingled with SPDY
> code.
>
> With the alternative proposal (after the WS opening handshake is
> finished), the WS code can stay essentially untouched: the bytes just not
> come from wire anymore, but from SPDY data frame payloads.
>

Right. Inter-layer coordination complicates things given one has already
implemented RFC6455. I understand your argument about clear layering.

It seems that there're different visions of how future transport for
WebSocket API should be. IIUC, Roberto is trying to make HTTP/2.0 versatile
transport for the web. He's thinking it's costly to carry over RFC6455 to
that future. In Google, as you know, we're deploying SPDY widely and
therefore SPDY is always there. If HTTP/2.0 is success, it'll be true also
for others.

--047d7bf109fe15e3fb04effd726f
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On W=
ed, Jan 15, 2014 at 4:07 PM, Tobias Oberstein <span dir=3D"ltr">&lt;<a href=
=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstein@=
tavendo.de</a>&gt;</span> wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div class=3D"im">&gt;&gt;&gt;In any layerin=
g scheme, we&#39;d need to use a HEADER frame one way or another because we=
 need to at least announce that we&#39;re connecting to a WS endpoint.<br>


<br>
&gt;&gt;No, the alternative WS/SPDY proposal uses SPDY HEADER only for the =
initial WS opening handshake, but not for the subsequent WS messaging traff=
ic.<br>
<br>
&gt;I think Roberto meant that we need to have code to parse HEADERS frame =
regardless of how layering scheme is. As you say, we&#39;ll exchange openin=
g handshake using HEADERS. I.e. a server implementing your proposal will ha=
ve HEADERS parser. So, use of HEADERS &gt;itself doesn&#39;t add cost.<br>


<br>
</div>Yes, SPDY HEADER parsing code needs to be there in both, but:<br>
<br>
The implementation cost for the current WS/SPDY proposal is for WS implemen=
tations that want to both support native WS and WS/SPDY:<br>
<br>
the WS message data processing code needs to be intermingled with SPDY code=
.<br>
<br>
With the alternative proposal (after the WS opening handshake is finished),=
 the WS code can stay essentially untouched: the bytes just not come from w=
ire anymore, but from SPDY data frame payloads.<br></blockquote><div><br>

</div><div>Right. Inter-layer coordination complicates things given one has=
 already implemented RFC6455. I understand your argument about clear layeri=
ng.</div><div><br></div><div>It seems that there&#39;re different visions o=
f how future transport for WebSocket API should be. IIUC, Roberto is trying=
 to make HTTP/2.0 versatile transport for the web. He&#39;s thinking it&#39=
;s costly to carry over RFC6455 to that future. In Google, as you know, we&=
#39;re deploying SPDY widely and therefore SPDY is always there. If HTTP/2.=
0 is success, it&#39;ll be true also for others.</div>

</div></div></div>

--047d7bf109fe15e3fb04effd726f--

From tobias.oberstein@tavendo.de  Wed Jan 15 08:01:25 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id BAF5A1AE3A7 for <hybi@ietfa.amsl.com>; Wed, 15 Jan 2014 08:01:25 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.3
X-Spam-Level: 
X-Spam-Status: No, score=-1.3 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, J_CHICKENPOX_36=0.6, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id czrRdN1KNTk1 for <hybi@ietfa.amsl.com>; Wed, 15 Jan 2014 08:01:24 -0800 (PST)
Received: from EXHUB020-1.exch020.serverdata.net (exhub020-1.exch020.serverdata.net [206.225.164.28]) by ietfa.amsl.com (Postfix) with ESMTP id 2FE501AE382 for <hybi@ietf.org>; Wed, 15 Jan 2014 08:01:24 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-1.exch020.serverdata.net ([206.225.164.28]) with mapi; Wed, 15 Jan 2014 08:01:12 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Adam Rice <ricea@google.com>
Date: Wed, 15 Jan 2014 08:01:10 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8RrB/pZv3Oy9DrS5GwR4m7iu07IgAXbPYg
Message-ID: <634914A010D0B943A035D226786325D4446BF98D0A@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CAHixhFpVywsZPCx1r_HYXEUQ10WN31NnVYM5XT6-n78uAcrjig@mail.gmail.com>
In-Reply-To: <CAHixhFpVywsZPCx1r_HYXEUQ10WN31NnVYM5XT6-n78uAcrjig@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 15 Jan 2014 16:01:25 -0000
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From joakim@intalio.com  Wed Jan 15 09:08:07 2014
Return-Path: <joakim@intalio.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id B48F41AE13A for <hybi@ietfa.amsl.com>; Wed, 15 Jan 2014 09:08:07 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.377
X-Spam-Level: 
X-Spam-Status: No, score=-1.377 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, J_CHICKENPOX_36=0.6, RCVD_IN_DNSWL_LOW=-0.7] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id S45JQvvVbSKJ for <hybi@ietfa.amsl.com>; Wed, 15 Jan 2014 09:08:06 -0800 (PST)
Received: from mail-ee0-f47.google.com (mail-ee0-f47.google.com [74.125.83.47]) by ietfa.amsl.com (Postfix) with ESMTP id BA0361AE0FC for <hybi@ietf.org>; Wed, 15 Jan 2014 09:08:05 -0800 (PST)
Received: by mail-ee0-f47.google.com with SMTP id d49so249282eek.20 for <hybi@ietf.org>; Wed, 15 Jan 2014 09:07:53 -0800 (PST)
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=q46X9rDfokmxZ0u3ERRpld06E++0io7oP9ozZl8WSFM=; b=Z9doZTng/7JdyzZZMEwzmVesaKlVVN3UzMMKwqW/5sfH6y4GjgXpSdrewErudLr03V sbdaiJzPpLuQn1gqv5ot0Zi5evglRC4fViOvJcS7JMx2FZJ7Y9nKmqtfzfVl6kDS1Pn6 qVhRwt9TKXTqrkgliZVD1zycnGrm0WTCeZpc0YUb1rwbV9AMZErzJ8nHkJrlSUB0y/c7 QQAjyi8YbZkvXL01aOVVIFvQmUOFXi6hipDKiHuDoAL+475qx1bVFw5LlV8pGrY+chYY RWafZM+OL85pUO4Mmx9VoNoHnsu6vfvn3wfV6gjer+xqSUwJhx/ngTDdWQbwhVmFG1la sbMw==
X-Gm-Message-State: ALoCoQmvmheaFiDgbUhTmKvnusMe9bqPVnnXsWlhZ0risXqjlquZasuKt++ttmxIoA1RxPtq9U4y
MIME-Version: 1.0
X-Received: by 10.15.32.73 with SMTP id z49mr5211643eeu.27.1389805673422; Wed, 15 Jan 2014 09:07:53 -0800 (PST)
Received: by 10.14.208.131 with HTTP; Wed, 15 Jan 2014 09:07:53 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BF98D0A@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CAHixhFpVywsZPCx1r_HYXEUQ10WN31NnVYM5XT6-n78uAcrjig@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF98D0A@EXVMBX020-12.exch020.serverdata.net>
Date: Wed, 15 Jan 2014 10:07:53 -0700
Message-ID: <CAG4zZZCemh=8zOg6zuvaSbX5iPMHzq7jTKHg_0VtyTA_q1QaYA@mail.gmail.com>
From: Joakim Erdfelt <joakim@intalio.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=089e0160c5cc78b30704f0055912
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 15 Jan 2014 17:08:07 -0000

--089e0160c5cc78b30704f0055912
Content-Type: text/plain; charset=ISO-8859-1

One minor niggle from Jetty on this (we are lurking and reading)

Knowing the size of a frame at the start of the frame, traditional
websocket behavior, is desired, as it allows us to fast fail for the 1009
(message too big) use case.
Having no prior knowledge of frame size is undesirable to us.
This is inline with what Tobias is asking for too.


--
Joakim Erdfelt <joakim@intalio.com>
webtide.com <http://www.webtide.com/> - intalio.com/jetty
Expert advice, services and support from from the Jetty & CometD experts
eclipse.org/jetty - cometd.org


On Wed, Jan 15, 2014 at 9:01 AM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> >>On 15 January 2014 04:09, Tobias Oberstein <tobias.oberstein@tavendo.de>
> wrote:
>
> >> * has significant wire-level overhead compared to native WS (and
> compared to the proposal below)
>
> >If you accept the argument that unencrypted WebSockets are unusable on
> heterogeneous networks due to interference from middleboxes, then the byte
> overhead is a red-herring. Everything has to be encrypted, and so will get
> padded out to the cipher block size >anyway.
>
> I did more readup on this one .. e.g. not all modes require padding
> (AES-GCM does not) - nevertheless, you are totally right: the overhead of
> TLS can be substantial.
>
> Here are measurements for Chrome, Firefox and IE11:
>
> http://tavendo.com/blog/post/dissecting-websocket-overhead/
>
> The overhead induced by TCP/IP and TLS in particular can dwarf the
> overhead of WebSocket itself.
>
> Thank you very much for pointing into this direction .. this is important.
>
> As an immediate consequence, we will introduce message batching into WAMP
> (a WebSocket subprotocol for RPC+PubSub -
> https://github.com/tavendo/WAMP/tree/master/spec )
>
> /Tobias
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>

--089e0160c5cc78b30704f0055912
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">One minor niggle from Jetty on this (we are lurking and re=
ading)<div><br></div><div>Knowing the size of a frame at the start of the f=
rame, traditional websocket behavior, is desired, as it allows us to fast f=
ail for the 1009 (message too big) use case.</div>
<div>Having no prior knowledge of frame size is undesirable to us.</div><di=
v>This is inline with what Tobias is asking for too.</div><div><br></div></=
div><div class=3D"gmail_extra"><br clear=3D"all"><div><div dir=3D"ltr"><div=
>--</div>
<div>Joakim Erdfelt &lt;<a href=3D"mailto:joakim@intalio.com" target=3D"_bl=
ank">joakim@intalio.com</a>&gt;</div><div><a href=3D"http://www.webtide.com=
/" target=3D"_blank">webtide.com</a>=A0- <a href=3D"http://intalio.com/jett=
y" target=3D"_blank">intalio.com/jetty</a></div>
<div><span>Expert advice, services and support from=A0</span><span>from the=
 Jetty &amp; CometD experts</span></div><div><a href=3D"http://eclipse.org/=
jetty/" target=3D"_blank">eclipse.org/jetty</a>=A0-=A0<a href=3D"http://com=
etd.org/" target=3D"_blank">cometd.org</a></div>
</div></div>
<br><br><div class=3D"gmail_quote">On Wed, Jan 15, 2014 at 9:01 AM, Tobias =
Oberstein <span dir=3D"ltr">&lt;<a href=3D"mailto:tobias.oberstein@tavendo.=
de" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt;</span> wrote:<br>=
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">
<div class=3D"im">&gt;&gt;On 15 January 2014 04:09, Tobias Oberstein &lt;<a=
 href=3D"mailto:tobias.oberstein@tavendo.de">tobias.oberstein@tavendo.de</a=
>&gt; wrote:<br>
<br>
&gt;&gt;=A0* has significant wire-level overhead compared to native WS (and=
 compared to the proposal below)<br>
<br>
&gt;If you accept the argument that unencrypted WebSockets are unusable on =
heterogeneous networks due to interference from middleboxes, then the byte =
overhead is a red-herring. Everything has to be encrypted, and so will get =
padded out to the cipher block size &gt;anyway.<br>

<br>
</div>I did more readup on this one .. e.g. not all modes require padding (=
AES-GCM does not) - nevertheless, you are totally right: the overhead of TL=
S can be substantial.<br>
<br>
Here are measurements for Chrome, Firefox and IE11:<br>
<br>
<a href=3D"http://tavendo.com/blog/post/dissecting-websocket-overhead/" tar=
get=3D"_blank">http://tavendo.com/blog/post/dissecting-websocket-overhead/<=
/a><br>
<br>
The overhead induced by TCP/IP and TLS in particular can dwarf the overhead=
 of WebSocket itself.<br>
<br>
Thank you very much for pointing into this direction .. this is important.<=
br>
<br>
As an immediate consequence, we will introduce message batching into WAMP (=
a WebSocket subprotocol for RPC+PubSub - <a href=3D"https://github.com/tave=
ndo/WAMP/tree/master/spec" target=3D"_blank">https://github.com/tavendo/WAM=
P/tree/master/spec</a> )<br>

<span class=3D"HOEnZb"><font color=3D"#888888"><br>
/Tobias<br>
</font></span><div class=3D"HOEnZb"><div class=3D"h5">_____________________=
__________________________<br>
hybi mailing list<br>
<a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br>
<a href=3D"https://www.ietf.org/mailman/listinfo/hybi" target=3D"_blank">ht=
tps://www.ietf.org/mailman/listinfo/hybi</a><br>
</div></div></blockquote></div><br></div>

--089e0160c5cc78b30704f0055912--

From webmaster@zaphoyd.com  Wed Jan 15 19:33:20 2014
Return-Path: <webmaster@zaphoyd.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id D8CEE1AE1D8 for <hybi@ietfa.amsl.com>; Wed, 15 Jan 2014 19:33:20 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.699
X-Spam-Level: 
X-Spam-Status: No, score=-0.699 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, J_CHICKENPOX_52=0.6, J_CHICKENPOX_64=0.6, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id zi62krQJYizN for <hybi@ietfa.amsl.com>; Wed, 15 Jan 2014 19:33:18 -0800 (PST)
Received: from authsmtp01.uchicago.edu (authsmtp01.uchicago.edu [128.135.12.121]) by ietfa.amsl.com (Postfix) with ESMTP id 7E7B61AE1CD for <hybi@ietf.org>; Wed, 15 Jan 2014 19:33:18 -0800 (PST)
Received: from [10.0.1.3] (c-68-51-76-178.hsd1.il.comcast.net [68.51.76.178]) (Authenticated sender: zaphoyd) by authsmtp01.uchicago.edu (Postfix) with ESMTP id F13A149809E; Wed, 15 Jan 2014 21:33:05 -0600 (CST)
Content-Type: multipart/alternative; boundary="Apple-Mail=_292F86ED-3727-4EA6-BAFE-69EC6D1C3F03"
Mime-Version: 1.0 (Mac OS X Mail 7.0 \(1822\))
From: Peter Thorson <webmaster@zaphoyd.com>
In-Reply-To: <CAH9hSJapTnYmrs8Q=ojhK9X63Ceo3cUM4gonvEvwJgK-yTf-tw@mail.gmail.com>
Date: Wed, 15 Jan 2014 21:33:08 -0600
Message-Id: <DEB610E7-28AE-4AC0-B126-24DA4A99EE63@zaphoyd.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CAGzyod7HAjdDHt1B4XRHT6h43DXnoupqGmr4buP6D+bQs86m+w@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF98B9D@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZVfNATVm2P6qGh8JAWEWq05Gn+7p81W7kHwqz85brRTA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF98BA1@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJapTnYmrs8Q=ojhK9X63Ceo3cUM4gonvEvwJgK-yTf-tw@mail.gmail.com>
To: Takeshi Yoshino <tyoshino@google.com>
X-Mailer: Apple Mail (2.1822)
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 16 Jan 2014 03:33:21 -0000

--Apple-Mail=_292F86ED-3727-4EA6-BAFE-69EC6D1C3F03
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=windows-1252

If RFC6455 semantics are to be preserved the only way that it makes =
sense to do that is in a fully tunneled method such that the resulting =
stream could be fed to an existing RFC6455 infrastructure. While =
conceptually simple, as mentioned a number of times in the past threads, =
this introduces a number of problems including significant overhead and =
reduced transparency to HTTP 2.0 intermediaries. I don=92t see any great =
benefit in this method.

I also don=92t see any benefit in doing a half and half style protocol =
where RFC6455 frames are encapsulated in HTTP 2.0 header+data frame =
pairs. This seems like the worst of both worlds. Not enough like RFC6455 =
to actually re-use existing infrastructure but enough like it to make =
the messaging component not mesh well with the rest of HTTP 2.0. The =
alternate method proposed earlier where HTTP 2.0 simply defines its own =
=93ordered bidirectional message stream=94 semantics independently of =
RFC6455 either via a specific frame type or via additional flags on the =
data frame seems to be an obvious choice to me.

HTTP 2.0 natively provides all of the core facilities that RFC6455 does: =
connection establishment, header flags, control messaging, close =
messaging, framing, etc. There is no need to redefine any of this in =
terms of RFC6455 opcodes or frame bits order to exchange messages =
bidirectionally over the existing connection between an HTTP 2.0 browser =
and server. Doing so only adds complexity for no real benefit. At best I =
can see including a method for mapping RFC6455 close codes to RST_STREAM =
error codes.

Using HTTP 2.0 for this sort of messaging would be less efficient than =
raw RFC6455 but likely more efficient overall for sites with a healthy =
mix of HTTP and WebSocket traffic to the same host. It would also solve =
any issues about unknown frame boundaries, make all messaging traffic a =
first class HTTP 2.0 citizen and understandable to intermediaries. It =
would not require HTTP 2.0 only programs to also have an RFC6455 parser. =
It would remove the need to try and fit any extensions designed for =
RFC6455 into HTTP 2.0 or restrict RFC6455 extension design to things =
that could fit into HTTP 2.0. There doesn=92t seem to be anything =
supported by the JavaScript WebSocket API that couldn=92t be easily =
transported using such a scheme.

That said, because of the differences between HTTP 2.0 and RFC6455 =
WebSockets I would hope to see them coexist in browsers rather than be a =
replacement. There are a number of cases that users of my WebSocket =
library have discussed with me where HTTP 2.0 is vast overkill and =
needlessly complex for what they need and RFC6455 fits the bill =
perfectly.

On the subject of WS MUX, given the existence of a protocol like HTTP =
2.0 with the above messaging feature, based on the experiences of my =
library users and clients, any use case for a muxed WebSocket-type =
messaging system would be satisfied by HTTP 2.0.


On Jan 15, 2014, at 1:41, Takeshi Yoshino <tyoshino@google.com> wrote:

> On Wed, Jan 15, 2014 at 4:07 PM, Tobias Oberstein =
<tobias.oberstein@tavendo.de> wrote:
> >>>In any layering scheme, we'd need to use a HEADER frame one way or =
another because we need to at least announce that we're connecting to a =
WS endpoint.
>=20
> >>No, the alternative WS/SPDY proposal uses SPDY HEADER only for the =
initial WS opening handshake, but not for the subsequent WS messaging =
traffic.
>=20
> >I think Roberto meant that we need to have code to parse HEADERS =
frame regardless of how layering scheme is. As you say, we'll exchange =
opening handshake using HEADERS. I.e. a server implementing your =
proposal will have HEADERS parser. So, use of HEADERS >itself doesn't =
add cost.
>=20
> Yes, SPDY HEADER parsing code needs to be there in both, but:
>=20
> The implementation cost for the current WS/SPDY proposal is for WS =
implementations that want to both support native WS and WS/SPDY:
>=20
> the WS message data processing code needs to be intermingled with SPDY =
code.
>=20
> With the alternative proposal (after the WS opening handshake is =
finished), the WS code can stay essentially untouched: the bytes just =
not come from wire anymore, but from SPDY data frame payloads.
>=20
> Right. Inter-layer coordination complicates things given one has =
already implemented RFC6455. I understand your argument about clear =
layering.
>=20
> It seems that there're different visions of how future transport for =
WebSocket API should be. IIUC, Roberto is trying to make HTTP/2.0 =
versatile transport for the web. He's thinking it's costly to carry over =
RFC6455 to that future. In Google, as you know, we're deploying SPDY =
widely and therefore SPDY is always there. If HTTP/2.0 is success, it'll =
be true also for others.
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi


--Apple-Mail=_292F86ED-3727-4EA6-BAFE-69EC6D1C3F03
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
	charset=windows-1252

<html><head><meta http-equiv=3D"Content-Type" content=3D"text/html =
charset=3Dwindows-1252"><meta http-equiv=3D"Content-Type" =
content=3D"text/html charset=3Dwindows-1252"></head><body =
style=3D"word-wrap: break-word; -webkit-nbsp-mode: space; =
-webkit-line-break: after-white-space;"><div><div style=3D"margin: =
0px;">If RFC6455 semantics are to be preserved the only way that it =
makes sense to do that is in a fully tunneled method such that the =
resulting stream could be fed to an existing RFC6455 infrastructure. =
While conceptually simple, as mentioned a number of times in the past =
threads, this introduces a number of problems including significant =
overhead and reduced transparency to HTTP 2.0 intermediaries. I don=92t =
see any great benefit in this method.</div><div style=3D"margin: 0px; =
min-height: 14px;"><br></div><div style=3D"margin: 0px;">I also don=92t =
see any benefit in doing a half and half style protocol where RFC6455 =
frames are encapsulated in HTTP 2.0 header+data frame pairs. This seems =
like the worst of both worlds. Not enough like RFC6455 to actually =
re-use existing infrastructure but enough like it to make the messaging =
component not mesh well with the rest of HTTP 2.0. The alternate method =
proposed earlier where HTTP 2.0 simply defines its own =93ordered =
bidirectional message stream=94 semantics independently of RFC6455 =
either via a specific frame type or via additional flags on the data =
frame seems to be an obvious choice to me.</div><div style=3D"margin: =
0px; min-height: 14px;"><br></div><div style=3D"margin: 0px;">HTTP 2.0 =
natively provides all of the core facilities that RFC6455 does: =
connection establishment, header flags, control messaging, close =
messaging, framing, etc. There is no need to redefine any of this in =
terms of RFC6455 opcodes or frame bits order to exchange messages =
bidirectionally over the existing connection between an HTTP 2.0 browser =
and server. Doing so only adds complexity for no real benefit. At best I =
can see including a method for mapping RFC6455 close codes to RST_STREAM =
error codes.</div><div style=3D"margin: 0px; min-height: =
14px;"><br></div><div style=3D"margin: 0px;">Using HTTP 2.0 for this =
sort of messaging would be less efficient than raw RFC6455 but likely =
more efficient overall for sites with a healthy mix of HTTP and =
WebSocket traffic to the same host. It would also solve any issues about =
unknown frame boundaries, make all messaging traffic a first class HTTP =
2.0 citizen and understandable to intermediaries. It would not require =
HTTP 2.0 only programs to also have an RFC6455 parser. It would remove =
the need to try and fit any extensions designed for RFC6455 into HTTP =
2.0 or restrict RFC6455 extension design to things that could fit into =
HTTP 2.0. There doesn=92t seem to be anything supported by the =
JavaScript WebSocket API that couldn=92t be easily transported using =
such a scheme.</div><div style=3D"margin: 0px; min-height: =
14px;"><br></div><div style=3D"margin: 0px;">That said, because of the =
differences between HTTP 2.0 and RFC6455 WebSockets I would hope to see =
them coexist in browsers rather than be a replacement. There are a =
number of cases that users of my WebSocket library have discussed with =
me where HTTP 2.0 is vast overkill and needlessly complex for what they =
need and RFC6455 fits the bill perfectly.</div><div style=3D"margin: =
0px; min-height: 14px;"><br></div><div style=3D"margin: 0px;">On the =
subject of WS MUX, given the existence of a protocol like HTTP 2.0 with =
the above messaging feature, based on the experiences of my library =
users and clients, any use case for a muxed WebSocket-type messaging =
system would be satisfied by HTTP =
2.0.</div></div><div><br></div><br><div><div>On Jan 15, 2014, at 1:41, =
Takeshi Yoshino &lt;<a =
href=3D"mailto:tyoshino@google.com">tyoshino@google.com</a>&gt; =
wrote:</div><br class=3D"Apple-interchange-newline"><blockquote =
type=3D"cite"><div dir=3D"ltr"><div class=3D"gmail_extra"><div =
class=3D"gmail_quote">On Wed, Jan 15, 2014 at 4:07 PM, Tobias Oberstein =
<span dir=3D"ltr">&lt;<a href=3D"mailto:tobias.oberstein@tavendo.de" =
target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt;</span> wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 =
.8ex;border-left:1px #ccc solid;padding-left:1ex"><div =
class=3D"im">&gt;&gt;&gt;In any layering scheme, we'd need to use a =
HEADER frame one way or another because we need to at least announce =
that we're connecting to a WS endpoint.<br>


<br>
&gt;&gt;No, the alternative WS/SPDY proposal uses SPDY HEADER only for =
the initial WS opening handshake, but not for the subsequent WS =
messaging traffic.<br>
<br>
&gt;I think Roberto meant that we need to have code to parse HEADERS =
frame regardless of how layering scheme is. As you say, we'll exchange =
opening handshake using HEADERS. I.e. a server implementing your =
proposal will have HEADERS parser. So, use of HEADERS &gt;itself doesn't =
add cost.<br>


<br>
</div>Yes, SPDY HEADER parsing code needs to be there in both, but:<br>
<br>
The implementation cost for the current WS/SPDY proposal is for WS =
implementations that want to both support native WS and WS/SPDY:<br>
<br>
the WS message data processing code needs to be intermingled with SPDY =
code.<br>
<br>
With the alternative proposal (after the WS opening handshake is =
finished), the WS code can stay essentially untouched: the bytes just =
not come from wire anymore, but from SPDY data frame =
payloads.<br></blockquote><div><br>

</div><div>Right. Inter-layer coordination complicates things given one =
has already implemented RFC6455. I understand your argument about clear =
layering.</div><div><br></div><div>It seems that there're different =
visions of how future transport for WebSocket API should be. IIUC, =
Roberto is trying to make HTTP/2.0 versatile transport for the web. He's =
thinking it's costly to carry over RFC6455 to that future. In Google, as =
you know, we're deploying SPDY widely and therefore SPDY is always =
there. If HTTP/2.0 is success, it'll be true also for others.</div>

</div></div></div>
_______________________________________________<br>hybi mailing =
list<br><a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br><a =
href=3D"https://www.ietf.org/mailman/listinfo/hybi">https://www.ietf.org/m=
ailman/listinfo/hybi</a><br></blockquote></div><br></body></html>=

--Apple-Mail=_292F86ED-3727-4EA6-BAFE-69EC6D1C3F03--

From yhirano@google.com  Wed Jan 15 21:13:27 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id AAE161AE1CA for <hybi@ietfa.amsl.com>; Wed, 15 Jan 2014 21:13:27 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.716
X-Spam-Level: 
X-Spam-Status: No, score=-0.716 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, J_CHICKENPOX_52=0.6, J_CHICKENPOX_64=0.6, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id yP4ES5KzH8nS for <hybi@ietfa.amsl.com>; Wed, 15 Jan 2014 21:13:25 -0800 (PST)
Received: from mail-qe0-x22e.google.com (mail-qe0-x22e.google.com [IPv6:2607:f8b0:400d:c02::22e]) by ietfa.amsl.com (Postfix) with ESMTP id 461151AE1BE for <hybi@ietf.org>; Wed, 15 Jan 2014 21:13:25 -0800 (PST)
Received: by mail-qe0-f46.google.com with SMTP id 8so2047863qea.5 for <hybi@ietf.org>; Wed, 15 Jan 2014 21:13:13 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=vk+0a1Zo0VrksEzFNRvvtNrCpM3o3ZWbKy2muveWOGU=; b=YOdohrI6ILp5ApwMyM05QXLz3FR3XjkqAlIsy9GyxZThITSb3NI0Ia2VnvamcURjtb IHH1asYpH4ZfYMKmm75/RvvWdP40jUkscm8FemQ5dlTCCU1YRL5pzxFvkDXu2BzXkEel iy7owC+A8PzkMtFPLnkEbV75Abl2rJZkMBnVcWsjE1Ov8qPZpGJt/GpNR4bzraQtQj/H k44aIEoCe2ee+efh4i0rmtBCvBa3eT6S5b0VRENnX7/AU0DxoSR8+DzFrwG+bRHRUw66 9eRoMOpaqJptm1mTjBnqbqWVJR3Qf9aQ7Oo5P3kN2kExMjC1ipQbHsEV761smHFBeKbb zfzA==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=vk+0a1Zo0VrksEzFNRvvtNrCpM3o3ZWbKy2muveWOGU=; b=Lx+LVRIVK3KkzYaHXGBUL7J/nb6qTtCykQx2LtV5BHAuJ1iNbchN6RWaqmls/Ws5Bs 92jnlHtmPnV2oJKi3FegxDKLgfoqUbxin53fNrI05PVmxz4PbHQk4V7cSSk3viGdvG2d tqNtRbaoqebULoQC/GY9iJqS6dfdN/BLf/nFo=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=vk+0a1Zo0VrksEzFNRvvtNrCpM3o3ZWbKy2muveWOGU=; b=Q5W6nGKCqQgCNBBJwiH6hUqjVYI9OaXVgn67/gAqsK66yPjLwrp+VGbmZ6CX100lY2 1XV3KT1297h+8hibJe7UOAArfpXS+SmWZV44nEG7IEYtTrNyAFCwSGc5jRfRUTxdvDRt dusC7irjCbCYTTAg2AumUPgATFQFozbf1IWMSXeZXIVK/xPqycbwpxsK4cfOWwSLWcEp E0cWp6Ao2YbIbhYnhesEIvvuRGq2dwGayUUT2NM4ZGm/SMafkEyDDowjwosNPWK6EJOG WSgUiLUaXgkdTa4qwDcNFhXt0vzjB2r1lJBf4mkqfISJ+bvxNz5WJeA6vIr2QcTpqjy2 jCvQ==
X-Gm-Message-State: ALoCoQk9ElKQvm1OWLVxcM88A/dCVtvqHOLzf8ASHXanhxpHUR//VbaKI6QacFnNQ9+MEq2iifzMN7bl90g0hm+KEwWGfKap1yC/tpMLwW1YgEM0XChgCaMWbNB3wZCtbe39bztpasdNq6ifut6/8apkp81S/4brxUFSNx2cpHqaQNAaY/AZNDWPXJRlgEoGeR9B01EH2U3c
MIME-Version: 1.0
X-Received: by 10.224.128.72 with SMTP id j8mr11937577qas.51.1389849193047; Wed, 15 Jan 2014 21:13:13 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.140.109.202 with HTTP; Wed, 15 Jan 2014 21:13:12 -0800 (PST)
In-Reply-To: <DEB610E7-28AE-4AC0-B126-24DA4A99EE63@zaphoyd.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CAGzyod7HAjdDHt1B4XRHT6h43DXnoupqGmr4buP6D+bQs86m+w@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF98B9D@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZVfNATVm2P6qGh8JAWEWq05Gn+7p81W7kHwqz85brRTA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF98BA1@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJapTnYmrs8Q=ojhK9X63Ceo3cUM4gonvEvwJgK-yTf-tw@mail.gmail.com> <DEB610E7-28AE-4AC0-B126-24DA4A99EE63@zaphoyd.com>
Date: Thu, 16 Jan 2014 14:13:12 +0900
X-Google-Sender-Auth: VTOcm267WFql6YHqAmdkm5hhA3s
Message-ID: <CABihn6EayGfJaDX0TYBPvXzghQr4D=wBiaNoGkGfrfnxyyt-LA@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: Peter Thorson <webmaster@zaphoyd.com>
Content-Type: multipart/alternative; boundary=001a11c1efa67177fa04f00f7b23
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 16 Jan 2014 05:13:27 -0000

--001a11c1efa67177fa04f00f7b23
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Peter,

Is the method you prefer "Plan B: Frame by frame mapping" in [1]?
If not, can you provide some pointers?

Thanks,

[1]
https://docs.google.com/a/google.com/document/d/1Dwd23OXidSRA0DupeWujYz2iPy=
VzvZT6RmEeg4vPY28/edit


On Thu, Jan 16, 2014 at 12:33 PM, Peter Thorson <webmaster@zaphoyd.com>wrot=
e:

> If RFC6455 semantics are to be preserved the only way that it makes sense
> to do that is in a fully tunneled method such that the resulting stream
> could be fed to an existing RFC6455 infrastructure. While conceptually
> simple, as mentioned a number of times in the past threads, this introduc=
es
> a number of problems including significant overhead and reduced
> transparency to HTTP 2.0 intermediaries. I don=E2=80=99t see any great be=
nefit in
> this method.
>
> I also don=E2=80=99t see any benefit in doing a half and half style proto=
col where
> RFC6455 frames are encapsulated in HTTP 2.0 header+data frame pairs. This
> seems like the worst of both worlds. Not enough like RFC6455 to actually
> re-use existing infrastructure but enough like it to make the messaging
> component not mesh well with the rest of HTTP 2.0. The alternate method
> proposed earlier where HTTP 2.0 simply defines its own =E2=80=9Cordered
> bidirectional message stream=E2=80=9D semantics independently of RFC6455 =
either via
> a specific frame type or via additional flags on the data frame seems to =
be
> an obvious choice to me.
>
> HTTP 2.0 natively provides all of the core facilities that RFC6455 does:
> connection establishment, header flags, control messaging, close messagin=
g,
> framing, etc. There is no need to redefine any of this in terms of RFC645=
5
> opcodes or frame bits order to exchange messages bidirectionally over the
> existing connection between an HTTP 2.0 browser and server. Doing so only
> adds complexity for no real benefit. At best I can see including a method
> for mapping RFC6455 close codes to RST_STREAM error codes.
>
> Using HTTP 2.0 for this sort of messaging would be less efficient than ra=
w
> RFC6455 but likely more efficient overall for sites with a healthy mix of
> HTTP and WebSocket traffic to the same host. It would also solve any issu=
es
> about unknown frame boundaries, make all messaging traffic a first class
> HTTP 2.0 citizen and understandable to intermediaries. It would not requi=
re
> HTTP 2.0 only programs to also have an RFC6455 parser. It would remove th=
e
> need to try and fit any extensions designed for RFC6455 into HTTP 2.0 or
> restrict RFC6455 extension design to things that could fit into HTTP 2.0.
> There doesn=E2=80=99t seem to be anything supported by the JavaScript Web=
Socket API
> that couldn=E2=80=99t be easily transported using such a scheme.
>
> That said, because of the differences between HTTP 2.0 and RFC6455
> WebSockets I would hope to see them coexist in browsers rather than be a
> replacement. There are a number of cases that users of my WebSocket libra=
ry
> have discussed with me where HTTP 2.0 is vast overkill and needlessly
> complex for what they need and RFC6455 fits the bill perfectly.
>
> On the subject of WS MUX, given the existence of a protocol like HTTP 2.0
> with the above messaging feature, based on the experiences of my library
> users and clients, any use case for a muxed WebSocket-type messaging syst=
em
> would be satisfied by HTTP 2.0.
>
>
> On Jan 15, 2014, at 1:41, Takeshi Yoshino <tyoshino@google.com> wrote:
>
> On Wed, Jan 15, 2014 at 4:07 PM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
>
>> >>>In any layering scheme, we'd need to use a HEADER frame one way or
>> another because we need to at least announce that we're connecting to a =
WS
>> endpoint.
>>
>> >>No, the alternative WS/SPDY proposal uses SPDY HEADER only for the
>> initial WS opening handshake, but not for the subsequent WS messaging
>> traffic.
>>
>> >I think Roberto meant that we need to have code to parse HEADERS frame
>> regardless of how layering scheme is. As you say, we'll exchange opening
>> handshake using HEADERS. I.e. a server implementing your proposal will h=
ave
>> HEADERS parser. So, use of HEADERS >itself doesn't add cost.
>>
>> Yes, SPDY HEADER parsing code needs to be there in both, but:
>>
>> The implementation cost for the current WS/SPDY proposal is for WS
>> implementations that want to both support native WS and WS/SPDY:
>>
>> the WS message data processing code needs to be intermingled with SPDY
>> code.
>>
>> With the alternative proposal (after the WS opening handshake is
>> finished), the WS code can stay essentially untouched: the bytes just no=
t
>> come from wire anymore, but from SPDY data frame payloads.
>>
>
> Right. Inter-layer coordination complicates things given one has already
> implemented RFC6455. I understand your argument about clear layering.
>
> It seems that there're different visions of how future transport for
> WebSocket API should be. IIUC, Roberto is trying to make HTTP/2.0 versati=
le
> transport for the web. He's thinking it's costly to carry over RFC6455 to
> that future. In Google, as you know, we're deploying SPDY widely and
> therefore SPDY is always there. If HTTP/2.0 is success, it'll be true als=
o
> for others.
>  _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>
>
>
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>
>

--001a11c1efa67177fa04f00f7b23
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div>Peter,</div><div><br></div><div>Is the method you pre=
fer &quot;Plan B: Frame by frame mapping&quot; in [1]?=C2=A0</div><div>If n=
ot, can you provide some pointers?</div><div><br></div><div>Thanks,</div><d=
iv><br>
</div>[1]=C2=A0<a href=3D"https://docs.google.com/a/google.com/document/d/1=
Dwd23OXidSRA0DupeWujYz2iPyVzvZT6RmEeg4vPY28/edit">https://docs.google.com/a=
/google.com/document/d/1Dwd23OXidSRA0DupeWujYz2iPyVzvZT6RmEeg4vPY28/edit</a=
><br>
</div><div class=3D"gmail_extra"><br><br><div class=3D"gmail_quote">On Thu,=
 Jan 16, 2014 at 12:33 PM, Peter Thorson <span dir=3D"ltr">&lt;<a href=3D"m=
ailto:webmaster@zaphoyd.com" target=3D"_blank">webmaster@zaphoyd.com</a>&gt=
;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div style=3D"word-wrap:break-word"><div><di=
v style=3D"margin:0px">If RFC6455 semantics are to be preserved the only wa=
y that it makes sense to do that is in a fully tunneled method such that th=
e resulting stream could be fed to an existing RFC6455 infrastructure. Whil=
e conceptually simple, as mentioned a number of times in the past threads, =
this introduces a number of problems including significant overhead and red=
uced transparency to HTTP 2.0 intermediaries. I don=E2=80=99t see any great=
 benefit in this method.</div>
<div style=3D"margin:0px;min-height:14px"><br></div><div style=3D"margin:0p=
x">I also don=E2=80=99t see any benefit in doing a half and half style prot=
ocol where RFC6455 frames are encapsulated in HTTP 2.0 header+data frame pa=
irs. This seems like the worst of both worlds. Not enough like RFC6455 to a=
ctually re-use existing infrastructure but enough like it to make the messa=
ging component not mesh well with the rest of HTTP 2.0. The alternate metho=
d proposed earlier where HTTP 2.0 simply defines its own =E2=80=9Cordered b=
idirectional message stream=E2=80=9D semantics independently of RFC6455 eit=
her via a specific frame type or via additional flags on the data frame see=
ms to be an obvious choice to me.</div>
<div style=3D"margin:0px;min-height:14px"><br></div><div style=3D"margin:0p=
x">HTTP 2.0 natively provides all of the core facilities that RFC6455 does:=
 connection establishment, header flags, control messaging, close messaging=
, framing, etc. There is no need to redefine any of this in terms of RFC645=
5 opcodes or frame bits order to exchange messages bidirectionally over the=
 existing connection between an HTTP 2.0 browser and server. Doing so only =
adds complexity for no real benefit. At best I can see including a method f=
or mapping RFC6455 close codes to RST_STREAM error codes.</div>
<div style=3D"margin:0px;min-height:14px"><br></div><div style=3D"margin:0p=
x">Using HTTP 2.0 for this sort of messaging would be less efficient than r=
aw RFC6455 but likely more efficient overall for sites with a healthy mix o=
f HTTP and WebSocket traffic to the same host. It would also solve any issu=
es about unknown frame boundaries, make all messaging traffic a first class=
 HTTP 2.0 citizen and understandable to intermediaries. It would not requir=
e HTTP 2.0 only programs to also have an RFC6455 parser. It would remove th=
e need to try and fit any extensions designed for RFC6455 into HTTP 2.0 or =
restrict RFC6455 extension design to things that could fit into HTTP 2.0. T=
here doesn=E2=80=99t seem to be anything supported by the JavaScript WebSoc=
ket API that couldn=E2=80=99t be easily transported using such a scheme.</d=
iv>
<div style=3D"margin:0px;min-height:14px"><br></div><div style=3D"margin:0p=
x">That said, because of the differences between HTTP 2.0 and RFC6455 WebSo=
ckets I would hope to see them coexist in browsers rather than be a replace=
ment. There are a number of cases that users of my WebSocket library have d=
iscussed with me where HTTP 2.0 is vast overkill and needlessly complex for=
 what they need and RFC6455 fits the bill perfectly.</div>
<div style=3D"margin:0px;min-height:14px"><br></div><div style=3D"margin:0p=
x">On the subject of WS MUX, given the existence of a protocol like HTTP 2.=
0 with the above messaging feature, based on the experiences of my library =
users and clients, any use case for a muxed WebSocket-type messaging system=
 would be satisfied by HTTP 2.0.</div>
</div><div><br></div><br><div><div><div class=3D"h5"><div>On Jan 15, 2014, =
at 1:41, Takeshi Yoshino &lt;<a href=3D"mailto:tyoshino@google.com" target=
=3D"_blank">tyoshino@google.com</a>&gt; wrote:</div><br></div></div><blockq=
uote type=3D"cite">
<div><div class=3D"h5"><div dir=3D"ltr"><div class=3D"gmail_extra"><div cla=
ss=3D"gmail_quote">On Wed, Jan 15, 2014 at 4:07 PM, Tobias Oberstein <span =
dir=3D"ltr">&lt;<a href=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_b=
lank">tobias.oberstein@tavendo.de</a>&gt;</span> wrote:<br>


<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div>&gt;&gt;&gt;In any layering scheme, we&=
#39;d need to use a HEADER frame one way or another because we need to at l=
east announce that we&#39;re connecting to a WS endpoint.<br>



<br>
&gt;&gt;No, the alternative WS/SPDY proposal uses SPDY HEADER only for the =
initial WS opening handshake, but not for the subsequent WS messaging traff=
ic.<br>
<br>
&gt;I think Roberto meant that we need to have code to parse HEADERS frame =
regardless of how layering scheme is. As you say, we&#39;ll exchange openin=
g handshake using HEADERS. I.e. a server implementing your proposal will ha=
ve HEADERS parser. So, use of HEADERS &gt;itself doesn&#39;t add cost.<br>



<br>
</div>Yes, SPDY HEADER parsing code needs to be there in both, but:<br>
<br>
The implementation cost for the current WS/SPDY proposal is for WS implemen=
tations that want to both support native WS and WS/SPDY:<br>
<br>
the WS message data processing code needs to be intermingled with SPDY code=
.<br>
<br>
With the alternative proposal (after the WS opening handshake is finished),=
 the WS code can stay essentially untouched: the bytes just not come from w=
ire anymore, but from SPDY data frame payloads.<br></blockquote><div><br>


</div><div>Right. Inter-layer coordination complicates things given one has=
 already implemented RFC6455. I understand your argument about clear layeri=
ng.</div><div><br></div><div>It seems that there&#39;re different visions o=
f how future transport for WebSocket API should be. IIUC, Roberto is trying=
 to make HTTP/2.0 versatile transport for the web. He&#39;s thinking it&#39=
;s costly to carry over RFC6455 to that future. In Google, as you know, we&=
#39;re deploying SPDY widely and therefore SPDY is always there. If HTTP/2.=
0 is success, it&#39;ll be true also for others.</div>


</div></div></div></div></div><div class=3D"im">
_______________________________________________<br>hybi mailing list<br><a =
href=3D"mailto:hybi@ietf.org" target=3D"_blank">hybi@ietf.org</a><br><a hre=
f=3D"https://www.ietf.org/mailman/listinfo/hybi" target=3D"_blank">https://=
www.ietf.org/mailman/listinfo/hybi</a><br>
</div></blockquote></div><br></div><br>____________________________________=
___________<br>
hybi mailing list<br>
<a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br>
<a href=3D"https://www.ietf.org/mailman/listinfo/hybi" target=3D"_blank">ht=
tps://www.ietf.org/mailman/listinfo/hybi</a><br>
<br></blockquote></div><br></div>

--001a11c1efa67177fa04f00f7b23--

From yhirano@google.com  Wed Jan 15 22:05:49 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id DD3A81AE1C0 for <hybi@ietfa.amsl.com>; Wed, 15 Jan 2014 22:05:49 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.316
X-Spam-Level: 
X-Spam-Status: No, score=-1.316 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, J_CHICKENPOX_36=0.6, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ZYFtrkVFXDwO for <hybi@ietfa.amsl.com>; Wed, 15 Jan 2014 22:05:48 -0800 (PST)
Received: from mail-qc0-x230.google.com (mail-qc0-x230.google.com [IPv6:2607:f8b0:400d:c01::230]) by ietfa.amsl.com (Postfix) with ESMTP id CB29C1ADEA6 for <hybi@ietf.org>; Wed, 15 Jan 2014 22:05:47 -0800 (PST)
Received: by mail-qc0-f176.google.com with SMTP id e16so1912639qcx.21 for <hybi@ietf.org>; Wed, 15 Jan 2014 22:05:35 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=xG+aRrv6uamdCN7JUwJqYCtncvqOwVPIABh2j1fo8XM=; b=JpTNI5qQPeujbsen0lppULYoCQfc2J1jA688V6q4ZBYdRyTfuz5U02GzK5QJDi/KvU +fGiIheJ0VAprzhyc9MdfnmOKjWKEXwUqn8Hjvx6LKZe2r7KrAMQYI9CqAjtIgqgSotn kurIWflaIT9BgO57ePcu7fRGamKPtiMtUibYx5cDjkhllYOV7oXZ9AbWNa84oCKNlJTM DRx/UETehp/nms6vkyX1244Oghw1aDy5oH/i0/6kss2HGo9mQ3UexgikzSJAY0Kpv51N oK/bg6byn2LvRKwl8zwBTTdQ+IgwAduz72/aRl2zEs26QPH+Rz8+IpLrlI0rFrPI4Yxj XkOQ==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=xG+aRrv6uamdCN7JUwJqYCtncvqOwVPIABh2j1fo8XM=; b=IlXVYSPsLNgnNCFVoMha0+hpzWEkWUTKQb6qPiEDABkD2ooIrFALqtB07e2N6sT5xi 2gf2rW/B6ckfoJPHWsBlIoPd3ll/CUSCsIfUlV+pZ2qoq+g1MK/HvPgr7udGAM6oIFKn mnltCns2rxntiW5sSQViEc1obu9DNIKOI7oPc=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=xG+aRrv6uamdCN7JUwJqYCtncvqOwVPIABh2j1fo8XM=; b=MiuJ512xqCaKzAer3D7XL5FteecWRkDrDIMMbN19DLNyjGyMEQg17wOoNmgG4aSf7s Kyhaiq0iJpdxxDxiUOKVnZCHVu6spa6pCxSOAfAQmTnkwBjKbgy5QMKFNdNZz+z9DdBS H/NMbpKHNX0A92r81WwFoxkfKhgDooAcPrR0PlwSCF5uTHJRNsurUJ5hkpxvQwUoM/Dq 8J2UM9mnsjGoj4AG/+LjcBRdOpi7eDLreIw7ZGlsgUuletljqSdRedRnpXvh8aDf28mU WsrgcbpNNZZUk1OH99TljEDaOVx+Xp7HApeKnI0f36OoSp5unElsfNcdE1aoIwsWJIyY ZWGw==
X-Gm-Message-State: ALoCoQnO5JU3/+zAmOvAuS2wW+1GCrWj3NddBprbBG37qY8VJ48K2D1K1BkeRXzRjJ9ArIen05m3YYUWeQtdZex7/EpeWCJWFa0qM7Nzj/hKJVDNZHJsGp1hh+DTyJ2xa8Pps1lCAyTgKTFzP6cWHr9XC/vBnAMFuY+z+mNVgZrWRCx3oZo/oMXqOnGsKgnT3chpGn3XNgqm
MIME-Version: 1.0
X-Received: by 10.140.31.247 with SMTP id f110mr4840282qgf.58.1389852335419; Wed, 15 Jan 2014 22:05:35 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.140.109.202 with HTTP; Wed, 15 Jan 2014 22:05:35 -0800 (PST)
In-Reply-To: <CAG4zZZCemh=8zOg6zuvaSbX5iPMHzq7jTKHg_0VtyTA_q1QaYA@mail.gmail.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CAHixhFpVywsZPCx1r_HYXEUQ10WN31NnVYM5XT6-n78uAcrjig@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF98D0A@EXVMBX020-12.exch020.serverdata.net> <CAG4zZZCemh=8zOg6zuvaSbX5iPMHzq7jTKHg_0VtyTA_q1QaYA@mail.gmail.com>
Date: Thu, 16 Jan 2014 15:05:35 +0900
X-Google-Sender-Auth: 0-Tv-taV55DTNJsS9dP9UxQFOfk
Message-ID: <CABihn6F-SDHPCetwjBaQoZBwaD5FzCty3m0OF+vOk7-XQw19fw@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: Joakim Erdfelt <joakim@intalio.com>
Content-Type: multipart/alternative; boundary=001a113a9c02be5a2f04f01036c3
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 16 Jan 2014 06:05:50 -0000

--001a113a9c02be5a2f04f01036c3
Content-Type: text/plain; charset=UTF-8

Hi Joakim,

Thanks for the comment.

Knowing the size of a frame at the start of the frame, traditional
> websocket behavior, is desired, as it allows us to fast fail for the 1009
> (message too big) use case.

Do you know how effective it is?

Having no prior knowledge of frame size is undesirable to us.

On the other hand, the frame sender have more freedom because they don't
have to know the size of the frame in front.
For example, it can construct and send a big WS data frame without minding
a control frame which may appear and be injected afterward.



On Thu, Jan 16, 2014 at 2:07 AM, Joakim Erdfelt <joakim@intalio.com> wrote:

> One minor niggle from Jetty on this (we are lurking and reading)
>
> Knowing the size of a frame at the start of the frame, traditional
> websocket behavior, is desired, as it allows us to fast fail for the 1009
> (message too big) use case.
> Having no prior knowledge of frame size is undesirable to us.
> This is inline with what Tobias is asking for too.
>
>
> --
> Joakim Erdfelt <joakim@intalio.com>
> webtide.com <http://www.webtide.com/> - intalio.com/jetty
> Expert advice, services and support from from the Jetty & CometD experts
> eclipse.org/jetty - cometd.org
>
>
> On Wed, Jan 15, 2014 at 9:01 AM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
>
>> >>On 15 January 2014 04:09, Tobias Oberstein <tobias.oberstein@tavendo.de>
>> wrote:
>>
>> >> * has significant wire-level overhead compared to native WS (and
>> compared to the proposal below)
>>
>> >If you accept the argument that unencrypted WebSockets are unusable on
>> heterogeneous networks due to interference from middleboxes, then the byte
>> overhead is a red-herring. Everything has to be encrypted, and so will get
>> padded out to the cipher block size >anyway.
>>
>> I did more readup on this one .. e.g. not all modes require padding
>> (AES-GCM does not) - nevertheless, you are totally right: the overhead of
>> TLS can be substantial.
>>
>> Here are measurements for Chrome, Firefox and IE11:
>>
>> http://tavendo.com/blog/post/dissecting-websocket-overhead/
>>
>> The overhead induced by TCP/IP and TLS in particular can dwarf the
>> overhead of WebSocket itself.
>>
>> Thank you very much for pointing into this direction .. this is important.
>>
>> As an immediate consequence, we will introduce message batching into WAMP
>> (a WebSocket subprotocol for RPC+PubSub -
>> https://github.com/tavendo/WAMP/tree/master/spec )
>>
>> /Tobias
>> _______________________________________________
>> hybi mailing list
>> hybi@ietf.org
>> https://www.ietf.org/mailman/listinfo/hybi
>>
>
>
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>
>

--001a113a9c02be5a2f04f01036c3
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Hi Joakim,<div><br></div><div>Thanks for the comment.<br><=
div><br></div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px=
 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left=
-style:solid;padding-left:1ex">
<span style=3D"font-family:arial,sans-serif;font-size:13px">Knowing the siz=
e of a frame at the start of the frame, traditional websocket behavior, is =
desired, as it allows us to fast fail for the 1009 (message too big) use ca=
se.</span></blockquote>
<div>Do you know how effective it is?</div><div><br></div><blockquote class=
=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;bo=
rder-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">
<span style=3D"font-family:arial,sans-serif;font-size:13px">Having no prior=
 knowledge of frame size is undesirable to us.</span></blockquote><div>On t=
he other hand, the frame sender have more freedom because they don&#39;t ha=
ve to know the size of the frame in front.</div>
<div>For example, it can construct and send a big WS data frame without min=
ding a control frame which may appear and be injected afterward.</div><div>=
<br></div></div></div><div class=3D"gmail_extra"><br><br><div class=3D"gmai=
l_quote">
On Thu, Jan 16, 2014 at 2:07 AM, Joakim Erdfelt <span dir=3D"ltr">&lt;<a hr=
ef=3D"mailto:joakim@intalio.com" target=3D"_blank">joakim@intalio.com</a>&g=
t;</span> wrote:<br><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0=
 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div dir=3D"ltr">One minor niggle from Jetty on this (we are lurking and re=
ading)<div><br></div><div>Knowing the size of a frame at the start of the f=
rame, traditional websocket behavior, is desired, as it allows us to fast f=
ail for the 1009 (message too big) use case.</div>

<div>Having no prior knowledge of frame size is undesirable to us.</div><di=
v>This is inline with what Tobias is asking for too.</div><div><br></div></=
div><div class=3D"gmail_extra"><br clear=3D"all"><div><div dir=3D"ltr"><div=
>
--</div>
<div>Joakim Erdfelt &lt;<a href=3D"mailto:joakim@intalio.com" target=3D"_bl=
ank">joakim@intalio.com</a>&gt;</div><div><a href=3D"http://www.webtide.com=
/" target=3D"_blank">webtide.com</a>=C2=A0- <a href=3D"http://intalio.com/j=
etty" target=3D"_blank">intalio.com/jetty</a></div>

<div><span>Expert advice, services and support from=C2=A0</span><span>from =
the Jetty &amp; CometD experts</span></div><div><a href=3D"http://eclipse.o=
rg/jetty/" target=3D"_blank">eclipse.org/jetty</a>=C2=A0-=C2=A0<a href=3D"h=
ttp://cometd.org/" target=3D"_blank">cometd.org</a></div>

</div></div>
<br><br><div class=3D"gmail_quote"><div><div class=3D"h5">On Wed, Jan 15, 2=
014 at 9:01 AM, Tobias Oberstein <span dir=3D"ltr">&lt;<a href=3D"mailto:to=
bias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstein@tavendo.de</a=
>&gt;</span> wrote:<br>
</div></div><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;bo=
rder-left:1px #ccc solid;padding-left:1ex"><div><div class=3D"h5">
<div>&gt;&gt;On 15 January 2014 04:09, Tobias Oberstein &lt;<a href=3D"mail=
to:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstein@tavendo.=
de</a>&gt; wrote:<br>
<br>
&gt;&gt;=C2=A0* has significant wire-level overhead compared to native WS (=
and compared to the proposal below)<br>
<br>
&gt;If you accept the argument that unencrypted WebSockets are unusable on =
heterogeneous networks due to interference from middleboxes, then the byte =
overhead is a red-herring. Everything has to be encrypted, and so will get =
padded out to the cipher block size &gt;anyway.<br>


<br>
</div>I did more readup on this one .. e.g. not all modes require padding (=
AES-GCM does not) - nevertheless, you are totally right: the overhead of TL=
S can be substantial.<br>
<br>
Here are measurements for Chrome, Firefox and IE11:<br>
<br>
<a href=3D"http://tavendo.com/blog/post/dissecting-websocket-overhead/" tar=
get=3D"_blank">http://tavendo.com/blog/post/dissecting-websocket-overhead/<=
/a><br>
<br>
The overhead induced by TCP/IP and TLS in particular can dwarf the overhead=
 of WebSocket itself.<br>
<br>
Thank you very much for pointing into this direction .. this is important.<=
br>
<br>
As an immediate consequence, we will introduce message batching into WAMP (=
a WebSocket subprotocol for RPC+PubSub - <a href=3D"https://github.com/tave=
ndo/WAMP/tree/master/spec" target=3D"_blank">https://github.com/tavendo/WAM=
P/tree/master/spec</a> )<br>


<span><font color=3D"#888888"><br>
/Tobias<br>
</font></span></div></div><div class=3D"im"><div><div>_____________________=
__________________________<br>
hybi mailing list<br>
<a href=3D"mailto:hybi@ietf.org" target=3D"_blank">hybi@ietf.org</a><br>
<a href=3D"https://www.ietf.org/mailman/listinfo/hybi" target=3D"_blank">ht=
tps://www.ietf.org/mailman/listinfo/hybi</a><br>
</div></div></div></blockquote></div><br></div>
<br>_______________________________________________<br>
hybi mailing list<br>
<a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br>
<a href=3D"https://www.ietf.org/mailman/listinfo/hybi" target=3D"_blank">ht=
tps://www.ietf.org/mailman/listinfo/hybi</a><br>
<br></blockquote></div><br></div>

--001a113a9c02be5a2f04f01036c3--

From duerst@it.aoyama.ac.jp  Wed Jan 15 22:28:03 2014
Return-Path: <duerst@it.aoyama.ac.jp>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C26651ADE7C for <hybi@ietfa.amsl.com>; Wed, 15 Jan 2014 22:28:03 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: 0.371
X-Spam-Level: 
X-Spam-Status: No, score=0.371 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HELO_EQ_JP=1.244, HOST_EQ_JP=1.265, MIME_8BIT_HEADER=0.3, RP_MATCHES_RCVD=-0.538] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Zcshv02MRxUG for <hybi@ietfa.amsl.com>; Wed, 15 Jan 2014 22:28:02 -0800 (PST)
Received: from scintmta01.scbb.aoyama.ac.jp (scintmta01.scbb.aoyama.ac.jp [133.2.253.33]) by ietfa.amsl.com (Postfix) with ESMTP id 0FEC51AE1C3 for <hybi@ietf.org>; Wed, 15 Jan 2014 22:28:00 -0800 (PST)
Received: from scmse02.scbb.aoyama.ac.jp ([133.2.253.231]) by scintmta01.scbb.aoyama.ac.jp (secret/secret) with SMTP id s0G6Rjqh013101; Thu, 16 Jan 2014 15:27:45 +0900
Received: from (unknown [133.2.206.134]) by scmse02.scbb.aoyama.ac.jp with smtp id 3c33_28fb_4fbfe256_7e77_11e3_b9e5_001e6722eec2; Thu, 16 Jan 2014 15:27:44 +0900
Received: from [IPv6:::1] (unknown [133.2.210.1]) by itmail2.it.aoyama.ac.jp (Postfix) with ESMTP id EA542BF539; Thu, 16 Jan 2014 15:27:44 +0900 (JST)
Message-ID: <52D77BD6.6080500@it.aoyama.ac.jp>
Date: Thu, 16 Jan 2014 15:27:34 +0900
From: =?UTF-8?B?Ik1hcnRpbiBKLiBEw7xyc3Qi?= <duerst@it.aoyama.ac.jp>
Organization: Aoyama Gakuin University
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.1.9) Gecko/20100722 Eudora/3.0.4
MIME-Version: 1.0
To: Yutaka Hirano <yhirano@chromium.org>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CAHixhFpVywsZPCx1r_HYXEUQ10WN31NnVYM5XT6-n78uAcrjig@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF98D0A@EXVMBX020-12.exch020.serverdata.net> <CAG4zZZCemh=8zOg6zuvaSbX5iPMHzq7jTKHg_0VtyTA_q1QaYA@mail.gmail.com> <CABihn6F-SDHPCetwjBaQoZBwaD5FzCty3m0OF+vOk7-XQw19fw@mail.gmail.com>
In-Reply-To: <CABihn6F-SDHPCetwjBaQoZBwaD5FzCty3m0OF+vOk7-XQw19fw@mail.gmail.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of	WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 16 Jan 2014 06:28:04 -0000

On 2014/01/16 15:05, Yutaka Hirano wrote:
> Hi Joakim,
>
> Thanks for the comment.
>
> Knowing the size of a frame at the start of the frame, traditional
>> websocket behavior, is desired, as it allows us to fast fail for the 1009
>> (message too big) use case.
>
> Do you know how effective it is?
>
> Having no prior knowledge of frame size is undesirable to us.
>
> On the other hand, the frame sender have more freedom because they don't
> have to know the size of the frame in front.

My understanding of why messages are split into frames was that this 
would allow to start a message without knowing its length. But if you 
don't know the size of a frame before sending it, then the right 
solution is to make it smaller. Frames are just arbitrary pieces of a 
message.

Regards,   Martin.

From tobias.oberstein@tavendo.de  Thu Jan 16 08:39:50 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9E72B1AE39A for <hybi@ietfa.amsl.com>; Thu, 16 Jan 2014 08:39:50 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id lKp6OwhAyqAD for <hybi@ietfa.amsl.com>; Thu, 16 Jan 2014 08:39:49 -0800 (PST)
Received: from EXHUB020-5.exch020.serverdata.net (exhub020-5.exch020.serverdata.net [206.225.164.32]) by ietfa.amsl.com (Postfix) with ESMTP id 708951AE3BD for <hybi@ietf.org>; Thu, 16 Jan 2014 08:39:45 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-5.exch020.serverdata.net ([206.225.164.32]) with mapi; Thu, 16 Jan 2014 08:39:32 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: "hybi@ietf.org" <hybi@ietf.org>
Date: Thu, 16 Jan 2014 08:39:29 -0800
Thread-Topic: permessage-priority extension draft
Thread-Index: Ac8S2TaopUz0OETkTu2tHIq/IN5VPQ==
Message-ID: <634914A010D0B943A035D226786325D4446BF991A9@EXVMBX020-12.exch020.serverdata.net>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
Subject: [hybi] permessage-priority extension draft
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 16 Jan 2014 16:39:50 -0000

Alright, I have written up a draft for a "permessage-priority" WebSocket ex=
tension:

https://github.com/oberstet/permessage-priority/blob/master/draft-oberstein=
-hybi-permessage-priority.txt

In writing this, I came to the conclusion that "message prioritization" on =
a single WebSocket connection
in fact is orthogonal to multiplexing multiple (logical) WebSocket connecti=
ons over a single physical one
(be it via WS-MUX or a variant of WS/SPDY).

So this proposal should not be seen as a competitive proposal to WS-MUX or =
WS/SPDY, but as ... something different;)

The motivation and use case is outlined in the introduction.

I'd appreciate any feedback, suggestions, corrections or even PRs (the draf=
t is on GitHub in a dedicated repo).

If the draft doesn't conform to RFC requirements or formatting rules, pleas=
e forgive - this is my first RFC.

Cheers,
/Tobias

PS: For one thing I don't know, how do I submit this draft with the IETF so=
 it can follow the proper way?


From tobias.oberstein@tavendo.de  Fri Jan 17 01:25:04 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 0C35D1ADDDA for <hybi@ietfa.amsl.com>; Fri, 17 Jan 2014 01:25:04 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 1nEMOqpMCnbJ for <hybi@ietfa.amsl.com>; Fri, 17 Jan 2014 01:25:02 -0800 (PST)
Received: from EXHUB020-5.exch020.serverdata.net (exhub020-5.exch020.serverdata.net [206.225.164.32]) by ietfa.amsl.com (Postfix) with ESMTP id A3EEB1ADF4F for <hybi@ietf.org>; Fri, 17 Jan 2014 01:25:00 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-5.exch020.serverdata.net ([206.225.164.32]) with mapi; Fri, 17 Jan 2014 01:24:48 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: "hybi@ietf.org" <hybi@ietf.org>
Date: Fri, 17 Jan 2014 01:24:45 -0800
Thread-Topic: WebSocket subprotocol parameters
Thread-Index: Ac8TZNT+ZxEt6fFfT4OtNh9SU04iIg==
Message-ID: <634914A010D0B943A035D226786325D4446BF9948F@EXVMBX020-12.exch020.serverdata.net>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
Subject: [hybi] WebSocket subprotocol parameters
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 17 Jan 2014 09:25:04 -0000

The "Sec-WebSocket-Extensions" allows for parameters for each element, eg (=
from "permessage-deflate"):

Sec-WebSocket-Extensions: permessage-deflate; client_max_window_bits; serve=
r_max_window_bits=3D10, permessage-deflate; client_max_window_bits

Similar seems to be disallowed for _subprotocols_.

RFC6455 says:

"""
The elements that comprise this value
MUST be non-empty strings with characters in the range U+0021 to
U+007E not including separator characters as defined in
[RFC2616] and MUST all be unique strings.  The ABNF for the
value of this header field is 1#token, where the definitions of
constructs and rules are as given in [RFC2616].
"""

and RFC2616 says:

"""
       token          =3D 1*<any CHAR except CTLs or separators>
       separators     =3D "(" | ")" | "<" | ">" | "@"
                      | "," | ";" | ":" | "\" | <">
                      | "/" | "[" | "]" | "?" | "=3D"
                      | "{" | "}" | SP | HAT
"""

So

Sec-WebSocket-Protocol: wamp; version=3D2; serialization=3Dmsgpack; batched=
, wamp; version=3D1

is disallowed, right?

This is unfortunate, since "simulating parameters" like

Sec-WebSocket-Protocol: wamp.2.msgpack.batched, wamp

is a hack, and more so, RFC6455 requires that all subprotocol name be regis=
tered via the " WebSocket Subprotocol Name Registry".

This would mean registering all subprotocol variants:

wamp
wamp.2.json
wamp.2.msgpack
wamp.2.json.batched
wamp.2.msgpack.batched
..

Was there any good reason disallowing parameters for "Sec-WebSocket-Protoco=
l"?

If not, could this be fixed via errata for RFC6455?

/Tobias


From julian.reschke@gmx.de  Fri Jan 17 03:10:58 2014
Return-Path: <julian.reschke@gmx.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 76EA01ADFE2 for <hybi@ietfa.amsl.com>; Fri, 17 Jan 2014 03:10:58 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, FREEMAIL_FROM=0.001, RCVD_IN_DNSWL_NONE=-0.0001, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id LB3KR_bgGSet for <hybi@ietfa.amsl.com>; Fri, 17 Jan 2014 03:10:56 -0800 (PST)
Received: from mout.gmx.net (mout.gmx.net [212.227.17.20]) by ietfa.amsl.com (Postfix) with ESMTP id 79E271ADFC6 for <hybi@ietf.org>; Fri, 17 Jan 2014 03:10:56 -0800 (PST)
Received: from [192.168.1.102] ([217.91.35.233]) by mail.gmx.com (mrgmx101) with ESMTPSA (Nemesis) id 0LyR1G-1VJcYw3YMo-015q8D for <hybi@ietf.org>; Fri, 17 Jan 2014 12:10:43 +0100
Message-ID: <52D90F99.6080205@gmx.de>
Date: Fri, 17 Jan 2014 12:10:17 +0100
From: Julian Reschke <julian.reschke@gmx.de>
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:24.0) Gecko/20100101 Thunderbird/24.2.0
MIME-Version: 1.0
To: Tobias Oberstein <tobias.oberstein@tavendo.de>,  "hybi@ietf.org" <hybi@ietf.org>
References: <634914A010D0B943A035D226786325D4446BF9948F@EXVMBX020-12.exch020.serverdata.net>
In-Reply-To: <634914A010D0B943A035D226786325D4446BF9948F@EXVMBX020-12.exch020.serverdata.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Provags-ID: V03:K0:5bqMjh/D/5Rlg3hSfxYte8Z7ca4G5NSHShQQllzhj+8j5ZOH6fC k702OJEG72FNVDq5KmTeISUXhJh089aWn5ZlfImlrLhSS/bwvrY/hsdMPpv3FE3UChJfVI+ CTs8LQR6oz93NUhHdv2Mh93pOgZEijfDuoR627aVyhzbNvZW3cImUJt6bYV4uPwc08tf+Oz 3pCMQV+9tCVIYw6G7zlGg==
Subject: Re: [hybi] WebSocket subprotocol parameters
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 17 Jan 2014 11:10:58 -0000

On 2014-01-17 10:24, Tobias Oberstein wrote:
> The "Sec-WebSocket-Extensions" allows for parameters for each element, eg (from "permessage-deflate"):
>
> Sec-WebSocket-Extensions: permessage-deflate; client_max_window_bits; server_max_window_bits=10, permessage-deflate; client_max_window_bits
>
> Similar seems to be disallowed for _subprotocols_.
>
> RFC6455 says:
>
> """
> The elements that comprise this value
> MUST be non-empty strings with characters in the range U+0021 to
> U+007E not including separator characters as defined in
> [RFC2616] and MUST all be unique strings.  The ABNF for the
> value of this header field is 1#token, where the definitions of
> constructs and rules are as given in [RFC2616].
> """
>
> and RFC2616 says:
>
> """
>         token          = 1*<any CHAR except CTLs or separators>
>         separators     = "(" | ")" | "<" | ">" | "@"
>                        | "," | ";" | ":" | "\" | <">
>                        | "/" | "[" | "]" | "?" | "="
>                        | "{" | "}" | SP | HAT
> """
>
> So
>
> Sec-WebSocket-Protocol: wamp; version=2; serialization=msgpack; batched, wamp; version=1
>
> is disallowed, right?
>
> This is unfortunate, since "simulating parameters" like
>
> Sec-WebSocket-Protocol: wamp.2.msgpack.batched, wamp
>
> is a hack, and more so, RFC6455 requires that all subprotocol name be registered via the " WebSocket Subprotocol Name Registry".
>
> This would mean registering all subprotocol variants:
>
> wamp
> wamp.2.json
> wamp.2.msgpack
> wamp.2.json.batched
> wamp.2.msgpack.batched
> ..
>
> Was there any good reason disallowing parameters for "Sec-WebSocket-Protocol"?

I think that was discussed and the decision was made on purpose. For 
details you'll have to visit the mailing list archives though.

> If not, could this be fixed via errata for RFC6455?

Um, no. This is not a bug but a change request (with the potential of 
breaking existing code).

Best regards, Julian


From tobias.oberstein@tavendo.de  Fri Jan 17 05:01:21 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 387E91AE0AD for <hybi@ietfa.amsl.com>; Fri, 17 Jan 2014 05:01:21 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Cn3sys6WHGLO for <hybi@ietfa.amsl.com>; Fri, 17 Jan 2014 05:01:20 -0800 (PST)
Received: from EXHUB020-4.exch020.serverdata.net (exhub020-4.exch020.serverdata.net [206.225.164.31]) by ietfa.amsl.com (Postfix) with ESMTP id E679B1AE0A6 for <hybi@ietf.org>; Fri, 17 Jan 2014 05:01:19 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-4.exch020.serverdata.net ([206.225.164.31]) with mapi; Fri, 17 Jan 2014 05:01:07 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Julian Reschke <julian.reschke@gmx.de>, "hybi@ietf.org" <hybi@ietf.org>
Date: Fri, 17 Jan 2014 05:01:05 -0800
Thread-Topic: [hybi] WebSocket subprotocol parameters
Thread-Index: Ac8TdMVAz2gO9MBiQtebOK6QNcwqsAADn3Ag
Message-ID: <634914A010D0B943A035D226786325D4446BF9949D@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF9948F@EXVMBX020-12.exch020.serverdata.net> <52D90F99.6080205@gmx.de>
In-Reply-To: <52D90F99.6080205@gmx.de>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
Subject: Re: [hybi] WebSocket subprotocol parameters
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 17 Jan 2014 13:01:21 -0000

> > Was there any good reason disallowing parameters for "Sec-WebSocket-
> Protocol"?
>=20
> I think that was discussed and the decision was made on purpose. For deta=
ils
> you'll have to visit the mailing list archives though.

I can't find it .. searching the Hybi archives is a pain ..

>=20
> > If not, could this be fixed via errata for RFC6455?
>=20
> Um, no. This is not a bug but a change request (with the potential of bre=
aking
> existing code).

Alright. I see.

It seems, WebKit is absolutely strict about allowed characters

https://trac.webkit.org/browser/trunk/Source/WebCore/Modules/websockets/Web=
Socket.cpp#L71

will only match "exactly"

https://trac.webkit.org/browser/trunk/Source/WebCore/Modules/websockets/Web=
SocketHandshake.cpp#L552

and, interestingly, will nevertheless escape any characters normally disall=
owed using Unicode escape sequences

https://trac.webkit.org/browser/trunk/Source/WebCore/Modules/websockets/Web=
Socket.cpp#L95

Thanks!
/Tobias

>=20
> Best regards, Julian


From derhoermi@gmx.net  Fri Jan 17 05:31:50 2014
Return-Path: <derhoermi@gmx.net>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id E11161AE076 for <hybi@ietfa.amsl.com>; Fri, 17 Jan 2014 05:31:50 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.438
X-Spam-Level: 
X-Spam-Status: No, score=-2.438 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, FREEMAIL_FROM=0.001, RCVD_IN_DNSWL_NONE=-0.0001, RP_MATCHES_RCVD=-0.538, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Aos8srmtlCpk for <hybi@ietfa.amsl.com>; Fri, 17 Jan 2014 05:31:49 -0800 (PST)
Received: from mout.gmx.net (mout.gmx.net [212.227.15.15]) by ietfa.amsl.com (Postfix) with ESMTP id 170CE1AE0B5 for <hybi@ietf.org>; Fri, 17 Jan 2014 05:31:49 -0800 (PST)
Received: from netb ([47.65.15.1]) by mail.gmx.com (mrgmx101) with ESMTPSA (Nemesis) id 0MIQlv-1W3LNA2YVL-004Cbs for <hybi@ietf.org>; Fri, 17 Jan 2014 14:31:35 +0100
From: Bjoern Hoehrmann <derhoermi@gmx.net>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Date: Fri, 17 Jan 2014 14:31:35 +0100
Message-ID: <drbid9t6cjhuf08qo6j7pe85lsttb75bid@hive.bjoern.hoehrmann.de>
References: <634914A010D0B943A035D226786325D4446BF9948F@EXVMBX020-12.exch020.serverdata.net> <52D90F99.6080205@gmx.de> <634914A010D0B943A035D226786325D4446BF9949D@EXVMBX020-12.exch020.serverdata.net>
In-Reply-To: <634914A010D0B943A035D226786325D4446BF9949D@EXVMBX020-12.exch020.serverdata.net>
X-Mailer: Forte Agent 3.3/32.846
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
X-Provags-ID: V03:K0:HiaH7qsNbPppzuumYJaMWVCpJlfVw9D6V1Kyw6oSfGEsa20sdwc e/p4itRxqfkA8YxKvNhstr25L7nu+Da1WT8wnZugd5urLYzMgXpwyWQXbD4n3+gx8SXSxC9 Z8omVUELSVErqki5K9eOZoXE5ztQk+wi3IpgKuKoO6geyD/6SpyvKaf9FzCpERKaWAToVDS w5ni2THY1g1vDVzjFBLqQ==
Cc: Julian Reschke <julian.reschke@gmx.de>, "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] WebSocket subprotocol parameters
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 17 Jan 2014 13:31:51 -0000

* Tobias Oberstein wrote:
>> > Was there any good reason disallowing parameters for "Sec-WebSocket-
>> Protocol"?
>> 
>> I think that was discussed and the decision was made on purpose. For details
>> you'll have to visit the mailing list archives though.
>
>I can't find it .. searching the Hybi archives is a pain ..

It is not as easy as taking the first result on 

  https://www.google.com/search?q=site:http://www.ietf.org/mail-archive/web/hybi/+%22Sec-WebSocket-Protocol%22+parameters

But e.g. http://www.ietf.org/mail-archive/web/hybi/current/msg07700.html
uses `Sec-WebSocket-Protocol` parameters in an example, so that might be
nearby, if anybody feels archaeologically inclined.
-- 
Björn Höhrmann · mailto:bjoern@hoehrmann.de · http://bjoern.hoehrmann.de
Am Badedeich 7 · Telefon: +49(0)160/4415681 · http://www.bjoernsworld.de
25899 Dagebüll · PGP Pub. KeyID: 0xA4357E78 · http://www.websitedev.de/ 

From tobias.oberstein@tavendo.de  Fri Jan 17 15:10:52 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id EEDED1ACCEC for <hybi@ietfa.amsl.com>; Fri, 17 Jan 2014 15:10:51 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id d8xmFzf6zzkb for <hybi@ietfa.amsl.com>; Fri, 17 Jan 2014 15:10:50 -0800 (PST)
Received: from EXHUB020-1.exch020.serverdata.net (exhub020-1.exch020.serverdata.net [206.225.164.28]) by ietfa.amsl.com (Postfix) with ESMTP id 2AE271ACCE6 for <hybi@ietf.org>; Fri, 17 Jan 2014 15:10:50 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-1.exch020.serverdata.net ([206.225.164.28]) with mapi; Fri, 17 Jan 2014 15:10:37 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: "hybi@ietf.org" <hybi@ietf.org>
Date: Fri, 17 Jan 2014 15:10:29 -0800
Thread-Topic: permessage-priority extension draft
Thread-Index: Ac8S2TaopUz0OETkTu2tHIq/IN5VPQA/hnBg
Message-ID: <634914A010D0B943A035D226786325D4446BF9987B@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF991A9@EXVMBX020-12.exch020.serverdata.net>
In-Reply-To: <634914A010D0B943A035D226786325D4446BF991A9@EXVMBX020-12.exch020.serverdata.net>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
Subject: Re: [hybi] permessage-priority extension draft
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 17 Jan 2014 23:10:52 -0000

There is a collection of open issues here:
https://github.com/oberstet/permessage-priority/issues

One of those is how to make permessage-priority compatible with permessage-=
deflate.

I guess it can be done by maintaining a separate compression context per pr=
iority class.
This allows intermediaries to prioritize and reorder frames without problem=
s. However,
it multiplies memory requirements.

Was there any discussion/plan about how permessage-deflate would have worke=
d with MUX
when compression was applied to the physical WebSocket connection?

I mean, when compression was applied to the logical MUXed WS connections, t=
here
would have been a separate compression context for each (logical) WS connec=
tion also,
and intermediaries can reMUX without recompressing.

But when applied to the physical WS?

/Tobias

> -----Urspr=FCngliche Nachricht-----
> Von: hybi [mailto:hybi-bounces@ietf.org] Im Auftrag von Tobias Oberstein
> Gesendet: Donnerstag, 16. Januar 2014 17:39
> An: hybi@ietf.org
> Betreff: [hybi] permessage-priority extension draft
>=20
> Alright, I have written up a draft for a "permessage-priority" WebSocket
> extension:
>=20
> https://github.com/oberstet/permessage-priority/blob/master/draft-oberste=
in-
> hybi-permessage-priority.txt
>=20
> In writing this, I came to the conclusion that "message prioritization" o=
n a
> single WebSocket connection in fact is orthogonal to multiplexing multipl=
e
> (logical) WebSocket connections over a single physical one (be it via WS-=
MUX
> or a variant of WS/SPDY).
>=20
> So this proposal should not be seen as a competitive proposal to WS-MUX o=
r
> WS/SPDY, but as ... something different;)
>=20
> The motivation and use case is outlined in the introduction.
>=20
> I'd appreciate any feedback, suggestions, corrections or even PRs (the dr=
aft is
> on GitHub in a dedicated repo).
>=20
> If the draft doesn't conform to RFC requirements or formatting rules, ple=
ase
> forgive - this is my first RFC.
>=20
> Cheers,
> /Tobias
>=20
> PS: For one thing I don't know, how do I submit this draft with the IETF =
so it can
> follow the proper way?
>=20
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi

From yhirano@google.com  Sun Jan 19 23:03:50 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9B54C1A005B for <hybi@ietfa.amsl.com>; Sun, 19 Jan 2014 23:03:50 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: 0.787
X-Spam-Level: 
X-Spam-Status: No, score=0.787 tagged_above=-999 required=5 tests=[BAYES_50=0.8, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 4ULIdSUaM5QJ for <hybi@ietfa.amsl.com>; Sun, 19 Jan 2014 23:03:46 -0800 (PST)
Received: from mail-qc0-x230.google.com (mail-qc0-x230.google.com [IPv6:2607:f8b0:400d:c01::230]) by ietfa.amsl.com (Postfix) with ESMTP id 1AFCB1A0058 for <hybi@ietf.org>; Sun, 19 Jan 2014 23:03:45 -0800 (PST)
Received: by mail-qc0-f176.google.com with SMTP id e16so5578700qcx.21 for <hybi@ietf.org>; Sun, 19 Jan 2014 23:03:46 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=OvAonTqGwKoGmXIqpAXQlnZNDw0LXZdIxsztCxnMqeA=; b=loEFzj9uVROD00UNK+PXRAvKeR0fBg5WYn3bk3dMcEaemt7gmQqciRI0mRmmnu9LyX GHNh/3k2uIDY6s95gn/ms6x4tCxAonqP9kuAFiW8+bCsknllMO5GFdsa+bDz2M0qsHro IotSMBVNEnMBattTJDrqUm0LpxXph4p+JogS2C1PXfRG2sTbc2M9RoE0esx2WukurkI3 hir//7sF7aOY019LT2HonDbvp8aezG1kswTpvdLM7maxQHtbfQaA0nhRKztg+8tRvIl9 QkDmyf7Erp9EkNpxl4bxPOwQGKFAD5kCyzcAL4WGnxo3uJ4bkCQe3ykZQzLrP+Vpk809 Rcdw==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=OvAonTqGwKoGmXIqpAXQlnZNDw0LXZdIxsztCxnMqeA=; b=Se/GggSdIBDht+GlRpmh/pUt0n4XsPHoOL1bv75tGQWD8EacdLbOm0MTg8myjaYFuK uJrvN6D2S/ae2CSWVL03dOd0EfgtKsg3SZbIBNWQFhOV2DYCUKPGuy7gQWrgRJx8Q9HD VOBk1j820nKX9/quuShQe79C9Bh37vaiTIPLA=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=OvAonTqGwKoGmXIqpAXQlnZNDw0LXZdIxsztCxnMqeA=; b=dbZ+tM3s/0BvJ/b820wkmDyZfV+aPCCrrtyQq3B30LRr/eQeUIWE4FfP7xL4Z3rHEJ KZOCkJto3vrYc3B3j7XUJaK5zveBkOjfxKGlbW3IetmK1u8O1ju0glzUcfV6n3ujGN6d /BBf8HBe0R7YnWR3i1UhL4jJX3AQWUiU3uEGaYetVpAIkz3/iJbJuy07dsnXn3N5yRl7 hrmfRtj9HfPmI60ZMAsH2lO8iNmNHcF5UlMvkU3hJk37NpIx3L6cGoNY5p3orilE/sfQ VxauZJNuKaJShxOcXyGWR+wxWdBTZVNsE1kR5CTTUHdkEessQq4OOGMixFQ6VBnLVZFE dPjA==
X-Gm-Message-State: ALoCoQkh2FXb458+pphJPwv+ZenwVEb9TNTkzco79LqN379I/CvN013hvCtmqqMhFsRBDp4tb11lX1R8M7HfzT33fMMTHpMI2fcVubeGNc8ShWEE9fudRkKgmo4ttB6HSwrI5fpM+f2tlSpvx/4YnxmnAJrzvVMzZr2/ETf2UwVCjPw+cPzapKuCcjhuJ3S//PvasE70zBzK
MIME-Version: 1.0
X-Received: by 10.224.72.11 with SMTP id k11mr24992724qaj.91.1390201426118; Sun, 19 Jan 2014 23:03:46 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.140.109.202 with HTTP; Sun, 19 Jan 2014 23:03:45 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net>
Date: Mon, 20 Jan 2014 16:03:45 +0900
X-Google-Sender-Auth: gsnDpH4F6Oc1O0eNZehAUh_tt8o
Message-ID: <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=047d7bea2f402bb47404f0617e69
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 07:03:50 -0000

--047d7bea2f402bb47404f0617e69
Content-Type: text/plain; charset=UTF-8

Tobias, let me confirm.


> I'd like to suggest a different WS/SPDY:
>
> Pack WebSocket frames _including_ full WS frame header as _payload_ of
> SPDY data frames only - no SPDY header frames at all.

My understanding is that by suggesting the above one you retracted the
below proposal you made before, is that right?

On Fri, Jan 10, 2014 at 5:37 PM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> Why not map WS messages directly into SPDY data frames like so:
> Have 2 bits in SPDY data frames:
> 0x03 = MSG_TYPE : 0 == text, 1 == binary
> 0x04 = MSG_DONE
> A WS text message would simply be mapped into a series of SPDY data frames:
> MSG_TYPE = 0, MSG_DONE = 0
> MSG_TYPE = 0, MSG_DONE = 0
> ...
> MSG_TYPE = 0, MSG_DONE = 1
> Similar for a WS binary message.
>

Doing like above will allow to preserve the current WS API in HTML5 (which
> needs messages, and needs 2 message types).


Thanks,



On Wed, Jan 15, 2014 at 4:09 AM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> Hi Roberto,
>
>
>
> not sure if you agree, but FWIW, here is a summary as I see it (currently):
>
>
>
> 1) WebSocket frame boundaries _can_ have semantic meaning in WebSocket
> _extensions_
>
>
>
> 2) WebSocket message and frame boundaries are indeed preserved in the
> current WS/SPDY proposal,
>
> however, non-final WebSocket frame ends cannot be _immediately_ detected.
>
>
>
> 3) The current WS/SPDY proposal
>
>   * is comfortable to implement for implementations that _only_ want to
> implement WS/SPDY, and not native WS at the same time
>
>  * makes things complicated for WS implementations that want to support
> both WS/SPDY _and_ native WS at the same time
>
>  * has significant wire-level overhead compared to native WS (and compared
> to the proposal below)
>
>
>
> I'd like to suggest a different WS/SPDY:
>
>
>
> Pack WebSocket frames _including_ full WS frame header as _payload_ of
> SPDY data frames only - no SPDY header frames at all.
>
>
>
> In the absence of negotiated WebSocket extensions (or when all WS
> extensions in place do not assign semantics to WS frame boundaries),
>
> fragment WebSocket payload at the WS level such that WS frame
> header+payload fits into the payload of SPDY data frames.
>
>
>
> In the latter case, this essentially adds a small SPDY data frame header
> (of fixed length) to WS frames. Easy to implement and efficient on wire.
>
>
>
> IMHO, this has advantages compared to current proposal and considering 3)
> above.
>
>
>
> The only disadvantage is for implementations that _only_ want to implement
> WS/SPDY and not native WS at the same time.
>
>
>
> ==
>
>
>
> Regarding "permessage-priority" with ordering guarantees: should I find
> time and if there is interest, I'll write something up.
>
> This could be a WebSocket extension. To be useful, the WS API in browsers
> would need to expose it though ..
>
>
>
> Just my 2cts,
>
> Tobias
>
>
>
>
>
>
>
>
>
> *Von:* Roberto Peon [mailto:fenix@google.com]
> *Gesendet:* Dienstag, 14. Januar 2014 19:30
> *An:* Yutaka Hirano
> *Cc:* Tobias Oberstein; Takeshi Yoshino; hybi@ietf.org
>
> *Betreff:* Re: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
>
> I've been lurking anyway :)
>
>
>
> As I understand the proposal, (and it seems that Takeshi and Yukata are
> covering this nicely, so what i say is probably redundant):
>
> Message boundaries, having semantic meaning, are preserved.
>
>
>
> Frame boundaries, having no semantic meaning, are not preserved. I don't
> think it makes sense to worry about this greatly when the ability to extend
> the protocol as implemented in HTTP2 is present. It doesn't make a lot of
> sense to signal the frame boundaries through everything unless there is a
> high-likelyhood and high-utility use-case for such. If such is found, then
> HTTP2 can be revised to include it. This is the approach we've been taking
> with most of the HTTP2 stuff, since there has been great effort placed on
> allowing new versions to be deployed in the future.
>
>
>
> The mapping of WS onto HTTP2 is fairly direct-- we reserved the MSG_DONE
> bit specifically for this kind of use-case (i.e. bidirectional,
> full-duplex, message-based).
>
> The observation made is that there are essentially 3 types of data that
> need to be preserved one way or another:
>
> 1) metadata: data about the data, e.g. text, binary, where it is supposed
> to be going, etc.
>
> 2) payload
>
> 3) boundaries between these.
>
>
>
> imho, metadata is handled with HEADERS (I'd have preferred it to be called
> METADATA, but was voted down at the time :) ), payload is handled with DATA
> frames, and the boundaries are delimited either through the use of HEADERS
> frames, or via the MSG_DONE or STREAM_DONE bits.
>
> In the presence of extensions, assuming that MSG_DONE is required to
> signal the end of a message, then HEADERS can be used to signal any
> boundary desired.
>
>
>
>
>
> I totally agree that it would be nice to expose the priority of various
> messages/channels with respect to each other. That is really an issue that
> is outside the WG's purvue, though.
>
>
>
>
>
> On Thu, Jan 9, 2014 at 7:54 PM, Yutaka Hirano <yhirano@chromium.org>
> wrote:
>
> +Roberto as an HTTP2 expert.
>
>
>
> 2) Ok, both WS frame and message boundaries are preserved.
>
> But: I cannot recognize the end of a WS frame until the next WS frame
> starts (for which I receive a HEADERS frame, which is the implicit signal
> that the previous WS frame ended), right?
>
> You are right, you can't.
>
>
>
>
> And if so, why is there no FRAME_DONE bit on the SPDY frame that
> corresponds to the end of a WS frame?
>
> Given that WebSocket frames "have no semantic meaning" as in RFC6455, I
> don't think the immediate recognition of the end of a WebSocket frame is
> requisite.
>
> I would like to introduce MSG_DONE in HTTP2 so that intermediaries that
> don't fully understand WS over HTTP2 can relay it.
>
> Since there are not so much space in FLAGS field, Introducing FRAME_DONE
> may decrease the extensiveness of the spec.
>
> Do you have a strong preference to introduce FRAME_DONE?
>
>
>
>
>
> On Thu, Jan 9, 2014 at 6:53 PM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
>
> Hi Yutaka,
>
>
>
> 1) Yes, I understand. Thanks!
>
>
>
> 2) Ok, both WS frame and message boundaries are preserved.
>
>
>
> But: I cannot recognize the end of a WS frame until the next WS frame
> starts (for which I receive a HEADERS frame, which is the implicit signal
> that the previous WS frame ended), right?
>
>
>
> And if so, why is there no FRAME_DONE bit on the SPDY frame that
> corresponds to the end of a WS frame?
>
>
>
> 3) Ok, great! Hope this works, since SPDY provides header compression, but
> no sophisticated bidirectional payload compression as "WS
> permessage-deflate" does.
>
>
>
> Cheers,
>
> /Tobias
>
>
>
>
>
>
>
> *Von:* yhirano@google.com [mailto:yhirano@google.com] *Im Auftrag von *Yutaka
> Hirano
> *Gesendet:* Donnerstag, 9. Januar 2014 10:37
> *An:* Tobias Oberstein
> *Cc:* Takeshi Yoshino; hybi@ietf.org
> *Betreff:* Re: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
>
> Hi Tobias,
>
> Thank you for the comments.
>
>
>
> 1) Frame Mapping
> """
> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
> following plural SPDY data frames. The SPDY HEADERS frame must contain
> WebSocket frame fields and following SPDY data frames contain payload data.
> These data frames are free to be reframed.
> """
>
> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
> So reframing of WebSocket frames (the payload) into multiple SPDY data
> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
> is free to choose the boundaries within WS frame payload at which to
> reframe into different SPDY data frames as it likes?
>
> A WebSocket frame MAY be split to multiple SPDY DATA frames.
>
> Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA frame
> into multiple DATA frames.
>
> Is this answering your question?
>
>
>
> 2) End of WS Frame
> The end of a WS message is signaled with a SPDY data frame having MSG_DONE
> set.
> How is the end of a WS frame signaled, given that the WS frame may have
> been reframed into multiple SPDY data frames?
>
> When you receive a HEADERS frame, you can recognize that the last SPDY
> frame is that last frame in a WebSocket frame.
>
> When you receive a SPDY frame with MSG_DONE, you can recognize that it is
> the last SPDY frame in a WebSocket message and hence the last SPDY frame in
> a WebSocket frame.
>
>
>
> 3) WS extension
> Are WS extensions fully supported? In particular, is "permessage-deflate"
> supported when running WS/SPDY?
>
> Yes. We have reserved bits and we can emulate the opening handshake, I
> think.
>
>
>
> On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
>
> Hi Takeshi,
>
> it's a pity, but I can understand the motivation to shove everything into
> HTTP2.
>
> Anyway, some questions:
>
> 1) Frame Mapping
> """
> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
> following plural SPDY data frames. The SPDY HEADERS frame must contain
> WebSocket frame fields and following SPDY data frames contain payload data.
> These data frames are free to be reframed.
> """
>
> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
>
> So reframing of WebSocket frames (the payload) into multiple SPDY data
> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
> is free to choose the boundaries within WS frame payload at which to
> reframe into different SPDY data frames as it likes?
>
> 2) End of WS Frame
>
> The end of a WS message is signaled with a SPDY data frame having MSG_DONE
> set.
>
> How is the end of a WS frame signaled, given that the WS frame may have
> been reframed into multiple SPDY data frames?
>
> 3) WS extension
>
> Are WS extensions fully supported? In particular, is "permessage-deflate"
> supported when running WS/SPDY?
>
> Thanks!
> /Tobias
>
>
> Von: hybi [mailto:hybi-bounces@ietf.org] Im Auftrag von Takeshi Yoshino
> Gesendet: Dienstag, 7. Januar 2014 08:36
> An: hybi@ietf.org
> Betreff: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
> Hi all,
>
> We've been working on mux extension to make WebSocket scalable for long
> time, but as SPDY got widely supported and adopted quickly, now there seems
> to be no longer any good reason to continue this work which is WebSocket
> dedicated.
>
> Scalability is still important issue to address. From now, I'd like to
> focus on WS/SPDY effort which we've been working on in parallel with the
> mux extension.
>
> Please take a look at the latest proposal in this post.
> https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc
>
> If you really want the dedicated mux extension than one over SPDY, sorry
> but please take over the work.
>
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>
>
>
>
>
>
>

--047d7bea2f402bb47404f0617e69
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div>Tobias, let me confirm.</div><div>=C2=A0</div><div><b=
lockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-le=
ft-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;pad=
ding-left:1ex">
<span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,7=
3,125)">I&#39;d like to suggest a different WS/SPDY:<br></span><span style=
=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,73,125)">=C2=
=A0<br>
</span>Pack WebSocket frames _including_ full WS frame header as _payload_ =
of SPDY data frames only - no SPDY header frames at all.</blockquote><p sty=
le=3D"font-family:arial,sans-serif;font-size:13px"><u></u></p></div><div>
My understanding is that by suggesting the above one you retracted the belo=
w proposal you made before, is that right?</div><div><br></div><div>On Fri,=
 Jan 10, 2014 at 5:37 PM, Tobias Oberstein=C2=A0<span dir=3D"ltr">&lt;<a hr=
ef=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstei=
n@tavendo.de</a>&gt;</span>=C2=A0wrote:<br>
</div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;b=
order-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:s=
olid;padding-left:1ex"><span style=3D"font-family:arial,sans-serif;font-siz=
e:13px">Why not map WS messages directly into SPDY data frames like so:</sp=
an><br style=3D"font-family:arial,sans-serif;font-size:13px">
<span style=3D"font-family:arial,sans-serif;font-size:13px">Have 2 bits in =
SPDY data frames:</span><br style=3D"font-family:arial,sans-serif;font-size=
:13px"><span style=3D"font-family:arial,sans-serif;font-size:13px">0x03 =3D=
 MSG_TYPE : 0 =3D=3D text, 1 =3D=3D binary<br>
</span><span style=3D"font-family:arial,sans-serif;font-size:13px">0x04 =3D=
 MSG_DONE</span><br style=3D"font-family:arial,sans-serif;font-size:13px"><=
span style=3D"font-family:arial,sans-serif;font-size:13px">A WS text messag=
e would simply be mapped into a series of SPDY data frames:</span><br style=
=3D"font-family:arial,sans-serif;font-size:13px">
<span style=3D"font-family:arial,sans-serif;font-size:13px">MSG_TYPE =3D 0,=
 MSG_DONE =3D 0<br></span><span style=3D"font-family:arial,sans-serif;font-=
size:13px">MSG_TYPE =3D 0, MSG_DONE =3D 0<br></span><span style=3D"font-fam=
ily:arial,sans-serif;font-size:13px">...<br>
</span><span style=3D"font-family:arial,sans-serif;font-size:13px">MSG_TYPE=
 =3D 0, MSG_DONE =3D 1</span><br style=3D"font-family:arial,sans-serif;font=
-size:13px"><span style=3D"font-family:arial,sans-serif;font-size:13px">Sim=
ilar for a WS binary message.<br>
</span><span style=3D"font-family:arial,sans-serif;font-size:13px">=C2=A0</=
span>=C2=A0</blockquote><blockquote class=3D"gmail_quote" style=3D"margin:0=
px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);b=
order-left-style:solid;padding-left:1ex">
<span style=3D"font-family:arial,sans-serif;font-size:13px"></span><span st=
yle=3D"font-family:arial,sans-serif;font-size:13px">Doing like above will a=
llow to preserve the current WS API in HTML5 (which needs messages, and nee=
ds 2 message types).</span></blockquote>
<div><br></div><div>Thanks,</div><div>=C2=A0</div></div><div class=3D"gmail=
_extra"><br><br><div class=3D"gmail_quote">On Wed, Jan 15, 2014 at 4:09 AM,=
 Tobias Oberstein <span dir=3D"ltr">&lt;<a href=3D"mailto:tobias.oberstein@=
tavendo.de" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt;</span> wr=
ote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div lang=3D"DE" link=3D"blue" vlink=3D"purp=
le"><div><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family=
:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Hi Roberto,<u></=
u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">not sure if you agr=
ee, but FWIW, here is a summary as I see it (currently):<u></u><u></u></spa=
n></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">1) WebSocket frame =
boundaries _can_ have semantic meaning in WebSocket _extensions_<u></u><u><=
/u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">2) WebSocket messag=
e and frame boundaries are indeed preserved in the current WS/SPDY proposal=
,<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">however, non-final WebSoc=
ket frame ends cannot be _immediately_ detected.<u></u><u></u></span></p><p=
 class=3D"MsoNormal">
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></span></p><p class=3D"MsoNo=
rmal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot=
;sans-serif&quot;;color:#1f497d">3) The current WS/SPDY proposal<u></u><u><=
/u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0 * is comfortable t=
o implement for implementations that _only_ want to implement WS/SPDY, and =
not native WS at the same time<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"> =C2=A0* makes things com=
plicated for WS implementations that want to support both WS/SPDY _and_ nat=
ive WS at the same time<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"> =C2=A0* has significant =
wire-level overhead compared to native WS (and compared to the proposal bel=
ow)<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">I&#39;d like to sug=
gest a different WS/SPDY:<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p>Pack WebSocket frames _including_ full WS frame header as _payload=
_ of SPDY data frames only - no SPDY header frames at all.<u></u><u></u></p=
>
<p><u></u>=C2=A0<u></u></p><p>In the absence of negotiated WebSocket extens=
ions (or when all WS extensions in place do not assign semantics to WS fram=
e boundaries),<u></u><u></u></p><p>fragment WebSocket payload at the WS lev=
el such that WS frame header+payload fits into the payload of SPDY data fra=
mes.<u></u><u></u></p>
<p><u></u>=C2=A0<u></u></p><p>In the latter case, this essentially adds a s=
mall SPDY data frame header (of fixed length) to WS frames. Easy to impleme=
nt and efficient on wire.<u></u><u></u></p><p><u></u>=C2=A0<u></u></p><p>IM=
HO, this has advantages compared to current proposal and considering 3) abo=
ve.<u></u><u></u></p>
<p><u></u>=C2=A0<u></u></p><p>The only disadvantage is for implementations =
that _only_ want to implement WS/SPDY and not native WS at the same time.<u=
></u><u></u></p><p><u></u>=C2=A0<u></u></p><p>=3D=3D<u></u><u></u></p><p><u=
></u>=C2=A0<u></u></p>
<p>Regarding &quot;permessage-priority&quot; with ordering guarantees: shou=
ld I find time and if there is interest, I&#39;ll write something up.<u></u=
><u></u></p><p>This could be a WebSocket extension. To be useful, the WS AP=
I in browsers would need to expose it though ..<u></u><u></u></p>
<p><u></u>=C2=A0<u></u></p><p>Just my 2cts,<u></u><u></u></p><p>Tobias<u></=
u><u></u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-fa=
mily:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=
=A0<u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u=
></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><div style=3D"border:none;border-left:solid blue 1.5pt;padding:0cm 0c=
m 0cm 4.0pt">
<div><div style=3D"border:none;border-top:solid #e1e1e1 1.0pt;padding:3.0pt=
 0cm 0cm 0cm"><p class=3D"MsoNormal"><b><span style=3D"font-size:11.0pt;fon=
t-family:&quot;Calibri&quot;,&quot;sans-serif&quot;">Von:</span></b><span s=
tyle=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&q=
uot;"> Roberto Peon [mailto:<a href=3D"mailto:fenix@google.com" target=3D"_=
blank">fenix@google.com</a>] <br>
<b>Gesendet:</b> Dienstag, 14. Januar 2014 19:30<br><b>An:</b> Yutaka Hiran=
o<br><b>Cc:</b> Tobias Oberstein; Takeshi Yoshino; <a href=3D"mailto:hybi@i=
etf.org" target=3D"_blank">hybi@ietf.org</a></span></p><div><div class=3D"h=
5">
<br><b>Betreff:</b> Re: [hybi] Discontinuation of mux standardizaton in fav=
or of WS/HTTP/2.0<u></u><u></u></div></div><p></p></div></div><div><div cla=
ss=3D"h5"><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><div><p class=3D"M=
soNormal">
I&#39;ve been lurking anyway :)<u></u><u></u></p><div><p class=3D"MsoNormal=
"><u></u>=C2=A0<u></u></p></div><div><p class=3D"MsoNormal">As I understand=
 the proposal, (and it seems that Takeshi and Yukata are covering this nice=
ly, so what i say is probably redundant):<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">Message boundaries, having semantic meani=
ng, are preserved.<u></u><u></u></p></div><div><p class=3D"MsoNormal"><u></=
u>=C2=A0<u></u></p></div><div><p class=3D"MsoNormal">Frame boundaries, havi=
ng no semantic meaning, are not preserved. I don&#39;t think it makes sense=
 to worry about this greatly when the ability to extend the protocol as imp=
lemented in HTTP2 is present. It doesn&#39;t make a lot of sense to signal =
the frame boundaries through everything unless there is a high-likelyhood a=
nd high-utility use-case for such. If such is found, then HTTP2 can be revi=
sed to include it. This is the approach we&#39;ve been taking with most of =
the HTTP2 stuff, since there has been great effort placed on allowing new v=
ersions to be deployed in the future.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div><p cla=
ss=3D"MsoNormal">The mapping of WS onto HTTP2 is fairly direct-- we reserve=
d the MSG_DONE bit specifically for this kind of use-case (i.e. bidirection=
al, full-duplex, message-based).<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">The observation made is that there are es=
sentially 3 types of data that need to be preserved one way or another:<u><=
/u><u></u></p></div><div><p class=3D"MsoNormal">1) metadata: data about the=
 data, e.g. text, binary, where it is supposed to be going, etc.<u></u><u><=
/u></p>
</div><div><p class=3D"MsoNormal">2) payload<u></u><u></u></p></div><div><p=
 class=3D"MsoNormal">3) boundaries between these.<u></u><u></u></p></div><d=
iv><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div><p class=3D"Ms=
oNormal">
imho, metadata is handled with HEADERS (I&#39;d have preferred it to be cal=
led METADATA, but was voted down at the time :) ), payload is handled with =
DATA frames, and the boundaries are delimited either through the use of HEA=
DERS frames, or via the MSG_DONE or STREAM_DONE bits.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">In the presence of extensions, assuming t=
hat MSG_DONE is required to signal the end of a message, then HEADERS can b=
e used to signal any boundary desired.<u></u><u></u></p></div><div><p class=
=3D"MsoNormal">
<u></u>=C2=A0<u></u></p></div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u><=
/u></p></div><div><p class=3D"MsoNormal">I totally agree that it would be n=
ice to expose the priority of various messages/channels with respect to eac=
h other. That is really an issue that is outside the WG&#39;s purvue, thoug=
h.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div></div><div>=
<p class=3D"MsoNormal" style=3D"margin-bottom:12.0pt"><u></u>=C2=A0<u></u><=
/p><div><p class=3D"MsoNormal">On Thu, Jan 9, 2014 at 7:54 PM, Yutaka Hiran=
o &lt;<a href=3D"mailto:yhirano@chromium.org" target=3D"_blank">yhirano@chr=
omium.org</a>&gt; wrote:<u></u><u></u></p>
<blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0c=
m 0cm 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-=
bottom:5.0pt"><div><div><p class=3D"MsoNormal">+Roberto as an HTTP2 expert.=
<u></u><u></u></p>
</div><div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><block=
quote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm =
0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom=
:5.0pt">
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">2) Ok, both WS frame and =
message boundaries are preserved.<br>=C2=A0<br>But: I cannot recognize the =
end of a WS frame until the next WS frame starts (for which I receive a HEA=
DERS frame, which is the implicit signal that the previous WS frame ended),=
 right?</span><u></u><u></u></p>
</blockquote></div><p class=3D"MsoNormal">You are right, you can&#39;t.<u><=
/u><u></u></p><div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></di=
v><blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:=
0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margi=
n-bottom:5.0pt">
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0<br>And if so, why =
is there no FRAME_DONE bit on the SPDY frame that corresponds to the end of=
 a WS frame?</span><u></u><u></u></p>
</blockquote></div><p class=3D"MsoNormal">Given that WebSocket frames &quot=
;have no semantic meaning&quot; as in RFC6455, I don&#39;t think the immedi=
ate recognition of the end of a WebSocket frame is requisite.<u></u><u></u>=
</p>
<div><p class=3D"MsoNormal">I would like to introduce MSG_DONE in HTTP2 so =
that intermediaries that don&#39;t fully understand WS over HTTP2 can relay=
 it.<u></u><u></u></p></div><div><p class=3D"MsoNormal">Since there are not=
 so much space in FLAGS field, Introducing FRAME_DONE may decrease the exte=
nsiveness of the spec.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">Do you have a strong preference to introd=
uce FRAME_DONE?<u></u><u></u></p></div><div><p class=3D"MsoNormal"><u></u>=
=C2=A0<u></u></p></div></div><div><div><div><p class=3D"MsoNormal" style=3D=
"margin-bottom:12.0pt">
<u></u>=C2=A0<u></u></p><div><p class=3D"MsoNormal">On Thu, Jan 9, 2014 at =
6:53 PM, Tobias Oberstein &lt;<a href=3D"mailto:tobias.oberstein@tavendo.de=
" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt; wrote:<u></u><u></u=
></p><blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;paddi=
ng:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;ma=
rgin-bottom:5.0pt">
<div><div><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-famil=
y:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Hi Yutaka,</spa=
n><u></u><u></u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;=
font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=
=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">1) Yes, I understand. Tha=
nks!</span><u></u><u></u></p><p class=3D"MsoNormal"><span style=3D"font-siz=
e:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f49=
7d">=C2=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">2) Ok, both WS frame and =
message boundaries are preserved.</span><u></u><u></u></p><p class=3D"MsoNo=
rmal">
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></u></p><p class=3D"MsoNo=
rmal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot=
;sans-serif&quot;;color:#1f497d">But: I cannot recognize the end of a WS fr=
ame until the next WS frame starts (for which I receive a HEADERS frame, wh=
ich is the implicit signal that the previous WS frame ended), right?</span>=
<u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">And if so, why is t=
here no FRAME_DONE bit on the SPDY frame that corresponds to the end of a W=
S frame?</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">3) Ok, great! Hope =
this works, since SPDY provides header compression, but no sophisticated bi=
directional payload compression as &quot;WS permessage-deflate&quot; does.<=
/span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Cheers,</span><u></=
u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">/Tobias</span><u></u><u><=
/u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&=
quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></=
u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u=
><u></u></p>
<div style=3D"border:none;border-left:solid blue 1.5pt;padding:0cm 0cm 0cm =
4.0pt"><div><div style=3D"border:none;border-top:solid #e1e1e1 1.0pt;paddin=
g:3.0pt 0cm 0cm 0cm"><p class=3D"MsoNormal"><b><span style=3D"font-size:11.=
0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;">Von:</span></b>=
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;"> </span><a href=3D"mailto:yhirano@google.com" target=3D"_blank=
"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;san=
s-serif&quot;">yhirano@google.com</span></a><span style=3D"font-size:11.0pt=
;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;"> [mailto:</span><a=
 href=3D"mailto:yhirano@google.com" target=3D"_blank"><span style=3D"font-s=
ize:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;">yhirano@=
google.com</span></a><span style=3D"font-size:11.0pt;font-family:&quot;Cali=
bri&quot;,&quot;sans-serif&quot;">] <b>Im Auftrag von </b>Yutaka Hirano<br>
<b>Gesendet:</b> Donnerstag, 9. Januar 2014 10:37<br><b>An:</b> Tobias Ober=
stein<br><b>Cc:</b> Takeshi Yoshino; </span><a href=3D"mailto:hybi@ietf.org=
" target=3D"_blank"><span style=3D"font-size:11.0pt;font-family:&quot;Calib=
ri&quot;,&quot;sans-serif&quot;">hybi@ietf.org</span></a><span style=3D"fon=
t-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;"><br>
<b>Betreff:</b> Re: [hybi] Discontinuation of mux standardizaton in favor o=
f WS/HTTP/2.0</span><u></u><u></u></p></div></div><div><div><p class=3D"Mso=
Normal">=C2=A0<u></u><u></u></p><div><p class=3D"MsoNormal">Hi Tobias,<u></=
u><u></u></p>
<div><p class=3D"MsoNormal">Thank you for the comments.<u></u><u></u></p></=
div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p><blockquote style=
=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm 6.0pt;m=
argin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom:5.0pt">
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ar=
ial&quot;,&quot;sans-serif&quot;">1) Frame Mapping<br>&quot;&quot;&quot;<br=
>One WebSocket frame is mapped into one preceding SPDY HEADERS frame and fo=
llowing plural SPDY data frames. The SPDY HEADERS frame must contain WebSoc=
ket frame fields and following SPDY data frames contain payload data. These=
 data frames are free to be reframed.<br>
&quot;&quot;&quot;<br></span><a href=3D"https://docs.google.com/document/d/=
124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank"><span =
style=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&qu=
ot;">https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw=
9aWzSHijk/edit</span></a><span style=3D"font-size:10.0pt;font-family:&quot;=
Arial&quot;,&quot;sans-serif&quot;"><br>
So reframing of WebSocket frames (the payload) into multiple SPDY data fram=
es MAY occur, but does not need to occur? And a WS/SPDY implementation is f=
ree to choose the boundaries within WS frame payload at which to reframe in=
to different SPDY data frames as it likes?</span><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal">A WebSocket frame MAY be split to =
multiple SPDY DATA frames.<u></u><u></u></p></div><div><p class=3D"MsoNorma=
l">Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA fram=
e into multiple DATA frames.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">Is this answering your question?<u></u><u=
></u></p></div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><b=
lockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm =
0cm 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bo=
ttom:5.0pt">
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ar=
ial&quot;,&quot;sans-serif&quot;">2) End of WS Frame<br>The end of a WS mes=
sage is signaled with a SPDY data frame having MSG_DONE set.<br>How is the =
end of a WS frame signaled, given that the WS frame may have been reframed =
into multiple SPDY data frames?</span><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal">When you receive a HEADERS frame, =
you can recognize that the last SPDY frame is that last frame in a WebSocke=
t frame.<u></u><u></u></p></div><div><p class=3D"MsoNormal">When you receiv=
e a SPDY frame with MSG_DONE, you can recognize that it is the last SPDY fr=
ame in a WebSocket message and hence the last SPDY frame in a WebSocket fra=
me.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><blockquote=
 style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm 6=
.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom:5.0p=
t"><p class=3D"MsoNormal">
<span style=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-se=
rif&quot;">3) WS extension<br>Are WS extensions fully supported? In particu=
lar, is &quot;permessage-deflate&quot; supported when running WS/SPDY?</spa=
n><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal">Yes. We have reserved bits and we =
can emulate the opening handshake, I think.<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><div><p class=3D"MsoNorma=
l">On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein &lt;<a href=3D"mailto:t=
obias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstein@tavendo.de</=
a>&gt; wrote:<u></u><u></u></p>
<blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0c=
m 0cm 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-=
bottom:5.0pt"><p class=3D"MsoNormal">Hi Takeshi,<br><br>it&#39;s a pity, bu=
t I can understand the motivation to shove everything into HTTP2.<br>
<br>Anyway, some questions:<br><br>1) Frame Mapping<br>&quot;&quot;&quot;<b=
r>One WebSocket frame is mapped into one preceding SPDY HEADERS frame and f=
ollowing plural SPDY data frames. The SPDY HEADERS frame must contain WebSo=
cket frame fields and following SPDY data frames contain payload data. Thes=
e data frames are free to be reframed.<br>
&quot;&quot;&quot;<br><a href=3D"https://docs.google.com/document/d/124vB5N=
V1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank">https://docs.=
google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit</a>=
<br>
<br>So reframing of WebSocket frames (the payload) into multiple SPDY data =
frames MAY occur, but does not need to occur? And a WS/SPDY implementation =
is free to choose the boundaries within WS frame payload at which to refram=
e into different SPDY data frames as it likes?<br>
<br>2) End of WS Frame<br><br>The end of a WS message is signaled with a SP=
DY data frame having MSG_DONE set.<br><br>How is the end of a WS frame sign=
aled, given that the WS frame may have been reframed into multiple SPDY dat=
a frames?<br>
<br>3) WS extension<br><br>Are WS extensions fully supported? In particular=
, is &quot;permessage-deflate&quot; supported when running WS/SPDY?<br><br>=
Thanks!<br>/Tobias<br><br><br>Von: hybi [mailto:<a href=3D"mailto:hybi-boun=
ces@ietf.org" target=3D"_blank">hybi-bounces@ietf.org</a>] Im Auftrag von T=
akeshi Yoshino<br>
Gesendet: Dienstag, 7. Januar 2014 08:36<br>An: <a href=3D"mailto:hybi@ietf=
.org" target=3D"_blank">hybi@ietf.org</a><br>Betreff: [hybi] Discontinuatio=
n of mux standardizaton in favor of WS/HTTP/2.0<u></u><u></u></p><div><div>
<p class=3D"MsoNormal"><br>Hi all,<br><br>We&#39;ve been working on mux ext=
ension to make WebSocket scalable for long time, but as SPDY got widely sup=
ported and adopted quickly, now there seems to be no longer any good reason=
 to continue this work which is WebSocket dedicated.<br>
<br>Scalability is still important issue to address. From now, I&#39;d like=
 to focus on WS/SPDY effort which we&#39;ve been working on in parallel wit=
h the mux extension.<br><br>Please take a look at the latest proposal in th=
is post.<br>
<a href=3D"https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc" ta=
rget=3D"_blank">https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3gl=
c</a><br><br>If you really want the dedicated mux extension than one over S=
PDY, sorry but please take over the work.<u></u><u></u></p>
</div></div><div><div><p class=3D"MsoNormal">______________________________=
_________________<br>hybi mailing list<br><a href=3D"mailto:hybi@ietf.org" =
target=3D"_blank">hybi@ietf.org</a><br><a href=3D"https://www.ietf.org/mail=
man/listinfo/hybi" target=3D"_blank">https://www.ietf.org/mailman/listinfo/=
hybi</a><u></u><u></u></p>
</div></div></blockquote></div><p class=3D"MsoNormal">=C2=A0<u></u><u></u><=
/p></div></div></div></div></div></div></div></blockquote></div><p class=3D=
"MsoNormal"><u></u>=C2=A0<u></u></p></div></div></div></blockquote></div><p=
 class=3D"MsoNormal">
<u></u>=C2=A0<u></u></p></div></div></div></div></div></div></blockquote></=
div><br></div>

--047d7bea2f402bb47404f0617e69--


From tyoshino@google.com  Mon Jan 20 00:35:58 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 095F41A009F for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 00:35:58 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.913
X-Spam-Level: 
X-Spam-Status: No, score=-1.913 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id tyz9rrq6PsDa for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 00:35:56 -0800 (PST)
Received: from mail-wg0-x234.google.com (mail-wg0-x234.google.com [IPv6:2a00:1450:400c:c00::234]) by ietfa.amsl.com (Postfix) with ESMTP id 09E0E1A00A4 for <hybi@ietf.org>; Mon, 20 Jan 2014 00:35:55 -0800 (PST)
Received: by mail-wg0-f52.google.com with SMTP id b13so6633766wgh.7 for <hybi@ietf.org>; Mon, 20 Jan 2014 00:35:55 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=KSZcIMgqda4VCuX6GVTvQ0QhyPOCBZ4GhwV1S34SJkw=; b=ZqO/7H81zIPH8eeN/zsObbbQNLEau5ir0Rk854qa2bcmKt4zLObP9Iff+dRsFck6V3 yXgpZ47LwwMseawjtq/7gk5CLZ77IcpYna71+79apX91HwcgIgxUDlKmTtbanSr8tjQI GNrKy2BPLNZ5dAfOfmq/dOgPThLQsro6NThBSTf9mTMah/pz5AJHIrO8ams2dV9x6TaA oTepns2QT2hdAfeQy5Gg3jyzYjoBzVK4ZWIY7eg1OkxMICsoYEMtRrMKe3E6DtGfiF4P JjypftGs0hq333kdudEcyoEXYqIUgD1zjDNDvY5KGM+LBSYGwC292MR1ey8a6ZgFIBGL /iRQ==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=KSZcIMgqda4VCuX6GVTvQ0QhyPOCBZ4GhwV1S34SJkw=; b=Vjxj+adNUhFTCO3aNae+aMa7x3eTBDYtecmuzL+Tt2L9aDwbY4OUxNbck3hJjsK7Wn IUbqKHFU/ec5NJVDYPk7Vjg44vF4NmhlRQLb0UrxuV81YKRCCl9lGXcI5AQMF+qWvG0M jV+Xif86Eb/ygDJLHMhrAd8CcuqKGsPHPcdWRtKmx6wbFmYRzjz3F1Wb0j1aSjyTb5qs 6idVEgA/wEOwj14oFF3mP/3zyzSqNSSDARMshR5N9geiTVqaB+QBvNpZxAkjHkBydyYw cWxCtvg40ChXLzsKDJlnGtmiX3VCHc6inNo5BoaOfSoME1KtsQ9vwkNX+zQUpYHM09CF cvJA==
X-Gm-Message-State: ALoCoQl1Jp86D1Vy/pQ5NrTVgWYvyJ4CGI0FR4p3o6s7L9jnpugZdct6DvNO5CrbVZMTg7RJWTxUC6AcZ64jVTL+gCPlU22I9+f0lmcyO7gsidCb0e7X4GKIG/zQqPibE9XtL8whsqxHJA5Kr75XzFoJT49TBgT/sq6cstKRKTz5DiU92ttTNHD8YxAVMUWQ1jCsqokweuRf
X-Received: by 10.194.63.228 with SMTP id j4mr13376984wjs.34.1390206955791; Mon, 20 Jan 2014 00:35:55 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Mon, 20 Jan 2014 00:35:34 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BF9949D@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF9948F@EXVMBX020-12.exch020.serverdata.net> <52D90F99.6080205@gmx.de> <634914A010D0B943A035D226786325D4446BF9949D@EXVMBX020-12.exch020.serverdata.net>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Mon, 20 Jan 2014 17:35:34 +0900
Message-ID: <CAH9hSJZQ3NxVW36PnZ0TpMF4tPPaJLr12M8NtbPb6pUejf1wcQ@mail.gmail.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=047d7ba97f3ac3cee504f062c71b
Cc: Julian Reschke <julian.reschke@gmx.de>, "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] WebSocket subprotocol parameters
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 08:35:58 -0000

--047d7ba97f3ac3cee504f062c71b
Content-Type: text/plain; charset=UTF-8

On Fri, Jan 17, 2014 at 10:01 PM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> > > Was there any good reason disallowing parameters for "Sec-WebSocket-
> > Protocol"?
> >
> > I think that was discussed and the decision was made on purpose. For
> details
> > you'll have to visit the mailing list archives though.
>
> I can't find it .. searching the Hybi archives is a pain ..


I can't recall such discussion.

In this thread, Brodie used an example similar to that, but the main topic
of the post was to update the spec to include multiple subprotocol offer
mechanism.
http://www.ietf.org/mail-archive/web/hybi/current/msg04467.html

In this long thread, what character should be used for subprotocol is
discussed. Some people preferred more liberal rule, but no one discussed
qvalue-like grammar.
http://www.ietf.org/mail-archive/web/hybi/current/msg00913.html

Subprotocol History FYI:

We added the constraint "U+0021 ..  U+007E" from
http://tools.ietf.org/html/draft-hixie-thewebsocketprotocol-31 -> 32

Single value -> multiple value (SP-separated)
http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-02 -> 03

We made it follow 1#(token | quoted-string) ABNF from
http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-03 -> 04

Changed from (token | quoted-string) to token
http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-07

--047d7ba97f3ac3cee504f062c71b
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On F=
ri, Jan 17, 2014 at 10:01 PM, Tobias Oberstein <span dir=3D"ltr">&lt;<a hre=
f=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstein=
@tavendo.de</a>&gt;</span> wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-=
left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;p=
adding-left:1ex"><div class=3D"im">&gt; &gt; Was there any good reason disa=
llowing parameters for &quot;Sec-WebSocket-<br>


&gt; Protocol&quot;?<br>
&gt;<br>
&gt; I think that was discussed and the decision was made on purpose. For d=
etails<br>
&gt; you&#39;ll have to visit the mailing list archives though.<br>
<br>
</div>I can&#39;t find it .. searching the Hybi archives is a pain ..</bloc=
kquote><div><br></div><div>I can&#39;t recall such discussion.</div><div><b=
r></div><div>In this thread, Brodie used an example similar to that, but th=
e main topic of the post was to update the spec to include multiple subprot=
ocol offer mechanism.</div>

<div><a href=3D"http://www.ietf.org/mail-archive/web/hybi/current/msg04467.=
html">http://www.ietf.org/mail-archive/web/hybi/current/msg04467.html</a><b=
r></div><div><br></div><div>In this long thread, what character should be u=
sed for subprotocol is discussed. Some people preferred more liberal rule, =
but no one discussed qvalue-like grammar.</div>

<div><a href=3D"http://www.ietf.org/mail-archive/web/hybi/current/msg00913.=
html">http://www.ietf.org/mail-archive/web/hybi/current/msg00913.html</a><b=
r></div><div><br></div><div>Subprotocol History FYI:</div><div><br></div>

<div>We added the constraint &quot;U+0021 .. =C2=A0U+007E&quot; from</div><=
a href=3D"http://tools.ietf.org/html/draft-hixie-thewebsocketprotocol-31">h=
ttp://tools.ietf.org/html/draft-hixie-thewebsocketprotocol-31</a> -&gt; 32<=
div>

<br></div><div>Single value -&gt; multiple value (SP-separated)</div><div><=
a href=3D"http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-0=
2">http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-02</a> -=
&gt; 03<br>

</div><div><br></div><div>We made it follow 1#(token | quoted-string) ABNF =
from</div><div><a href=3D"http://tools.ietf.org/html/draft-ietf-hybi-theweb=
socketprotocol-03">http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketp=
rotocol-03</a> -&gt; 04</div>

<div><br></div><div>Changed from (token | quoted-string) to token</div><div=
><a href=3D"http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol=
-07">http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-07</a>=
=C2=A0</div>

</div></div></div>

--047d7ba97f3ac3cee504f062c71b--

From yhirano@google.com  Mon Jan 20 02:37:30 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 00CA61A00F6 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 02:37:30 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.913
X-Spam-Level: 
X-Spam-Status: No, score=-1.913 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id LPFUVVpxB0aR for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 02:37:25 -0800 (PST)
Received: from mail-qa0-x22e.google.com (mail-qa0-x22e.google.com [IPv6:2607:f8b0:400d:c00::22e]) by ietfa.amsl.com (Postfix) with ESMTP id 288701A00F5 for <hybi@ietf.org>; Mon, 20 Jan 2014 02:37:25 -0800 (PST)
Received: by mail-qa0-f46.google.com with SMTP id ii20so5311479qab.19 for <hybi@ietf.org>; Mon, 20 Jan 2014 02:37:25 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=3HqrYRzUFBBQ8A6rwQ6D84zAdXF3kEUa8tJK/mjMHH4=; b=nl7U2YxhLRVijzJCvqvhsZ68TrOgwhUy2ghGst5oF9t3I0eKpAIIhR8W4UZrJbHn6d WQzFYsq+VSmRWJcifi0j2HV0Vm/RpLPXsoRXEkW+ID96Xlgtx8uW6IOwgP63c2ngBTW8 WxNznhQqnyPV3TL+LGBBCsymEsnWgFHviiOorLJriCd3u806LSWkGF2SmXAn33qSx7Bc B9gvUlO5YoQ9IjA/DA9D28Ll1fFfox/sPYH+h5ZeQ3uH3q+H6Xd0F4wWpevVHbnNNZrh FkvWnSO9WvSs6rQ/NKk7c+4En3hq+c7ojpmNMNDgnCVRKxG7xc6BgoCe5OqkZVczKM87 faJg==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=3HqrYRzUFBBQ8A6rwQ6D84zAdXF3kEUa8tJK/mjMHH4=; b=XHK75Hjt6YRZh2TqsOG3p4eCyeTr22OdhG8NKlFuMXOvl9tsBmZBTxijun4HhHMlfu Afgve5Z2ISdwkJrnF7afrpdu7Q2dO8AxIh5PAv+W9B/hcUyJk07vnJJSEGRFa4djU3Wg ktvVkr0JS3hm4WM/Xd/X9le40xz/o6Os34gnY=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=3HqrYRzUFBBQ8A6rwQ6D84zAdXF3kEUa8tJK/mjMHH4=; b=T394wtTiLEb9KGYsd8Qh8GmHYUTWSnhwj1Z/FLDazR7bsIUaF3OhvjIqFq9rpQEjf5 3VWlgRgzvW79bquRma8JHvJzN0i7TLNLjdLO+AB+rJH47ZGVRj9bu3n469yjTJ1bch3r 7OJGw0KLJ33EIXmvnwzF6MTSTk4c18xpgJtH3yjjw7oeN9WIeGwOYGGQeexBMSYxVvNW gSQl7pWvxMZYj1smLWQFDzUmlq3M+F7RZVuwq20oe/VPl+Ur93hxm5JODpHPyi/92t34 ds41VR3bEF2x/SkAWIU8T+a6HPY07keqj+JY0MAxJ1vUsuOwhWi2gpzSLOKw3R7T9mBH pjdw==
X-Gm-Message-State: ALoCoQm8PUJ78QzTGl5Wj6JkFyxuvvHfYRMM2IOjw9f+85hnsyCP9QGvLSWOIKfNy8ppejQ/KfkbOPFgT6GTldTwEbHfReIS43laAPgtyXMymBl42fwZW1AWNMRGDP12t/OpzV3SV6MptOmmLgjprPrpUvwZzWV/b9G4KApBbO7c6rstc6Qwdmi0ekWloRS65U+WHoQu9Vxc
MIME-Version: 1.0
X-Received: by 10.224.72.72 with SMTP id l8mr9721543qaj.51.1390214245194; Mon, 20 Jan 2014 02:37:25 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.140.109.202 with HTTP; Mon, 20 Jan 2014 02:37:25 -0800 (PST)
In-Reply-To: <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com>
Date: Mon, 20 Jan 2014 19:37:25 +0900
X-Google-Sender-Auth: k20IrmqXMWOZSeA4xAiBCWJ8RIo
Message-ID: <CABihn6GpedG+2m69BK3znx+19jxU-sfo+At_RTnQHuvqw_ec3g@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: Roberto Peon <fenix@google.com>
Content-Type: multipart/alternative; boundary=047d7bf0e1f63f322504f0647a26
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 10:37:30 -0000

--047d7bf0e1f63f322504f0647a26
Content-Type: text/plain; charset=UTF-8

Hi Roberto,

The mapping of WS onto HTTP2 is fairly direct-- we reserved the MSG_DONE
> bit specifically for this kind of use-case (i.e. bidirectional,
> full-duplex, message-based).
> The observation made is that there are essentially 3 types of data that
> need to be preserved one way or another:
> 1) metadata: data about the data, e.g. text, binary, where it is supposed
> to be going, etc.
> 2) payload
> 3) boundaries between these.

Given that in WebSocket metadata belongs to WS Frame not to WS Message, I'm
wondering if it is better to use MSG_DONE as the boundary between WS Frames.
In addition to that, we can simply map a WS Frame to a HTTP2 HEADERS and
DATAs, but WS Message is fairly complicated (e.g. control message injection
in a data message).
I believe that we should split the WebSocket spec into abstract parts and
concrete parts, define concrete parts for HTTP2 and share abstract parts
between WS and WS/HTTP2.
I think the boundary of them should be WebSocket Frame, i.e. WS/HTTP2
specific framing spec should not know about WS Message.
What do you think?



On Wed, Jan 15, 2014 at 3:30 AM, Roberto Peon <fenix@google.com> wrote:

> I've been lurking anyway :)
>
> As I understand the proposal, (and it seems that Takeshi and Yukata are
> covering this nicely, so what i say is probably redundant):
> Message boundaries, having semantic meaning, are preserved.
>
> Frame boundaries, having no semantic meaning, are not preserved. I don't
> think it makes sense to worry about this greatly when the ability to extend
> the protocol as implemented in HTTP2 is present. It doesn't make a lot of
> sense to signal the frame boundaries through everything unless there is a
> high-likelyhood and high-utility use-case for such. If such is found, then
> HTTP2 can be revised to include it. This is the approach we've been taking
> with most of the HTTP2 stuff, since there has been great effort placed on
> allowing new versions to be deployed in the future.
>
> The mapping of WS onto HTTP2 is fairly direct-- we reserved the MSG_DONE
> bit specifically for this kind of use-case (i.e. bidirectional,
> full-duplex, message-based).
> The observation made is that there are essentially 3 types of data that
> need to be preserved one way or another:
> 1) metadata: data about the data, e.g. text, binary, where it is supposed
> to be going, etc.
> 2) payload
> 3) boundaries between these.
>
> imho, metadata is handled with HEADERS (I'd have preferred it to be called
> METADATA, but was voted down at the time :) ), payload is handled with DATA
> frames, and the boundaries are delimited either through the use of HEADERS
> frames, or via the MSG_DONE or STREAM_DONE bits.
> In the presence of extensions, assuming that MSG_DONE is required to
> signal the end of a message, then HEADERS can be used to signal any
> boundary desired.
>
>
> I totally agree that it would be nice to expose the priority of various
> messages/channels with respect to each other. That is really an issue that
> is outside the WG's purvue, though.
>
>
>
> On Thu, Jan 9, 2014 at 7:54 PM, Yutaka Hirano <yhirano@chromium.org>wrote:
>
>> +Roberto as an HTTP2 expert.
>>
>>
>>> 2) Ok, both WS frame and message boundaries are preserved.
>>>
>>> But: I cannot recognize the end of a WS frame until the next WS frame
>>> starts (for which I receive a HEADERS frame, which is the implicit signal
>>> that the previous WS frame ended), right?
>>>
>> You are right, you can't.
>>
>>
>>>
>>> And if so, why is there no FRAME_DONE bit on the SPDY frame that
>>> corresponds to the end of a WS frame?
>>
>>  Given that WebSocket frames "have no semantic meaning" as in RFC6455, I
>> don't think the immediate recognition of the end of a WebSocket frame is
>> requisite.
>> I would like to introduce MSG_DONE in HTTP2 so that intermediaries that
>> don't fully understand WS over HTTP2 can relay it.
>> Since there are not so much space in FLAGS field, Introducing FRAME_DONE
>> may decrease the extensiveness of the spec.
>> Do you have a strong preference to introduce FRAME_DONE?
>>
>>
>>
>> On Thu, Jan 9, 2014 at 6:53 PM, Tobias Oberstein <
>> tobias.oberstein@tavendo.de> wrote:
>>
>>> Hi Yutaka,
>>>
>>>
>>>
>>> 1) Yes, I understand. Thanks!
>>>
>>>
>>>
>>> 2) Ok, both WS frame and message boundaries are preserved.
>>>
>>>
>>>
>>> But: I cannot recognize the end of a WS frame until the next WS frame
>>> starts (for which I receive a HEADERS frame, which is the implicit signal
>>> that the previous WS frame ended), right?
>>>
>>>
>>>
>>> And if so, why is there no FRAME_DONE bit on the SPDY frame that
>>> corresponds to the end of a WS frame?
>>>
>>>
>>>
>>> 3) Ok, great! Hope this works, since SPDY provides header compression,
>>> but no sophisticated bidirectional payload compression as "WS
>>> permessage-deflate" does.
>>>
>>>
>>>
>>> Cheers,
>>>
>>> /Tobias
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> *Von:* yhirano@google.com [mailto:yhirano@google.com] *Im Auftrag von *Yutaka
>>> Hirano
>>> *Gesendet:* Donnerstag, 9. Januar 2014 10:37
>>> *An:* Tobias Oberstein
>>> *Cc:* Takeshi Yoshino; hybi@ietf.org
>>> *Betreff:* Re: [hybi] Discontinuation of mux standardizaton in favor of
>>> WS/HTTP/2.0
>>>
>>>
>>>
>>> Hi Tobias,
>>>
>>> Thank you for the comments.
>>>
>>>
>>>
>>> 1) Frame Mapping
>>> """
>>> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
>>> following plural SPDY data frames. The SPDY HEADERS frame must contain
>>> WebSocket frame fields and following SPDY data frames contain payload data.
>>> These data frames are free to be reframed.
>>> """
>>>
>>> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
>>> So reframing of WebSocket frames (the payload) into multiple SPDY data
>>> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
>>> is free to choose the boundaries within WS frame payload at which to
>>> reframe into different SPDY data frames as it likes?
>>>
>>> A WebSocket frame MAY be split to multiple SPDY DATA frames.
>>>
>>> Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA
>>> frame into multiple DATA frames.
>>>
>>> Is this answering your question?
>>>
>>>
>>>
>>> 2) End of WS Frame
>>> The end of a WS message is signaled with a SPDY data frame having
>>> MSG_DONE set.
>>> How is the end of a WS frame signaled, given that the WS frame may have
>>> been reframed into multiple SPDY data frames?
>>>
>>> When you receive a HEADERS frame, you can recognize that the last SPDY
>>> frame is that last frame in a WebSocket frame.
>>>
>>> When you receive a SPDY frame with MSG_DONE, you can recognize that it
>>> is the last SPDY frame in a WebSocket message and hence the last SPDY frame
>>> in a WebSocket frame.
>>>
>>>
>>>
>>> 3) WS extension
>>> Are WS extensions fully supported? In particular, is
>>> "permessage-deflate" supported when running WS/SPDY?
>>>
>>> Yes. We have reserved bits and we can emulate the opening handshake, I
>>> think.
>>>
>>>
>>>
>>> On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein <
>>> tobias.oberstein@tavendo.de> wrote:
>>>
>>> Hi Takeshi,
>>>
>>> it's a pity, but I can understand the motivation to shove everything
>>> into HTTP2.
>>>
>>> Anyway, some questions:
>>>
>>> 1) Frame Mapping
>>> """
>>> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
>>> following plural SPDY data frames. The SPDY HEADERS frame must contain
>>> WebSocket frame fields and following SPDY data frames contain payload data.
>>> These data frames are free to be reframed.
>>> """
>>>
>>> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
>>>
>>> So reframing of WebSocket frames (the payload) into multiple SPDY data
>>> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
>>> is free to choose the boundaries within WS frame payload at which to
>>> reframe into different SPDY data frames as it likes?
>>>
>>> 2) End of WS Frame
>>>
>>> The end of a WS message is signaled with a SPDY data frame having
>>> MSG_DONE set.
>>>
>>> How is the end of a WS frame signaled, given that the WS frame may have
>>> been reframed into multiple SPDY data frames?
>>>
>>> 3) WS extension
>>>
>>> Are WS extensions fully supported? In particular, is
>>> "permessage-deflate" supported when running WS/SPDY?
>>>
>>> Thanks!
>>> /Tobias
>>>
>>>
>>> Von: hybi [mailto:hybi-bounces@ietf.org] Im Auftrag von Takeshi Yoshino
>>> Gesendet: Dienstag, 7. Januar 2014 08:36
>>> An: hybi@ietf.org
>>> Betreff: [hybi] Discontinuation of mux standardizaton in favor of
>>> WS/HTTP/2.0
>>>
>>>
>>> Hi all,
>>>
>>> We've been working on mux extension to make WebSocket scalable for long
>>> time, but as SPDY got widely supported and adopted quickly, now there seems
>>> to be no longer any good reason to continue this work which is WebSocket
>>> dedicated.
>>>
>>> Scalability is still important issue to address. From now, I'd like to
>>> focus on WS/SPDY effort which we've been working on in parallel with the
>>> mux extension.
>>>
>>> Please take a look at the latest proposal in this post.
>>> https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc
>>>
>>> If you really want the dedicated mux extension than one over SPDY, sorry
>>> but please take over the work.
>>>
>>> _______________________________________________
>>> hybi mailing list
>>> hybi@ietf.org
>>> https://www.ietf.org/mailman/listinfo/hybi
>>>
>>>
>>>
>>
>>
>

--047d7bf0e1f63f322504f0647a26
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Hi Roberto,<div><br></div><div><blockquote class=3D"gmail_=
quote" style=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-=
color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">The mappin=
g of WS onto HTTP2 is fairly direct-- we reserved the MSG_DONE bit specific=
ally for this kind of use-case (i.e. bidirectional, full-duplex, message-ba=
sed).<br>
The observation made is that there are essentially 3 types of data that nee=
d to be preserved one way or another:<br>1) metadata: data about the data, =
e.g. text, binary, where it is supposed to be going, etc.<br>2) payload<br>
3) boundaries between these.</blockquote><div>Given that in WebSocket metad=
ata belongs to WS Frame not to WS Message, I&#39;m wondering if it is bette=
r to use MSG_DONE as the boundary between WS Frames.</div><div>In addition =
to that, we can simply map a WS Frame to a HTTP2 HEADERS and DATAs, but WS =
Message is fairly complicated (e.g. control message injection in a data mes=
sage).<br>
</div><div>I believe that we should split the WebSocket spec into abstract =
parts and concrete parts, define concrete parts for HTTP2 and share abstrac=
t parts between WS and WS/HTTP2.</div><div>I think the boundary of them sho=
uld be WebSocket Frame, i.e. WS/HTTP2 specific framing spec should not know=
 about WS Message.</div>
<div>What do you think?<br></div><div><br></div></div></div><div class=3D"g=
mail_extra"><br><br><div class=3D"gmail_quote">On Wed, Jan 15, 2014 at 3:30=
 AM, Roberto Peon <span dir=3D"ltr">&lt;<a href=3D"mailto:fenix@google.com"=
 target=3D"_blank">fenix@google.com</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div dir=3D"ltr">I&#39;ve been lurking anywa=
y :)<div><br></div><div>As I understand the proposal, (and it seems that Ta=
keshi and Yukata are covering this nicely, so what i say is probably redund=
ant):</div>
<div>Message boundaries, having semantic meaning, are preserved.</div>
<div><br></div><div>Frame boundaries, having no semantic meaning, are not p=
reserved. I don&#39;t think it makes sense to worry about this greatly when=
 the ability to extend the protocol as implemented in HTTP2 is present. It =
doesn&#39;t make a lot of sense to signal the frame boundaries through ever=
ything unless there is a high-likelyhood and high-utility use-case for such=
. If such is found, then HTTP2 can be revised to include it. This is the ap=
proach we&#39;ve been taking with most of the HTTP2 stuff, since there has =
been great effort placed on allowing new versions to be deployed in the fut=
ure.</div>

<div><br></div><div>The mapping of WS onto HTTP2 is fairly direct-- we rese=
rved the MSG_DONE bit specifically for this kind of use-case (i.e. bidirect=
ional, full-duplex, message-based).</div><div>The observation made is that =
there are essentially 3 types of data that need to be preserved one way or =
another:</div>

<div>1) metadata: data about the data, e.g. text, binary, where it is suppo=
sed to be going, etc.</div><div>2) payload</div><div>3) boundaries between =
these.</div><div><br></div><div>imho, metadata is handled with HEADERS (I&#=
39;d have preferred it to be called METADATA, but was voted down at the tim=
e :) ), payload is handled with DATA frames, and the boundaries are delimit=
ed either through the use of HEADERS frames, or via the MSG_DONE or STREAM_=
DONE bits.</div>

<div>In the presence of extensions, assuming that MSG_DONE is required to s=
ignal the end of a message, then HEADERS can be used to signal any boundary=
 desired.</div><div><br></div><div><br></div><div>I totally agree that it w=
ould be nice to expose the priority of various messages/channels with respe=
ct to each other. That is really an issue that is outside the WG&#39;s purv=
ue, though.</div>

<div><br></div></div><div class=3D"HOEnZb"><div class=3D"h5"><div class=3D"=
gmail_extra"><br><br><div class=3D"gmail_quote">On Thu, Jan 9, 2014 at 7:54=
 PM, Yutaka Hirano <span dir=3D"ltr">&lt;<a href=3D"mailto:yhirano@chromium=
.org" target=3D"_blank">yhirano@chromium.org</a>&gt;</span> wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div dir=3D"ltr"><div>+Roberto as an HTTP2 e=
xpert.</div><div><div>=C2=A0</div><blockquote class=3D"gmail_quote" style=
=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(20=
4,204,204);border-left-style:solid;padding-left:1ex">


<span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,7=
3,125)">2) Ok, both WS frame and message boundaries are preserved.<br></spa=
n><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31=
,73,125)">=C2=A0<br>


</span><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:r=
gb(31,73,125)">But: I cannot recognize the end of a WS frame until the next=
 WS frame starts (for which I receive a HEADERS frame, which is the implici=
t signal that the previous WS frame ended), right?<br>


</span><span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:r=
gb(31,73,125)"></span></blockquote></div>You are right, you can&#39;t.<div>=
<div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px =
0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-l=
eft-style:solid;padding-left:1ex">


<span style=3D"font-size:11pt;font-family:Calibri,sans-serif;color:rgb(31,7=
3,125)">=C2=A0<br></span><span style=3D"font-size:11pt;font-family:Calibri,=
sans-serif;color:rgb(31,73,125)">And if so, why is there no FRAME_DONE bit =
on the SPDY frame that corresponds to the end of a WS frame?</span></blockq=
uote>

</div>
Given that WebSocket frames &quot;have no semantic meaning&quot; as in RFC6=
455, I don&#39;t think the immediate recognition of the end of a WebSocket =
frame is requisite.<div>I would like to introduce MSG_DONE in HTTP2 so that=
 intermediaries that don&#39;t fully understand WS over HTTP2 can relay it.=
</div>


<div>Since there are not so much space in FLAGS field, Introducing FRAME_DO=
NE may decrease the extensiveness of the spec.</div><div>Do you have a stro=
ng preference to introduce FRAME_DONE?</div><div><br></div></div><div>

<div><div class=3D"gmail_extra">
<br><br><div class=3D"gmail_quote">On Thu, Jan 9, 2014 at 6:53 PM, Tobias O=
berstein <span dir=3D"ltr">&lt;<a href=3D"mailto:tobias.oberstein@tavendo.d=
e" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt;</span> wrote:<br><=
blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px=
 #ccc solid;padding-left:1ex">


<div lang=3D"DE" link=3D"blue" vlink=3D"purple"><div><p class=3D"MsoNormal"=
><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans=
-serif&quot;;color:#1f497d">Hi Yutaka,<u></u><u></u></span></p><p class=3D"=
MsoNormal">


<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></span></p><p class=3D"MsoNo=
rmal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot=
;sans-serif&quot;;color:#1f497d">1) Yes, I understand. Thanks!<u></u><u></u=
></span></p>


<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">2) Ok, both WS fram=
e and message boundaries are preserved.<u></u><u></u></span></p>


<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">But: I cannot recog=
nize the end of a WS frame until the next WS frame starts (for which I rece=
ive a HEADERS frame, which is the implicit signal that the previous WS fram=
e ended), right?<u></u><u></u></span></p>


<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">And if so, why is t=
here no FRAME_DONE bit on the SPDY frame that corresponds to the end of a W=
S frame?<u></u><u></u></span></p>


<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">3) Ok, great! Hope =
this works, since SPDY provides header compression, but no sophisticated bi=
directional payload compression as &quot;WS permessage-deflate&quot; does.<=
u></u><u></u></span></p>


<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Cheers,<u></u><u></=
u></span></p>


<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">/Tobias<u></u><u></u></sp=
an></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&=
quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></=
u></span></p>


<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u=
></span></p>


<div style=3D"border:none;border-left:solid blue 1.5pt;padding:0cm 0cm 0cm =
4.0pt"><div><div style=3D"border:none;border-top:solid #e1e1e1 1.0pt;paddin=
g:3.0pt 0cm 0cm 0cm"><p class=3D"MsoNormal"><b><span style=3D"font-size:11.=
0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;">Von:</span></b>=
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;"> <a href=3D"mailto:yhirano@google.com" target=3D"_blank">yhira=
no@google.com</a> [mailto:<a href=3D"mailto:yhirano@google.com" target=3D"_=
blank">yhirano@google.com</a>] <b>Im Auftrag von </b>Yutaka Hirano<br>


<b>Gesendet:</b> Donnerstag, 9. Januar 2014 10:37<br><b>An:</b> Tobias Ober=
stein<br><b>Cc:</b> Takeshi Yoshino; <a href=3D"mailto:hybi@ietf.org" targe=
t=3D"_blank">hybi@ietf.org</a><br><b>Betreff:</b> Re: [hybi] Discontinuatio=
n of mux standardizaton in favor of WS/HTTP/2.0<u></u><u></u></span></p>


</div></div><div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><div><=
p class=3D"MsoNormal">Hi Tobias,<u></u><u></u></p><div><p class=3D"MsoNorma=
l">Thank you for the comments.<u></u><u></u></p></div><div><p class=3D"MsoN=
ormal">
<u></u>=C2=A0<u></u></p><blockquote style=3D"border:none;border-left:solid =
#cccccc 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm"=
><p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;A=
rial&quot;,&quot;sans-serif&quot;">1) Frame Mapping<br>


&quot;&quot;&quot;<br>One WebSocket frame is mapped into one preceding SPDY=
 HEADERS frame and following plural SPDY data frames. The SPDY HEADERS fram=
e must contain WebSocket frame fields and following SPDY data frames contai=
n payload data. These data frames are free to be reframed.<br>


&quot;&quot;&quot;<br></span><a href=3D"https://docs.google.com/document/d/=
124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank"><span =
style=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&qu=
ot;">https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw=
9aWzSHijk/edit</span></a><span style=3D"font-size:10.0pt;font-family:&quot;=
Arial&quot;,&quot;sans-serif&quot;"><br>


So reframing of WebSocket frames (the payload) into multiple SPDY data fram=
es MAY occur, but does not need to occur? And a WS/SPDY implementation is f=
ree to choose the boundaries within WS frame payload at which to reframe in=
to different SPDY data frames as it likes?</span><u></u><u></u></p>


</blockquote><div><p class=3D"MsoNormal">A WebSocket frame MAY be split to =
multiple SPDY DATA frames.<u></u><u></u></p></div><div><p class=3D"MsoNorma=
l">Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA fram=
e into multiple DATA frames.<u></u><u></u></p>


</div><div><p class=3D"MsoNormal">Is this answering your question?<u></u><u=
></u></p></div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><b=
lockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm =
0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm">


<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ar=
ial&quot;,&quot;sans-serif&quot;">2) End of WS Frame<br>The end of a WS mes=
sage is signaled with a SPDY data frame having MSG_DONE set.<br>How is the =
end of a WS frame signaled, given that the WS frame may have been reframed =
into multiple SPDY data frames?</span><u></u><u></u></p>


</blockquote><div><p class=3D"MsoNormal">When you receive a HEADERS frame, =
you can recognize that the last SPDY frame is that last frame in a WebSocke=
t frame.<u></u><u></u></p></div><div><p class=3D"MsoNormal">When you receiv=
e a SPDY frame with MSG_DONE, you can recognize that it is the last SPDY fr=
ame in a WebSocket message and hence the last SPDY frame in a WebSocket fra=
me.<u></u><u></u></p>


</div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><blockquote=
 style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm 6=
.0pt;margin-left:4.8pt;margin-right:0cm"><p class=3D"MsoNormal"><span style=
=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&quot;">=
3) WS extension<br>


Are WS extensions fully supported? In particular, is &quot;permessage-defla=
te&quot; supported when running WS/SPDY?</span><u></u><u></u></p></blockquo=
te><div><p class=3D"MsoNormal">Yes. We have reserved bits and we can emulat=
e the opening handshake, I think.<u></u><u></u></p>


</div><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div><p cla=
ss=3D"MsoNormal">On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein &lt;<a hr=
ef=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstei=
n@tavendo.de</a>&gt; wrote:<u></u><u></u></p>


<blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0c=
m 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm"><p class=3D"MsoNormal">=
Hi Takeshi,<br><br>it&#39;s a pity, but I can understand the motivation to =
shove everything into HTTP2.<br>


<br>Anyway, some questions:<br><br>1) Frame Mapping<br>&quot;&quot;&quot;<b=
r>One WebSocket frame is mapped into one preceding SPDY HEADERS frame and f=
ollowing plural SPDY data frames. The SPDY HEADERS frame must contain WebSo=
cket frame fields and following SPDY data frames contain payload data. Thes=
e data frames are free to be reframed.<br>


&quot;&quot;&quot;<br><a href=3D"https://docs.google.com/document/d/124vB5N=
V1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank">https://docs.=
google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit</a>=
<br>


<br>So reframing of WebSocket frames (the payload) into multiple SPDY data =
frames MAY occur, but does not need to occur? And a WS/SPDY implementation =
is free to choose the boundaries within WS frame payload at which to refram=
e into different SPDY data frames as it likes?<br>


<br>2) End of WS Frame<br><br>The end of a WS message is signaled with a SP=
DY data frame having MSG_DONE set.<br><br>How is the end of a WS frame sign=
aled, given that the WS frame may have been reframed into multiple SPDY dat=
a frames?<br>


<br>3) WS extension<br><br>Are WS extensions fully supported? In particular=
, is &quot;permessage-deflate&quot; supported when running WS/SPDY?<br><br>=
Thanks!<br>/Tobias<br><br><br>Von: hybi [mailto:<a href=3D"mailto:hybi-boun=
ces@ietf.org" target=3D"_blank">hybi-bounces@ietf.org</a>] Im Auftrag von T=
akeshi Yoshino<br>


Gesendet: Dienstag, 7. Januar 2014 08:36<br>An: <a href=3D"mailto:hybi@ietf=
.org" target=3D"_blank">hybi@ietf.org</a><br>Betreff: [hybi] Discontinuatio=
n of mux standardizaton in favor of WS/HTTP/2.0<u></u><u></u></p><div><div>


<p class=3D"MsoNormal"><br>Hi all,<br><br>We&#39;ve been working on mux ext=
ension to make WebSocket scalable for long time, but as SPDY got widely sup=
ported and adopted quickly, now there seems to be no longer any good reason=
 to continue this work which is WebSocket dedicated.<br>


<br>Scalability is still important issue to address. From now, I&#39;d like=
 to focus on WS/SPDY effort which we&#39;ve been working on in parallel wit=
h the mux extension.<br><br>Please take a look at the latest proposal in th=
is post.<br>


<a href=3D"https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc" ta=
rget=3D"_blank">https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3gl=
c</a><br><br>If you really want the dedicated mux extension than one over S=
PDY, sorry but please take over the work.<u></u><u></u></p>


</div></div><div><div><p class=3D"MsoNormal">______________________________=
_________________<br>hybi mailing list<br><a href=3D"mailto:hybi@ietf.org" =
target=3D"_blank">hybi@ietf.org</a><br><a href=3D"https://www.ietf.org/mail=
man/listinfo/hybi" target=3D"_blank">https://www.ietf.org/mailman/listinfo/=
hybi</a><u></u><u></u></p>


</div></div></blockquote></div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u><=
/p></div></div></div></div></div></div></div></blockquote></div><br></div>
</div></div></blockquote></div><br></div>
</div></div></blockquote></div><br></div>

--047d7bf0e1f63f322504f0647a26--

From tobias.oberstein@tavendo.de  Mon Jan 20 03:16:26 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 51CC61A0120 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 03:16:26 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.899
X-Spam-Level: 
X-Spam-Status: No, score=-1.899 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id iuT2rY7tXwOb for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 03:16:22 -0800 (PST)
Received: from EXHUB020-2.exch020.serverdata.net (exhub020-2.exch020.serverdata.net [206.225.164.29]) by ietfa.amsl.com (Postfix) with ESMTP id 14E541A011F for <hybi@ietf.org>; Mon, 20 Jan 2014 03:16:22 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-2.exch020.serverdata.net ([206.225.164.29]) with mapi; Mon, 20 Jan 2014 03:16:21 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Yutaka Hirano <yhirano@chromium.org>
Date: Mon, 20 Jan 2014 03:16:17 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8VrcSQ0ltilFsnS8qmz7HGH75YfQAIjftw
Message-ID: <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com>
In-Reply-To: <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: multipart/alternative; boundary="_000_634914A010D0B943A035D226786325D4446BF99976EXVMBX02012ex_"
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>, "Peter H Thorson \(webmaster@zaphoyd.com\)" <webmaster@zaphoyd.com>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 11:16:26 -0000

--_000_634914A010D0B943A035D226786325D4446BF99976EXVMBX02012ex_
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_634914A010D0B943A035D226786325D4446BF99976EXVMBX02012ex_
Content-Type: text/html; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_634914A010D0B943A035D226786325D4446BF99976EXVMBX02012ex_--

From tobias.oberstein@tavendo.de  Mon Jan 20 03:26:08 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 6FBC71A0122 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 03:26:08 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.899
X-Spam-Level: 
X-Spam-Status: No, score=-1.899 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id C2rPJ8yyspqa for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 03:26:06 -0800 (PST)
Received: from EXHUB020-3.exch020.serverdata.net (exhub020-3.exch020.serverdata.net [206.225.164.30]) by ietfa.amsl.com (Postfix) with ESMTP id 107FD1A0128 for <hybi@ietf.org>; Mon, 20 Jan 2014 03:26:06 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-3.exch020.serverdata.net ([206.225.164.30]) with mapi; Mon, 20 Jan 2014 03:26:06 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Takeshi Yoshino <tyoshino@google.com>
Date: Mon, 20 Jan 2014 03:26:03 -0800
Thread-Topic: [hybi] WebSocket subprotocol parameters
Thread-Index: Ac8VuqQgIB5XX0N2T2usfyvZTAcXWwAFtW9Q
Message-ID: <634914A010D0B943A035D226786325D4446BF99977@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF9948F@EXVMBX020-12.exch020.serverdata.net> <52D90F99.6080205@gmx.de> <634914A010D0B943A035D226786325D4446BF9949D@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZQ3NxVW36PnZ0TpMF4tPPaJLr12M8NtbPb6pUejf1wcQ@mail.gmail.com>
In-Reply-To: <CAH9hSJZQ3NxVW36PnZ0TpMF4tPPaJLr12M8NtbPb6pUejf1wcQ@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: multipart/alternative; boundary="_000_634914A010D0B943A035D226786325D4446BF99977EXVMBX02012ex_"
MIME-Version: 1.0
Cc: Julian Reschke <julian.reschke@gmx.de>, "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] WebSocket subprotocol parameters
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 11:26:08 -0000

--_000_634914A010D0B943A035D226786325D4446BF99977EXVMBX02012ex_
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_634914A010D0B943A035D226786325D4446BF99977EXVMBX02012ex_
Content-Type: text/html; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_634914A010D0B943A035D226786325D4446BF99977EXVMBX02012ex_--

From yhirano@google.com  Mon Jan 20 04:07:17 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 3C1DD1A0133 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 04:07:17 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.913
X-Spam-Level: 
X-Spam-Status: No, score=-1.913 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id k_vlP3iOLypa for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 04:07:11 -0800 (PST)
Received: from mail-qc0-x22f.google.com (mail-qc0-x22f.google.com [IPv6:2607:f8b0:400d:c01::22f]) by ietfa.amsl.com (Postfix) with ESMTP id 16F771A0113 for <hybi@ietf.org>; Mon, 20 Jan 2014 04:07:09 -0800 (PST)
Received: by mail-qc0-f175.google.com with SMTP id x13so5891261qcv.34 for <hybi@ietf.org>; Mon, 20 Jan 2014 04:07:09 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=4bX8EHl1ckSTyeUQUotktTGTO2cPjzlNiBox5k7hLO8=; b=j27DSrll3f01m2UiD+AkiDfg2NTSf2GBlEVuOHlDRmI49+2PT2i+fyA16SuitBT+bB mZaMTfaY+IZ/Hm1rWi2qZnhomiwIpDX3Wzyzw05DPJurCBhNz91cuNqiSuE0ftMfup6h DmTZ0XydywEg9BvVuU6p/l9r5ipOyt7kqXFqhIc7KWaU2FyuLUFNHRrGP36nNK5tXmmU ZJjQ+h6v/LRQiVFZl7nyXIlHW4huNcVLwbTELi0HPuCvsAPfxdfINJXrvfPp+3DcTkOB ijtZv3EenbkVyEfNKn1AHXAD7fmj4nsFivzODcKPKYB2UNz7Ses7phw66sUT4XXLa2aS YGPA==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=4bX8EHl1ckSTyeUQUotktTGTO2cPjzlNiBox5k7hLO8=; b=Tv1gnaCt6v4XDiBttpFKjktkoeQedEgBBdd8sha8BZHXZIAGpQjFflKmsFnDusuQ2e qSNpBkj+iITcMwicQBYMim/pd3JvSN39zJAgleA/Ces+l55OVkrQlYcrD0nJxSSwopKG 8O38Hbx8imTLMIxcbchJWyKDxNUj7Mjhwx/Qg=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=4bX8EHl1ckSTyeUQUotktTGTO2cPjzlNiBox5k7hLO8=; b=X2FXKyB5Sc5dXhOIHnRs3wInvnbj2iRPWINCxQfDvk8bs757VxsiLrfGcxneFVtA69 qJlUBJh58enJou4d0GA3pAPZ6nVanHksKSDpWY/MC/2PlY5v4+Re9A7raBn4DHJ0ddBW gaz/eSooE8ebZ3LCSART7gdlqNYzZ0dLIApvNP8lm4xU0ec8sW76/mYiMWgFqoEDFHJe dxmtMPhgS7HvY+VntEw7Q1n3U9tgKHadIdbCXCljLKUzdi5i+vVXY5nmQ67mpl5zm8V2 iDqz8fNYDE2RX1nESyfL7tbD4+RM8z9kbENhoCyIFkoVBX8Fi9JhhMBFIfDD8fmyoDTk Dozg==
X-Gm-Message-State: ALoCoQmTDuFffJhp0rixRHeFk5koDvnu7ZQqpsVXXUciAvcq5qvolHDmbbyPSOKUAy74FphW8CuxbxEbEKg8jSeqHp3emF9K9RaoSbsSFuue+GTkzWgV4yH617g+qMVndpD7QhoQEtxLsyEwsYKxkr9Cj+VWQbPqCzHulz0A/sBcntjuJfr9BzrxViwII034lhx0223ELTsv
MIME-Version: 1.0
X-Received: by 10.224.68.70 with SMTP id u6mr26833857qai.5.1390219629120; Mon, 20 Jan 2014 04:07:09 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.140.109.202 with HTTP; Mon, 20 Jan 2014 04:07:08 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net>
Date: Mon, 20 Jan 2014 21:07:08 +0900
X-Google-Sender-Auth: OJ0F55xP-gPHkiOHecQuy65gUKo
Message-ID: <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=001a11c2f58e277c1d04f065bb95
Cc: "hybi@ietf.org" <hybi@ietf.org>, "Peter H Thorson \(webmaster@zaphoyd.com\)" <webmaster@zaphoyd.com>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 12:07:17 -0000

--001a11c2f58e277c1d04f065bb95
Content-Type: text/plain; charset=UTF-8

Thank you for the explanation, finally I may understand...
Your proposal (SPDY/Messaging) is a brand-new full-duplex messaging
protocol on SPDY and the only constraint is compatibility with the W3C
WebSocket API, right?



On Mon, Jan 20, 2014 at 8:16 PM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> Hi Yutaka,
>
>
>
> in fact, I completely agree with Peter (
> http://www.ietf.org/mail-archive/web/hybi/current/msg10310.html) - either
> have
>
>
>
> 1) full WebSocket tunneled over SPDY ("SPDY/WS-Tunnel")
>
> => this is the proposal to transport full WS frames as payload of SPDY
> data frames describe in my other post
>
>
>
> or
>
>
>
> 2) define a SPDY messaging, independent of WS ("SPDY/Messaging")
>
> => This is the proposal below (Fri, Jan 10, 2014 at 5:37 PM). Or
> something in this direction.
>
>
>
> The current SPDY/WS proposal is different from both 1) and 2) above, and
> again I agree with Peter, that this is the least desirable of all.
>
>
>
> My order of preference would be:
>
>
>
> ++1:       SPDY/Messaging
>
> +1:         SPDY/WS-Tunnel
>
> 0:            SPDY/WS
>
>
>
> /Tobias
>
>
>
>
>
>
>
>
>
> *Von:* yhirano@google.com [mailto:yhirano@google.com] *Im Auftrag von *Yutaka
> Hirano
> *Gesendet:* Montag, 20. Januar 2014 08:04
> *An:* Tobias Oberstein
> *Cc:* Roberto Peon; Takeshi Yoshino; hybi@ietf.org
>
> *Betreff:* Re: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
>
> Tobias, let me confirm.
>
>
>
> I'd like to suggest a different WS/SPDY:
>
> Pack WebSocket frames _including_ full WS frame header as _payload_ of
> SPDY data frames only - no SPDY header frames at all.
>
> My understanding is that by suggesting the above one you retracted the
> below proposal you made before, is that right?
>
>
>
> On Fri, Jan 10, 2014 at 5:37 PM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
>
> Why not map WS messages directly into SPDY data frames like so:
> Have 2 bits in SPDY data frames:
> 0x03 = MSG_TYPE : 0 == text, 1 == binary
> 0x04 = MSG_DONE
> A WS text message would simply be mapped into a series of SPDY data frames:
> MSG_TYPE = 0, MSG_DONE = 0
> MSG_TYPE = 0, MSG_DONE = 0
> ...
> MSG_TYPE = 0, MSG_DONE = 1
> Similar for a WS binary message.
>
>
> Doing like above will allow to preserve the current WS API in HTML5 (which
> needs messages, and needs 2 message types).
>
>
>
> Thanks,
>
>
>
>
>
> On Wed, Jan 15, 2014 at 4:09 AM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
>
> Hi Roberto,
>
>
>
> not sure if you agree, but FWIW, here is a summary as I see it (currently):
>
>
>
> 1) WebSocket frame boundaries _can_ have semantic meaning in WebSocket
> _extensions_
>
>
>
> 2) WebSocket message and frame boundaries are indeed preserved in the
> current WS/SPDY proposal,
>
> however, non-final WebSocket frame ends cannot be _immediately_ detected.
>
>
>
> 3) The current WS/SPDY proposal
>
>   * is comfortable to implement for implementations that _only_ want to
> implement WS/SPDY, and not native WS at the same time
>
>  * makes things complicated for WS implementations that want to support
> both WS/SPDY _and_ native WS at the same time
>
>  * has significant wire-level overhead compared to native WS (and compared
> to the proposal below)
>
>
>
> I'd like to suggest a different WS/SPDY:
>
>
>
> Pack WebSocket frames _including_ full WS frame header as _payload_ of
> SPDY data frames only - no SPDY header frames at all.
>
>
>
> In the absence of negotiated WebSocket extensions (or when all WS
> extensions in place do not assign semantics to WS frame boundaries),
>
> fragment WebSocket payload at the WS level such that WS frame
> header+payload fits into the payload of SPDY data frames.
>
>
>
> In the latter case, this essentially adds a small SPDY data frame header
> (of fixed length) to WS frames. Easy to implement and efficient on wire.
>
>
>
> IMHO, this has advantages compared to current proposal and considering 3)
> above.
>
>
>
> The only disadvantage is for implementations that _only_ want to implement
> WS/SPDY and not native WS at the same time.
>
>
>
> ==
>
>
>
> Regarding "permessage-priority" with ordering guarantees: should I find
> time and if there is interest, I'll write something up.
>
> This could be a WebSocket extension. To be useful, the WS API in browsers
> would need to expose it though ..
>
>
>
> Just my 2cts,
>
> Tobias
>
>
>
>
>
>
>
>
>
> *Von:* Roberto Peon [mailto:fenix@google.com]
> *Gesendet:* Dienstag, 14. Januar 2014 19:30
> *An:* Yutaka Hirano
> *Cc:* Tobias Oberstein; Takeshi Yoshino; hybi@ietf.org
>
>
> *Betreff:* Re: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
>
> I've been lurking anyway :)
>
>
>
> As I understand the proposal, (and it seems that Takeshi and Yukata are
> covering this nicely, so what i say is probably redundant):
>
> Message boundaries, having semantic meaning, are preserved.
>
>
>
> Frame boundaries, having no semantic meaning, are not preserved. I don't
> think it makes sense to worry about this greatly when the ability to extend
> the protocol as implemented in HTTP2 is present. It doesn't make a lot of
> sense to signal the frame boundaries through everything unless there is a
> high-likelyhood and high-utility use-case for such. If such is found, then
> HTTP2 can be revised to include it. This is the approach we've been taking
> with most of the HTTP2 stuff, since there has been great effort placed on
> allowing new versions to be deployed in the future.
>
>
>
> The mapping of WS onto HTTP2 is fairly direct-- we reserved the MSG_DONE
> bit specifically for this kind of use-case (i.e. bidirectional,
> full-duplex, message-based).
>
> The observation made is that there are essentially 3 types of data that
> need to be preserved one way or another:
>
> 1) metadata: data about the data, e.g. text, binary, where it is supposed
> to be going, etc.
>
> 2) payload
>
> 3) boundaries between these.
>
>
>
> imho, metadata is handled with HEADERS (I'd have preferred it to be called
> METADATA, but was voted down at the time :) ), payload is handled with DATA
> frames, and the boundaries are delimited either through the use of HEADERS
> frames, or via the MSG_DONE or STREAM_DONE bits.
>
> In the presence of extensions, assuming that MSG_DONE is required to
> signal the end of a message, then HEADERS can be used to signal any
> boundary desired.
>
>
>
>
>
> I totally agree that it would be nice to expose the priority of various
> messages/channels with respect to each other. That is really an issue that
> is outside the WG's purvue, though.
>
>
>
>
>
> On Thu, Jan 9, 2014 at 7:54 PM, Yutaka Hirano <yhirano@chromium.org>
> wrote:
>
> +Roberto as an HTTP2 expert.
>
>
>
> 2) Ok, both WS frame and message boundaries are preserved.
>
> But: I cannot recognize the end of a WS frame until the next WS frame
> starts (for which I receive a HEADERS frame, which is the implicit signal
> that the previous WS frame ended), right?
>
> You are right, you can't.
>
>
>
>
> And if so, why is there no FRAME_DONE bit on the SPDY frame that
> corresponds to the end of a WS frame?
>
> Given that WebSocket frames "have no semantic meaning" as in RFC6455, I
> don't think the immediate recognition of the end of a WebSocket frame is
> requisite.
>
> I would like to introduce MSG_DONE in HTTP2 so that intermediaries that
> don't fully understand WS over HTTP2 can relay it.
>
> Since there are not so much space in FLAGS field, Introducing FRAME_DONE
> may decrease the extensiveness of the spec.
>
> Do you have a strong preference to introduce FRAME_DONE?
>
>
>
>
>
> On Thu, Jan 9, 2014 at 6:53 PM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
>
> Hi Yutaka,
>
>
>
> 1) Yes, I understand. Thanks!
>
>
>
> 2) Ok, both WS frame and message boundaries are preserved.
>
>
>
> But: I cannot recognize the end of a WS frame until the next WS frame
> starts (for which I receive a HEADERS frame, which is the implicit signal
> that the previous WS frame ended), right?
>
>
>
> And if so, why is there no FRAME_DONE bit on the SPDY frame that
> corresponds to the end of a WS frame?
>
>
>
> 3) Ok, great! Hope this works, since SPDY provides header compression, but
> no sophisticated bidirectional payload compression as "WS
> permessage-deflate" does.
>
>
>
> Cheers,
>
> /Tobias
>
>
>
>
>
>
>
> *Von:* yhirano@google.com [mailto:yhirano@google.com] *Im Auftrag von *Yutaka
> Hirano
> *Gesendet:* Donnerstag, 9. Januar 2014 10:37
> *An:* Tobias Oberstein
> *Cc:* Takeshi Yoshino; hybi@ietf.org
> *Betreff:* Re: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
>
> Hi Tobias,
>
> Thank you for the comments.
>
>
>
> 1) Frame Mapping
> """
> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
> following plural SPDY data frames. The SPDY HEADERS frame must contain
> WebSocket frame fields and following SPDY data frames contain payload data.
> These data frames are free to be reframed.
> """
>
> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
> So reframing of WebSocket frames (the payload) into multiple SPDY data
> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
> is free to choose the boundaries within WS frame payload at which to
> reframe into different SPDY data frames as it likes?
>
> A WebSocket frame MAY be split to multiple SPDY DATA frames.
>
> Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA frame
> into multiple DATA frames.
>
> Is this answering your question?
>
>
>
> 2) End of WS Frame
> The end of a WS message is signaled with a SPDY data frame having MSG_DONE
> set.
> How is the end of a WS frame signaled, given that the WS frame may have
> been reframed into multiple SPDY data frames?
>
> When you receive a HEADERS frame, you can recognize that the last SPDY
> frame is that last frame in a WebSocket frame.
>
> When you receive a SPDY frame with MSG_DONE, you can recognize that it is
> the last SPDY frame in a WebSocket message and hence the last SPDY frame in
> a WebSocket frame.
>
>
>
> 3) WS extension
> Are WS extensions fully supported? In particular, is "permessage-deflate"
> supported when running WS/SPDY?
>
> Yes. We have reserved bits and we can emulate the opening handshake, I
> think.
>
>
>
> On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
>
> Hi Takeshi,
>
> it's a pity, but I can understand the motivation to shove everything into
> HTTP2.
>
> Anyway, some questions:
>
> 1) Frame Mapping
> """
> One WebSocket frame is mapped into one preceding SPDY HEADERS frame and
> following plural SPDY data frames. The SPDY HEADERS frame must contain
> WebSocket frame fields and following SPDY data frames contain payload data.
> These data frames are free to be reframed.
> """
>
> https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit
>
> So reframing of WebSocket frames (the payload) into multiple SPDY data
> frames MAY occur, but does not need to occur? And a WS/SPDY implementation
> is free to choose the boundaries within WS frame payload at which to
> reframe into different SPDY data frames as it likes?
>
> 2) End of WS Frame
>
> The end of a WS message is signaled with a SPDY data frame having MSG_DONE
> set.
>
> How is the end of a WS frame signaled, given that the WS frame may have
> been reframed into multiple SPDY data frames?
>
> 3) WS extension
>
> Are WS extensions fully supported? In particular, is "permessage-deflate"
> supported when running WS/SPDY?
>
> Thanks!
> /Tobias
>
>
> Von: hybi [mailto:hybi-bounces@ietf.org] Im Auftrag von Takeshi Yoshino
> Gesendet: Dienstag, 7. Januar 2014 08:36
> An: hybi@ietf.org
> Betreff: [hybi] Discontinuation of mux standardizaton in favor of
> WS/HTTP/2.0
>
>
> Hi all,
>
> We've been working on mux extension to make WebSocket scalable for long
> time, but as SPDY got widely supported and adopted quickly, now there seems
> to be no longer any good reason to continue this work which is WebSocket
> dedicated.
>
> Scalability is still important issue to address. From now, I'd like to
> focus on WS/SPDY effort which we've been working on in parallel with the
> mux extension.
>
> Please take a look at the latest proposal in this post.
> https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc
>
> If you really want the dedicated mux extension than one over SPDY, sorry
> but please take over the work.
>
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>
>
>
>
>
>
>
>
>

--001a11c2f58e277c1d04f065bb95
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Thank you for the explanation, finally I may understand...=
<div>Your proposal (SPDY/Messaging) is a brand-new full-duplex messaging pr=
otocol on SPDY and the only constraint is compatibility with the W3C WebSoc=
ket API, right?<br>
</div><div><br></div></div><div class=3D"gmail_extra"><br><br><div class=3D=
"gmail_quote">On Mon, Jan 20, 2014 at 8:16 PM, Tobias Oberstein <span dir=
=3D"ltr">&lt;<a href=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blan=
k">tobias.oberstein@tavendo.de</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div lang=3D"DE" link=3D"blue" vlink=3D"purp=
le"><div><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family=
:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Hi Yutaka,<u></u=
><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">in fact, I complete=
ly agree with Peter (<a href=3D"http://www.ietf.org/mail-archive/web/hybi/c=
urrent/msg10310.html" target=3D"_blank">http://www.ietf.org/mail-archive/we=
b/hybi/current/msg10310.html</a>) - either have <u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">1) full WebSocket t=
unneled over SPDY (&quot;SPDY/WS-Tunnel&quot;)<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=3D&gt; this is the propo=
sal to transport full WS frames as payload of SPDY data frames describe in =
my other post<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">or<u></u><u></u></s=
pan></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">2) define a SPDY me=
ssaging, independent of WS (&quot;SPDY/Messaging&quot;)<u></u><u></u></span=
></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=3D&gt; This is the propo=
sal below (</span>Fri, Jan 10, 2014 at 5:37 PM)<span style=3D"font-size:11.=
0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">.=
 Or something in this direction.<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">The current SPDY/WS=
 proposal is different from both 1) and 2) above, and again I agree with Pe=
ter, that this is the least desirable of all.<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">My order of prefere=
nce would be:<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">++1:=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0=C2=A0 SPDY/Messaging<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">+1:=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0=C2=A0=C2=A0 SPDY/WS-Tunnel<u></u><u></u></span></p><p class=
=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&qu=
ot;,&quot;sans-serif&quot;;color:#1f497d">0:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 SPDY/WS<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">/Tobias<u></u><u></=
u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u=
></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u=
></span></p>
<div style=3D"border:none;border-left:solid blue 1.5pt;padding:0cm 0cm 0cm =
4.0pt"><div><div style=3D"border:none;border-top:solid #e1e1e1 1.0pt;paddin=
g:3.0pt 0cm 0cm 0cm"><p class=3D"MsoNormal"><b><span style=3D"font-size:11.=
0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;">Von:</span></b>=
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;"> <a href=3D"mailto:yhirano@google.com" target=3D"_blank">yhira=
no@google.com</a> [mailto:<a href=3D"mailto:yhirano@google.com" target=3D"_=
blank">yhirano@google.com</a>] <b>Im Auftrag von </b>Yutaka Hirano<br>
<b>Gesendet:</b> Montag, 20. Januar 2014 08:04<br><b>An:</b> Tobias Oberste=
in<br><b>Cc:</b> Roberto Peon; Takeshi Yoshino; <a href=3D"mailto:hybi@ietf=
.org" target=3D"_blank">hybi@ietf.org</a></span></p><div><div class=3D"h5">=
<br>
<b>Betreff:</b> Re: [hybi] Discontinuation of mux standardizaton in favor o=
f WS/HTTP/2.0<u></u><u></u></div></div><p></p></div></div><div><div class=
=3D"h5"><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><div><div><p class=
=3D"MsoNormal">
Tobias, let me confirm.<u></u><u></u></p></div><div><p class=3D"MsoNormal">=
=C2=A0<u></u><u></u></p></div><div><blockquote style=3D"border:none;border-=
left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin=
-right:0cm">
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">I&#39;d like to suggest a=
 different WS/SPDY:<br>=C2=A0<br></span>Pack WebSocket frames _including_ f=
ull WS frame header as _payload_ of SPDY data frames only - no SPDY header =
frames at all.<u></u><u></u></p>
</blockquote></div><div><p class=3D"MsoNormal">My understanding is that by =
suggesting the above one you retracted the below proposal you made before, =
is that right?<u></u><u></u></p></div><div><p class=3D"MsoNormal"><u></u>=
=C2=A0<u></u></p>
</div><div><p class=3D"MsoNormal">On Fri, Jan 10, 2014 at 5:37 PM, Tobias O=
berstein=C2=A0&lt;<a href=3D"mailto:tobias.oberstein@tavendo.de" target=3D"=
_blank">tobias.oberstein@tavendo.de</a>&gt;=C2=A0wrote:<u></u><u></u></p></=
div><blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;paddin=
g:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm">
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ar=
ial&quot;,&quot;sans-serif&quot;">Why not map WS messages directly into SPD=
Y data frames like so:<br>Have 2 bits in SPDY data frames:<br>0x03 =3D MSG_=
TYPE : 0 =3D=3D text, 1 =3D=3D binary<br>
0x04 =3D MSG_DONE<br>A WS text message would simply be mapped into a series=
 of SPDY data frames:<br>MSG_TYPE =3D 0, MSG_DONE =3D 0<br>MSG_TYPE =3D 0, =
MSG_DONE =3D 0<br>...<br>MSG_TYPE =3D 0, MSG_DONE =3D 1<br>Similar for a WS=
 binary message.<br>
=C2=A0</span>=C2=A0<u></u><u></u></p></blockquote><blockquote style=3D"bord=
er:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-le=
ft:4.8pt;margin-right:0cm"><p class=3D"MsoNormal"><span style=3D"font-size:=
10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&quot;">Doing like abo=
ve will allow to preserve the current WS API in HTML5 (which needs messages=
, and needs 2 message types).</span><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div><div=
><p class=3D"MsoNormal">Thanks,<u></u><u></u></p></div><div><p class=3D"Mso=
Normal">=C2=A0<u></u><u></u></p></div></div><div><p class=3D"MsoNormal" sty=
le=3D"margin-bottom:12.0pt">
<u></u>=C2=A0<u></u></p><div><p class=3D"MsoNormal">On Wed, Jan 15, 2014 at=
 4:09 AM, Tobias Oberstein &lt;<a href=3D"mailto:tobias.oberstein@tavendo.d=
e" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt; wrote:<u></u><u></=
u></p><blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padd=
ing:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm">
<div><div><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-famil=
y:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Hi Roberto,</sp=
an><u></u><u></u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt=
;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=
=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">not sure if you agree, bu=
t FWIW, here is a summary as I see it (currently):</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">1) WebSocket frame =
boundaries _can_ have semantic meaning in WebSocket _extensions_</span><u><=
/u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">2) WebSocket messag=
e and frame boundaries are indeed preserved in the current WS/SPDY proposal=
,</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">however, non-final WebSoc=
ket frame ends cannot be _immediately_ detected.</span><u></u><u></u></p><p=
 class=3D"MsoNormal">
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></u></p><p class=3D"MsoNo=
rmal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot=
;sans-serif&quot;;color:#1f497d">3) The current WS/SPDY proposal</span><u><=
/u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0 * is comfortable t=
o implement for implementations that _only_ want to implement WS/SPDY, and =
not native WS at the same time</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0* makes things comp=
licated for WS implementations that want to support both WS/SPDY _and_ nati=
ve WS at the same time</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0* has significant w=
ire-level overhead compared to native WS (and compared to the proposal belo=
w)</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">I&#39;d like to sug=
gest a different WS/SPDY:</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p>Pack WebSocket frames _including_ full WS frame header as _payload=
_ of SPDY data frames only - no SPDY header frames at all.<u></u><u></u></p=
>
<p>=C2=A0<u></u><u></u></p><p>In the absence of negotiated WebSocket extens=
ions (or when all WS extensions in place do not assign semantics to WS fram=
e boundaries),<u></u><u></u></p><p>fragment WebSocket payload at the WS lev=
el such that WS frame header+payload fits into the payload of SPDY data fra=
mes.<u></u><u></u></p>
<p>=C2=A0<u></u><u></u></p><p>In the latter case, this essentially adds a s=
mall SPDY data frame header (of fixed length) to WS frames. Easy to impleme=
nt and efficient on wire.<u></u><u></u></p><p>=C2=A0<u></u><u></u></p><p>IM=
HO, this has advantages compared to current proposal and considering 3) abo=
ve.<u></u><u></u></p>
<p>=C2=A0<u></u><u></u></p><p>The only disadvantage is for implementations =
that _only_ want to implement WS/SPDY and not native WS at the same time.<u=
></u><u></u></p><p>=C2=A0<u></u><u></u></p><p>=3D=3D<u></u><u></u></p><p>=
=C2=A0<u></u><u></u></p>
<p>Regarding &quot;permessage-priority&quot; with ordering guarantees: shou=
ld I find time and if there is interest, I&#39;ll write something up.<u></u=
><u></u></p><p>This could be a WebSocket extension. To be useful, the WS AP=
I in browsers would need to expose it though ..<u></u><u></u></p>
<p>=C2=A0<u></u><u></u></p><p>Just my 2cts,<u></u><u></u></p><p>Tobias<u></=
u><u></u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-fa=
mily:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span=
><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u=
><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><div style=3D"border:none;border-left:solid blue 1.5pt;padding:0cm 0c=
m 0cm 4.0pt">
<div><div style=3D"border:none;border-top:solid #e1e1e1 1.0pt;padding:3.0pt=
 0cm 0cm 0cm"><p class=3D"MsoNormal"><b><span style=3D"font-size:11.0pt;fon=
t-family:&quot;Calibri&quot;,&quot;sans-serif&quot;">Von:</span></b><span s=
tyle=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&q=
uot;"> Roberto Peon [mailto:<a href=3D"mailto:fenix@google.com" target=3D"_=
blank">fenix@google.com</a>] <br>
<b>Gesendet:</b> Dienstag, 14. Januar 2014 19:30<br><b>An:</b> Yutaka Hiran=
o<br><b>Cc:</b> Tobias Oberstein; Takeshi Yoshino; <a href=3D"mailto:hybi@i=
etf.org" target=3D"_blank">hybi@ietf.org</a></span><u></u><u></u></p><div><=
div>
<p class=3D"MsoNormal"><br><b>Betreff:</b> Re: [hybi] Discontinuation of mu=
x standardizaton in favor of WS/HTTP/2.0<u></u><u></u></p></div></div></div=
></div><div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p><div><p cla=
ss=3D"MsoNormal">
I&#39;ve been lurking anyway :)<u></u><u></u></p><div><p class=3D"MsoNormal=
">=C2=A0<u></u><u></u></p></div><div><p class=3D"MsoNormal">As I understand=
 the proposal, (and it seems that Takeshi and Yukata are covering this nice=
ly, so what i say is probably redundant):<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">Message boundaries, having semantic meani=
ng, are preserved.<u></u><u></u></p></div><div><p class=3D"MsoNormal">=C2=
=A0<u></u><u></u></p></div><div><p class=3D"MsoNormal">Frame boundaries, ha=
ving no semantic meaning, are not preserved. I don&#39;t think it makes sen=
se to worry about this greatly when the ability to extend the protocol as i=
mplemented in HTTP2 is present. It doesn&#39;t make a lot of sense to signa=
l the frame boundaries through everything unless there is a high-likelyhood=
 and high-utility use-case for such. If such is found, then HTTP2 can be re=
vised to include it. This is the approach we&#39;ve been taking with most o=
f the HTTP2 stuff, since there has been great effort placed on allowing new=
 versions to be deployed in the future.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><div><p cla=
ss=3D"MsoNormal">The mapping of WS onto HTTP2 is fairly direct-- we reserve=
d the MSG_DONE bit specifically for this kind of use-case (i.e. bidirection=
al, full-duplex, message-based).<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">The observation made is that there are es=
sentially 3 types of data that need to be preserved one way or another:<u><=
/u><u></u></p></div><div><p class=3D"MsoNormal">1) metadata: data about the=
 data, e.g. text, binary, where it is supposed to be going, etc.<u></u><u><=
/u></p>
</div><div><p class=3D"MsoNormal">2) payload<u></u><u></u></p></div><div><p=
 class=3D"MsoNormal">3) boundaries between these.<u></u><u></u></p></div><d=
iv><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><div><p class=3D"Ms=
oNormal">
imho, metadata is handled with HEADERS (I&#39;d have preferred it to be cal=
led METADATA, but was voted down at the time :) ), payload is handled with =
DATA frames, and the boundaries are delimited either through the use of HEA=
DERS frames, or via the MSG_DONE or STREAM_DONE bits.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">In the presence of extensions, assuming t=
hat MSG_DONE is required to signal the end of a message, then HEADERS can b=
e used to signal any boundary desired.<u></u><u></u></p></div><div><p class=
=3D"MsoNormal">
=C2=A0<u></u><u></u></p></div><div><p class=3D"MsoNormal">=C2=A0<u></u><u><=
/u></p></div><div><p class=3D"MsoNormal">I totally agree that it would be n=
ice to expose the priority of various messages/channels with respect to eac=
h other. That is really an issue that is outside the WG&#39;s purvue, thoug=
h.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div></div><div>=
<p class=3D"MsoNormal" style=3D"margin-bottom:12.0pt">=C2=A0<u></u><u></u><=
/p><div><p class=3D"MsoNormal">On Thu, Jan 9, 2014 at 7:54 PM, Yutaka Hiran=
o &lt;<a href=3D"mailto:yhirano@chromium.org" target=3D"_blank">yhirano@chr=
omium.org</a>&gt; wrote:<u></u><u></u></p>
<blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0c=
m 0cm 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-=
bottom:5.0pt"><div><div><p class=3D"MsoNormal">+Roberto as an HTTP2 expert.=
<u></u><u></u></p>
</div><div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><block=
quote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm =
0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom=
:5.0pt">
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">2) Ok, both WS frame and =
message boundaries are preserved.<br>=C2=A0<br>But: I cannot recognize the =
end of a WS frame until the next WS frame starts (for which I receive a HEA=
DERS frame, which is the implicit signal that the previous WS frame ended),=
 right?</span><u></u><u></u></p>
</blockquote></div><p class=3D"MsoNormal">You are right, you can&#39;t.<u><=
/u><u></u></p><div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></di=
v><blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:=
0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margi=
n-bottom:5.0pt">
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0<br>And if so, why =
is there no FRAME_DONE bit on the SPDY frame that corresponds to the end of=
 a WS frame?</span><u></u><u></u></p>
</blockquote></div><p class=3D"MsoNormal">Given that WebSocket frames &quot=
;have no semantic meaning&quot; as in RFC6455, I don&#39;t think the immedi=
ate recognition of the end of a WebSocket frame is requisite.<u></u><u></u>=
</p>
<div><p class=3D"MsoNormal">I would like to introduce MSG_DONE in HTTP2 so =
that intermediaries that don&#39;t fully understand WS over HTTP2 can relay=
 it.<u></u><u></u></p></div><div><p class=3D"MsoNormal">Since there are not=
 so much space in FLAGS field, Introducing FRAME_DONE may decrease the exte=
nsiveness of the spec.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">Do you have a strong preference to introd=
uce FRAME_DONE?<u></u><u></u></p></div><div><p class=3D"MsoNormal">=C2=A0<u=
></u><u></u></p></div></div><div><div><div><p class=3D"MsoNormal" style=3D"=
margin-bottom:12.0pt">
=C2=A0<u></u><u></u></p><div><p class=3D"MsoNormal">On Thu, Jan 9, 2014 at =
6:53 PM, Tobias Oberstein &lt;<a href=3D"mailto:tobias.oberstein@tavendo.de=
" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt; wrote:<u></u><u></u=
></p><blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;paddi=
ng:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;ma=
rgin-bottom:5.0pt">
<div><div><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-famil=
y:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Hi Yutaka,</spa=
n><u></u><u></u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;=
font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=
=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">1) Yes, I understand. Tha=
nks!</span><u></u><u></u></p><p class=3D"MsoNormal"><span style=3D"font-siz=
e:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f49=
7d">=C2=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">2) Ok, both WS frame and =
message boundaries are preserved.</span><u></u><u></u></p><p class=3D"MsoNo=
rmal">
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></u></p><p class=3D"MsoNo=
rmal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot=
;sans-serif&quot;;color:#1f497d">But: I cannot recognize the end of a WS fr=
ame until the next WS frame starts (for which I receive a HEADERS frame, wh=
ich is the implicit signal that the previous WS frame ended), right?</span>=
<u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">And if so, why is t=
here no FRAME_DONE bit on the SPDY frame that corresponds to the end of a W=
S frame?</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">3) Ok, great! Hope =
this works, since SPDY provides header compression, but no sophisticated bi=
directional payload compression as &quot;WS permessage-deflate&quot; does.<=
/span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">Cheers,</span><u></=
u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">/Tobias</span><u></u><u><=
/u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&=
quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></=
u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">=C2=A0</span><u></u=
><u></u></p>
<div style=3D"border:none;border-left:solid blue 1.5pt;padding:0cm 0cm 0cm =
4.0pt"><div><div style=3D"border:none;border-top:solid #e1e1e1 1.0pt;paddin=
g:3.0pt 0cm 0cm 0cm"><p class=3D"MsoNormal"><b><span style=3D"font-size:11.=
0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;">Von:</span></b>=
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;"> </span><a href=3D"mailto:yhirano@google.com" target=3D"_blank=
"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;san=
s-serif&quot;">yhirano@google.com</span></a><span style=3D"font-size:11.0pt=
;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;"> [mailto:</span><a=
 href=3D"mailto:yhirano@google.com" target=3D"_blank"><span style=3D"font-s=
ize:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;">yhirano@=
google.com</span></a><span style=3D"font-size:11.0pt;font-family:&quot;Cali=
bri&quot;,&quot;sans-serif&quot;">] <b>Im Auftrag von </b>Yutaka Hirano<br>
<b>Gesendet:</b> Donnerstag, 9. Januar 2014 10:37<br><b>An:</b> Tobias Ober=
stein<br><b>Cc:</b> Takeshi Yoshino; </span><a href=3D"mailto:hybi@ietf.org=
" target=3D"_blank"><span style=3D"font-size:11.0pt;font-family:&quot;Calib=
ri&quot;,&quot;sans-serif&quot;">hybi@ietf.org</span></a><span style=3D"fon=
t-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;"><br>
<b>Betreff:</b> Re: [hybi] Discontinuation of mux standardizaton in favor o=
f WS/HTTP/2.0</span><u></u><u></u></p></div></div><div><div><p class=3D"Mso=
Normal">=C2=A0<u></u><u></u></p><div><p class=3D"MsoNormal">Hi Tobias,<u></=
u><u></u></p>
<div><p class=3D"MsoNormal">Thank you for the comments.<u></u><u></u></p></=
div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p><blockquote style=
=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm 6.0pt;m=
argin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom:5.0pt">
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ar=
ial&quot;,&quot;sans-serif&quot;">1) Frame Mapping<br>&quot;&quot;&quot;<br=
>One WebSocket frame is mapped into one preceding SPDY HEADERS frame and fo=
llowing plural SPDY data frames. The SPDY HEADERS frame must contain WebSoc=
ket frame fields and following SPDY data frames contain payload data. These=
 data frames are free to be reframed.<br>
&quot;&quot;&quot;<br></span><a href=3D"https://docs.google.com/document/d/=
124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank"><span =
style=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&qu=
ot;">https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw=
9aWzSHijk/edit</span></a><span style=3D"font-size:10.0pt;font-family:&quot;=
Arial&quot;,&quot;sans-serif&quot;"><br>
So reframing of WebSocket frames (the payload) into multiple SPDY data fram=
es MAY occur, but does not need to occur? And a WS/SPDY implementation is f=
ree to choose the boundaries within WS frame payload at which to reframe in=
to different SPDY data frames as it likes?</span><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal">A WebSocket frame MAY be split to =
multiple SPDY DATA frames.<u></u><u></u></p></div><div><p class=3D"MsoNorma=
l">Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA fram=
e into multiple DATA frames.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">Is this answering your question?<u></u><u=
></u></p></div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><b=
lockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm =
0cm 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bo=
ttom:5.0pt">
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ar=
ial&quot;,&quot;sans-serif&quot;">2) End of WS Frame<br>The end of a WS mes=
sage is signaled with a SPDY data frame having MSG_DONE set.<br>How is the =
end of a WS frame signaled, given that the WS frame may have been reframed =
into multiple SPDY data frames?</span><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal">When you receive a HEADERS frame, =
you can recognize that the last SPDY frame is that last frame in a WebSocke=
t frame.<u></u><u></u></p></div><div><p class=3D"MsoNormal">When you receiv=
e a SPDY frame with MSG_DONE, you can recognize that it is the last SPDY fr=
ame in a WebSocket message and hence the last SPDY frame in a WebSocket fra=
me.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><blockquote=
 style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm 6=
.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom:5.0p=
t"><p class=3D"MsoNormal">
<span style=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-se=
rif&quot;">3) WS extension<br>Are WS extensions fully supported? In particu=
lar, is &quot;permessage-deflate&quot; supported when running WS/SPDY?</spa=
n><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal">Yes. We have reserved bits and we =
can emulate the opening handshake, I think.<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">=C2=A0<u></u><u></u></p></div><div><p class=3D"MsoNorma=
l">On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein &lt;<a href=3D"mailto:t=
obias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstein@tavendo.de</=
a>&gt; wrote:<u></u><u></u></p>
<blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0c=
m 0cm 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-=
bottom:5.0pt"><p class=3D"MsoNormal">Hi Takeshi,<br><br>it&#39;s a pity, bu=
t I can understand the motivation to shove everything into HTTP2.<br>
<br>Anyway, some questions:<br><br>1) Frame Mapping<br>&quot;&quot;&quot;<b=
r>One WebSocket frame is mapped into one preceding SPDY HEADERS frame and f=
ollowing plural SPDY data frames. The SPDY HEADERS frame must contain WebSo=
cket frame fields and following SPDY data frames contain payload data. Thes=
e data frames are free to be reframed.<br>
&quot;&quot;&quot;<br><a href=3D"https://docs.google.com/document/d/124vB5N=
V1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit" target=3D"_blank">https://docs.=
google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWzSHijk/edit</a>=
<br>
<br>So reframing of WebSocket frames (the payload) into multiple SPDY data =
frames MAY occur, but does not need to occur? And a WS/SPDY implementation =
is free to choose the boundaries within WS frame payload at which to refram=
e into different SPDY data frames as it likes?<br>
<br>2) End of WS Frame<br><br>The end of a WS message is signaled with a SP=
DY data frame having MSG_DONE set.<br><br>How is the end of a WS frame sign=
aled, given that the WS frame may have been reframed into multiple SPDY dat=
a frames?<br>
<br>3) WS extension<br><br>Are WS extensions fully supported? In particular=
, is &quot;permessage-deflate&quot; supported when running WS/SPDY?<br><br>=
Thanks!<br>/Tobias<br><br><br>Von: hybi [mailto:<a href=3D"mailto:hybi-boun=
ces@ietf.org" target=3D"_blank">hybi-bounces@ietf.org</a>] Im Auftrag von T=
akeshi Yoshino<br>
Gesendet: Dienstag, 7. Januar 2014 08:36<br>An: <a href=3D"mailto:hybi@ietf=
.org" target=3D"_blank">hybi@ietf.org</a><br>Betreff: [hybi] Discontinuatio=
n of mux standardizaton in favor of WS/HTTP/2.0<u></u><u></u></p><div><div>
<p class=3D"MsoNormal"><br>Hi all,<br><br>We&#39;ve been working on mux ext=
ension to make WebSocket scalable for long time, but as SPDY got widely sup=
ported and adopted quickly, now there seems to be no longer any good reason=
 to continue this work which is WebSocket dedicated.<br>
<br>Scalability is still important issue to address. From now, I&#39;d like=
 to focus on WS/SPDY effort which we&#39;ve been working on in parallel wit=
h the mux extension.<br><br>Please take a look at the latest proposal in th=
is post.<br>
<a href=3D"https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc" ta=
rget=3D"_blank">https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3gl=
c</a><br><br>If you really want the dedicated mux extension than one over S=
PDY, sorry but please take over the work.<u></u><u></u></p>
</div></div><div><div><p class=3D"MsoNormal">______________________________=
_________________<br>hybi mailing list<br><a href=3D"mailto:hybi@ietf.org" =
target=3D"_blank">hybi@ietf.org</a><br><a href=3D"https://www.ietf.org/mail=
man/listinfo/hybi" target=3D"_blank">https://www.ietf.org/mailman/listinfo/=
hybi</a><u></u><u></u></p>
</div></div></blockquote></div><p class=3D"MsoNormal">=C2=A0<u></u><u></u><=
/p></div></div></div></div></div></div></div></blockquote></div><p class=3D=
"MsoNormal">=C2=A0<u></u><u></u></p></div></div></div></blockquote></div><p=
 class=3D"MsoNormal">
=C2=A0<u></u><u></u></p></div></div></div></div></div></div></blockquote></=
div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p></div></div></div></div>=
</div></div></blockquote></div><br></div>

--001a11c2f58e277c1d04f065bb95--


From tobias.oberstein@tavendo.de  Mon Jan 20 04:17:35 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C53591A012B for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 04:17:35 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ViJnFWlMDfSL for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 04:17:34 -0800 (PST)
Received: from EXHUB020-2.exch020.serverdata.net (exhub020-2.exch020.serverdata.net [206.225.164.29]) by ietfa.amsl.com (Postfix) with ESMTP id D0A5A1A0123 for <hybi@ietf.org>; Mon, 20 Jan 2014 04:17:33 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-2.exch020.serverdata.net ([206.225.164.29]) with mapi; Mon, 20 Jan 2014 04:17:33 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Yutaka Hirano <yhirano@chromium.org>
Date: Mon, 20 Jan 2014 04:17:31 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8V2CahuJX21kmRTFWTvSepVX5S0gAADdqg
Message-ID: <634914A010D0B943A035D226786325D4446BF9997A@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com>
In-Reply-To: <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>, "Peter H Thorson \(webmaster@zaphoyd.com\)" <webmaster@zaphoyd.com>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 12:17:36 -0000
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From webmaster@zaphoyd.com  Mon Jan 20 07:31:30 2014
Return-Path: <webmaster@zaphoyd.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id D73A01A01B9 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 07:31:30 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.899
X-Spam-Level: 
X-Spam-Status: No, score=-1.899 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Ad4oCkCj3cOC for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 07:31:23 -0800 (PST)
Received: from authsmtp01.uchicago.edu (authsmtp01.uchicago.edu [128.135.12.121]) by ietfa.amsl.com (Postfix) with ESMTP id 10A881A01B4 for <hybi@ietf.org>; Mon, 20 Jan 2014 07:31:22 -0800 (PST)
Received: from [10.0.1.3] (c-68-51-76-178.hsd1.il.comcast.net [68.51.76.178]) (Authenticated sender: zaphoyd) by authsmtp01.uchicago.edu (Postfix) with ESMTP id 28678498098; Mon, 20 Jan 2014 09:31:21 -0600 (CST)
Content-Type: multipart/alternative; boundary="Apple-Mail=_A1C8F556-3822-4BCA-938D-030D010F3104"
Mime-Version: 1.0 (Mac OS X Mail 7.0 \(1822\))
From: Peter Thorson <webmaster@zaphoyd.com>
In-Reply-To: <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com>
Date: Mon, 20 Jan 2014 09:34:28 -0600
Message-Id: <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com>
To: Yutaka Hirano <yhirano@chromium.org>
X-Mailer: Apple Mail (2.1822)
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 15:31:31 -0000

--Apple-Mail=_A1C8F556-3822-4BCA-938D-030D010F3104
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=windows-1252

Hi all,

Sorry about the delay in clarifying my previous message.

> Your proposal (SPDY/Messaging) is a brand-new full-duplex messaging =
protocol on SPDY and the only constraint is compatibility with the W3C =
WebSocket API, right?


That statement describes my point in a nutshell.=20

My reasoning and more details/questions:

The goal [roughly] of WebSocket was to provide a protocol for =
establishing a persistent, ordered, bidirectional messaging channel with =
minimal framing overhead. This channel=92s most important user would be =
JavaScript applications in web browsers so browser related security =
concerns and the API there are significant. HTTP 1.1 does not have the =
functionality required to implement any semblance of these requirements. =
As such, RFC6455 defines a handshake, framing layer, error handling =
layer, and guidelines for extensions for a standalone TCP connection =
that fulfills these requirements.

HTTP 2.0 on the other hand already provides all of the underlying =
plumbing to provide a persistent, ordered, bidirectional messaging =
channel. =46rom a browser=92s prospective, whether the WebSocket channel =
that it opens goes through a separate TCP connection, a hybrid =
=93websocket frames encapsulated within HTTP 2.0 frames, or native HTTP =
2.0 frames really doesn=92t matter as long as HTTP 2.0 is able to match =
the functionality of the WebSocket browser API.

A simple proposal to support WebSocket API functionality might be =
something like:

- HTTP 2.0 HEADERS message already provides the functionality necessary =
to negotiate a given HTTP 2.0 channel as a =93WebSocket=94 channel along =
with sending subprotocol, cookies, etc.
- An HTTP 2.0 WEBSOCKET frame type could be used to deliver WebSocket =
messages. You=92d need a FIN flag for the message boundaries, frame =
boundaries would be just standard HTTP 2.0 frames. No need to repackage =
or reframe or make a distinction between =93websocket=94 frames and HTTP =
2.0 frames.
- Depending on how closely we want HTTP 2.0 WebSockets to match RFC6455 =
WebSockets additional support could be added.. for example, define the =
WEBSOCKET frame type=92s flag bits as follows (full compatibility flag =
configuration):

   0     1     2     3     4     5     6     7
+-----+-----+-----+-----+-----+-----+-----+-----+
| FIN |     RFC6455 OPCODE    |RSV1 |RSV2 |RSV3 |
+-----+-----+-----+-----+-----+-----+-----+-----+

This provides the ability to keep RFC6455 opcodes and RSV bits without. =
Logical WebSocket stream ending would be handled exactly as in RFC6455 =
by exchanging close frames. This would potentially allow an easier =
migration for RFC6455 extensions by including the RSV bits.

A more simple frame structure could also be imagined without as much of =
the RFC6455 compatibility considering the following assumptions:=20
- Half of the RFC6455 opcodes are reserved for control messages that so =
far are entirely provided by HTTP 2.0 natively (ping/pong, stream ending =
flags, etc).
- So far only two =93data=94 opcodes have been defined.
- RFC6455 extensions have proven to be fairly invasive to base =
implementations and require a lot of careful thought about how they will =
interact with each other and the rest of the protocol. Raw WebSocket and =
HTTP 2.0 may have different enough framing layers that extensions for =
one may not be directly translatable in the same way the framing isn=92t =
directly translatable. It may be sufficient to adapt the interfaces of =
popular WebSocket extensions over HTTP 2.0 mechanisms rather than =
literally try and implement WS extensions in HTTP 2.0. Take for example =
permessage-deflate extensions. HTTP 2.0 by default will have deflate =
support and will be maintaining compression contexts for header =
compression. Multiple different streams within the connection (either =
HTTP or other WebSockets) might also be interested in compression. Would =
a more generic compression system that could share state among these =
connections make sense here rather than individually stored and =
negotiated compression parameters per logical connection? It certainly =
would not make sense for raw WebSocket (sans MUX at least).

Example of a potentially simpler WEBSOCKET frame flags (limited =
compatibility flag configuration):

   0     1     2     3     4     5     6     7
+-----+-----+-----+-----+-----+-----+-----+-----+
| SFIN| MFIN|UTF8 |     |     |RSV1*|RSV2*|RSV3*|
+-----+-----+-----+-----+-----+-----+-----+-----+

SFIN =3D stream fin
MFIN =3D message fin
UTF8 =3D UTF8 message, validate as such. Alternatively bits 2-4 could =
represent a =93data opcode=94 similar to the lower 3 bits of RFC6455=92s =
opcode (ie drop the control opcodes).
RSV1-3 could be included as mapped WebSocket extension bits or just used =
for HTTP 2.0 websocket message bits.

This configuration would also open the option of reusing the straight up =
DATA frame as that already has the SFIN bit, though the =93reserved=94 =
#2 bit might be a problem depending on how reserved it actually is. =
There is certainly enough information in the headers to allow an =
implementation to know whether a channel is a WebSocket one without a =
separate message type. I still think a separate message is a good idea =
because the types of flags for these types of messages will be different =
and the frame processing logic will be different and i=92ve found in my =
applications at least it is more straight forward to switch on frame =
type than to keep state about how to interpret a generic =93DATA=94 =
frame. HTTP 2.0 has a significantly larger range for frame types than =
RFC6455 did as well.

On Jan 20, 2014, at 6:07, Yutaka Hirano <yhirano@chromium.org> wrote:

> Thank you for the explanation, finally I may understand...
> Your proposal (SPDY/Messaging) is a brand-new full-duplex messaging =
protocol on SPDY and the only constraint is compatibility with the W3C =
WebSocket API, right?
>=20
>=20
>=20
> On Mon, Jan 20, 2014 at 8:16 PM, Tobias Oberstein =
<tobias.oberstein@tavendo.de> wrote:
> Hi Yutaka,
>=20
> =20
>=20
> in fact, I completely agree with Peter =
(http://www.ietf.org/mail-archive/web/hybi/current/msg10310.html) - =
either have
>=20
> =20
>=20
> 1) full WebSocket tunneled over SPDY ("SPDY/WS-Tunnel")
>=20
> =3D> this is the proposal to transport full WS frames as payload of =
SPDY data frames describe in my other post
>=20
> =20
>=20
> or
>=20
> =20
>=20
> 2) define a SPDY messaging, independent of WS ("SPDY/Messaging")
>=20
> =3D> This is the proposal below (Fri, Jan 10, 2014 at 5:37 PM). Or =
something in this direction.
>=20
> =20
>=20
> The current SPDY/WS proposal is different from both 1) and 2) above, =
and again I agree with Peter, that this is the least desirable of all.
>=20
> =20
>=20
> My order of preference would be:
>=20
> =20
>=20
> ++1:       SPDY/Messaging
>=20
> +1:         SPDY/WS-Tunnel
>=20
> 0:            SPDY/WS
>=20
> =20
>=20
> /Tobias
>=20
> =20
>=20
> =20
>=20
> =20
>=20
> =20
>=20
> Von: yhirano@google.com [mailto:yhirano@google.com] Im Auftrag von =
Yutaka Hirano
> Gesendet: Montag, 20. Januar 2014 08:04
> An: Tobias Oberstein
> Cc: Roberto Peon; Takeshi Yoshino; hybi@ietf.org
>=20
>=20
> Betreff: Re: [hybi] Discontinuation of mux standardizaton in favor of =
WS/HTTP/2.0
>=20
> =20
>=20
> Tobias, let me confirm.
>=20
> =20
>=20
> I'd like to suggest a different WS/SPDY:
> =20
> Pack WebSocket frames _including_ full WS frame header as _payload_ of =
SPDY data frames only - no SPDY header frames at all.
>=20
> My understanding is that by suggesting the above one you retracted the =
below proposal you made before, is that right?
>=20
> =20
>=20
> On Fri, Jan 10, 2014 at 5:37 PM, Tobias Oberstein =
<tobias.oberstein@tavendo.de> wrote:
>=20
> Why not map WS messages directly into SPDY data frames like so:
> Have 2 bits in SPDY data frames:
> 0x03 =3D MSG_TYPE : 0 =3D=3D text, 1 =3D=3D binary
> 0x04 =3D MSG_DONE
> A WS text message would simply be mapped into a series of SPDY data =
frames:
> MSG_TYPE =3D 0, MSG_DONE =3D 0
> MSG_TYPE =3D 0, MSG_DONE =3D 0
> ...
> MSG_TYPE =3D 0, MSG_DONE =3D 1
> Similar for a WS binary message.
>  =20
>=20
> Doing like above will allow to preserve the current WS API in HTML5 =
(which needs messages, and needs 2 message types).
>=20
> =20
>=20
> Thanks,
>=20
> =20
>=20
> =20
>=20
> On Wed, Jan 15, 2014 at 4:09 AM, Tobias Oberstein =
<tobias.oberstein@tavendo.de> wrote:
>=20
> Hi Roberto,
>=20
> =20
>=20
> not sure if you agree, but FWIW, here is a summary as I see it =
(currently):
>=20
> =20
>=20
> 1) WebSocket frame boundaries _can_ have semantic meaning in WebSocket =
_extensions_
>=20
> =20
>=20
> 2) WebSocket message and frame boundaries are indeed preserved in the =
current WS/SPDY proposal,
>=20
> however, non-final WebSocket frame ends cannot be _immediately_ =
detected.
>=20
> =20
>=20
> 3) The current WS/SPDY proposal
>=20
>   * is comfortable to implement for implementations that _only_ want =
to implement WS/SPDY, and not native WS at the same time
>=20
>  * makes things complicated for WS implementations that want to =
support both WS/SPDY _and_ native WS at the same time
>=20
>  * has significant wire-level overhead compared to native WS (and =
compared to the proposal below)
>=20
> =20
>=20
> I'd like to suggest a different WS/SPDY:
>=20
> =20
>=20
> Pack WebSocket frames _including_ full WS frame header as _payload_ of =
SPDY data frames only - no SPDY header frames at all.
>=20
> =20
>=20
> In the absence of negotiated WebSocket extensions (or when all WS =
extensions in place do not assign semantics to WS frame boundaries),
>=20
> fragment WebSocket payload at the WS level such that WS frame =
header+payload fits into the payload of SPDY data frames.
>=20
> =20
>=20
> In the latter case, this essentially adds a small SPDY data frame =
header (of fixed length) to WS frames. Easy to implement and efficient =
on wire.
>=20
> =20
>=20
> IMHO, this has advantages compared to current proposal and considering =
3) above.
>=20
> =20
>=20
> The only disadvantage is for implementations that _only_ want to =
implement WS/SPDY and not native WS at the same time.
>=20
> =20
>=20
> =3D=3D
>=20
> =20
>=20
> Regarding "permessage-priority" with ordering guarantees: should I =
find time and if there is interest, I'll write something up.
>=20
> This could be a WebSocket extension. To be useful, the WS API in =
browsers would need to expose it though ..
>=20
> =20
>=20
> Just my 2cts,
>=20
> Tobias
>=20
> =20
>=20
> =20
>=20
> =20
>=20
> =20
>=20
> Von: Roberto Peon [mailto:fenix@google.com]=20
> Gesendet: Dienstag, 14. Januar 2014 19:30
> An: Yutaka Hirano
> Cc: Tobias Oberstein; Takeshi Yoshino; hybi@ietf.org
>=20
>=20
> Betreff: Re: [hybi] Discontinuation of mux standardizaton in favor of =
WS/HTTP/2.0
>=20
> =20
>=20
> I've been lurking anyway :)
>=20
> =20
>=20
> As I understand the proposal, (and it seems that Takeshi and Yukata =
are covering this nicely, so what i say is probably redundant):
>=20
> Message boundaries, having semantic meaning, are preserved.
>=20
> =20
>=20
> Frame boundaries, having no semantic meaning, are not preserved. I =
don't think it makes sense to worry about this greatly when the ability =
to extend the protocol as implemented in HTTP2 is present. It doesn't =
make a lot of sense to signal the frame boundaries through everything =
unless there is a high-likelyhood and high-utility use-case for such. If =
such is found, then HTTP2 can be revised to include it. This is the =
approach we've been taking with most of the HTTP2 stuff, since there has =
been great effort placed on allowing new versions to be deployed in the =
future.
>=20
> =20
>=20
> The mapping of WS onto HTTP2 is fairly direct-- we reserved the =
MSG_DONE bit specifically for this kind of use-case (i.e. bidirectional, =
full-duplex, message-based).
>=20
> The observation made is that there are essentially 3 types of data =
that need to be preserved one way or another:
>=20
> 1) metadata: data about the data, e.g. text, binary, where it is =
supposed to be going, etc.
>=20
> 2) payload
>=20
> 3) boundaries between these.
>=20
> =20
>=20
> imho, metadata is handled with HEADERS (I'd have preferred it to be =
called METADATA, but was voted down at the time :) ), payload is handled =
with DATA frames, and the boundaries are delimited either through the =
use of HEADERS frames, or via the MSG_DONE or STREAM_DONE bits.
>=20
> In the presence of extensions, assuming that MSG_DONE is required to =
signal the end of a message, then HEADERS can be used to signal any =
boundary desired.
>=20
> =20
>=20
> =20
>=20
> I totally agree that it would be nice to expose the priority of =
various messages/channels with respect to each other. That is really an =
issue that is outside the WG's purvue, though.
>=20
> =20
>=20
> =20
>=20
> On Thu, Jan 9, 2014 at 7:54 PM, Yutaka Hirano <yhirano@chromium.org> =
wrote:
>=20
> +Roberto as an HTTP2 expert.
>=20
> =20
>=20
> 2) Ok, both WS frame and message boundaries are preserved.
> =20
> But: I cannot recognize the end of a WS frame until the next WS frame =
starts (for which I receive a HEADERS frame, which is the implicit =
signal that the previous WS frame ended), right?
>=20
> You are right, you can't.
>=20
> =20
>=20
> =20
> And if so, why is there no FRAME_DONE bit on the SPDY frame that =
corresponds to the end of a WS frame?
>=20
> Given that WebSocket frames "have no semantic meaning" as in RFC6455, =
I don't think the immediate recognition of the end of a WebSocket frame =
is requisite.
>=20
> I would like to introduce MSG_DONE in HTTP2 so that intermediaries =
that don't fully understand WS over HTTP2 can relay it.
>=20
> Since there are not so much space in FLAGS field, Introducing =
FRAME_DONE may decrease the extensiveness of the spec.
>=20
> Do you have a strong preference to introduce FRAME_DONE?
>=20
> =20
>=20
> =20
>=20
> On Thu, Jan 9, 2014 at 6:53 PM, Tobias Oberstein =
<tobias.oberstein@tavendo.de> wrote:
>=20
> Hi Yutaka,
>=20
> =20
>=20
> 1) Yes, I understand. Thanks!
>=20
> =20
>=20
> 2) Ok, both WS frame and message boundaries are preserved.
>=20
> =20
>=20
> But: I cannot recognize the end of a WS frame until the next WS frame =
starts (for which I receive a HEADERS frame, which is the implicit =
signal that the previous WS frame ended), right?
>=20
> =20
>=20
> And if so, why is there no FRAME_DONE bit on the SPDY frame that =
corresponds to the end of a WS frame?
>=20
> =20
>=20
> 3) Ok, great! Hope this works, since SPDY provides header compression, =
but no sophisticated bidirectional payload compression as "WS =
permessage-deflate" does.
>=20
> =20
>=20
> Cheers,
>=20
> /Tobias
>=20
> =20
>=20
> =20
>=20
> =20
>=20
> Von: yhirano@google.com [mailto:yhirano@google.com] Im Auftrag von =
Yutaka Hirano
> Gesendet: Donnerstag, 9. Januar 2014 10:37
> An: Tobias Oberstein
> Cc: Takeshi Yoshino; hybi@ietf.org
> Betreff: Re: [hybi] Discontinuation of mux standardizaton in favor of =
WS/HTTP/2.0
>=20
> =20
>=20
> Hi Tobias,
>=20
> Thank you for the comments.
>=20
> =20
>=20
> 1) Frame Mapping
> """
> One WebSocket frame is mapped into one preceding SPDY HEADERS frame =
and following plural SPDY data frames. The SPDY HEADERS frame must =
contain WebSocket frame fields and following SPDY data frames contain =
payload data. These data frames are free to be reframed.
> """
> =
https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWz=
SHijk/edit
> So reframing of WebSocket frames (the payload) into multiple SPDY data =
frames MAY occur, but does not need to occur? And a WS/SPDY =
implementation is free to choose the boundaries within WS frame payload =
at which to reframe into different SPDY data frames as it likes?
>=20
> A WebSocket frame MAY be split to multiple SPDY DATA frames.
>=20
> Intermediaries MAY merge SPDY DATA frames and MAY split a SPDY DATA =
frame into multiple DATA frames.
>=20
> Is this answering your question?
>=20
> =20
>=20
> 2) End of WS Frame
> The end of a WS message is signaled with a SPDY data frame having =
MSG_DONE set.
> How is the end of a WS frame signaled, given that the WS frame may =
have been reframed into multiple SPDY data frames?
>=20
> When you receive a HEADERS frame, you can recognize that the last SPDY =
frame is that last frame in a WebSocket frame.
>=20
> When you receive a SPDY frame with MSG_DONE, you can recognize that it =
is the last SPDY frame in a WebSocket message and hence the last SPDY =
frame in a WebSocket frame.
>=20
> =20
>=20
> 3) WS extension
> Are WS extensions fully supported? In particular, is =
"permessage-deflate" supported when running WS/SPDY?
>=20
> Yes. We have reserved bits and we can emulate the opening handshake, I =
think.
>=20
> =20
>=20
> On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein =
<tobias.oberstein@tavendo.de> wrote:
>=20
> Hi Takeshi,
>=20
> it's a pity, but I can understand the motivation to shove everything =
into HTTP2.
>=20
> Anyway, some questions:
>=20
> 1) Frame Mapping
> """
> One WebSocket frame is mapped into one preceding SPDY HEADERS frame =
and following plural SPDY data frames. The SPDY HEADERS frame must =
contain WebSocket frame fields and following SPDY data frames contain =
payload data. These data frames are free to be reframed.
> """
> =
https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7ukw9aWz=
SHijk/edit
>=20
> So reframing of WebSocket frames (the payload) into multiple SPDY data =
frames MAY occur, but does not need to occur? And a WS/SPDY =
implementation is free to choose the boundaries within WS frame payload =
at which to reframe into different SPDY data frames as it likes?
>=20
> 2) End of WS Frame
>=20
> The end of a WS message is signaled with a SPDY data frame having =
MSG_DONE set.
>=20
> How is the end of a WS frame signaled, given that the WS frame may =
have been reframed into multiple SPDY data frames?
>=20
> 3) WS extension
>=20
> Are WS extensions fully supported? In particular, is =
"permessage-deflate" supported when running WS/SPDY?
>=20
> Thanks!
> /Tobias
>=20
>=20
> Von: hybi [mailto:hybi-bounces@ietf.org] Im Auftrag von Takeshi =
Yoshino
> Gesendet: Dienstag, 7. Januar 2014 08:36
> An: hybi@ietf.org
> Betreff: [hybi] Discontinuation of mux standardizaton in favor of =
WS/HTTP/2.0
>=20
>=20
> Hi all,
>=20
> We've been working on mux extension to make WebSocket scalable for =
long time, but as SPDY got widely supported and adopted quickly, now =
there seems to be no longer any good reason to continue this work which =
is WebSocket dedicated.
>=20
> Scalability is still important issue to address. =46rom now, I'd like =
to focus on WS/SPDY effort which we've been working on in parallel with =
the mux extension.
>=20
> Please take a look at the latest proposal in this post.
> https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc
>=20
> If you really want the dedicated mux extension than one over SPDY, =
sorry but please take over the work.
>=20
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>=20
> =20
>=20
> =20
>=20
> =20
>=20
> =20
>=20
>=20


--Apple-Mail=_A1C8F556-3822-4BCA-938D-030D010F3104
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
	charset=windows-1252

<html><head><meta http-equiv=3D"Content-Type" content=3D"text/html =
charset=3Dwindows-1252"><meta http-equiv=3D"Content-Type" =
content=3D"text/html charset=3Dwindows-1252"></head><body =
style=3D"word-wrap: break-word; -webkit-nbsp-mode: space; =
-webkit-line-break: after-white-space;">Hi all,<div><br></div><div>Sorry =
about the delay in clarifying my previous =
message.<br><div><br></div><div><blockquote type=3D"cite">Your proposal =
(SPDY/Messaging) is a brand-new full-duplex messaging protocol on SPDY =
and the only constraint is compatibility with the W3C WebSocket API, =
right?</blockquote></div><div><br></div><div>That statement describes my =
point in a nutshell.&nbsp;</div><div><br></div><div>My reasoning and =
more details/questions:</div><div><br></div><div>The goal [roughly] of =
WebSocket was to provide a protocol for establishing a persistent, =
ordered, bidirectional messaging channel with minimal framing overhead. =
This channel=92s most important user would be JavaScript applications in =
web browsers so browser related security concerns and the API there are =
significant. HTTP 1.1 does not have the functionality required to =
implement any semblance of these requirements. As such, RFC6455 defines =
a handshake, framing layer, error handling layer, and guidelines for =
extensions for a standalone TCP connection that fulfills these =
requirements.</div><div><br></div><div>HTTP 2.0 on the other hand =
already provides all of the underlying plumbing to provide a persistent, =
ordered, bidirectional messaging channel. =46rom a browser=92s =
prospective, whether the WebSocket channel that it opens goes through a =
separate TCP connection, a hybrid =93websocket frames encapsulated =
within HTTP 2.0 frames, or native HTTP 2.0 frames really doesn=92t =
matter as long as HTTP 2.0 is able to match the functionality of the =
WebSocket browser API.</div><div><br></div><div>A simple proposal to =
support WebSocket API functionality might be something =
like:</div><div><br></div><div>- HTTP 2.0 HEADERS message already =
provides the functionality necessary to negotiate a given HTTP 2.0 =
channel as a =93WebSocket=94 channel along with sending subprotocol, =
cookies, etc.</div><div>- An HTTP 2.0 WEBSOCKET frame type could be used =
to deliver WebSocket messages. You=92d need a FIN flag for the message =
boundaries, frame boundaries would be just standard HTTP 2.0 frames. No =
need to repackage or reframe or make a distinction between =93websocket=94=
 frames and HTTP 2.0 frames.</div><div>- Depending on how closely we =
want HTTP 2.0 WebSockets to match RFC6455 WebSockets additional support =
could be added.. for example, define the WEBSOCKET frame type=92s flag =
bits as follows (full compatibility flag =
configuration):</div><div><br></div><div><div><font face=3D"Courier =
New">&nbsp; &nbsp;0 &nbsp; &nbsp; 1 &nbsp; &nbsp; 2 &nbsp; &nbsp; 3 =
&nbsp; &nbsp; 4 &nbsp; &nbsp; 5 &nbsp; &nbsp; 6 &nbsp; &nbsp; =
7</font></div><div><font face=3D"Courier =
New">+-----+-----+-----+-----+-----+-----+-----+-----+</font></div><div><f=
ont face=3D"Courier New">| FIN | &nbsp; &nbsp; RFC6455 OPCODE &nbsp; =
&nbsp;|RSV1 |RSV2 |RSV3 |</font></div><div><font face=3D"Courier =
New">+-----+-----+-----+-----+-----+-----+-----+-----+</font></div></div><=
div><br></div><div>This provides the ability to keep RFC6455 opcodes and =
RSV bits without. Logical WebSocket stream ending would be handled =
exactly as in RFC6455 by exchanging close frames. This would potentially =
allow an easier migration for RFC6455 extensions by including the RSV =
bits.</div><div><br></div><div>A more simple frame structure could also =
be imagined without as much of the RFC6455 compatibility considering the =
following assumptions:&nbsp;</div><div>- Half of the RFC6455 opcodes are =
reserved for control messages that so far are entirely provided by HTTP =
2.0 natively (ping/pong, stream ending flags, etc).</div><div>- So far =
only two =93data=94 opcodes have been defined.</div><div>- RFC6455 =
extensions have proven to be fairly invasive to base implementations and =
require a lot of careful thought about how they will interact with each =
other and the rest of the protocol. Raw WebSocket and HTTP 2.0 may have =
different enough framing layers that extensions for one may not be =
directly translatable in the same way the framing isn=92t directly =
translatable. It may be sufficient to adapt the interfaces of popular =
WebSocket extensions over HTTP 2.0 mechanisms rather than literally try =
and implement WS extensions in HTTP 2.0. Take for example =
permessage-deflate extensions. HTTP 2.0 by default will have deflate =
support and will be maintaining compression contexts for header =
compression. Multiple different streams within the connection (either =
HTTP or other WebSockets) might also be interested in compression. Would =
a more generic compression system that could share state among these =
connections make sense here rather than individually stored and =
negotiated compression parameters per logical connection? It certainly =
would not make sense for raw WebSocket (sans MUX at =
least).</div><div><br></div><div>Example of a potentially simpler =
WEBSOCKET frame flags (limited compatibility flag =
configuration):</div><div><br></div><div><div><font face=3D"Courier =
New">&nbsp; &nbsp;0 &nbsp; &nbsp; 1 &nbsp; &nbsp; 2 &nbsp; &nbsp; 3 =
&nbsp; &nbsp; 4 &nbsp; &nbsp; 5 &nbsp; &nbsp; 6 &nbsp; &nbsp; =
7</font></div><div><font face=3D"Courier =
New">+-----+-----+-----+-----+-----+-----+-----+-----+</font></div><div><f=
ont face=3D"Courier New">| SFIN| MFIN|UTF8 | &nbsp; &nbsp; | &nbsp; =
&nbsp; |RSV1*|RSV2*|</font><span style=3D"font-family: 'Courier =
New';">RSV3*</span><span style=3D"font-family: 'Courier =
New';">|</span></div><div><font face=3D"Courier =
New">+-----+-----+-----+-----+-----+-----+-----+-----+</font></div></div><=
div><br></div><div>SFIN =3D stream fin</div><div>MFIN =3D message =
fin</div><div>UTF8 =3D UTF8 message, validate as such. Alternatively =
bits 2-4 could represent a =93data opcode=94 similar to the lower 3 bits =
of RFC6455=92s opcode (ie drop the control opcodes).</div><div>RSV1-3 =
could be included as mapped WebSocket extension bits or just used for =
HTTP 2.0 websocket message bits.</div><div><br></div><div>This =
configuration would also open the option of reusing the straight up DATA =
frame as that already has the SFIN bit, though the =93reserved=94 #2 bit =
might be a problem depending on how reserved it actually is. There is =
certainly enough information in the headers to allow an implementation =
to know whether a channel is a WebSocket one without a separate message =
type. I still think a separate message is a good idea because the types =
of flags for these types of messages will be different and the frame =
processing logic will be different and i=92ve found in my applications =
at least it is more straight forward to switch on frame type than to =
keep state about how to interpret a generic =93DATA=94 frame. HTTP 2.0 =
has a significantly larger range for frame types than RFC6455 did as =
well.</div><div><br><div><div>On Jan 20, 2014, at 6:07, Yutaka Hirano =
&lt;<a href=3D"mailto:yhirano@chromium.org">yhirano@chromium.org</a>&gt; =
wrote:</div><br class=3D"Apple-interchange-newline"><blockquote =
type=3D"cite"><div dir=3D"ltr">Thank you for the explanation, finally I =
may understand...<div>Your proposal (SPDY/Messaging) is a brand-new =
full-duplex messaging protocol on SPDY and the only constraint is =
compatibility with the W3C WebSocket API, right?<br>
</div><div><br></div></div><div class=3D"gmail_extra"><br><br><div =
class=3D"gmail_quote">On Mon, Jan 20, 2014 at 8:16 PM, Tobias Oberstein =
<span dir=3D"ltr">&lt;<a href=3D"mailto:tobias.oberstein@tavendo.de" =
target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 =
.8ex;border-left:1px #ccc solid;padding-left:1ex"><div lang=3D"DE" =
link=3D"blue" vlink=3D"purple"><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">Hi Yutaka,<u></u><u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d"><u></u>&nbsp;<u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">in fact, I completely agree with Peter (<a =
href=3D"http://www.ietf.org/mail-archive/web/hybi/current/msg10310.html" =
target=3D"_blank">http://www.ietf.org/mail-archive/web/hybi/current/msg103=
10.html</a>) - either have <u></u><u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d"><u></u>&nbsp;<u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">1) full WebSocket tunneled over SPDY =
("SPDY/WS-Tunnel")<u></u><u></u></span></p><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">=3D&gt; this is the proposal to transport full WS =
frames as payload of SPDY data frames describe in my other =
post<u></u><u></u></span></p><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d"><u></u>&nbsp;<u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">or<u></u><u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d"><u></u>&nbsp;<u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">2) define a SPDY messaging, independent of WS =
("SPDY/Messaging")<u></u><u></u></span></p><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">=3D&gt; This is the proposal below (</span>Fri, =
Jan 10, 2014 at 5:37 PM)<span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">. Or something in this =
direction.<u></u><u></u></span></p><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d"><u></u>&nbsp;<u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">The current SPDY/WS proposal is different from =
both 1) and 2) above, and again I agree with Peter, that this is the =
least desirable of all.<u></u><u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d"><u></u>&nbsp;<u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">My order of preference would =
be:<u></u><u></u></span></p><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d"><u></u>&nbsp;<u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">++1:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
SPDY/Messaging<u></u><u></u></span></p><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">+1:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
SPDY/WS-Tunnel<u></u><u></u></span></p><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">0:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&n=
bsp;&nbsp;&nbsp; SPDY/WS<u></u><u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d"><u></u>&nbsp;<u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">/Tobias<u></u><u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d"><u></u>&nbsp;<u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d"><u></u>&nbsp;<u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d"><u></u>&nbsp;<u></u></span></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d"><u></u>&nbsp;<u></u></span></p>
<div style=3D"border:none;border-left:solid blue 1.5pt;padding:0cm 0cm =
0cm 4.0pt"><div><div style=3D"border:none;border-top:solid #e1e1e1 =
1.0pt;padding:3.0pt 0cm 0cm 0cm"><p class=3D"MsoNormal"><b><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;">Von:</span></b><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;"> <a href=3D"mailto:yhirano@google.com" =
target=3D"_blank">yhirano@google.com</a> [mailto:<a =
href=3D"mailto:yhirano@google.com" =
target=3D"_blank">yhirano@google.com</a>] <b>Im Auftrag von </b>Yutaka =
Hirano<br>
<b>Gesendet:</b> Montag, 20. Januar 2014 08:04<br><b>An:</b> Tobias =
Oberstein<br><b>Cc:</b> Roberto Peon; Takeshi Yoshino; <a =
href=3D"mailto:hybi@ietf.org" =
target=3D"_blank">hybi@ietf.org</a></span></p><div><div class=3D"h5"><br>
<b>Betreff:</b> Re: [hybi] Discontinuation of mux standardizaton in =
favor of WS/HTTP/2.0<u></u><u></u></div></div><div><br =
class=3D"webkit-block-placeholder"></div></div></div><div><div =
class=3D"h5"><p class=3D"MsoNormal"><u></u>&nbsp;<u></u></p><div><div><p =
class=3D"MsoNormal">
Tobias, let me confirm.<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p></div><div><blockquote =
style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm =
6.0pt;margin-left:4.8pt;margin-right:0cm"><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">I'd like to suggest a different =
WS/SPDY:<br>&nbsp;<br></span>Pack WebSocket frames _including_ full WS =
frame header as _payload_ of SPDY data frames only - no SPDY header =
frames at all.<u></u><u></u></p>
</blockquote></div><div><p class=3D"MsoNormal">My understanding is that =
by suggesting the above one you retracted the below proposal you made =
before, is that right?<u></u><u></u></p></div><div><p =
class=3D"MsoNormal"><u></u>&nbsp;<u></u></p>
</div><div><p class=3D"MsoNormal">On Fri, Jan 10, 2014 at 5:37 PM, =
Tobias Oberstein&nbsp;&lt;<a href=3D"mailto:tobias.oberstein@tavendo.de" =
target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt;&nbsp;wrote:<u></u><u=
></u></p></div><blockquote style=3D"border:none;border-left:solid =
#cccccc 1.0pt;padding:0cm 0cm 0cm =
6.0pt;margin-left:4.8pt;margin-right:0cm"><p class=3D"MsoNormal"><span =
style=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&q=
uot;">Why not map WS messages directly into SPDY data frames like =
so:<br>Have 2 bits in SPDY data frames:<br>0x03 =3D MSG_TYPE : 0 =3D=3D =
text, 1 =3D=3D binary<br>
0x04 =3D MSG_DONE<br>A WS text message would simply be mapped into a =
series of SPDY data frames:<br>MSG_TYPE =3D 0, MSG_DONE =3D =
0<br>MSG_TYPE =3D 0, MSG_DONE =3D 0<br>...<br>MSG_TYPE =3D 0, MSG_DONE =3D=
 1<br>Similar for a WS binary message.<br>
&nbsp;</span>&nbsp;<u></u><u></u></p></blockquote><blockquote =
style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm =
6.0pt;margin-left:4.8pt;margin-right:0cm"><p class=3D"MsoNormal"><span =
style=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&q=
uot;">Doing like above will allow to preserve the current WS API in =
HTML5 (which needs messages, and needs 2 message =
types).</span><u></u><u></u></p>
</blockquote><div><p =
class=3D"MsoNormal"><u></u>&nbsp;<u></u></p></div><div><p =
class=3D"MsoNormal">Thanks,<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p></div></div><div><p =
class=3D"MsoNormal" style=3D"margin-bottom:12.0pt">
<u></u>&nbsp;<u></u></p><div><p class=3D"MsoNormal">On Wed, Jan 15, 2014 =
at 4:09 AM, Tobias Oberstein &lt;<a =
href=3D"mailto:tobias.oberstein@tavendo.de" =
target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt; =
wrote:<u></u><u></u></p><blockquote style=3D"border:none;border-left:solid=
 #cccccc 1.0pt;padding:0cm 0cm 0cm =
6.0pt;margin-left:4.8pt;margin-right:0cm">
<div><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">Hi Roberto,</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">not sure if you agree, but FWIW, here is a summary =
as I see it (currently):</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">1) WebSocket frame boundaries _can_ have semantic =
meaning in WebSocket _extensions_</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">2) WebSocket message and frame boundaries are =
indeed preserved in the current WS/SPDY =
proposal,</span><u></u><u></u></p><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">however, non-final WebSocket frame ends cannot be =
_immediately_ detected.</span><u></u><u></u></p><p class=3D"MsoNormal">
<span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">3) The current WS/SPDY =
proposal</span><u></u><u></u></p><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp; * is comfortable to implement for =
implementations that _only_ want to implement WS/SPDY, and not native WS =
at the same time</span><u></u><u></u></p><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;* makes things complicated for WS =
implementations that want to support both WS/SPDY _and_ native WS at the =
same time</span><u></u><u></u></p><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;* has significant wire-level overhead =
compared to native WS (and compared to the proposal =
below)</span><u></u><u></u></p><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">I'd like to suggest a different =
WS/SPDY:</span><u></u><u></u></p><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p>Pack WebSocket =
frames _including_ full WS frame header as _payload_ of SPDY data frames =
only - no SPDY header frames at =
all.<u></u><u></u></p><p>&nbsp;<u></u><u></u></p><p>In the absence of =
negotiated WebSocket extensions (or when all WS extensions in place do =
not assign semantics to WS frame =
boundaries),<u></u><u></u></p><p>fragment WebSocket payload at the WS =
level such that WS frame header+payload fits into the payload of SPDY =
data frames.<u></u><u></u></p><p>&nbsp;<u></u><u></u></p><p>In the =
latter case, this essentially adds a small SPDY data frame header (of =
fixed length) to WS frames. Easy to implement and efficient on =
wire.<u></u><u></u></p><p>&nbsp;<u></u><u></u></p><p>IMHO, this has =
advantages compared to current proposal and considering 3) =
above.<u></u><u></u></p><p>&nbsp;<u></u><u></u></p><p>The only =
disadvantage is for implementations that _only_ want to implement =
WS/SPDY and not native WS at the same =
time.<u></u><u></u></p><p>&nbsp;<u></u><u></u></p><p>=3D=3D<u></u><u></u><=
/p><p>&nbsp;<u></u><u></u></p><p>Regarding "permessage-priority" with =
ordering guarantees: should I find time and if there is interest, I'll =
write something up.<u></u><u></u></p><p>This could be a WebSocket =
extension. To be useful, the WS API in browsers would need to expose it =
though ..<u></u><u></u></p><p>&nbsp;<u></u><u></u></p><p>Just my =
2cts,<u></u><u></u></p><p>Tobias<u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><div =
style=3D"border:none;border-left:solid blue 1.5pt;padding:0cm 0cm 0cm =
4.0pt">
<div><div style=3D"border:none;border-top:solid #e1e1e1 =
1.0pt;padding:3.0pt 0cm 0cm 0cm"><p class=3D"MsoNormal"><b><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;">Von:</span></b><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;"> Roberto Peon [mailto:<a href=3D"mailto:fenix@google.com" =
target=3D"_blank">fenix@google.com</a>] <br>
<b>Gesendet:</b> Dienstag, 14. Januar 2014 19:30<br><b>An:</b> Yutaka =
Hirano<br><b>Cc:</b> Tobias Oberstein; Takeshi Yoshino; <a =
href=3D"mailto:hybi@ietf.org" =
target=3D"_blank">hybi@ietf.org</a></span><u></u><u></u></p><div><p =
class=3D"MsoNormal"><br><b>Betreff:</b> Re: [hybi] Discontinuation of =
mux standardizaton in favor of =
WS/HTTP/2.0<u></u><u></u></p></div></div></div><div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p><div><p class=3D"MsoNormal">
I've been lurking anyway :)<u></u><u></u></p><div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">As I understand the proposal, (and it seems that =
Takeshi and Yukata are covering this nicely, so what i say is probably =
redundant):<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">Message boundaries, having semantic =
meaning, are preserved.<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">Frame boundaries, having no semantic meaning, are =
not preserved. I don't think it makes sense to worry about this greatly =
when the ability to extend the protocol as implemented in HTTP2 is =
present. It doesn't make a lot of sense to signal the frame boundaries =
through everything unless there is a high-likelyhood and high-utility =
use-case for such. If such is found, then HTTP2 can be revised to =
include it. This is the approach we've been taking with most of the =
HTTP2 stuff, since there has been great effort placed on allowing new =
versions to be deployed in the future.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">&nbsp;<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">The mapping of WS onto HTTP2 is fairly direct-- we =
reserved the MSG_DONE bit specifically for this kind of use-case (i.e. =
bidirectional, full-duplex, message-based).<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">The observation made is that there are =
essentially 3 types of data that need to be preserved one way or =
another:<u></u><u></u></p></div><div><p class=3D"MsoNormal">1) metadata: =
data about the data, e.g. text, binary, where it is supposed to be =
going, etc.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">2) =
payload<u></u><u></u></p></div><div><p class=3D"MsoNormal">3) boundaries =
between these.<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">
imho, metadata is handled with HEADERS (I'd have preferred it to be =
called METADATA, but was voted down at the time :) ), payload is handled =
with DATA frames, and the boundaries are delimited either through the =
use of HEADERS frames, or via the MSG_DONE or STREAM_DONE =
bits.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">In the presence of extensions, =
assuming that MSG_DONE is required to signal the end of a message, then =
HEADERS can be used to signal any boundary =
desired.<u></u><u></u></p></div><div><p class=3D"MsoNormal">
&nbsp;<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">I totally agree that it would be nice to expose the =
priority of various messages/channels with respect to each other. That =
is really an issue that is outside the WG's purvue, =
though.<u></u><u></u></p>
</div><div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p></div></div><div><p =
class=3D"MsoNormal" =
style=3D"margin-bottom:12.0pt">&nbsp;<u></u><u></u></p><div><p =
class=3D"MsoNormal">On Thu, Jan 9, 2014 at 7:54 PM, Yutaka Hirano &lt;<a =
href=3D"mailto:yhirano@chromium.org" =
target=3D"_blank">yhirano@chromium.org</a>&gt; wrote:<u></u><u></u></p>
<blockquote style=3D"border:none;border-left:solid #cccccc =
1.0pt;padding:0cm 0cm 0cm =
6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom:5.=
0pt"><div><div><p class=3D"MsoNormal">+Roberto as an HTTP2 =
expert.<u></u><u></u></p>
</div><div><div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p></div><blockquote =
style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm =
6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom:5.=
0pt"><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">2) Ok, both WS frame and message boundaries are =
preserved.<br>&nbsp;<br>But: I cannot recognize the end of a WS frame =
until the next WS frame starts (for which I receive a HEADERS frame, =
which is the implicit signal that the previous WS frame ended), =
right?</span><u></u><u></u></p>
</blockquote></div><p class=3D"MsoNormal">You are right, you =
can't.<u></u><u></u></p><div><div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p></div><blockquote =
style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm =
6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom:5.=
0pt"><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;<br>And if so, why is there no FRAME_DONE =
bit on the SPDY frame that corresponds to the end of a WS =
frame?</span><u></u><u></u></p>
</blockquote></div><p class=3D"MsoNormal">Given that WebSocket frames =
"have no semantic meaning" as in RFC6455, I don't think the immediate =
recognition of the end of a WebSocket frame is =
requisite.<u></u><u></u></p>
<div><p class=3D"MsoNormal">I would like to introduce MSG_DONE in HTTP2 =
so that intermediaries that don't fully understand WS over HTTP2 can =
relay it.<u></u><u></u></p></div><div><p class=3D"MsoNormal">Since there =
are not so much space in FLAGS field, Introducing FRAME_DONE may =
decrease the extensiveness of the spec.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">Do you have a strong preference to =
introduce FRAME_DONE?<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p></div></div><div><p =
class=3D"MsoNormal" style=3D"margin-bottom:12.0pt">
&nbsp;<u></u><u></u></p><div><p class=3D"MsoNormal">On Thu, Jan 9, 2014 =
at 6:53 PM, Tobias Oberstein &lt;<a =
href=3D"mailto:tobias.oberstein@tavendo.de" =
target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt; =
wrote:<u></u><u></u></p><blockquote style=3D"border:none;border-left:solid=
 #cccccc 1.0pt;padding:0cm 0cm 0cm =
6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom:5.=
0pt">
<div><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">Hi Yutaka,</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">1) Yes, I understand. =
Thanks!</span><u></u><u></u></p><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">2) Ok, both WS frame and message boundaries are =
preserved.</span><u></u><u></u></p><p class=3D"MsoNormal">
<span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">But: I cannot recognize the end of a WS frame =
until the next WS frame starts (for which I receive a HEADERS frame, =
which is the implicit signal that the previous WS frame ended), =
right?</span><u></u><u></u></p><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">And if so, why is there no FRAME_DONE bit on the =
SPDY frame that corresponds to the end of a WS =
frame?</span><u></u><u></u></p><p class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">3) Ok, great! Hope this works, since SPDY provides =
header compression, but no sophisticated bidirectional payload =
compression as "WS permessage-deflate" does.</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">Cheers,</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">/Tobias</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p><p =
class=3D"MsoNormal"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;;color:#1f497d">&nbsp;</span><u></u><u></u></p>
<div style=3D"border:none;border-left:solid blue 1.5pt;padding:0cm 0cm =
0cm 4.0pt"><div><div style=3D"border:none;border-top:solid #e1e1e1 =
1.0pt;padding:3.0pt 0cm 0cm 0cm"><p class=3D"MsoNormal"><b><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;">Von:</span></b><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;"> </span><a href=3D"mailto:yhirano@google.com" =
target=3D"_blank"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;">yhirano@google.com</span></a><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;"> [mailto:</span><a href=3D"mailto:yhirano@google.com" =
target=3D"_blank"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;">yhirano@google.com</span></a><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;">] <b>Im Auftrag von </b>Yutaka Hirano<br>
<b>Gesendet:</b> Donnerstag, 9. Januar 2014 10:37<br><b>An:</b> Tobias =
Oberstein<br><b>Cc:</b> Takeshi Yoshino; </span><a =
href=3D"mailto:hybi@ietf.org" target=3D"_blank"><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;">hybi@ietf.org</span></a><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif=
&quot;"><br>
<b>Betreff:</b> Re: [hybi] Discontinuation of mux standardizaton in =
favor of WS/HTTP/2.0</span><u></u><u></u></p></div></div><div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p><div><p class=3D"MsoNormal">Hi=
 Tobias,<u></u><u></u></p>
<div><p class=3D"MsoNormal">Thank you for the =
comments.<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p><blockquote =
style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm =
6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom:5.=
0pt"><p class=3D"MsoNormal"><span =
style=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&q=
uot;">1) Frame Mapping<br>"""<br>One WebSocket frame is mapped into one =
preceding SPDY HEADERS frame and following plural SPDY data frames. The =
SPDY HEADERS frame must contain WebSocket frame fields and following =
SPDY data frames contain payload data. These data frames are free to be =
reframed.<br>
"""<br></span><a =
href=3D"https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc=
7ukw9aWzSHijk/edit" target=3D"_blank"><span =
style=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&q=
uot;">https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc7u=
kw9aWzSHijk/edit</span></a><span =
style=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&q=
uot;"><br>
So reframing of WebSocket frames (the payload) into multiple SPDY data =
frames MAY occur, but does not need to occur? And a WS/SPDY =
implementation is free to choose the boundaries within WS frame payload =
at which to reframe into different SPDY data frames as it =
likes?</span><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal">A WebSocket frame MAY be split =
to multiple SPDY DATA frames.<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">Intermediaries MAY merge SPDY DATA frames and MAY =
split a SPDY DATA frame into multiple DATA frames.<u></u><u></u></p>
</div><div><p class=3D"MsoNormal">Is this answering your =
question?<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p></div><blockquote =
style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm =
6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom:5.=
0pt"><p class=3D"MsoNormal"><span =
style=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&q=
uot;">2) End of WS Frame<br>The end of a WS message is signaled with a =
SPDY data frame having MSG_DONE set.<br>How is the end of a WS frame =
signaled, given that the WS frame may have been reframed into multiple =
SPDY data frames?</span><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal">When you receive a HEADERS =
frame, you can recognize that the last SPDY frame is that last frame in =
a WebSocket frame.<u></u><u></u></p></div><div><p class=3D"MsoNormal">When=
 you receive a SPDY frame with MSG_DONE, you can recognize that it is =
the last SPDY frame in a WebSocket message and hence the last SPDY frame =
in a WebSocket frame.<u></u><u></u></p>
</div><div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p></div><blockquote =
style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm =
6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom:5.=
0pt"><p class=3D"MsoNormal">
<span =
style=3D"font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&q=
uot;">3) WS extension<br>Are WS extensions fully supported? In =
particular, is "permessage-deflate" supported when running =
WS/SPDY?</span><u></u><u></u></p>
</blockquote><div><p class=3D"MsoNormal">Yes. We have reserved bits and =
we can emulate the opening handshake, I =
think.<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p></div><div><p =
class=3D"MsoNormal">On Thu, Jan 9, 2014 at 6:11 PM, Tobias Oberstein =
&lt;<a href=3D"mailto:tobias.oberstein@tavendo.de" =
target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt; =
wrote:<u></u><u></u></p>
<blockquote style=3D"border:none;border-left:solid #cccccc =
1.0pt;padding:0cm 0cm 0cm =
6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom:5.=
0pt"><p class=3D"MsoNormal">Hi Takeshi,<br><br>it's a pity, but I can =
understand the motivation to shove everything into HTTP2.<br>
<br>Anyway, some questions:<br><br>1) Frame Mapping<br>"""<br>One =
WebSocket frame is mapped into one preceding SPDY HEADERS frame and =
following plural SPDY data frames. The SPDY HEADERS frame must contain =
WebSocket frame fields and following SPDY data frames contain payload =
data. These data frames are free to be reframed.<br>
"""<br><a =
href=3D"https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXaUW0VICmCfc=
7ukw9aWzSHijk/edit" =
target=3D"_blank">https://docs.google.com/document/d/124vB5NV1p2N55wg-MdXa=
UW0VICmCfc7ukw9aWzSHijk/edit</a><br>
<br>So reframing of WebSocket frames (the payload) into multiple SPDY =
data frames MAY occur, but does not need to occur? And a WS/SPDY =
implementation is free to choose the boundaries within WS frame payload =
at which to reframe into different SPDY data frames as it likes?<br>
<br>2) End of WS Frame<br><br>The end of a WS message is signaled with a =
SPDY data frame having MSG_DONE set.<br><br>How is the end of a WS frame =
signaled, given that the WS frame may have been reframed into multiple =
SPDY data frames?<br>
<br>3) WS extension<br><br>Are WS extensions fully supported? In =
particular, is "permessage-deflate" supported when running =
WS/SPDY?<br><br>Thanks!<br>/Tobias<br><br><br>Von: hybi [mailto:<a =
href=3D"mailto:hybi-bounces@ietf.org" =
target=3D"_blank">hybi-bounces@ietf.org</a>] Im Auftrag von Takeshi =
Yoshino<br>
Gesendet: Dienstag, 7. Januar 2014 08:36<br>An: <a =
href=3D"mailto:hybi@ietf.org" =
target=3D"_blank">hybi@ietf.org</a><br>Betreff: [hybi] Discontinuation =
of mux standardizaton in favor of WS/HTTP/2.0<u></u><u></u></p><div><p =
class=3D"MsoNormal"><br>Hi all,<br><br>We've been working on mux =
extension to make WebSocket scalable for long time, but as SPDY got =
widely supported and adopted quickly, now there seems to be no longer =
any good reason to continue this work which is WebSocket dedicated.<br>
<br>Scalability is still important issue to address. =46rom now, I'd =
like to focus on WS/SPDY effort which we've been working on in parallel =
with the mux extension.<br><br>Please take a look at the latest proposal =
in this post.<br>
<a href=3D"https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax3glc" =
target=3D"_blank">https://groups.google.com/forum/#!topic/spdy-dev/aJut2ax=
3glc</a><br><br>If you really want the dedicated mux extension than one =
over SPDY, sorry but please take over the work.<u></u><u></u></p>
</div><div><p =
class=3D"MsoNormal">_______________________________________________<br>hyb=
i mailing list<br><a href=3D"mailto:hybi@ietf.org" =
target=3D"_blank">hybi@ietf.org</a><br><a =
href=3D"https://www.ietf.org/mailman/listinfo/hybi" =
target=3D"_blank">https://www.ietf.org/mailman/listinfo/hybi</a><u></u><u>=
</u></p>
</div></blockquote></div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p></div></div></div></div></div>=
</blockquote></div><p =
class=3D"MsoNormal">&nbsp;<u></u><u></u></p></div></blockquote></div><p =
class=3D"MsoNormal">
&nbsp;<u></u><u></u></p></div></div></div></div></blockquote></div><p =
class=3D"MsoNormal"><u></u>&nbsp;<u></u></p></div></div></div></div></div>=
</blockquote></div><br></div>
</blockquote></div><br></div></div></body></html>=

--Apple-Mail=_A1C8F556-3822-4BCA-938D-030D010F3104--

From webmaster@zaphoyd.com  Mon Jan 20 07:38:14 2014
Return-Path: <webmaster@zaphoyd.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 5CEAA1A017A for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 07:38:14 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id mD1Ys6c7CXmq for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 07:38:12 -0800 (PST)
Received: from authsmtp00.uchicago.edu (authsmtp00.uchicago.edu [128.135.12.120]) by ietfa.amsl.com (Postfix) with ESMTP id AC6601A0176 for <hybi@ietf.org>; Mon, 20 Jan 2014 07:38:12 -0800 (PST)
Received: from [10.0.1.3] (c-68-51-76-178.hsd1.il.comcast.net [68.51.76.178]) (Authenticated sender: zaphoyd) by authsmtp00.uchicago.edu (Postfix) with ESMTP id 018BE4980A5; Mon, 20 Jan 2014 09:38:11 -0600 (CST)
Content-Type: text/plain; charset=windows-1252
Mime-Version: 1.0 (Mac OS X Mail 7.0 \(1822\))
From: Peter Thorson <webmaster@zaphoyd.com>
In-Reply-To: <634914A010D0B943A035D226786325D4446BF9997A@EXVMBX020-12.exch020.serverdata.net>
Date: Mon, 20 Jan 2014 09:41:19 -0600
Content-Transfer-Encoding: quoted-printable
Message-Id: <9B286350-A5BB-4A3E-9E5C-61F9FAABEBBD@zaphoyd.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF9997A@EXVMBX020-12.exch020.serverdata.net>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
X-Mailer: Apple Mail (2.1822)
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 15:38:14 -0000

In such a setup I would actually prefer that the browser be even more =
transparent about whether it is using raw WebSocket or HTTP 2.0 =
=93WebSocket=94

A WebSocket opened (via ws:// scheme) to the same host/port as the HTTP =
page itself (or perhaps to any other host/port that already has an HTTP =
2.0 connection active) should automatically go over the HTTP 2.0 =
WebSocket interface if possible and fall back to a new raw WebSocket =
connection if not. If an application wants to guarantee the use of raw =
websocket it should use a separate DNS name or port that is only running =
RFC6455.

On Jan 20, 2014, at 6:17, Tobias Oberstein <tobias.oberstein@tavendo.de> =
wrote:

>> Thank you for the explanation, finally I may understand...
>> Your proposal (SPDY/Messaging) is a brand-new full-duplex messaging =
protocol on SPDY and the only constraint is compatibility with the W3C =
WebSocket API, right?
>=20
> Yes, exactly.=20
>=20
> The only constraint and goal is compatibility with the W3C WebSocket =
API.
>=20
> E.g.
>=20
> // open WebSocket connection
> socket =3D new window.WebSocket("ws://myserver.com");
>=20
> // open SPDY/Messaging connection
> socket =3D new window.WebSocket("http://myserver.com");
>=20
> =3D> "socket" should have identical API in both.
>=20
> A browser might also automatically fall back to probing the other =
protocol, if the protocol
> requested by the app isn't available:
>=20
> // first try WS. If that fails, try SPDY/Messaging automatically.
> socket =3D new window.WebSocket("ws://myserver.com");
>=20
> // first try SPDY/Messaging. If that fails, try WS automatically.
> socket =3D new window.WebSocket("http://myserver.com");
>=20
> /Tobias


From tobias.oberstein@tavendo.de  Mon Jan 20 07:51:41 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 36C181A01B1 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 07:51:41 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id VcAEBQj0C_5i for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 07:51:37 -0800 (PST)
Received: from EXHUB020-3.exch020.serverdata.net (exhub020-3.exch020.serverdata.net [206.225.164.30]) by ietfa.amsl.com (Postfix) with ESMTP id CFAC01A01A9 for <hybi@ietf.org>; Mon, 20 Jan 2014 07:51:36 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-3.exch020.serverdata.net ([206.225.164.30]) with mapi; Mon, 20 Jan 2014 07:51:37 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Peter Thorson <webmaster@zaphoyd.com>, Yutaka Hirano <yhirano@chromium.org>
Date: Mon, 20 Jan 2014 07:51:34 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8V9K5jLsznY3LUQQOf/D4G61sLMgAAY7HQ
Message-ID: <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com>
In-Reply-To: <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 15:51:41 -0000

Hi Peter

>A simple proposal to support WebSocket API functionality might be somethin=
g like:
>- HTTP 2.0 HEADERS message already provides the functionality necessary to=
 negotiate a given HTTP 2.0 channel as a "WebSocket" channel along with sen=
ding subprotocol, cookies, etc.

Yep.

>- An HTTP 2.0 WEBSOCKET frame type could be used to deliver WebSocket mess=
ages. You'd need a FIN flag for the message boundaries, frame boundaries wo=
uld be just standard HTTP 2.0 frames. No need to repackage or reframe or ma=
ke a distinction between >"websocket" frames and HTTP 2.0 frames.

Why a separate SPDY frame type? Why not SPDY/DATA frames? After negotiation=
, the whole SPDY _stream_ will be "messaging" only anyway ..
=20
>- Depending on how closely we want HTTP 2.0 WebSockets to match RFC6455 We=
bSockets additional support could be added.. for example, define the WEBSOC=
KET frame type's flag bits as follows (full compatibility flag configuratio=
n):

>From the goal of W3C WebSocket API only compatibility, the only other infor=
mation (besides message boundaries) needed in SPDY/DATA frames would be the=
 payload type: UTF8 vs binary.

So the minimum appears to be 2 bits in SPDY/DATA frame headers:

1 bit for MSG_DONE on the last SPDY/DATA frame and
1 bit for MSG_TYPE (UTF8/binary) on the first SPDY/DATA frame that starts a=
 new msg

Am I missing something?

/Tobias

From webmaster@zaphoyd.com  Mon Jan 20 07:55:25 2014
Return-Path: <webmaster@zaphoyd.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 094721A01A9 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 07:55:25 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.899
X-Spam-Level: 
X-Spam-Status: No, score=-1.899 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 2c0iNOpzn-if for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 07:55:23 -0800 (PST)
Received: from authsmtp00.uchicago.edu (authsmtp00.uchicago.edu [128.135.12.120]) by ietfa.amsl.com (Postfix) with ESMTP id E590B1A0191 for <hybi@ietf.org>; Mon, 20 Jan 2014 07:55:22 -0800 (PST)
Received: from [10.0.1.3] (c-68-51-76-178.hsd1.il.comcast.net [68.51.76.178]) (Authenticated sender: zaphoyd) by authsmtp00.uchicago.edu (Postfix) with ESMTP id 6B36C49806D; Mon, 20 Jan 2014 09:55:22 -0600 (CST)
Content-Type: multipart/alternative; boundary="Apple-Mail=_5F76D0B9-EEFF-4342-8D4F-D9447743D12F"
Mime-Version: 1.0 (Mac OS X Mail 7.0 \(1822\))
From: Peter Thorson <webmaster@zaphoyd.com>
In-Reply-To: <634914A010D0B943A035D226786325D4446BF99977@EXVMBX020-12.exch020.serverdata.net>
Date: Mon, 20 Jan 2014 09:58:30 -0600
Message-Id: <B8ED6F18-B710-44AE-829B-EDE5859B2C5B@zaphoyd.com>
References: <634914A010D0B943A035D226786325D4446BF9948F@EXVMBX020-12.exch020.serverdata.net> <52D90F99.6080205@gmx.de> <634914A010D0B943A035D226786325D4446BF9949D@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZQ3NxVW36PnZ0TpMF4tPPaJLr12M8NtbPb6pUejf1wcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99977@EXVMBX020-12.exch020.serverdata.net>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
X-Mailer: Apple Mail (2.1822)
Cc: Julian Reschke <julian.reschke@gmx.de>, "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] WebSocket subprotocol parameters
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 15:55:25 -0000

--Apple-Mail=_5F76D0B9-EEFF-4342-8D4F-D9447743D12F
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=windows-1252

My impression was that subprotocol primarily changed the interpretation =
of messages payloads by the end applications themselves. WebSocket =
messages allow essentially arbitrary data so you can define whatever =
parameter setting mechanism you like with whatever grammar makes sense =
for your subprotocol and send the parameters in the opening =93hello=94 =
message. Defining a grammar in the socket spec to allow application =
protocols to configure themselves seems limiting and out of scope.

Extensions on the other hand needed a way to communicate with themselves =
as defined and thus needed an explicitly provided grammar.

On Jan 20, 2014, at 5:26, Tobias Oberstein <tobias.oberstein@tavendo.de> =
wrote:

> Hi Takeshi,
> =20
> thanks alot for digging out these posts!
> =20
> In retrospect, I think it's a pity that we did not specify =
"qvalue-like" grammer for Sec-WebSocket-Protocol as we did for =
Sec-WebSocket-Extension.
> =20
> The latter provides a lot of useful flexibility =85 eg as used in =
"permessage-deflate".
> =20
> Hacks like having subprotocols like this
> =20
> wamp.2.json
> wamp.2.msgpack
> wamp.2.json.batched
> wamp.2.msgpack.batched
> ..
> =20
> negotiated only works for "discrete value parameters" (json/msgpack, =
batched/unbatched), but breaks down quite quickly.
> =20
> As Julian pointed out, making Sec-WebSocket-Protocol using the same =
"qvalue-grammer" as Sec-WebSocket-Extension
> would be a significant change to the RFC6455, and break existing code.
> =20
> It would also possibly require changes to the WS API in JavaScript to =
be useful.
> =20
> Mmh. Too bad;(
> =20
> /Tobias
> =20
> =20
> =20
> =20
> Von: Takeshi Yoshino [mailto:tyoshino@google.com]=20
> Gesendet: Montag, 20. Januar 2014 09:36
> An: Tobias Oberstein
> Cc: Julian Reschke; hybi@ietf.org
> Betreff: Re: [hybi] WebSocket subprotocol parameters
> =20
> On Fri, Jan 17, 2014 at 10:01 PM, Tobias Oberstein =
<tobias.oberstein@tavendo.de> wrote:
> > > Was there any good reason disallowing parameters for =
"Sec-WebSocket-
> > Protocol"?
> >
> > I think that was discussed and the decision was made on purpose. For =
details
> > you'll have to visit the mailing list archives though.
>=20
> I can't find it .. searching the Hybi archives is a pain ..
> =20
> I can't recall such discussion.
> =20
> In this thread, Brodie used an example similar to that, but the main =
topic of the post was to update the spec to include multiple subprotocol =
offer mechanism.
> http://www.ietf.org/mail-archive/web/hybi/current/msg04467.html
> =20
> In this long thread, what character should be used for subprotocol is =
discussed. Some people preferred more liberal rule, but no one discussed =
qvalue-like grammar.
> http://www.ietf.org/mail-archive/web/hybi/current/msg00913.html
> =20
> Subprotocol History FYI:
> =20
> We added the constraint "U+0021 ..  U+007E" from
> http://tools.ietf.org/html/draft-hixie-thewebsocketprotocol-31 -> 32
> =20
> Single value -> multiple value (SP-separated)
> http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-02 -> =
03
> =20
> We made it follow 1#(token | quoted-string) ABNF from
> http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-03 -> =
04
> =20
> Changed from (token | quoted-string) to token
> http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-07=20
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi


--Apple-Mail=_5F76D0B9-EEFF-4342-8D4F-D9447743D12F
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
	charset=windows-1252

<html><head><meta http-equiv=3D"Content-Type" content=3D"text/html =
charset=3Dwindows-1252"></head><body style=3D"word-wrap: break-word; =
-webkit-nbsp-mode: space; -webkit-line-break: =
after-white-space;"><div>My impression was that subprotocol primarily =
changed the interpretation of messages payloads by the end applications =
themselves. WebSocket messages allow essentially arbitrary data so you =
can define whatever parameter setting mechanism you like with whatever =
grammar makes sense for your subprotocol and send the parameters in the =
opening =93hello=94 message. Defining a grammar in the socket spec to =
allow application protocols to configure themselves seems limiting and =
out of scope.</div><div><br></div><div>Extensions on the other hand =
needed a way to communicate with themselves as defined and thus needed =
an explicitly provided grammar.</div><br><div><div>On Jan 20, 2014, at =
5:26, Tobias Oberstein &lt;<a =
href=3D"mailto:tobias.oberstein@tavendo.de">tobias.oberstein@tavendo.de</a=
>&gt; wrote:</div><br class=3D"Apple-interchange-newline"><blockquote =
type=3D"cite"><div lang=3D"DE" link=3D"blue" vlink=3D"purple" =
style=3D"font-family: Helvetica; font-size: 12px; font-style: normal; =
font-variant: normal; font-weight: normal; letter-spacing: normal; =
line-height: normal; orphans: auto; text-align: start; text-indent: 0px; =
text-transform: none; white-space: normal; widows: auto; word-spacing: =
0px; -webkit-text-stroke-width: 0px;"><div class=3D"WordSection1" =
style=3D"page: WordSection1;"><div style=3D"margin: 0cm 0cm 0.0001pt; =
font-size: 12pt; font-family: 'Times New Roman', serif;"><span =
style=3D"font-size: 11pt; font-family: Calibri, sans-serif; color: =
rgb(31, 73, 125);">Hi Takeshi,<o:p></o:p></span></div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, 125);">&nbsp;</span></div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, 125);">thanks alot for digging =
out these posts!<o:p></o:p></span></div><div style=3D"margin: 0cm 0cm =
0.0001pt; font-size: 12pt; font-family: 'Times New Roman', serif;"><span =
style=3D"font-size: 11pt; font-family: Calibri, sans-serif; color: =
rgb(31, 73, 125);">&nbsp;</span></div><div style=3D"margin: 0cm 0cm =
0.0001pt; font-size: 12pt; font-family: 'Times New Roman', serif;"><span =
style=3D"font-size: 11pt; font-family: Calibri, sans-serif; color: =
rgb(31, 73, 125);">In retrospect, I think it's a pity that we did not =
specify "qvalue-like" grammer for Sec-WebSocket-Protocol as we did for =
Sec-WebSocket-Extension.<o:p></o:p></span></div><div style=3D"margin: =
0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times New Roman', =
serif;"><span style=3D"font-size: 11pt; font-family: Calibri, =
sans-serif; color: rgb(31, 73, 125);">&nbsp;</span></div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, 125);">The latter provides a lot =
of useful flexibility =85 eg as used in =
"permessage-deflate".<o:p></o:p></span></div><div style=3D"margin: 0cm =
0cm 0.0001pt; font-size: 12pt; font-family: 'Times New Roman', =
serif;"><span style=3D"font-size: 11pt; font-family: Calibri, =
sans-serif; color: rgb(31, 73, 125);">&nbsp;</span></div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, 125);">Hacks like having =
subprotocols like this<o:p></o:p></span></div><div style=3D"margin: 0cm =
0cm 0.0001pt; font-size: 12pt; font-family: 'Times New Roman', =
serif;"><span style=3D"font-size: 11pt; font-family: Calibri, =
sans-serif; color: rgb(31, 73, 125);">&nbsp;</span></div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, =
125);">wamp.2.json<o:p></o:p></span></div><div style=3D"margin: 0cm 0cm =
0.0001pt; font-size: 12pt; font-family: 'Times New Roman', serif;"><span =
style=3D"font-size: 11pt; font-family: Calibri, sans-serif; color: =
rgb(31, 73, 125);">wamp.2.msgpack<o:p></o:p></span></div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, =
125);">wamp.2.json.batched<o:p></o:p></span></div><div style=3D"margin: =
0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times New Roman', =
serif;"><span style=3D"font-size: 11pt; font-family: Calibri, =
sans-serif; color: rgb(31, 73, =
125);">wamp.2.msgpack.batched<o:p></o:p></span></div><div style=3D"margin:=
 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times New Roman', =
serif;"><span style=3D"font-size: 11pt; font-family: Calibri, =
sans-serif; color: rgb(31, 73, 125);">..<o:p></o:p></span></div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, 125);">&nbsp;</span></div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, 125);">negotiated only works for =
"discrete value parameters" (json/msgpack, batched/unbatched), but =
breaks down quite quickly.<o:p></o:p></span></div><div style=3D"margin: =
0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times New Roman', =
serif;"><span style=3D"font-size: 11pt; font-family: Calibri, =
sans-serif; color: rgb(31, 73, 125);">&nbsp;</span></div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, 125);">As Julian pointed out, =
making Sec-WebSocket-Protocol using the same "qvalue-grammer" as =
Sec-WebSocket-Extension<o:p></o:p></span></div><div style=3D"margin: 0cm =
0cm 0.0001pt; font-size: 12pt; font-family: 'Times New Roman', =
serif;"><span style=3D"font-size: 11pt; font-family: Calibri, =
sans-serif; color: rgb(31, 73, 125);">would be a significant change to =
the RFC6455, and break existing code.<o:p></o:p></span></div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, 125);">&nbsp;</span></div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, 125);">It would also possibly =
require changes to the WS API in JavaScript to be =
useful.<o:p></o:p></span></div><div style=3D"margin: 0cm 0cm 0.0001pt; =
font-size: 12pt; font-family: 'Times New Roman', serif;"><span =
style=3D"font-size: 11pt; font-family: Calibri, sans-serif; color: =
rgb(31, 73, 125);">&nbsp;</span></div><div style=3D"margin: 0cm 0cm =
0.0001pt; font-size: 12pt; font-family: 'Times New Roman', serif;"><span =
style=3D"font-size: 11pt; font-family: Calibri, sans-serif; color: =
rgb(31, 73, 125);">Mmh. Too bad;(<o:p></o:p></span></div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, 125);">&nbsp;</span></div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, =
125);">/Tobias<o:p></o:p></span></div><div style=3D"margin: 0cm 0cm =
0.0001pt; font-size: 12pt; font-family: 'Times New Roman', serif;"><span =
style=3D"font-size: 11pt; font-family: Calibri, sans-serif; color: =
rgb(31, 73, 125);">&nbsp;</span></div><div style=3D"margin: 0cm 0cm =
0.0001pt; font-size: 12pt; font-family: 'Times New Roman', serif;"><span =
style=3D"font-size: 11pt; font-family: Calibri, sans-serif; color: =
rgb(31, 73, 125);">&nbsp;</span></div><div style=3D"margin: 0cm 0cm =
0.0001pt; font-size: 12pt; font-family: 'Times New Roman', serif;"><span =
style=3D"font-size: 11pt; font-family: Calibri, sans-serif; color: =
rgb(31, 73, 125);">&nbsp;</span></div><div style=3D"margin: 0cm 0cm =
0.0001pt; font-size: 12pt; font-family: 'Times New Roman', serif;"><span =
style=3D"font-size: 11pt; font-family: Calibri, sans-serif; color: =
rgb(31, 73, 125);">&nbsp;</span></div><div style=3D"border-style: none =
none none solid; border-left-color: blue; border-left-width: 1.5pt; =
padding: 0cm 0cm 0cm 4pt;"><div><div style=3D"border-style: solid none =
none; border-top-color: rgb(225, 225, 225); border-top-width: 1pt; =
padding: 3pt 0cm 0cm;"><div style=3D"margin: 0cm 0cm 0.0001pt; =
font-size: 12pt; font-family: 'Times New Roman', serif;"><b><span =
style=3D"font-size: 11pt; font-family: Calibri, =
sans-serif;">Von:</span></b><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif;"><span =
class=3D"Apple-converted-space">&nbsp;</span>Takeshi Yoshino [<a =
href=3D"mailto:tyoshino@google.com">mailto:tyoshino@google.com</a>]<span =
class=3D"Apple-converted-space">&nbsp;</span><br><b>Gesendet:</b><span =
class=3D"Apple-converted-space">&nbsp;</span>Montag, 20. Januar 2014 =
09:36<br><b>An:</b><span =
class=3D"Apple-converted-space">&nbsp;</span>Tobias =
Oberstein<br><b>Cc:</b><span =
class=3D"Apple-converted-space">&nbsp;</span>Julian Reschke; <a =
href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br><b>Betreff:</b><span =
class=3D"Apple-converted-space">&nbsp;</span>Re: [hybi] WebSocket =
subprotocol parameters<o:p></o:p></span></div></div></div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><o:p>&nbsp;</o:p></div><div><div style=3D"margin: =
0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times New Roman', =
serif;">On Fri, Jan 17, 2014 at 10:01 PM, Tobias Oberstein &lt;<a =
href=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank" =
style=3D"color: purple; text-decoration: =
underline;">tobias.oberstein@tavendo.de</a>&gt; =
wrote:<o:p></o:p></div><blockquote style=3D"border-style: none none none =
solid; border-left-color: rgb(204, 204, 204); border-left-width: 1pt; =
padding: 0cm 0cm 0cm 6pt; margin-left: 4.8pt; margin-right: =
0cm;"><div><p class=3D"MsoNormal" style=3D"margin: 0cm 0cm 12pt; =
font-size: 12pt; font-family: 'Times New Roman', serif;">&gt; &gt; Was =
there any good reason disallowing parameters for "Sec-WebSocket-<br>&gt; =
Protocol"?<br>&gt;<br>&gt; I think that was discussed and the decision =
was made on purpose. For details<br>&gt; you'll have to visit the =
mailing list archives though.<o:p></o:p></p></div><div style=3D"margin: =
0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times New Roman', =
serif;">I can't find it .. searching the Hybi archives is a pain =
..<o:p></o:p></div></blockquote><div><div style=3D"margin: 0cm 0cm =
0.0001pt; font-size: 12pt; font-family: 'Times New Roman', =
serif;"><o:p>&nbsp;</o:p></div></div><div><div style=3D"margin: 0cm 0cm =
0.0001pt; font-size: 12pt; font-family: 'Times New Roman', serif;">I =
can't recall such discussion.<o:p></o:p></div></div><div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><o:p>&nbsp;</o:p></div></div><div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;">In this thread, Brodie used an example similar to =
that, but the main topic of the post was to update the spec to include =
multiple subprotocol offer mechanism.<o:p></o:p></div></div><div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><a =
href=3D"http://www.ietf.org/mail-archive/web/hybi/current/msg04467.html" =
style=3D"color: purple; text-decoration: =
underline;">http://www.ietf.org/mail-archive/web/hybi/current/msg04467.htm=
l</a><o:p></o:p></div></div><div><div style=3D"margin: 0cm 0cm 0.0001pt; =
font-size: 12pt; font-family: 'Times New Roman', =
serif;"><o:p>&nbsp;</o:p></div></div><div><div style=3D"margin: 0cm 0cm =
0.0001pt; font-size: 12pt; font-family: 'Times New Roman', serif;">In =
this long thread, what character should be used for subprotocol is =
discussed. Some people preferred more liberal rule, but no one discussed =
qvalue-like grammar.<o:p></o:p></div></div><div><div style=3D"margin: =
0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times New Roman', =
serif;"><a =
href=3D"http://www.ietf.org/mail-archive/web/hybi/current/msg00913.html" =
style=3D"color: purple; text-decoration: =
underline;">http://www.ietf.org/mail-archive/web/hybi/current/msg00913.htm=
l</a><o:p></o:p></div></div><div><div style=3D"margin: 0cm 0cm 0.0001pt; =
font-size: 12pt; font-family: 'Times New Roman', =
serif;"><o:p>&nbsp;</o:p></div></div><div><div style=3D"margin: 0cm 0cm =
0.0001pt; font-size: 12pt; font-family: 'Times New Roman', =
serif;">Subprotocol History FYI:<o:p></o:p></div></div><div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;"><o:p>&nbsp;</o:p></div></div><div><div =
style=3D"margin: 0cm 0cm 0.0001pt; font-size: 12pt; font-family: 'Times =
New Roman', serif;">We added the constraint "U+0021 .. &nbsp;U+007E" =
from<o:p></o:p></div></div><div style=3D"margin: 0cm 0cm 0.0001pt; =
font-size: 12pt; font-family: 'Times New Roman', serif;"><a =
href=3D"http://tools.ietf.org/html/draft-hixie-thewebsocketprotocol-31" =
style=3D"color: purple; text-decoration: =
underline;">http://tools.ietf.org/html/draft-hixie-thewebsocketprotocol-31=
</a><span class=3D"Apple-converted-space">&nbsp;</span>-&gt; =
32<o:p></o:p></div><div><div style=3D"margin: 0cm 0cm 0.0001pt; =
font-size: 12pt; font-family: 'Times New Roman', =
serif;"><o:p>&nbsp;</o:p></div></div><div><div style=3D"margin: 0cm 0cm =
0.0001pt; font-size: 12pt; font-family: 'Times New Roman', =
serif;">Single value -&gt; multiple value =
(SP-separated)<o:p></o:p></div></div><div><div style=3D"margin: 0cm 0cm =
0.0001pt; font-size: 12pt; font-family: 'Times New Roman', serif;"><a =
href=3D"http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-02=
" style=3D"color: purple; text-decoration: =
underline;">http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotoco=
l-02</a><span class=3D"Apple-converted-space">&nbsp;</span>-&gt; =
03<o:p></o:p></div></div><div><div style=3D"margin: 0cm 0cm 0.0001pt; =
font-size: 12pt; font-family: 'Times New Roman', =
serif;"><o:p>&nbsp;</o:p></div></div><div><div style=3D"margin: 0cm 0cm =
0.0001pt; font-size: 12pt; font-family: 'Times New Roman', serif;">We =
made it follow 1#(token | quoted-string) ABNF =
from<o:p></o:p></div></div><div><div style=3D"margin: 0cm 0cm 0.0001pt; =
font-size: 12pt; font-family: 'Times New Roman', serif;"><a =
href=3D"http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-03=
" style=3D"color: purple; text-decoration: =
underline;">http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotoco=
l-03</a><span class=3D"Apple-converted-space">&nbsp;</span>-&gt; =
04<o:p></o:p></div></div><div><div style=3D"margin: 0cm 0cm 0.0001pt; =
font-size: 12pt; font-family: 'Times New Roman', =
serif;"><o:p>&nbsp;</o:p></div></div><div><div style=3D"margin: 0cm 0cm =
0.0001pt; font-size: 12pt; font-family: 'Times New Roman', =
serif;">Changed from (token | quoted-string) to =
token<o:p></o:p></div></div><div><div style=3D"margin: 0cm 0cm 0.0001pt; =
font-size: 12pt; font-family: 'Times New Roman', serif;"><a =
href=3D"http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-07=
" style=3D"color: purple; text-decoration: =
underline;">http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotoco=
l-07</a>&nbsp;<o:p></o:p></div></div></div></div></div>___________________=
____________________________<br>hybi mailing list<br><a =
href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br>https://www.ietf.org/ma=
ilman/listinfo/hybi</div></blockquote></div><br></body></html>=

--Apple-Mail=_5F76D0B9-EEFF-4342-8D4F-D9447743D12F--

From tobias.oberstein@tavendo.de  Mon Jan 20 08:13:09 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id EE6501A01BD for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 08:13:08 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.899
X-Spam-Level: 
X-Spam-Status: No, score=-1.899 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id vZJCkQGqldTN for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 08:13:05 -0800 (PST)
Received: from EXHUB020-4.exch020.serverdata.net (exhub020-4.exch020.serverdata.net [206.225.164.31]) by ietfa.amsl.com (Postfix) with ESMTP id BEFBE1A0162 for <hybi@ietf.org>; Mon, 20 Jan 2014 08:13:04 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-4.exch020.serverdata.net ([206.225.164.31]) with mapi; Mon, 20 Jan 2014 08:13:04 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Peter Thorson <webmaster@zaphoyd.com>
Date: Mon, 20 Jan 2014 08:13:02 -0800
Thread-Topic: [hybi] WebSocket subprotocol parameters
Thread-Index: Ac8V+AgPnCXQDYtNRqaOl8GOCCVVCAAAe1sg
Message-ID: <634914A010D0B943A035D226786325D4446BF99ACB@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF9948F@EXVMBX020-12.exch020.serverdata.net> <52D90F99.6080205@gmx.de> <634914A010D0B943A035D226786325D4446BF9949D@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZQ3NxVW36PnZ0TpMF4tPPaJLr12M8NtbPb6pUejf1wcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99977@EXVMBX020-12.exch020.serverdata.net> <B8ED6F18-B710-44AE-829B-EDE5859B2C5B@zaphoyd.com>
In-Reply-To: <B8ED6F18-B710-44AE-829B-EDE5859B2C5B@zaphoyd.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: multipart/alternative; boundary="_000_634914A010D0B943A035D226786325D4446BF99ACBEXVMBX02012ex_"
MIME-Version: 1.0
Cc: Julian Reschke <julian.reschke@gmx.de>, "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] WebSocket subprotocol parameters
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 16:13:09 -0000

--_000_634914A010D0B943A035D226786325D4446BF99ACBEXVMBX02012ex_
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

Take an example:

Client sends:
Sec-WebSocket-Protocol: foo; max_param1=3D10

Server answers:
Sec-WebSocket-Protocol: foo; max_param1=3D8

There are 2 problems:

1) ";" is not allowed as a character in subprotocols
2) the client would reject the WS connection anyway since " foo; max_param1=
=3D8" does not match _exactly_ (one of the) subprotocols announced: "foo; m=
ax_param1=3D10"

"""
   6.  If the response includes a |Sec-WebSocket-Protocol| header field
       and this header field indicates the use of a subprotocol that was
       not present in the client's handshake (the server has indicated a
       subprotocol not requested by the client), the client MUST _Fail
       the WebSocket Connection_.
"""


Von: Peter Thorson [mailto:webmaster@zaphoyd.com]
Gesendet: Montag, 20. Januar 2014 16:59
An: Tobias Oberstein
Cc: Takeshi Yoshino; Julian Reschke; hybi@ietf.org
Betreff: Re: [hybi] WebSocket subprotocol parameters

My impression was that subprotocol primarily changed the interpretation of =
messages payloads by the end applications themselves. WebSocket messages al=
low essentially arbitrary data so you can define whatever parameter setting=
 mechanism you like with whatever grammar makes sense for your subprotocol =
and send the parameters in the opening "hello" message. Defining a grammar =
in the socket spec to allow application protocols to configure themselves s=
eems limiting and out of scope.

Extensions on the other hand needed a way to communicate with themselves as=
 defined and thus needed an explicitly provided grammar.

On Jan 20, 2014, at 5:26, Tobias Oberstein <tobias.oberstein@tavendo.de<mai=
lto:tobias.oberstein@tavendo.de>> wrote:


Hi Takeshi,

thanks alot for digging out these posts!

In retrospect, I think it's a pity that we did not specify "qvalue-like" gr=
ammer for Sec-WebSocket-Protocol as we did for Sec-WebSocket-Extension.

The latter provides a lot of useful flexibility ... eg as used in "permessa=
ge-deflate".

Hacks like having subprotocols like this

wamp.2.json
wamp.2.msgpack
wamp.2.json.batched
wamp.2.msgpack.batched
..

negotiated only works for "discrete value parameters" (json/msgpack, batche=
d/unbatched), but breaks down quite quickly.

As Julian pointed out, making Sec-WebSocket-Protocol using the same "qvalue=
-grammer" as Sec-WebSocket-Extension
would be a significant change to the RFC6455, and break existing code.

It would also possibly require changes to the WS API in JavaScript to be us=
eful.

Mmh. Too bad;(

/Tobias




Von: Takeshi Yoshino [mailto:tyoshino@google.com]
Gesendet: Montag, 20. Januar 2014 09:36
An: Tobias Oberstein
Cc: Julian Reschke; hybi@ietf.org<mailto:hybi@ietf.org>
Betreff: Re: [hybi] WebSocket subprotocol parameters

On Fri, Jan 17, 2014 at 10:01 PM, Tobias Oberstein <tobias.oberstein@tavend=
o.de<mailto:tobias.oberstein@tavendo.de>> wrote:
> > Was there any good reason disallowing parameters for "Sec-WebSocket-
> Protocol"?
>
> I think that was discussed and the decision was made on purpose. For deta=
ils
> you'll have to visit the mailing list archives though.
I can't find it .. searching the Hybi archives is a pain ..

I can't recall such discussion.

In this thread, Brodie used an example similar to that, but the main topic =
of the post was to update the spec to include multiple subprotocol offer me=
chanism.
http://www.ietf.org/mail-archive/web/hybi/current/msg04467.html

In this long thread, what character should be used for subprotocol is discu=
ssed. Some people preferred more liberal rule, but no one discussed qvalue-=
like grammar.
http://www.ietf.org/mail-archive/web/hybi/current/msg00913.html

Subprotocol History FYI:

We added the constraint "U+0021 ..  U+007E" from
http://tools.ietf.org/html/draft-hixie-thewebsocketprotocol-31 -> 32

Single value -> multiple value (SP-separated)
http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-02 -> 03

We made it follow 1#(token | quoted-string) ABNF from
http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-03 -> 04

Changed from (token | quoted-string) to token
http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-07
_______________________________________________
hybi mailing list
hybi@ietf.org<mailto:hybi@ietf.org>
https://www.ietf.org/mailman/listinfo/hybi


--_000_634914A010D0B943A035D226786325D4446BF99ACBEXVMBX02012ex_
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-micr=
osoft-com:office:office" xmlns:w=3D"urn:schemas-microsoft-com:office:word" =
xmlns:m=3D"http://schemas.microsoft.com/office/2004/12/omml" 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"Micros=
oft Word 15 (filtered medium)"><style><!--
/* Font Definitions */
@font-face
	{font-family:Helvetica;
	panose-1:2 11 6 4 2 2 2 2 2 4;}
@font-face
	{font-family:"Cambria Math";
	panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
	{font-family:Calibri;
	panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0cm;
	margin-bottom:.0001pt;
	font-size:12.0pt;
	font-family:"Times New Roman","serif";}
a:link, span.MsoHyperlink
	{mso-style-priority:99;
	color:blue;
	text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
	{mso-style-priority:99;
	color:purple;
	text-decoration:underline;}
pre
	{mso-style-priority:99;
	mso-style-link:"HTML Vorformatiert Zchn";
	margin:0cm;
	margin-bottom:.0001pt;
	font-size:10.0pt;
	font-family:"Courier New";}
span.apple-converted-space
	{mso-style-name:apple-converted-space;}
span.E-MailFormatvorlage18
	{mso-style-type:personal-reply;
	font-family:"Calibri","sans-serif";
	color:#1F497D;}
span.HTMLVorformatiertZchn
	{mso-style-name:"HTML Vorformatiert Zchn";
	mso-style-priority:99;
	mso-style-link:"HTML Vorformatiert";
	font-family:"Courier New";}
.MsoChpDefault
	{mso-style-type:export-only;
	font-size:10.0pt;}
@page WordSection1
	{size:612.0pt 792.0pt;
	margin:70.85pt 70.85pt 2.0cm 70.85pt;}
div.WordSection1
	{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext=3D"edit" spidmax=3D"1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext=3D"edit">
<o:idmap v:ext=3D"edit" data=3D"1" />
</o:shapelayout></xml><![endif]--></head><body lang=3DDE link=3Dblue vlink=
=3Dpurple><div class=3DWordSection1><p class=3DMsoNormal><span style=3D'fon=
t-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D;mso-fareast-=
language:EN-US'>Take an example:<o:p></o:p></span></p><p class=3DMsoNormal>=
<span style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1=
F497D;mso-fareast-language:EN-US'><o:p>&nbsp;</o:p></span></p><p class=3DMs=
oNormal><span style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";=
color:#1F497D;mso-fareast-language:EN-US'>Client sends:<o:p></o:p></span></=
p><p class=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Calibri=
","sans-serif";color:#1F497D;mso-fareast-language:EN-US'>Sec-WebSocket-Prot=
ocol: foo; max_param1=3D10<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D;=
mso-fareast-language:EN-US'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNorma=
l><span style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:=
#1F497D;mso-fareast-language:EN-US'>Server answers:<o:p></o:p></span></p><p=
 class=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Calibri","s=
ans-serif";color:#1F497D;mso-fareast-language:EN-US'>Sec-WebSocket-Protocol=
: foo; max_param1=3D8<o:p></o:p></span></p><p class=3DMsoNormal><span style=
=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D;mso-f=
areast-language:EN-US'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><sp=
an style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F49=
7D;mso-fareast-language:EN-US'>There are 2 problems:<o:p></o:p></span></p><=
p class=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Calibri","=
sans-serif";color:#1F497D;mso-fareast-language:EN-US'><o:p>&nbsp;</o:p></sp=
an></p><p class=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Ca=
libri","sans-serif";color:#1F497D;mso-fareast-language:EN-US'>1) &quot;;&qu=
ot; is not allowed as a character in subprotocols<o:p></o:p></span></p><p c=
lass=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Calibri","san=
s-serif";color:#1F497D;mso-fareast-language:EN-US'>2) the client would reje=
ct the WS connection anyway since &quot; foo; max_param1=3D8&quot; does not=
 match _exactly_ (one of the) subprotocols announced: &quot;foo; max_param1=
=3D10&quot;<o:p></o:p></span></p><p class=3DMsoNormal><span style=3D'font-s=
ize:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D;mso-fareast-lan=
guage:EN-US'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span style=
=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D;mso-f=
areast-language:EN-US'>&quot;&quot;&quot;<o:p></o:p></span></p><p class=3DM=
soNormal><span style=3D'font-size:10.0pt;font-family:"Courier New"'>&nbsp;&=
nbsp; 6.&nbsp; If the response includes a |Sec-WebSocket-Protocol| header f=
ield<o:p></o:p></span></p><p class=3DMsoNormal><span style=3D'font-size:10.=
0pt;font-family:"Courier New"'>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; and thi=
s header field indicates the use of a subprotocol that was<o:p></o:p></span=
></p><p class=3DMsoNormal><span style=3D'font-size:10.0pt;font-family:"Cour=
ier New"'>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; not present in the client's =
handshake (the server has indicated a<o:p></o:p></span></p><p class=3DMsoNo=
rmal><span style=3D'font-size:10.0pt;font-family:"Courier New"'>&nbsp;&nbsp=
;&nbsp;&nbsp;&nbsp;&nbsp; subprotocol not requested by the client), the cli=
ent MUST _Fail<o:p></o:p></span></p><p class=3DMsoNormal><span style=3D'fon=
t-size:10.0pt;font-family:"Courier New"'>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p; the WebSocket Connection_.<o:p></o:p></span></p><p class=3DMsoNormal><sp=
an style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F49=
7D;mso-fareast-language:EN-US'>&quot;&quot;&quot;<o:p></o:p></span></p><p c=
lass=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Calibri","san=
s-serif";color:#1F497D;mso-fareast-language:EN-US'><o:p>&nbsp;</o:p></span>=
</p><p class=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Calib=
ri","sans-serif";color:#1F497D;mso-fareast-language:EN-US'><o:p>&nbsp;</o:p=
></span></p><div style=3D'border:none;border-left:solid blue 1.5pt;padding:=
0cm 0cm 0cm 4.0pt'><div><div style=3D'border:none;border-top:solid #E1E1E1 =
1.0pt;padding:3.0pt 0cm 0cm 0cm'><p class=3DMsoNormal><b><span style=3D'fon=
t-size:11.0pt;font-family:"Calibri","sans-serif"'>Von:</span></b><span styl=
e=3D'font-size:11.0pt;font-family:"Calibri","sans-serif"'> Peter Thorson [m=
ailto:webmaster@zaphoyd.com] <br><b>Gesendet:</b> Montag, 20. Januar 2014 1=
6:59<br><b>An:</b> Tobias Oberstein<br><b>Cc:</b> Takeshi Yoshino; Julian R=
eschke; hybi@ietf.org<br><b>Betreff:</b> Re: [hybi] WebSocket subprotocol p=
arameters<o:p></o:p></span></p></div></div><p class=3DMsoNormal><o:p>&nbsp;=
</o:p></p><div><p class=3DMsoNormal>My impression was that subprotocol prim=
arily changed the interpretation of messages payloads by the end applicatio=
ns themselves. WebSocket messages allow essentially arbitrary data so you c=
an define whatever parameter setting mechanism you like with whatever gramm=
ar makes sense for your subprotocol and send the parameters in the opening =
&#8220;hello&#8221; message. Defining a grammar in the socket spec to allow=
 application protocols to configure themselves seems limiting and out of sc=
ope.<o:p></o:p></p></div><div><p class=3DMsoNormal><o:p>&nbsp;</o:p></p></d=
iv><div><p class=3DMsoNormal>Extensions on the other hand needed a way to c=
ommunicate with themselves as defined and thus needed an explicitly provide=
d grammar.<o:p></o:p></p></div><p class=3DMsoNormal><o:p>&nbsp;</o:p></p><d=
iv><div><p class=3DMsoNormal>On Jan 20, 2014, at 5:26, Tobias Oberstein &lt=
;<a href=3D"mailto:tobias.oberstein@tavendo.de">tobias.oberstein@tavendo.de=
</a>&gt; wrote:<o:p></o:p></p></div><p class=3DMsoNormal><br><br><o:p></o:p=
></p><blockquote style=3D'margin-top:5.0pt;margin-bottom:5.0pt'><div><div><=
p class=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Calibri","=
sans-serif";color:#1F497D'>Hi Takeshi,</span><o:p></o:p></p></div><div><p c=
lass=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Calibri","san=
s-serif";color:#1F497D'>&nbsp;</span><o:p></o:p></p></div><div><p class=3DM=
soNormal><span style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif"=
;color:#1F497D'>thanks alot for digging out these posts!</span><o:p></o:p><=
/p></div><div><p class=3DMsoNormal><span style=3D'font-size:11.0pt;font-fam=
ily:"Calibri","sans-serif";color:#1F497D'>&nbsp;</span><o:p></o:p></p></div=
><div><p class=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Cal=
ibri","sans-serif";color:#1F497D'>In retrospect, I think it's a pity that w=
e did not specify &quot;qvalue-like&quot; grammer for Sec-WebSocket-Protoco=
l as we did for Sec-WebSocket-Extension.</span><o:p></o:p></p></div><div><p=
 class=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Calibri","s=
ans-serif";color:#1F497D'>&nbsp;</span><o:p></o:p></p></div><div><p class=
=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Calibri","sans-se=
rif";color:#1F497D'>The latter provides a lot of useful flexibility &#8230;=
 eg as used in &quot;permessage-deflate&quot;.</span><o:p></o:p></p></div><=
div><p class=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Calib=
ri","sans-serif";color:#1F497D'>&nbsp;</span><o:p></o:p></p></div><div><p c=
lass=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Calibri","san=
s-serif";color:#1F497D'>Hacks like having subprotocols like this</span><o:p=
></o:p></p></div><div><p class=3DMsoNormal><span style=3D'font-size:11.0pt;=
font-family:"Calibri","sans-serif";color:#1F497D'>&nbsp;</span><o:p></o:p><=
/p></div><div><p class=3DMsoNormal><span style=3D'font-size:11.0pt;font-fam=
ily:"Calibri","sans-serif";color:#1F497D'>wamp.2.json</span><o:p></o:p></p>=
</div><div><p class=3DMsoNormal><span style=3D'font-size:11.0pt;font-family=
:"Calibri","sans-serif";color:#1F497D'>wamp.2.msgpack</span><o:p></o:p></p>=
</div><div><p class=3DMsoNormal><span style=3D'font-size:11.0pt;font-family=
:"Calibri","sans-serif";color:#1F497D'>wamp.2.json.batched</span><o:p></o:p=
></p></div><div><p class=3DMsoNormal><span style=3D'font-size:11.0pt;font-f=
amily:"Calibri","sans-serif";color:#1F497D'>wamp.2.msgpack.batched</span><o=
:p></o:p></p></div><div><p class=3DMsoNormal><span style=3D'font-size:11.0p=
t;font-family:"Calibri","sans-serif";color:#1F497D'>..</span><o:p></o:p></p=
></div><div><p class=3DMsoNormal><span style=3D'font-size:11.0pt;font-famil=
y:"Calibri","sans-serif";color:#1F497D'>&nbsp;</span><o:p></o:p></p></div><=
div><p class=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Calib=
ri","sans-serif";color:#1F497D'>negotiated only works for &quot;discrete va=
lue parameters&quot; (json/msgpack, batched/unbatched), but breaks down qui=
te quickly.</span><o:p></o:p></p></div><div><p class=3DMsoNormal><span styl=
e=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>&nb=
sp;</span><o:p></o:p></p></div><div><p class=3DMsoNormal><span style=3D'fon=
t-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>As Julian p=
ointed out, making Sec-WebSocket-Protocol using the same &quot;qvalue-gramm=
er&quot; as Sec-WebSocket-Extension</span><o:p></o:p></p></div><div><p clas=
s=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Calibri","sans-s=
erif";color:#1F497D'>would be a significant change to the RFC6455, and brea=
k existing code.</span><o:p></o:p></p></div><div><p class=3DMsoNormal><span=
 style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D=
'>&nbsp;</span><o:p></o:p></p></div><div><p class=3DMsoNormal><span style=
=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>It w=
ould also possibly require changes to the WS API in JavaScript to be useful=
.</span><o:p></o:p></p></div><div><p class=3DMsoNormal><span style=3D'font-=
size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>&nbsp;</span>=
<o:p></o:p></p></div><div><p class=3DMsoNormal><span style=3D'font-size:11.=
0pt;font-family:"Calibri","sans-serif";color:#1F497D'>Mmh. Too bad;(</span>=
<o:p></o:p></p></div><div><p class=3DMsoNormal><span style=3D'font-size:11.=
0pt;font-family:"Calibri","sans-serif";color:#1F497D'>&nbsp;</span><o:p></o=
:p></p></div><div><p class=3DMsoNormal><span style=3D'font-size:11.0pt;font=
-family:"Calibri","sans-serif";color:#1F497D'>/Tobias</span><o:p></o:p></p>=
</div><div><p class=3DMsoNormal><span style=3D'font-size:11.0pt;font-family=
:"Calibri","sans-serif";color:#1F497D'>&nbsp;</span><o:p></o:p></p></div><d=
iv><p class=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Calibr=
i","sans-serif";color:#1F497D'>&nbsp;</span><o:p></o:p></p></div><div><p cl=
ass=3DMsoNormal><span style=3D'font-size:11.0pt;font-family:"Calibri","sans=
-serif";color:#1F497D'>&nbsp;</span><o:p></o:p></p></div><div><p class=3DMs=
oNormal><span style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";=
color:#1F497D'>&nbsp;</span><o:p></o:p></p></div><div style=3D'border:none;=
border-left:solid blue 1.5pt;padding:0cm 0cm 0cm 4.0pt'><div><div style=3D'=
border:none;border-top:solid #E1E1E1 1.0pt;padding:3.0pt 0cm 0cm 0cm'><div>=
<p class=3DMsoNormal><b><span style=3D'font-size:11.0pt;font-family:"Calibr=
i","sans-serif"'>Von:</span></b><span class=3Dapple-converted-space><span s=
tyle=3D'font-size:11.0pt;font-family:"Calibri","sans-serif"'>&nbsp;</span><=
/span><span style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif"'>T=
akeshi Yoshino [<a href=3D"mailto:tyoshino@google.com">mailto:tyoshino@goog=
le.com</a>]<span class=3Dapple-converted-space>&nbsp;</span><br><b>Gesendet=
:</b><span class=3Dapple-converted-space>&nbsp;</span>Montag, 20. Januar 20=
14 09:36<br><b>An:</b><span class=3Dapple-converted-space>&nbsp;</span>Tobi=
as Oberstein<br><b>Cc:</b><span class=3Dapple-converted-space>&nbsp;</span>=
Julian Reschke; <a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br><b>Be=
treff:</b><span class=3Dapple-converted-space>&nbsp;</span>Re: [hybi] WebSo=
cket subprotocol parameters</span><o:p></o:p></p></div></div></div><div><p =
class=3DMsoNormal>&nbsp;<o:p></o:p></p></div><div><div><p class=3DMsoNormal=
>On Fri, Jan 17, 2014 at 10:01 PM, Tobias Oberstein &lt;<a href=3D"mailto:t=
obias.oberstein@tavendo.de" target=3D"_blank"><span style=3D'color:purple'>=
tobias.oberstein@tavendo.de</span></a>&gt; wrote:<o:p></o:p></p></div><bloc=
kquote style=3D'border:none;border-left:solid #CCCCCC 1.0pt;padding:0cm 0cm=
 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-botto=
m:5.0pt'><div><p class=3DMsoNormal style=3D'margin-bottom:12.0pt'>&gt; &gt;=
 Was there any good reason disallowing parameters for &quot;Sec-WebSocket-<=
br>&gt; Protocol&quot;?<br>&gt;<br>&gt; I think that was discussed and the =
decision was made on purpose. For details<br>&gt; you'll have to visit the =
mailing list archives though.<o:p></o:p></p></div><div><p class=3DMsoNormal=
>I can't find it .. searching the Hybi archives is a pain ..<o:p></o:p></p>=
</div></blockquote><div><div><p class=3DMsoNormal>&nbsp;<o:p></o:p></p></di=
v></div><div><div><p class=3DMsoNormal>I can't recall such discussion.<o:p>=
</o:p></p></div></div><div><div><p class=3DMsoNormal>&nbsp;<o:p></o:p></p><=
/div></div><div><div><p class=3DMsoNormal>In this thread, Brodie used an ex=
ample similar to that, but the main topic of the post was to update the spe=
c to include multiple subprotocol offer mechanism.<o:p></o:p></p></div></di=
v><div><div><p class=3DMsoNormal><a href=3D"http://www.ietf.org/mail-archiv=
e/web/hybi/current/msg04467.html"><span style=3D'color:purple'>http://www.i=
etf.org/mail-archive/web/hybi/current/msg04467.html</span></a><o:p></o:p></=
p></div></div><div><div><p class=3DMsoNormal>&nbsp;<o:p></o:p></p></div></d=
iv><div><div><p class=3DMsoNormal>In this long thread, what character shoul=
d be used for subprotocol is discussed. Some people preferred more liberal =
rule, but no one discussed qvalue-like grammar.<o:p></o:p></p></div></div><=
div><div><p class=3DMsoNormal><a href=3D"http://www.ietf.org/mail-archive/w=
eb/hybi/current/msg00913.html"><span style=3D'color:purple'>http://www.ietf=
.org/mail-archive/web/hybi/current/msg00913.html</span></a><o:p></o:p></p><=
/div></div><div><div><p class=3DMsoNormal>&nbsp;<o:p></o:p></p></div></div>=
<div><div><p class=3DMsoNormal>Subprotocol History FYI:<o:p></o:p></p></div=
></div><div><div><p class=3DMsoNormal>&nbsp;<o:p></o:p></p></div></div><div=
><div><p class=3DMsoNormal>We added the constraint &quot;U+0021 .. &nbsp;U+=
007E&quot; from<o:p></o:p></p></div></div><div><p class=3DMsoNormal><a href=
=3D"http://tools.ietf.org/html/draft-hixie-thewebsocketprotocol-31"><span s=
tyle=3D'color:purple'>http://tools.ietf.org/html/draft-hixie-thewebsocketpr=
otocol-31</span></a><span class=3Dapple-converted-space>&nbsp;</span>-&gt; =
32<o:p></o:p></p></div><div><div><p class=3DMsoNormal>&nbsp;<o:p></o:p></p>=
</div></div><div><div><p class=3DMsoNormal>Single value -&gt; multiple valu=
e (SP-separated)<o:p></o:p></p></div></div><div><div><p class=3DMsoNormal><=
a href=3D"http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-0=
2"><span style=3D'color:purple'>http://tools.ietf.org/html/draft-ietf-hybi-=
thewebsocketprotocol-02</span></a><span class=3Dapple-converted-space>&nbsp=
;</span>-&gt; 03<o:p></o:p></p></div></div><div><div><p class=3DMsoNormal>&=
nbsp;<o:p></o:p></p></div></div><div><div><p class=3DMsoNormal>We made it f=
ollow 1#(token | quoted-string) ABNF from<o:p></o:p></p></div></div><div><d=
iv><p class=3DMsoNormal><a href=3D"http://tools.ietf.org/html/draft-ietf-hy=
bi-thewebsocketprotocol-03"><span style=3D'color:purple'>http://tools.ietf.=
org/html/draft-ietf-hybi-thewebsocketprotocol-03</span></a><span class=3Dap=
ple-converted-space>&nbsp;</span>-&gt; 04<o:p></o:p></p></div></div><div><d=
iv><p class=3DMsoNormal>&nbsp;<o:p></o:p></p></div></div><div><div><p class=
=3DMsoNormal>Changed from (token | quoted-string) to token<o:p></o:p></p></=
div></div><div><div><p class=3DMsoNormal><a href=3D"http://tools.ietf.org/h=
tml/draft-ietf-hybi-thewebsocketprotocol-07"><span style=3D'color:purple'>h=
ttp://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-07</span></a=
>&nbsp;<o:p></o:p></p></div></div></div></div><p class=3DMsoNormal><span st=
yle=3D'font-size:9.0pt;font-family:"Helvetica","sans-serif"'>______________=
_________________________________<br>hybi mailing list<br><a href=3D"mailto=
:hybi@ietf.org">hybi@ietf.org</a><br><a href=3D"https://www.ietf.org/mailma=
n/listinfo/hybi">https://www.ietf.org/mailman/listinfo/hybi</a><o:p></o:p><=
/span></p></div></blockquote></div><p class=3DMsoNormal><o:p>&nbsp;</o:p></=
p></div></div></body></html>=

--_000_634914A010D0B943A035D226786325D4446BF99ACBEXVMBX02012ex_--

From webmaster@zaphoyd.com  Mon Jan 20 08:13:10 2014
Return-Path: <webmaster@zaphoyd.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id CFDC41A0162 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 08:13:10 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id InjoWHRYBcvE for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 08:13:09 -0800 (PST)
Received: from authsmtp00.uchicago.edu (authsmtp00.uchicago.edu [128.135.12.120]) by ietfa.amsl.com (Postfix) with ESMTP id D35C11A01B5 for <hybi@ietf.org>; Mon, 20 Jan 2014 08:13:08 -0800 (PST)
Received: from [10.0.1.3] (c-68-51-76-178.hsd1.il.comcast.net [68.51.76.178]) (Authenticated sender: zaphoyd) by authsmtp00.uchicago.edu (Postfix) with ESMTP id 31FBA49809F; Mon, 20 Jan 2014 10:13:08 -0600 (CST)
Content-Type: text/plain; charset=windows-1252
Mime-Version: 1.0 (Mac OS X Mail 7.0 \(1822\))
From: Peter Thorson <webmaster@zaphoyd.com>
In-Reply-To: <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net>
Date: Mon, 20 Jan 2014 10:16:17 -0600
Content-Transfer-Encoding: quoted-printable
Message-Id: <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
X-Mailer: Apple Mail (2.1822)
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 16:13:11 -0000

On Jan 20, 2014, at 9:51, Tobias Oberstein <tobias.oberstein@tavendo.de> =
wrote:

> Hi Peter
>=20
>> A simple proposal to support WebSocket API functionality might be =
something like:
>> - HTTP 2.0 HEADERS message already provides the functionality =
necessary to negotiate a given HTTP 2.0 channel as a "WebSocket" channel =
along with sending subprotocol, cookies, etc.
>=20
> Yep.
>=20
>> - An HTTP 2.0 WEBSOCKET frame type could be used to deliver WebSocket =
messages. You'd need a FIN flag for the message boundaries, frame =
boundaries would be just standard HTTP 2.0 frames. No need to repackage =
or reframe or make a distinction between >"websocket" frames and HTTP =
2.0 frames.
>=20
> Why a separate SPDY frame type? Why not SPDY/DATA frames? After =
negotiation, the whole SPDY _stream_ will be "messaging" only anyway ..
>=20
>> - Depending on how closely we want HTTP 2.0 WebSockets to match =
RFC6455 WebSockets additional support could be added.. for example, =
define the WEBSOCKET frame type's flag bits as follows (full =
compatibility flag configuration):
>=20
> =46rom the goal of W3C WebSocket API only compatibility, the only =
other information (besides message boundaries) needed in SPDY/DATA =
frames would be the payload type: UTF8 vs binary.
>=20
> So the minimum appears to be 2 bits in SPDY/DATA frame headers:
>=20
> 1 bit for MSG_DONE on the last SPDY/DATA frame and
> 1 bit for MSG_TYPE (UTF8/binary) on the first SPDY/DATA frame that =
starts a new msg
>=20
> Am I missing something?
>=20
> /Tobias


Using straight up DATA frames would certainly be an option. I suggest a =
separate frame type because I think the types of flags you would set for =
generic requests (like the HTTP 1.1 requests over HTTP 2.0 transport) =
would be different than the types you would set for the short websocket =
like messages. Take the UTF8/binary flag (or the RFC6455 opcode in the =
more general case) for example, does that make sense for a request for =
an image?

You could argue that you could reinterpret the flags in the DATA frame =
based on whether or not you have negotiated websocket mode but this =
seems ripe for confusion and exactly the point of having a frame type. A =
frame type lets you define what flags mean for that frame type and what =
the rules for the processing of that frame type would be. It also =
simplifies implementation logic that might want to logically switch =
processing branches on frame type.

------
Re: minimum bits

I definitely agree that as written today only message done, stream done, =
and utf8/binary are necessary. Stream done is present already in the =
DATA frame type. This does seem to ignore the fact that RFC6455 defines =
8 data message types though and might limit compatibility between the =
two in the future. That said, I am not really sure yet whether that =
matters. On one hand, text and binary covers quite a few cases and I =
can=92t really think of a good use for more data types visible at the =
application/WS API level since applications are free via =
subprotocols/convention to define whatever message types they want. On =
the other hand, =93because I can=92t think of a good use case=94 doesn=92t=
 seem like a strong argument.=

From webmaster@zaphoyd.com  Mon Jan 20 08:25:29 2014
Return-Path: <webmaster@zaphoyd.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9A2C81A01BF for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 08:25:29 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.899
X-Spam-Level: 
X-Spam-Status: No, score=-1.899 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id mCBaejUoAmIG for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 08:25:28 -0800 (PST)
Received: from authsmtp01.uchicago.edu (authsmtp01.uchicago.edu [128.135.12.121]) by ietfa.amsl.com (Postfix) with ESMTP id CCE001A01A5 for <hybi@ietf.org>; Mon, 20 Jan 2014 08:25:27 -0800 (PST)
Received: from [10.0.1.3] (c-68-51-76-178.hsd1.il.comcast.net [68.51.76.178]) (Authenticated sender: zaphoyd) by authsmtp01.uchicago.edu (Postfix) with ESMTP id 58A8649806B; Mon, 20 Jan 2014 10:25:27 -0600 (CST)
Content-Type: multipart/alternative; boundary="Apple-Mail=_9746B7A6-14E1-4125-B2E4-44DCE1A4C331"
Mime-Version: 1.0 (Mac OS X Mail 7.0 \(1822\))
From: Peter Thorson <webmaster@zaphoyd.com>
In-Reply-To: <634914A010D0B943A035D226786325D4446BF99ACB@EXVMBX020-12.exch020.serverdata.net>
Date: Mon, 20 Jan 2014 10:28:36 -0600
Message-Id: <82A1484B-1CDC-45BD-A2D2-6BED128F1448@zaphoyd.com>
References: <634914A010D0B943A035D226786325D4446BF9948F@EXVMBX020-12.exch020.serverdata.net> <52D90F99.6080205@gmx.de> <634914A010D0B943A035D226786325D4446BF9949D@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZQ3NxVW36PnZ0TpMF4tPPaJLr12M8NtbPb6pUejf1wcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99977@EXVMBX020-12.exch020.serverdata.net> <B8ED6F18-B710-44AE-829B-EDE5859B2C5B@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99ACB@EXVMBX020-12.exch020.serverdata.net>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
X-Mailer: Apple Mail (2.1822)
Cc: Julian Reschke <julian.reschke@gmx.de>, "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] WebSocket subprotocol parameters
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 16:25:29 -0000

--Apple-Mail=_9746B7A6-14E1-4125-B2E4-44DCE1A4C331
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=windows-1252


On Jan 20, 2014, at 10:13, Tobias Oberstein =
<tobias.oberstein@tavendo.de> wrote:

> Take an example:
> =20
> Client sends:
> Sec-WebSocket-Protocol: foo; max_param1=3D10
> =20
> Server answers:
> Sec-WebSocket-Protocol: foo; max_param1=3D8
> =20
> There are 2 problems:
> =20
> 1) ";" is not allowed as a character in subprotocols
> 2) the client would reject the WS connection anyway since " foo; =
max_param1=3D8" does not match _exactly_ (one of the) subprotocols =
announced: "foo; max_param1=3D10"
> =20
> """
>    6.  If the response includes a |Sec-WebSocket-Protocol| header =
field
>        and this header field indicates the use of a subprotocol that =
was
>        not present in the client's handshake (the server has indicated =
a
>        subprotocol not requested by the client), the client MUST _Fail
>        the WebSocket Connection_.=93
> ""

What do you gain by negotiating max_param1 at the subprotocol level vs =
negotiating =93foo=94 as your subprotocol and having the foo protocol =
define how to negotiate its own settings via regular websocket messages?=

--Apple-Mail=_9746B7A6-14E1-4125-B2E4-44DCE1A4C331
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
	charset=windows-1252

<html><head><meta http-equiv=3D"Content-Type" content=3D"text/html =
charset=3Dwindows-1252"></head><body style=3D"word-wrap: break-word; =
-webkit-nbsp-mode: space; -webkit-line-break: =
after-white-space;"><br><div><div>On Jan 20, 2014, at 10:13, Tobias =
Oberstein &lt;<a =
href=3D"mailto:tobias.oberstein@tavendo.de">tobias.oberstein@tavendo.de</a=
>&gt; wrote:</div><br class=3D"Apple-interchange-newline"><blockquote =
type=3D"cite"><div lang=3D"DE" link=3D"blue" vlink=3D"purple" =
style=3D"orphans: auto; text-align: start; text-indent: 0px; widows: =
auto;"><div class=3D"WordSection1" style=3D"page: WordSection1;"><div =
style=3D"font-family: 'Times New Roman', serif; font-size: 12pt; =
font-style: normal; font-variant: normal; font-weight: normal; =
letter-spacing: normal; line-height: normal; text-transform: none; =
white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; =
margin: 0cm 0cm 0.0001pt;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, 125);">Take an =
example:<o:p></o:p></span></div><div style=3D"font-family: 'Times New =
Roman', serif; font-size: 12pt; font-style: normal; font-variant: =
normal; font-weight: normal; letter-spacing: normal; line-height: =
normal; text-transform: none; white-space: normal; word-spacing: 0px; =
-webkit-text-stroke-width: 0px; margin: 0cm 0cm 0.0001pt;"><span =
style=3D"font-size: 11pt; font-family: Calibri, sans-serif; color: =
rgb(31, 73, 125);">&nbsp;</span></div><div style=3D"font-family: 'Times =
New Roman', serif; font-size: 12pt; font-style: normal; font-variant: =
normal; font-weight: normal; letter-spacing: normal; line-height: =
normal; text-transform: none; white-space: normal; word-spacing: 0px; =
-webkit-text-stroke-width: 0px; margin: 0cm 0cm 0.0001pt;"><span =
style=3D"font-size: 11pt; font-family: Calibri, sans-serif; color: =
rgb(31, 73, 125);">Client sends:<o:p></o:p></span></div><div =
style=3D"font-family: 'Times New Roman', serif; font-size: 12pt; =
font-style: normal; font-variant: normal; font-weight: normal; =
letter-spacing: normal; line-height: normal; text-transform: none; =
white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; =
margin: 0cm 0cm 0.0001pt;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, 125);">Sec-WebSocket-Protocol: =
foo; max_param1=3D10<o:p></o:p></span></div><div style=3D"font-family: =
'Times New Roman', serif; font-size: 12pt; font-style: normal; =
font-variant: normal; font-weight: normal; letter-spacing: normal; =
line-height: normal; text-transform: none; white-space: normal; =
word-spacing: 0px; -webkit-text-stroke-width: 0px; margin: 0cm 0cm =
0.0001pt;"><span style=3D"font-size: 11pt; font-family: Calibri, =
sans-serif; color: rgb(31, 73, 125);">&nbsp;</span></div><div =
style=3D"font-family: 'Times New Roman', serif; font-size: 12pt; =
font-style: normal; font-variant: normal; font-weight: normal; =
letter-spacing: normal; line-height: normal; text-transform: none; =
white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; =
margin: 0cm 0cm 0.0001pt;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, 125);">Server =
answers:<o:p></o:p></span></div><div style=3D"font-family: 'Times New =
Roman', serif; font-size: 12pt; font-style: normal; font-variant: =
normal; font-weight: normal; letter-spacing: normal; line-height: =
normal; text-transform: none; white-space: normal; word-spacing: 0px; =
-webkit-text-stroke-width: 0px; margin: 0cm 0cm 0.0001pt;"><span =
style=3D"font-size: 11pt; font-family: Calibri, sans-serif; color: =
rgb(31, 73, 125);">Sec-WebSocket-Protocol: foo; =
max_param1=3D8<o:p></o:p></span></div><div style=3D"font-family: 'Times =
New Roman', serif; font-size: 12pt; font-style: normal; font-variant: =
normal; font-weight: normal; letter-spacing: normal; line-height: =
normal; text-transform: none; white-space: normal; word-spacing: 0px; =
-webkit-text-stroke-width: 0px; margin: 0cm 0cm 0.0001pt;"><span =
style=3D"font-size: 11pt; font-family: Calibri, sans-serif; color: =
rgb(31, 73, 125);">&nbsp;</span></div><div style=3D"font-family: 'Times =
New Roman', serif; font-size: 12pt; font-style: normal; font-variant: =
normal; font-weight: normal; letter-spacing: normal; line-height: =
normal; text-transform: none; white-space: normal; word-spacing: 0px; =
-webkit-text-stroke-width: 0px; margin: 0cm 0cm 0.0001pt;"><span =
style=3D"font-size: 11pt; font-family: Calibri, sans-serif; color: =
rgb(31, 73, 125);">There are 2 problems:<o:p></o:p></span></div><div =
style=3D"font-family: 'Times New Roman', serif; font-size: 12pt; =
font-style: normal; font-variant: normal; font-weight: normal; =
letter-spacing: normal; line-height: normal; text-transform: none; =
white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; =
margin: 0cm 0cm 0.0001pt;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, 125);">&nbsp;</span></div><div =
style=3D"font-family: 'Times New Roman', serif; font-size: 12pt; =
font-style: normal; font-variant: normal; font-weight: normal; =
letter-spacing: normal; line-height: normal; text-transform: none; =
white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; =
margin: 0cm 0cm 0.0001pt;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, 125);">1) ";" is not allowed as =
a character in subprotocols<o:p></o:p></span></div><div =
style=3D"font-family: 'Times New Roman', serif; font-size: 12pt; =
font-style: normal; font-variant: normal; font-weight: normal; =
letter-spacing: normal; line-height: normal; text-transform: none; =
white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; =
margin: 0cm 0cm 0.0001pt;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, 125);">2) the client would =
reject the WS connection anyway since " foo; max_param1=3D8" does not =
match _exactly_ (one of the) subprotocols announced: "foo; =
max_param1=3D10"<o:p></o:p></span></div><div style=3D"font-family: =
'Times New Roman', serif; font-size: 12pt; font-style: normal; =
font-variant: normal; font-weight: normal; letter-spacing: normal; =
line-height: normal; text-transform: none; white-space: normal; =
word-spacing: 0px; -webkit-text-stroke-width: 0px; margin: 0cm 0cm =
0.0001pt;"><span style=3D"font-size: 11pt; font-family: Calibri, =
sans-serif; color: rgb(31, 73, 125);">&nbsp;</span></div><div =
style=3D"font-family: 'Times New Roman', serif; font-size: 12pt; =
font-style: normal; font-variant: normal; font-weight: normal; =
letter-spacing: normal; line-height: normal; text-transform: none; =
white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; =
margin: 0cm 0cm 0.0001pt;"><span style=3D"font-size: 11pt; font-family: =
Calibri, sans-serif; color: rgb(31, 73, =
125);">"""<o:p></o:p></span></div><div style=3D"font-family: 'Times New =
Roman', serif; font-size: 12pt; font-style: normal; font-variant: =
normal; font-weight: normal; letter-spacing: normal; line-height: =
normal; text-transform: none; white-space: normal; word-spacing: 0px; =
-webkit-text-stroke-width: 0px; margin: 0cm 0cm 0.0001pt;"><span =
style=3D"font-size: 10pt; font-family: 'Courier New';">&nbsp;&nbsp; =
6.&nbsp; If the response includes a |Sec-WebSocket-Protocol| header =
field<o:p></o:p></span></div><div style=3D"font-family: 'Times New =
Roman', serif; font-size: 12pt; font-style: normal; font-variant: =
normal; font-weight: normal; letter-spacing: normal; line-height: =
normal; text-transform: none; white-space: normal; word-spacing: 0px; =
-webkit-text-stroke-width: 0px; margin: 0cm 0cm 0.0001pt;"><span =
style=3D"font-size: 10pt; font-family: 'Courier =
New';">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; and this header field =
indicates the use of a subprotocol that was<o:p></o:p></span></div><div =
style=3D"font-family: 'Times New Roman', serif; font-size: 12pt; =
font-style: normal; font-variant: normal; font-weight: normal; =
letter-spacing: normal; line-height: normal; text-transform: none; =
white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; =
margin: 0cm 0cm 0.0001pt;"><span style=3D"font-size: 10pt; font-family: =
'Courier New';">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; not present in the =
client's handshake (the server has indicated =
a<o:p></o:p></span></div><div style=3D"font-family: 'Times New Roman', =
serif; font-size: 12pt; font-style: normal; font-variant: normal; =
font-weight: normal; letter-spacing: normal; line-height: normal; =
text-transform: none; white-space: normal; word-spacing: 0px; =
-webkit-text-stroke-width: 0px; margin: 0cm 0cm 0.0001pt;"><span =
style=3D"font-size: 10pt; font-family: 'Courier =
New';">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; subprotocol not requested by =
the client), the client MUST _Fail<o:p></o:p></span></div><div =
style=3D"margin: 0cm 0cm 0.0001pt;"><span style=3D"font-family: 'Courier =
New'; font-size: 10pt; font-style: normal; font-variant: normal; =
font-weight: normal; letter-spacing: normal; line-height: normal; =
text-transform: none; white-space: normal; word-spacing: 0px; =
-webkit-text-stroke-width: 0px;">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
the WebSocket Connection_.<font color=3D"#1f497d" face=3D"Calibri, =
sans-serif"><span style=3D"font-size: =
15px;"><o:p></o:p></span></font></span><font color=3D"#1f497d" =
face=3D"Calibri, sans-serif"><span style=3D"font-size: =
15px;">=93<br>""</span></font></div></div></div></blockquote><br></div><di=
v>What do you gain by negotiating max_param1 at the subprotocol level vs =
negotiating =93foo=94 as your subprotocol and having the foo protocol =
define how to negotiate its own settings via regular websocket =
messages?</div></body></html>=

--Apple-Mail=_9746B7A6-14E1-4125-B2E4-44DCE1A4C331--

From tobias.oberstein@tavendo.de  Mon Jan 20 09:12:53 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 17C5C1A01D5 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 09:12:53 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id okqIGCLrnDBQ for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 09:12:51 -0800 (PST)
Received: from EXHUB020-3.exch020.serverdata.net (exhub020-3.exch020.serverdata.net [206.225.164.30]) by ietfa.amsl.com (Postfix) with ESMTP id 6DC101A0162 for <hybi@ietf.org>; Mon, 20 Jan 2014 09:12:51 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-3.exch020.serverdata.net ([206.225.164.30]) with mapi; Mon, 20 Jan 2014 09:12:51 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Peter Thorson <webmaster@zaphoyd.com>
Date: Mon, 20 Jan 2014 09:12:48 -0800
Thread-Topic: [hybi] WebSocket subprotocol parameters
Thread-Index: Ac8V/D1Ae5jPSiwTRGOJCbF2SnIQEQABj9aQ
Message-ID: <634914A010D0B943A035D226786325D4446BF99B57@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF9948F@EXVMBX020-12.exch020.serverdata.net> <52D90F99.6080205@gmx.de> <634914A010D0B943A035D226786325D4446BF9949D@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZQ3NxVW36PnZ0TpMF4tPPaJLr12M8NtbPb6pUejf1wcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99977@EXVMBX020-12.exch020.serverdata.net> <B8ED6F18-B710-44AE-829B-EDE5859B2C5B@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99ACB@EXVMBX020-12.exch020.serverdata.net> <82A1484B-1CDC-45BD-A2D2-6BED128F1448@zaphoyd.com>
In-Reply-To: <82A1484B-1CDC-45BD-A2D2-6BED128F1448@zaphoyd.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
Cc: Julian Reschke <julian.reschke@gmx.de>, "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] WebSocket subprotocol parameters
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 17:12:53 -0000

>What do you gain by negotiating max_param1 at the subprotocol level vs neg=
otiating "foo" as your subprotocol and having the foo protocol define how t=
o negotiate its own settings via regular websocket messages?

If "param1" would influence how to parse raw WS payload into "foo" protocol=
, that would not be possible.

But I agree in general: as long as "param1" does not influence the applicat=
ion protocol parsing, the app protocol could do (any further) parametrizati=
on/negotiation itself and no WS subprotocol level negotiation is necessary.

/Tobias

From tobias.oberstein@tavendo.de  Mon Jan 20 09:36:03 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 666401A0214 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 09:36:03 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id wM2gQZ0e79n7 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 09:36:01 -0800 (PST)
Received: from EXHUB020-2.exch020.serverdata.net (exhub020-2.exch020.serverdata.net [206.225.164.29]) by ietfa.amsl.com (Postfix) with ESMTP id AE05D1A0213 for <hybi@ietf.org>; Mon, 20 Jan 2014 09:36:01 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-2.exch020.serverdata.net ([206.225.164.29]) with mapi; Mon, 20 Jan 2014 09:36:01 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Peter Thorson <webmaster@zaphoyd.com>
Date: Mon, 20 Jan 2014 09:35:58 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8V+oM863LiA9S3R22ycqnjSI6MXAACo4Ww
Message-ID: <634914A010D0B943A035D226786325D4446C0826EC@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net> <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com>
In-Reply-To: <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 17:36:03 -0000

> Using straight up DATA frames would certainly be an option. I suggest a
> separate frame type because I think the types of flags you would set for
> generic requests (like the HTTP 1.1 requests over HTTP 2.0 transport) wou=
ld
> be different than the types you would set for the short websocket like
> messages. Take the UTF8/binary flag (or the RFC6455 opcode in the more
> general case) for example, does that make sense for a request for an imag=
e?

I agree, MSG_TYPE bit would probably not be useful in general.

>=20
> You could argue that you could reinterpret the flags in the DATA frame ba=
sed
> on whether or not you have negotiated websocket mode but this seems ripe
> for confusion and exactly the point of having a frame type. A frame type =
lets
> you define what flags mean for that frame type and what the rules for the
> processing of that frame type would be. It also simplifies implementation
> logic that might want to logically switch processing branches on frame ty=
pe.

However, an implementation would probably want to remember if a SPDY stream=
 ID=20
is SPDY/Messaging or not, if just to check that all frames on that stream c=
onform to
SPDY/Messaging requirements: either frame opcode or MSG_TYPE flag, and payl=
oad
conformance.

Another consideration are intermediaries: do they want simplified handling =
(all SPDY/DATA
frames handled alike) or specific handling: SPDY/MESSAGE frame differently.

>=20
> ------
> Re: minimum bits
>=20
> I definitely agree that as written today only message done, stream done, =
and
> utf8/binary are necessary. Stream done is present already in the DATA fra=
me
> type. This does seem to ignore the fact that RFC6455 defines 8 data messa=
ge
> types though and might limit compatibility between the two in the future.

Yep, that's the idea. Ignore other WS opcodes, since those are not exposed =
in the W3C API
anyway, and compatibility with the latter only was the supposed goal.

There is 1 more API thing: W3C API lets me close a WS with "code" and "reas=
on".

How do would this be mapped?

> That said, I am not really sure yet whether that matters. On one hand, te=
xt
> and binary covers quite a few cases and I can't really think of a good us=
e for
> more data types visible at the application/WS API level since application=
s are
> free via subprotocols/convention to define whatever message types they
> want. On the other hand, "because I can't think of a good use case" doesn=
't
> seem like a strong argument.

Why not? I think thats a good argument in fact;) If we can't come up with a=
 use case (and
I cannot imagine one also), then why pre-plan for a yet-to-be-imagined chim=
era?

/Tobias

From webmaster@zaphoyd.com  Mon Jan 20 09:49:40 2014
Return-Path: <webmaster@zaphoyd.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 1C5BB1A0212 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 09:49:40 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id g-RyBn_2v_1L for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 09:49:39 -0800 (PST)
Received: from authsmtp00.uchicago.edu (authsmtp00.uchicago.edu [128.135.12.120]) by ietfa.amsl.com (Postfix) with ESMTP id E846F1A0169 for <hybi@ietf.org>; Mon, 20 Jan 2014 09:49:38 -0800 (PST)
Received: from [10.0.1.3] (c-68-51-76-178.hsd1.il.comcast.net [68.51.76.178]) (Authenticated sender: zaphoyd) by authsmtp00.uchicago.edu (Postfix) with ESMTP id AE92949809F; Mon, 20 Jan 2014 11:49:38 -0600 (CST)
Content-Type: text/plain; charset=windows-1252
Mime-Version: 1.0 (Mac OS X Mail 7.0 \(1822\))
From: Peter Thorson <webmaster@zaphoyd.com>
In-Reply-To: <634914A010D0B943A035D226786325D4446C0826EC@EXVMBX020-12.exch020.serverdata.net>
Date: Mon, 20 Jan 2014 11:52:50 -0600
Content-Transfer-Encoding: quoted-printable
Message-Id: <6652CC7E-7E92-40EF-BD50-2DDD39E15533@zaphoyd.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net> <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com> <634914A010D0B943A035D226786325D4446C0826EC@EXVMBX020-12. exch020.serverdata.net>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
X-Mailer: Apple Mail (2.1822)
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Jan 2014 17:49:40 -0000

On Jan 20, 2014, at 11:35, Tobias Oberstein =
<tobias.oberstein@tavendo.de> wrote:

>> Re: minimum bits
>>=20
>> I definitely agree that as written today only message done, stream =
done, and
>> utf8/binary are necessary. Stream done is present already in the DATA =
frame
>> type. This does seem to ignore the fact that RFC6455 defines 8 data =
message
>> types though and might limit compatibility between the two in the =
future.
>=20
> Yep, that's the idea. Ignore other WS opcodes, since those are not =
exposed in the W3C API
> anyway, and compatibility with the latter only was the supposed goal.
>=20
> There is 1 more API thing: W3C API lets me close a WS with "code" and =
"reason".
>=20
> How do would this be mapped?

I=92d probably map it exactly the way it is now in RFC6455. The payload =
of an HTTP 2.0 frame with stream fin bit set would be the close code + =
reason.=

From tyoshino@google.com  Mon Jan 20 19:53:32 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 56C931A0296 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 19:53:32 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.913
X-Spam-Level: 
X-Spam-Status: No, score=-1.913 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id kSqwthmc_Bb8 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 19:53:30 -0800 (PST)
Received: from mail-wi0-x22a.google.com (mail-wi0-x22a.google.com [IPv6:2a00:1450:400c:c05::22a]) by ietfa.amsl.com (Postfix) with ESMTP id 2BE401A0295 for <hybi@ietf.org>; Mon, 20 Jan 2014 19:53:29 -0800 (PST)
Received: by mail-wi0-f170.google.com with SMTP id ex4so5085389wid.3 for <hybi@ietf.org>; Mon, 20 Jan 2014 19:53:29 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=+l0cTwjt9arIH2UZla48W/nm+kbbDqtms7BSRMuANbo=; b=hu1CX9u/pIH4ANvwqKgyv7krMrsjIhtAexIGmObQBPBKs9gE+Rfk7cnW2gFQ2Ix2Lf lhwk4L2kYJA+QhQiAebbLbKLIUBAuUMcMqmTMB63AvhOCyL7BvxxT9x5GoFuqZS3afFe j2Z5W2faeBYflvJmM4fra9vbF/TYjGjBv2j6mLEsQn73RaZ1HqlCfF86WUw873wtv7QE 1dKsrfgwOFk5S6T67n81iJZmXxpdMBoT+ymI63b0i+NUXz2JWluBp5HAKPzoszrwx8GT c9Nse1yEb3gH3MZVTAoHf5BdSRW6rBsWYXOA9jpsI3WShQ9R4WmZwMitXGB4EVQzQvaQ F2Zw==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=+l0cTwjt9arIH2UZla48W/nm+kbbDqtms7BSRMuANbo=; b=UJ0ylNyya1r0y6X6kVvvl/YkilIthPXJg2D5SXll0Wxj9NZmG8xPuxDzFQPK6vkmsm Hy0or54JOP0akf+A+dE97LmS9KoaNK4tTK4XhCSIvterXLYarUTqGbo5Dji/gMr6PX6m tHIpFf6+d+jmnvMxQ61gVd2q/F7TrnSl3rD0VO1EeTgdHSKEK8VMi9hQ3G9xSPBXYAz+ A/D1U6Hj/IlscBrKXW3gjYMi4ntkMrvmVREADDWINDBLikC0kTpc6KYo/Ak5USSVsB0u riDFVFR9HekOW05YMH6WPtC0TfuoyqLg12HuzYvuIeGabdLE57kLgdCRa/8z9TesDq91 gvaA==
X-Gm-Message-State: ALoCoQlxyv9QqfBP3hpuwaARqBHXF99If1kh6Q9ihGkQXFL1IAPbUQHmtOtBMzu6Qji+3F+vTKEP5hG7HKMje9Vu2iAPBWsa+Kj8nZTF4wdHLYFpamsAtHMvayhSDH+5sIo3TFK0bhedYzpFlv+jcI1lDzi/CNttbUhHDLsvcLirouyBD7wktfueBChfdQ/dMBv74FEQoHBs
X-Received: by 10.181.13.11 with SMTP id eu11mr12327342wid.30.1390276409670; Mon, 20 Jan 2014 19:53:29 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Mon, 20 Jan 2014 19:53:09 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BF99B57@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF9948F@EXVMBX020-12.exch020.serverdata.net> <52D90F99.6080205@gmx.de> <634914A010D0B943A035D226786325D4446BF9949D@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZQ3NxVW36PnZ0TpMF4tPPaJLr12M8NtbPb6pUejf1wcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99977@EXVMBX020-12.exch020.serverdata.net> <B8ED6F18-B710-44AE-829B-EDE5859B2C5B@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99ACB@EXVMBX020-12.exch020.serverdata.net> <82A1484B-1CDC-45BD-A2D2-6BED128F1448@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99B57@EXVMBX020-12.exch020.serverdata.net>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Tue, 21 Jan 2014 12:53:09 +0900
Message-ID: <CAH9hSJYhEh3ux1zoJOG7kz4dP6U=YhAJ2HsWnciKLDCXoZhRcQ@mail.gmail.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=f46d043c7de689df3e04f072f3e3
Cc: Julian Reschke <julian.reschke@gmx.de>, "hybi@ietf.org" <hybi@ietf.org>, Peter Thorson <webmaster@zaphoyd.com>
Subject: Re: [hybi] WebSocket subprotocol parameters
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 21 Jan 2014 03:53:32 -0000

--f46d043c7de689df3e04f072f3e3
Content-Type: text/plain; charset=UTF-8

On Tue, Jan 21, 2014 at 2:12 AM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> >What do you gain by negotiating max_param1 at the subprotocol level vs
> negotiating "foo" as your subprotocol and having the foo protocol define
> how to negotiate its own settings via regular websocket messages?
>
> If "param1" would influence how to parse raw WS payload into "foo"
> protocol, that would not be possible.
>
> But I agree in general: as long as "param1" does not influence the
> application protocol parsing, the app protocol could do (any further)
> parametrization/negotiation itself and no WS subprotocol level negotiation
> is necessary.
>

Yes, subprotocol negotiation can be continued while extension negotiation
needs to be completed in opening handshake before handing off the channel
for the app layer.

The only difference between them is latency. It takes 1 more RTT if param1
is not encoded into subprotocol (or url) but sent as a frame.

http://tools.ietf.org/html/rfc6455#section-4.1
>   Once the client's opening handshake has been sent, the client MUST
>   wait for a response from the server before sending any further data.
>   The client MUST validate the server's response as follows:

In practice, I'd recommend embedding offered subprotocol parameters into
subprotocol (which fits in token ABNF) or url if setup latency is
important. If you want to reply to coolproto.param1.10 with param1=8, you
could just acknowedge coolproto.param1.10 but send param1=8 in the first
WebSocket message.

--f46d043c7de689df3e04f072f3e3
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On T=
ue, Jan 21, 2014 at 2:12 AM, Tobias Oberstein <span dir=3D"ltr">&lt;<a href=
=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstein@=
tavendo.de</a>&gt;</span> wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-=
left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;p=
adding-left:1ex"><div class=3D"im">&gt;What do you gain by negotiating max_=
param1 at the subprotocol level vs negotiating &quot;foo&quot; as your subp=
rotocol and having the foo protocol define how to negotiate its own setting=
s via regular websocket messages?<br>


<br>
</div>If &quot;param1&quot; would influence how to parse raw WS payload int=
o &quot;foo&quot; protocol, that would not be possible.<br>
<br>
But I agree in general: as long as &quot;param1&quot; does not influence th=
e application protocol parsing, the app protocol could do (any further) par=
ametrization/negotiation itself and no WS subprotocol level negotiation is =
necessary.<br>

</blockquote><div><br></div><div>Yes, subprotocol negotiation can be contin=
ued while extension negotiation needs to be completed in opening handshake =
before handing off the channel for the app layer.</div><div><br></div>
<div>
The only difference between them is latency. It takes 1 more RTT if param1 =
is not encoded into subprotocol (or url) but sent as a frame.</div><div><br=
></div><div><a href=3D"http://tools.ietf.org/html/rfc6455#section-4.1">http=
://tools.ietf.org/html/rfc6455#section-4.1</a><br>

</div><div><div>&gt; =C2=A0 Once the client&#39;s opening handshake has bee=
n sent, the client MUST</div><div>&gt; =C2=A0 wait for a response from the =
server before sending any further data.</div><div>&gt; =C2=A0 The client MU=
ST validate the server&#39;s response as follows:</div>

</div><div><br></div><div>In practice, I&#39;d recommend embedding offered =
subprotocol parameters into subprotocol (which fits in token ABNF) or url i=
f setup latency is important. If you want to reply to coolproto.param1.10 w=
ith param1=3D8, you could just acknowedge coolproto.param1.10 but send para=
m1=3D8 in the first WebSocket message.</div>

</div></div></div>

--f46d043c7de689df3e04f072f3e3--

From tyoshino@google.com  Mon Jan 20 20:20:18 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 3B2AA1A0037 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 20:20:18 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.913
X-Spam-Level: 
X-Spam-Status: No, score=-1.913 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 3QqOomf5q017 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 20:20:16 -0800 (PST)
Received: from mail-wg0-x234.google.com (mail-wg0-x234.google.com [IPv6:2a00:1450:400c:c00::234]) by ietfa.amsl.com (Postfix) with ESMTP id 641FC1A002F for <hybi@ietf.org>; Mon, 20 Jan 2014 20:20:16 -0800 (PST)
Received: by mail-wg0-f52.google.com with SMTP id b13so7623520wgh.7 for <hybi@ietf.org>; Mon, 20 Jan 2014 20:20:16 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:from:date:message-id:subject:to:content-type; bh=FkHWDvVSccgMdsa93+JHVQdjCHRXfb4lnR2zrgcphaI=; b=ok3m0uEFn8XijMyve31Su72PUuUpRmahz1gMf6eskblRTBjp6507fl5NLt11S1HF86 84monrzKO+o6da+VAvEpmw9kn1/Bys3uzaDGTup8Sz8D0MeFb8jMAoNnVYY0P414fNgH lMY0Vn4TQ8cvbjoS0ybXY6kjIhAHYJrqOT5/1kQ9TPlnJJj1z4biz3sYw2kSvu+o8QfV svDNTgx2Uxkhxj1eT9VdUpe/3B7yQqKO/R5ZHlxhm2YuMAsFYa+Opezqn5e1WuzpfwOX Mtd7rdzNo7UAYmJs3LYBxl8mw4u4KpHJDHYg/LCqZVxU+mNPKEnOqvzZ1q9oqCN571jI w6vg==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:from:date:message-id:subject:to :content-type; bh=FkHWDvVSccgMdsa93+JHVQdjCHRXfb4lnR2zrgcphaI=; b=ezqop++y4IH1EbNe4/PrCxHfDszpmXUQ28YT9+a8FxXCuUrCV55k/7+D3yKQ5JpeZp bbZeJFQGQeIaKdJHTYcc+NtKD3VHfNP8R8QPCl1ag6+GMOsy8RZmar1qtYbf+wXLUawe 91bU5lybcXe+o1ziTTWyIbRIhvd3mfOME9WsmYCXr6aDy2bLSzLwctSxl4oQFxPjy1g0 wuN56xtKqdeZtMqguXF/iLqKYKBCuTtoJxbWKvXBrPetiz4SePZM+04ee39cpZ6Ye9+5 /7JWAYiynyJYuI6083iBi2B7tvqFGRSBY/U0HJIBIM1i6hIzv6aSbrFod3Huq7ss0szl 2Cog==
X-Gm-Message-State: ALoCoQkNRNIiG+NDF8Fk04Tl8jjoI9jRaAfG2tf4INP8ce9C3iQRFv3YpGd2f00kHV2pMfAjIQ8KR5K2Y5UDGW/VCWjYyQGkRNFcdriYaMdHD+fVZiGNIyLB58P4WAjzA/zEFLUsLA0l0cdUr2GsidNO8BqG3ztrELAv8+cJPmtamu9UBsYzWo7YpbYH2Q59Op/8nKqASv66
X-Received: by 10.180.10.105 with SMTP id h9mr11523931wib.11.1390278015852; Mon, 20 Jan 2014 20:20:15 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Mon, 20 Jan 2014 20:19:55 -0800 (PST)
From: Takeshi Yoshino <tyoshino@google.com>
Date: Tue, 21 Jan 2014 13:19:55 +0900
Message-ID: <CAH9hSJZ=1QLP0kNFHwKFsj0hpHmnfaEKJmE3sVa73Rcg73k-NA@mail.gmail.com>
To: "hybi@ietf.org" <hybi@ietf.org>
Content-Type: multipart/alternative; boundary=001a11c26366462f6704f073534f
Subject: [hybi] Consideration about CRIME like attacks against WebSocket
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 21 Jan 2014 04:20:18 -0000

--001a11c26366462f6704f073534f
Content-Type: text/plain; charset=UTF-8

As announced, Chrome 32 supports permessage-deflate.

We consider that the extension won't help leak any information such as
cookie string in the opening handshake since only payload data is subject
to compression, and so CRIME won't work for permessage-deflate.

However, as new CRIME like attacks such as BREACH were discovered in last
year, it's not impossible that new exploits based on length information
that affect WebSocket will be revealed. Please keep your eyes on findings
from security researches.

Re: possible BREACH variant for WebSocket, you just need to make sure you
don't respond with any sensitive info to WebSocket with an unknown Origin
value.

--001a11c26366462f6704f073534f
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">As announced, Chrome 32 supports permessage-deflate.<div><=
br></div><div>We consider that the extension won&#39;t help leak any inform=
ation such as cookie string in the opening handshake since only payload dat=
a is subject to compression, and so CRIME won&#39;t work for permessage-def=
late.</div>


<div><br></div><div>However, as new CRIME like attacks such as BREACH were =
discovered in last year, it&#39;s not impossible that new exploits based on=
 length information that affect WebSocket will be revealed. Please keep you=
r eyes on findings from security researches.</div>


<div><br></div><div>Re: possible BREACH variant for WebSocket, you just nee=
d to make sure you don&#39;t respond with any sensitive info to WebSocket w=
ith an unknown Origin value.</div></div>

--001a11c26366462f6704f073534f--

From tyoshino@google.com  Mon Jan 20 23:48:27 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 2F1A61A0054 for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 23:48:27 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.913
X-Spam-Level: 
X-Spam-Status: No, score=-1.913 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id K1YOhfWHH5eB for <hybi@ietfa.amsl.com>; Mon, 20 Jan 2014 23:48:25 -0800 (PST)
Received: from mail-we0-x22c.google.com (mail-we0-x22c.google.com [IPv6:2a00:1450:400c:c03::22c]) by ietfa.amsl.com (Postfix) with ESMTP id 961E61A0020 for <hybi@ietf.org>; Mon, 20 Jan 2014 23:48:25 -0800 (PST)
Received: by mail-we0-f172.google.com with SMTP id q58so7841421wes.17 for <hybi@ietf.org>; Mon, 20 Jan 2014 23:48:25 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=NKak44evUdnGw498hdTsEbwswDt3ivwbmrDlVRQ5Hck=; b=FLU/qQSMna7+XDVBzDRreyIgzTNQYubbrwUTilHpJ4JHGKPPwVxLC+SFC+nz/YevbW q/tf3gJ3jyXw0UjGF7tVFpxqYQaVSh4whyVt3DdLYS/Fpt1eD49Jm8JWJ9oKcq4TZmRS Yun83mv5a3bwUpYaro93huqV//giegRFTEk03TpC8umvcVa4SVWbwIIQIWhX8QMmmmTV Fx+uAEvF219BCiKrWCS65boWPIkku62qez08NLLnqbNP4M7nRoYExI8zfFlUdo/JMbFi rMDK2jUvspz8yKJc8LFj638gybDhf1ZGvojavTdo/UBfXuQMios+dDBQrnhW8Q6wxauY uxYQ==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=NKak44evUdnGw498hdTsEbwswDt3ivwbmrDlVRQ5Hck=; b=LVDqANvdbocGLKkJDn4hFUCKh8kqCxrEMt0Q34ySy9dvf4ldSo7dy5qLYIu3NvIuiU iMD68PKfKHNEn79r+YDqp0iXFuKQcRLbJrlUlvDrT96yu5Tuuv+BguhGndSt23aMuyAL gLIu6bRjxcFYHB+SOjrvmr9ImQjKfHQ0tBtdbkOFJj7bdSs6Yw6VahSyBTIqgNeWcjmI 0tUtorJ8hvhE0rmB41/XFmZMADpRYmFwOlyuLdmK/pYXrNnRUUwBE8JSeILdkg4sENTx TIEbW7//P+wL5ue/t2ez5xbdybt9+kokfSiBbKbQaTnvnNzbZYvdUWcrebFXVPkMskKR WhJw==
X-Gm-Message-State: ALoCoQkFU5ueKclJZ7m4u9/Dy02omzZCH5j8lD0oLG4Dvv1yyz5mFOYz+BPZqBFNDUGjTMyWIF849suq8QtPvGbAGWOV83E3wj30i2UIfmxFsgQ33JAwm0ZSJY1fvBHgmoSuiFYKl69YPBx/DRK4JD+d+VRPzw6AQeGaMtrN8YwGrX8jGN49iC3VNEMbYC6UoEr2d9H4u0mo
X-Received: by 10.180.10.105 with SMTP id h9mr12127996wib.11.1390290505033; Mon, 20 Jan 2014 23:48:25 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Mon, 20 Jan 2014 23:48:04 -0800 (PST)
In-Reply-To: <003c01cf0ddf$985d4070$c917c150$@noemax.com>
References: <003c01cf0ddf$985d4070$c917c150$@noemax.com>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Tue, 21 Jan 2014 16:48:04 +0900
Message-ID: <CAH9hSJZQ6_wo8urUTqsi6ofEo-uAi2_Pkejrg+jUWQzV+ujOaw@mail.gmail.com>
To: Arman Djusupov <arman@noemax.com>
Content-Type: multipart/alternative; boundary=001a11c26366b00e6204f0763b48
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Permessage-deflate final frame
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 21 Jan 2014 07:48:27 -0000

--001a11c26366b00e6204f0763b48
Content-Type: text/plain; charset=UTF-8

I posted a reply to your question here.
http://www.ietf.org/mail-archive/web/hybi/current/msg10286.html
Just in case. Thanks.

--001a11c26366b00e6204f0763b48
Content-Type: text/html; charset=UTF-8

<div dir="ltr"><div>I posted a reply to your question here.</div><a href="http://www.ietf.org/mail-archive/web/hybi/current/msg10286.html" target="_blank">http://www.ietf.org/mail-archive/web/hybi/current/msg10286.html</a><br>

<div>Just in case. Thanks.</div></div>

--001a11c26366b00e6204f0763b48--

From duerst@it.aoyama.ac.jp  Tue Jan 21 01:17:43 2014
Return-Path: <duerst@it.aoyama.ac.jp>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 6674B1A00A1 for <hybi@ietfa.amsl.com>; Tue, 21 Jan 2014 01:17:43 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.926
X-Spam-Level: 
X-Spam-Status: No, score=-1.926 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HELO_EQ_JP=1.244, HOST_EQ_JP=1.265, MIME_8BIT_HEADER=0.3, RCVD_IN_DNSWL_MED=-2.3, RP_MATCHES_RCVD=-0.535] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id d1jeeLY8NJbC for <hybi@ietfa.amsl.com>; Tue, 21 Jan 2014 01:17:41 -0800 (PST)
Received: from scintmta02.scbb.aoyama.ac.jp (scintmta02.scbb.aoyama.ac.jp [133.2.253.34]) by ietfa.amsl.com (Postfix) with ESMTP id 0B65F1A00A2 for <hybi@ietf.org>; Tue, 21 Jan 2014 01:17:39 -0800 (PST)
Received: from scmse02.scbb.aoyama.ac.jp ([133.2.253.231]) by scintmta02.scbb.aoyama.ac.jp (secret/secret) with SMTP id s0L9HSS2008242; Tue, 21 Jan 2014 18:17:29 +0900
Received: from (unknown [133.2.206.134]) by scmse02.scbb.aoyama.ac.jp with smtp id 6b24_316e_d9756f6c_827c_11e3_a1b5_001e6722eec2; Tue, 21 Jan 2014 18:17:28 +0900
Received: from [IPv6:::1] (unknown [133.2.210.1]) by itmail2.it.aoyama.ac.jp (Postfix) with ESMTP id E70F2BF4E2; Tue, 21 Jan 2014 18:17:27 +0900 (JST)
Message-ID: <52DE3B1B.1010804@it.aoyama.ac.jp>
Date: Tue, 21 Jan 2014 18:17:15 +0900
From: =?UTF-8?B?Ik1hcnRpbiBKLiBEw7xyc3Qi?= <duerst@it.aoyama.ac.jp>
Organization: Aoyama Gakuin University
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.1.9) Gecko/20100722 Eudora/3.0.4
MIME-Version: 1.0
To: Peter Thorson <webmaster@zaphoyd.com>
References: <634914A010D0B943A035D226786325D4446BF9948F@EXVMBX020-12.exch020.serverdata.net> <52D90F99.6080205@gmx.de> <634914A010D0B943A035D226786325D4446BF9949D@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZQ3NxVW36PnZ0TpMF4tPPaJLr12M8NtbPb6pUejf1wcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99977@EXVMBX020-12.exch020.serverdata.net> <B8ED6F18-B710-44AE-829B-EDE5859B2C5B@zaphoyd.com>
In-Reply-To: <B8ED6F18-B710-44AE-829B-EDE5859B2C5B@zaphoyd.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: quoted-printable
Cc: Julian Reschke <julian.reschke@gmx.de>, "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] WebSocket subprotocol parameters
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 21 Jan 2014 09:17:43 -0000

On 2014/01/21 0:58, Peter Thorson wrote:
> My impression was that subprotocol primarily changed the interpretation=
 of messages payloads by the end applications themselves. WebSocket messa=
ges allow essentially arbitrary data so you can define whatever parameter=
 setting mechanism you like with whatever grammar makes sense for your su=
bprotocol and send the parameters in the opening =E2=80=9Chello=E2=80=9D =
message. Defining a grammar in the socket spec to allow application proto=
cols to configure themselves seems limiting and out of scope.

Yes indeed. It has always been my impression that even subprotocols=20
themselves are rather unnecessary. On the server side, just use a=20
different URI/IRI as endpoint. On the client side, it's all JS anyway.

Regards,   Martin.

From tobias.oberstein@tavendo.de  Tue Jan 21 02:12:19 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id EEF2B1A009E for <hybi@ietfa.amsl.com>; Tue, 21 Jan 2014 02:12:18 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.4
X-Spam-Level: 
X-Spam-Status: No, score=-0.4 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, J_CHICKENPOX_34=0.6, J_CHICKENPOX_37=0.6, MIME_8BIT_HEADER=0.3, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id AY-lPpV_OQu7 for <hybi@ietfa.amsl.com>; Tue, 21 Jan 2014 02:12:18 -0800 (PST)
Received: from EXHUB020-3.exch020.serverdata.net (exhub020-3.exch020.serverdata.net [206.225.164.30]) by ietfa.amsl.com (Postfix) with ESMTP id C9F991A009D for <hybi@ietf.org>; Tue, 21 Jan 2014 02:12:17 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-3.exch020.serverdata.net ([206.225.164.30]) with mapi; Tue, 21 Jan 2014 02:12:17 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: =?utf-8?B?Ik1hcnRpbiBKLiBEw7xyc3Qi?= <duerst@it.aoyama.ac.jp>, Peter Thorson <webmaster@zaphoyd.com>
Date: Tue, 21 Jan 2014 02:12:14 -0800
Thread-Topic: [hybi] WebSocket subprotocol parameters
Thread-Index: Ac8WiadLG+rI7ZtES6iZWjxzi55plQAAS0rw
Message-ID: <634914A010D0B943A035D226786325D4446C082A0D@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF9948F@EXVMBX020-12.exch020.serverdata.net> <52D90F99.6080205@gmx.de> <634914A010D0B943A035D226786325D4446BF9949D@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZQ3NxVW36PnZ0TpMF4tPPaJLr12M8NtbPb6pUejf1wcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99977@EXVMBX020-12.exch020.serverdata.net> <B8ED6F18-B710-44AE-829B-EDE5859B2C5B@zaphoyd.com> <52DE3B1B.1010804@it.aoyama.ac.jp>
In-Reply-To: <52DE3B1B.1010804@it.aoyama.ac.jp>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: Julian Reschke <julian.reschke@gmx.de>, "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] WebSocket subprotocol parameters
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 21 Jan 2014 10:12:19 -0000
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From arman@noemax.com  Tue Jan 21 04:44:13 2014
Return-Path: <arman@noemax.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 4BA3D1A00D6 for <hybi@ietfa.amsl.com>; Tue, 21 Jan 2014 04:44:13 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.636
X-Spam-Level: 
X-Spam-Status: No, score=-0.636 tagged_above=-999 required=5 tests=[BAYES_20=-0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 8srLpSscanDn for <hybi@ietfa.amsl.com>; Tue, 21 Jan 2014 04:44:11 -0800 (PST)
Received: from mail.noemax.com (mail.noemax.com [74.208.113.37]) by ietfa.amsl.com (Postfix) with ESMTP id 07FC41A00C7 for <hybi@ietf.org>; Tue, 21 Jan 2014 04:44:10 -0800 (PST)
DKIM-Signature: a=rsa-sha1; t=1390308205; x=1390913005; s=m2048; d=noemax.com; c=relaxed/relaxed; v=1; bh=oLWXSb+JWWwwhH1aG7lyABkDmXg=; h=From:Subject:Date:Message-ID:To:Cc:MIME-Version:Content-Type:In-Reply-To:References; b=KMyIbtuptXK+fkN30YXmhCzVu7jyQBkttRBkKMMC5fyKKc7bYLpoNOHA5gSNCzL5ZnPBPge60XBwfwZ5r9zvQfsXJPZDk+9kACE9hnJEznlSsRH3UC2jO+vujuOXQwPM1ZVYmX+IAIqMjdnjkx1Jgbp64Ucyp8u9mfAiSeWvfaqwJ4xJdY/MRVWgR7H46Aad1qOcaZI0wc4a/MH8MWBeKstu5RN+YchJRMbc5w6RSs/9M8wWTDB0fj/jql5mgEBQ0iyfJtAgRl7sDUhSS34svr4Uuruj/NfJh3OwK0/rg+KnIOwBd+umhqLe94JhUXwpbFa5CAYfVafliP0jk3YU2g==
Received: from mail.noemax.com by mail.noemax.com (Noemax Mail Server) with ASMTP (SSL) id 201401211443237331; Tue, 21 Jan 2014 14:43:23 +0200
From: "Arman Djusupov" <arman@noemax.com>
To: "'Takeshi Yoshino'" <tyoshino@google.com>
References: <CAH9hSJbeY5VOY_iuwrdBq-KYcoVkW_8KArPp70hP4tdZj6eQfg@mail.gmail.com>
In-Reply-To: <CAH9hSJbeY5VOY_iuwrdBq-KYcoVkW_8KArPp70hP4tdZj6eQfg@mail.gmail.com>
Date: Tue, 21 Jan 2014 14:44:05 +0200
Message-ID: <006501cf16a6$7a94d470$6fbe7d50$@noemax.com>
MIME-Version: 1.0
Content-Type: multipart/alternative; boundary="----=_NextPart_000_0066_01CF16B7.3E201570"
X-Mailer: Microsoft Outlook 14.0
Thread-Index: AQG5HshGVuJS4QKFNVB3Zyo3aQvC9Zq7DuAg
Content-Language: en-us
Cc: hybi@ietf.org
Subject: Re: [hybi] [permessage-deflate] Compressing fragmented data
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 21 Jan 2014 12:44:13 -0000

This is a multipart message in MIME format.

------=_NextPart_000_0066_01CF16B7.3E201570
Content-Type: text/plain;
	charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Hello Takeshi,

=20

If the empty-block-in-final-frame solution is the preferred one then I =
think it would be useful to add an example of such an empty block for =
the cases when there is no compressed data to be sent in the final frame =
due to the compression stream having been already flushed and byte =
aligned. DEFLATE implementations like Zlib do not normally produce an =
empty block encoding when there is no data to flush, so developers will =
need to have some understanding of the DEFLATE format in order to =
produce this block artificially. This can result in mistakes. Since it =
is rare to encounter cases with data aligned in such a manner that there =
is nothing to send in the final frame, such mistakes might not be =
detected during development/testing but might leak in production and =
result in the occasional communication failures. So IMHO adding an =
example would be useful.

=20

Thanks!

=20

With best regards,

Arman

=20

From: Takeshi Yoshino [mailto:tyoshino@google.com]=20
Sent: Tuesday, January 14, 2014 6:05 AM
To: Arman Djusupov
Cc: hybi@ietf.org
Subject: [permessage-deflate] Compressing fragmented data

=20

Hi Arman,

=20

Let me change the subject.

=20

On Thu, Jan 9, 2014 at 11:46 PM, Arman Djusupov <arman@noemax.com> =
wrote:

Hello Takeshi,

=20

In the current draft, the description of the compression process states =
:

=20
   1.  Compress all the octets of the payload of the message using
       DEFLATE.

=20

This doesn=E2=80=99t take into account outbound messages of arbitrary =
size that NEED to be fragmented, for which it is more favorable to use =
fragmentation and to compress & send fragments until the end of the =
message is reached.

=20

OK. I'll add some text to explain how this works for fragmented =
messages.

=20

=20

In such cases the implementation buffers the input from its source and =
then flushes it out into a compressed fragment, repeating this process =
until the end of the source data is reached. At that point it flushes =
any remaining bytes buffered into a final frame, or flushes a final =
frame of 0 length.

=20

=20

When the implementation produces compressed fragments, it periodically =
produces frames with 0x00 0x00 0xFF 0xFF at the end of the frame due to =
flushing DEFLATE. But because of the following requirements:

=20

   2. If the resulting data does not end with an empty DEFLATE block

       with no compression (the "BTYPE" bits is set to 00), append an

       empty DEFLATE block with no compression to the tail end.

=20

   3.  Remove 4 octets (that are 0x00 0x00 0xff 0xff) from the tail end.

       After this step, the last octet of the compressed data contains

       (possibly part of) the DEFLATE header bits with the "BTYPE" bits

       set to 00.

=20

the implementation must ensure that the message ends empty blocks =
without a trailing 0x0000FFFF, so the implementation must keep track on =
what type of block was at the end of the frame that was sent last. If =
the last frame sent ends with 0x0000FFFF then the implementation cannot =
remove those 4 bytes from the wire, but it must artificially produce an =
empty DEFLATE block to send as the final frame. This is only due to the =
requirement to remove 0x0000FFFF.

=20

Yes. It's just two byte long. DEFLATE block header (3bit) + Fixed =
Huffman code for end of block symbol (7bit) + padding (6bit).

=20

=20

=20

Wouldn=E2=80=99t it be easier to make the removal 0x00 0x00 0xFF 0xFF =
optional, while at the same time requiring that the receiving side =
appends a 0x00 0x00 0xFF 0xFF to the final frame in case when its =
missing?

=20

=20

I think it complicates decoders. Code which doesn't know DEFLATE =
decompressor's state cannot check if 0x00 0x00 0xFF 0xFF at the end is =
really an uncompressed block body or not. It could be something else, =
e.g. body of compressed block.

=20

It is much simpler to just check the final bytes of the message instead =
of having to remember the state of the previous frame.


------=_NextPart_000_0066_01CF16B7.3E201570
Content-Type: text/html;
	charset="UTF-8"
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:m=3D"http://schemas.microsoft.com/office/2004/12/omml" =
xmlns=3D"http://www.w3.org/TR/REC-html40"><head><meta =
http-equiv=3DContent-Type content=3D"text/html; charset=3Dutf-8"><meta =
name=3DGenerator content=3D"Microsoft Word 14 (filtered =
medium)"><style><!--
/* Font Definitions */
@font-face
	{font-family:"Cambria Math";
	panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
	{font-family:Calibri;
	panose-1:2 15 5 2 2 2 4 3 2 4;}
@font-face
	{font-family:Tahoma;
	panose-1:2 11 6 4 3 5 4 4 2 4;}
@font-face
	{font-family:Consolas;
	panose-1:2 11 6 9 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0in;
	margin-bottom:.0001pt;
	font-size:12.0pt;
	font-family:"Times New Roman","serif";}
a:link, span.MsoHyperlink
	{mso-style-priority:99;
	color:blue;
	text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
	{mso-style-priority:99;
	color:purple;
	text-decoration:underline;}
pre
	{mso-style-priority:99;
	mso-style-link:"HTML Preformatted Char";
	margin:0in;
	margin-bottom:.0001pt;
	font-size:10.0pt;
	font-family:"Courier New";}
span.HTMLPreformattedChar
	{mso-style-name:"HTML Preformatted Char";
	mso-style-priority:99;
	mso-style-link:"HTML Preformatted";
	font-family:Consolas;
	mso-ligatures:none;}
span.EmailStyle19
	{mso-style-type:personal-reply;
	font-family:"Calibri","sans-serif";
	color:#44546A;}
.MsoChpDefault
	{mso-style-type:export-only;
	font-size:10.0pt;}
@page WordSection1
	{size:8.5in 11.0in;
	margin:1.0in 1.0in 1.0in 1.0in;}
div.WordSection1
	{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext=3D"edit" spidmax=3D"1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext=3D"edit">
<o:idmap v:ext=3D"edit" data=3D"1" />
</o:shapelayout></xml><![endif]--></head><body lang=3DEN-US link=3Dblue =
vlink=3Dpurple><div class=3DWordSection1><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>Hello Takeshi,<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>If the empty-block-in-final-frame solution is the preferred one then =
I think it would be useful to add an example of such an empty block for =
the cases when there is no compressed data to be sent in the final frame =
due to the compression stream having been already flushed and byte =
aligned. DEFLATE implementations like Zlib do not normally produce an =
empty block encoding when there is no data to flush, so developers will =
need to have some understanding of the DEFLATE format in order to =
produce this block artificially. This can result in mistakes. Since it =
is rare to encounter cases with data aligned in such a manner that there =
is nothing to send in the final frame, such mistakes might not be =
detected during development/testing but might leak in production and =
result in the occasional communication failures. So IMHO adding an =
example would be useful.<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>Thanks!<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>With best regards,<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>Arman<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><b><span =
style=3D'font-size:10.0pt;font-family:"Tahoma","sans-serif"'>From:</span>=
</b><span style=3D'font-size:10.0pt;font-family:"Tahoma","sans-serif"'> =
Takeshi Yoshino [mailto:tyoshino@google.com] <br><b>Sent:</b> Tuesday, =
January 14, 2014 6:05 AM<br><b>To:</b> Arman Djusupov<br><b>Cc:</b> =
hybi@ietf.org<br><b>Subject:</b> [permessage-deflate] Compressing =
fragmented data<o:p></o:p></span></p><p =
class=3DMsoNormal><o:p>&nbsp;</o:p></p><div><div><div><p =
class=3DMsoNormal>Hi Arman,<o:p></o:p></p></div><div><p =
class=3DMsoNormal><o:p>&nbsp;</o:p></p></div><div><p =
class=3DMsoNormal>Let me change the subject.<o:p></o:p></p></div><div><p =
class=3DMsoNormal><o:p>&nbsp;</o:p></p></div><div><p =
class=3DMsoNormal>On Thu, Jan 9, 2014 at 11:46 PM, Arman Djusupov &lt;<a =
href=3D"mailto:arman@noemax.com" =
target=3D"_blank">arman@noemax.com</a>&gt; =
wrote:<o:p></o:p></p><div><div><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>Hello Takeshi,</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>&nbsp;</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>In the current draft, t</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>he description of </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>compression process states =
:</span><o:p></o:p></p><pre>&nbsp;<o:p></o:p></pre><pre>&nbsp;&nbsp; =
1.&nbsp; Compress all the octets of the payload of the message =
using<o:p></o:p></pre><pre>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
DEFLATE.<o:p></o:p></pre><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>&nbsp;</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>T</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>his doesn=E2=80=99t</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'> take</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> into account </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>outbound </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>messages of arbitrary size that NEED to be fragmented, </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>for which it is</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> more favorable to use fragmentation and </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>to </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>compress</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'> &amp; send</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> fragments until the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>end of the message is reached</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>.</span><o:p></o:p></p></div></div><div><p =
class=3DMsoNormal><o:p>&nbsp;</o:p></p></div><div><p =
class=3DMsoNormal>OK. I'll add some text to explain how this works for =
fragmented messages.<o:p></o:p></p></div><div><p =
class=3DMsoNormal>&nbsp;<o:p></o:p></p></div><blockquote =
style=3D'border:none;border-left:solid #CCCCCC 1.0pt;padding:0in 0in 0in =
6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0in;margin-bottom:5=
.0pt'><div><div><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>&nbsp;</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>In such cases the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>implementation buffers </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the in</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>put </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>from its source </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>and then flushes it out into a compressed fragment</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>, repeating this process</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> until the end of the source data </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>is </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>reached</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>. At that point </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>it </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>flushes any remaining bytes buffered into a final frame, or =
</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>flushes </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>a</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> final frame of 0 length.</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>&nbsp;</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>&nbsp;</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>When </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>implementation produces compressed </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>fragments,</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> it periodically produces frames with 0x00 0x00 0xFF 0xFF at the end =
of the frame</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'> due to flushing DEFLATE</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>. But </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>because of the</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> following requirements:</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>&nbsp;</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:10.0pt;font-family:"Courier New"'>&nbsp;&nbsp; 2. If =
the resulting data does not end with an empty DEFLATE =
block</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:10.0pt;font-family:"Courier =
New"'>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; with no compression (the =
&quot;BTYPE&quot; bits is set to 00), append an</span><o:p></o:p></p><p =
class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:10.0pt;font-family:"Courier =
New"'>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; empty DEFLATE block with no =
compression to the tail end.</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:10.0pt;font-family:"Courier =
New"'>&nbsp;</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:10.0pt;font-family:"Courier New"'>&nbsp;&nbsp; =
3.&nbsp; Remove 4 octets (that are 0x00 0x00 0xff 0xff) from the tail =
end.</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:10.0pt;font-family:"Courier =
New"'>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; After this step, the last =
octet of the compressed data contains</span><o:p></o:p></p><p =
class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:10.0pt;font-family:"Courier =
New"'>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; (possibly part of) the =
DEFLATE header bits with the &quot;BTYPE&quot; =
bits</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:10.0pt;font-family:"Courier =
New"'>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; set to =
00.</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>&nbsp;</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>implementation must </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>e</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>nsure that the message ends empty block</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>s</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> without </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>a trailing</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>&nbsp;0x0000FFFF, so </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>implementation must keep track on what type of block was at the end =
of the frame that was sent last. If the last frame sent ends with =
0x0000FFFF</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'> then</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>implementation cannot remove those 4 bytes from the wire</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>, but </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>it must artificially produce an empty DEFLATE block to send =
</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>as the</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> final frame. </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>This is only due to the requirement to remove </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>0x0000FFFF.</span><o:p></o:p></p></div></div></blockquote><div><p =
class=3DMsoNormal><o:p>&nbsp;</o:p></p></div><div><p =
class=3DMsoNormal>Yes. It's just two byte long. DEFLATE block header =
(3bit) + Fixed Huffman code for end of block symbol (7bit) + padding =
(6bit).<o:p></o:p></p></div><div><p =
class=3DMsoNormal>&nbsp;<o:p></o:p></p></div><blockquote =
style=3D'border:none;border-left:solid #CCCCCC 1.0pt;padding:0in 0in 0in =
6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0in;margin-bottom:5=
.0pt'><div><div><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'>&nbsp;<o:p><=
/o:p></p></div></div></blockquote><blockquote =
style=3D'border:none;border-left:solid #CCCCCC 1.0pt;padding:0in 0in 0in =
6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0in;margin-bottom:5=
.0pt'><div><div><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>&nbsp;</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>Wouldn=E2=80=99t it be easier to make </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>remov</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>al</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> 0x00 0x00 0xFF 0xFF optional</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>, while at the same time requiring that the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>receiving side a</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>ppends a</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> 0x00 0x00 0xFF 0xFF </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>to</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> the final frame in case when </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>its</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> missing</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>?</span><o:p></o:p></p><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>&nbsp;</span><o:p></o:p></p></div></div></blockquote><div><p =
class=3DMsoNormal><o:p>&nbsp;</o:p></p></div><div><p class=3DMsoNormal>I =
think it complicates decoders. Code which doesn't know DEFLATE =
decompressor's state cannot check if 0x00 0x00 0xFF 0xFF at the end is =
really an uncompressed block body or not. It could be something else, =
e.g. body of compressed block.<o:p></o:p></p></div><div><p =
class=3DMsoNormal>&nbsp;<o:p></o:p></p></div><blockquote =
style=3D'border:none;border-left:solid #CCCCCC 1.0pt;padding:0in 0in 0in =
6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0in;margin-bottom:5=
.0pt'><div><p class=3DMsoNormal =
style=3D'mso-margin-top-alt:auto;mso-margin-bottom-alt:auto'><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>It </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>is </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>much simpler to </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>just</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> check </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>final bytes of the message </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>instead of</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'> </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>having to </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>remember the state of </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>the </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>previous =
frame.</span><o:p></o:p></p></div></blockquote></div></div></div></div></=
body></html>
------=_NextPart_000_0066_01CF16B7.3E201570--


From tyoshino@google.com  Tue Jan 21 19:38:55 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 643C31A0125 for <hybi@ietfa.amsl.com>; Tue, 21 Jan 2014 19:38:55 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.913
X-Spam-Level: 
X-Spam-Status: No, score=-1.913 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id u5_IckSA-nkF for <hybi@ietfa.amsl.com>; Tue, 21 Jan 2014 19:38:53 -0800 (PST)
Received: from mail-wi0-x235.google.com (mail-wi0-x235.google.com [IPv6:2a00:1450:400c:c05::235]) by ietfa.amsl.com (Postfix) with ESMTP id EFD291A01BF for <hybi@ietf.org>; Tue, 21 Jan 2014 19:38:52 -0800 (PST)
Received: by mail-wi0-f181.google.com with SMTP id hi8so151859wib.14 for <hybi@ietf.org>; Tue, 21 Jan 2014 19:38:52 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=dM6euHqdmeptYDd8X/E+tq8fRwZMSlHhEQmAN/BVs1I=; b=eb9KP21g92CLoHxM3uHvOnzoQLpX/33/Fs5t4FwH2wag6r3I4ARCtU2lJ92BrIovmc oGCsGy/4JH2Icqm6AQ2u047o5ZlrUO1dP01XSDtdf9lRyGFbTKMHKXrPvwmXIu+PdqKP Q6cimvTjqLx6foOvPCmfRdJPk/1AOML/iD4KVSHtPHHKJ4Kxk9Y22zPbF1fgZ5xjN6zc ImdEJdrVvPPujNyewK1KvSZrqNpClIwjDG/ehPeX0svxRTy4lzqHHzOxtaxQGrfbJ5RK 0HGha9vtweQuahb6HIk3D5ggwDhxSH5jhQIdAuB/5a9TTmdXrd0vA8buWKK6lENUgm6N ls1w==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=dM6euHqdmeptYDd8X/E+tq8fRwZMSlHhEQmAN/BVs1I=; b=DYXbqrn2xepnI6RbqKaHHEQXCQ/ImiBnr9s/25ywmXsgHVtpxCd0JLAutSroH/9uRu yDUJXDPUDv1Old9s15J1OIcrstp9t2xkcLcY0grMoBmPvVJrS+2iGApcm4yPCmNYpwAb jdhyPIqjyzYqGgzWmYFaR0QGe2e5N7SKoGX13u0OujwJj6u5muscAoMrjP+SEmdHD+7w qegrzXR+uFJdO0ig9NFc/AP7ycLzvJV4OGSSNGps07EOW35VKdb4mWTipdlAk3Plzdsg BYzDveQqhSmssKoVn8egMfDcwDIgXydLJB/9GwZbI9L84Vtjxwo/Dax9J7N/fst7HfPz 71nA==
X-Gm-Message-State: ALoCoQnyA9TJthoL7da3D5Im2ha7rChddRG+1342hPFFiBtWmCLy8tBwEtwTFhdpvlRjCjS1OxrBaU8rvHQnnb4aMXHXrYEgV9Idch1sKyGDQJR4WOvvyJK6sSQKvZjxCsFLOOwS5cYcikxslHle4UV/U5uDWCToCGr0OToXdFLu8wePo6QRU/nGx5++W3YS+7cryWS8j/K+
X-Received: by 10.180.188.197 with SMTP id gc5mr780976wic.30.1390361932079; Tue, 21 Jan 2014 19:38:52 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Tue, 21 Jan 2014 19:38:31 -0800 (PST)
In-Reply-To: <006501cf16a6$7a94d470$6fbe7d50$@noemax.com>
References: <CAH9hSJbeY5VOY_iuwrdBq-KYcoVkW_8KArPp70hP4tdZj6eQfg@mail.gmail.com> <006501cf16a6$7a94d470$6fbe7d50$@noemax.com>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Wed, 22 Jan 2014 12:38:31 +0900
Message-ID: <CAH9hSJZT-wuNVcir9VtQUWeWtCUP-Wz4AVA_26quPjmyue09gQ@mail.gmail.com>
To: Arman Djusupov <arman@noemax.com>
Content-Type: multipart/alternative; boundary=001a11c3841a12452f04f086dd46
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] [permessage-deflate] Compressing fragmented data
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 22 Jan 2014 03:38:55 -0000

--001a11c3841a12452f04f086dd46
Content-Type: text/plain; charset=UTF-8

On Tue, Jan 21, 2014 at 9:44 PM, Arman Djusupov <arman@noemax.com> wrote:

> Hello Takeshi,
>
>
>
> If the empty-block-in-final-frame solution is the preferred one then I
> think it would be useful to add an example of such an empty block for the
> cases when there is no compressed data to be sent in the final frame due to
> the compression stream having been already flushed and byte aligned.
> DEFLATE implementations like Zlib do not normally produce an empty block
> encoding when there is no data to flush, so developers will need to have
> some understanding of the DEFLATE format in order to produce this block
> artificially.
>

Good point!


>  This can result in mistakes. Since it is rare to encounter cases with
> data aligned in such a manner that there is nothing to send in the final
> frame, such mistakes might not be detected during development/testing but
> might leak in production and result in the occasional communication
> failures. So IMHO adding an example would be useful.
>
>
Will add. Thanks.


> When the implementation produces compressed fragments, it periodically
> produces frames with 0x00 0x00 0xFF 0xFF at the end of the frame due to
> flushing DEFLATE. But because of the following requirements:
>
>
>
>    2. If the resulting data does not end with an empty DEFLATE block
>
>        with no compression (the "BTYPE" bits is set to 00), append an
>
>        empty DEFLATE block with no compression to the tail end.
>
>
>
>    3.  Remove 4 octets (that are 0x00 0x00 0xff 0xff) from the tail end.
>
>        After this step, the last octet of the compressed data contains
>
>        (possibly part of) the DEFLATE header bits with the "BTYPE" bits
>
>        set to 00.
>
>
>
> the implementation must ensure that the message ends empty blocks without a
> trailing 0x0000FFFF, so the implementation must keep track on what type
> of block was at the end of the frame that was sent last. If the last frame
> sent ends with 0x0000FFFF then the implementation cannot remove those 4
> bytes from the wire, but it must artificially produce an empty DEFLATE
> block to send as the final frame. This is only due to the requirement to
> remove 0x0000FFFF.
>
>
>
> Yes. It's just two byte long. DEFLATE block header (3bit) + Fixed Huffman
> code for end of block symbol (7bit) + padding (6bit).
>

Sorry. This comment is (partially) wrong. An artificial uncompressed
DEFLATE block must be used instead. It'll be 1 octet long. Please check a
new example in new draft when published.

Thanks

--001a11c3841a12452f04f086dd46
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On T=
ue, Jan 21, 2014 at 9:44 PM, Arman Djusupov <span dir=3D"ltr">&lt;<a href=
=3D"mailto:arman@noemax.com" target=3D"_blank">arman@noemax.com</a>&gt;</sp=
an> wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div lang=3D"EN-US" link=3D"blue" vlink=3D"p=
urple"><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&=
quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a">Hello Takeshi,<u><=
/u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#44546a"><u></u>=C2=A0<u></u></spa=
n></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a">If the empty-block-=
in-final-frame solution is the preferred one then I think it would be usefu=
l to add an example of such an empty block for the cases when there is no c=
ompressed data to be sent in the final frame due to the compression stream =
having been already flushed and byte aligned. DEFLATE implementations like =
Zlib do not normally produce an empty block encoding when there is no data =
to flush, so developers will need to have some understanding of the DEFLATE=
 format in order to produce this block artificially.</span></p>

</div></blockquote><div><br></div><div>Good point!</div><div>=C2=A0</div><b=
lockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px =
#ccc solid;padding-left:1ex"><div lang=3D"EN-US" link=3D"blue" vlink=3D"pur=
ple"><p class=3D"MsoNormal">

<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#44546a"> This can result in mistakes. Since it is rare t=
o encounter cases with data aligned in such a manner that there is nothing =
to send in the final frame, such mistakes might not be detected during deve=
lopment/testing but might leak in production and result in the occasional c=
ommunication failures. So IMHO adding an example would be useful.<u></u><u>=
</u></span></p>

<div><div class=3D"h5"><div><div><blockquote style=3D"border:none;border-le=
ft:solid #cccccc 1.0pt;padding:0in 0in 0in 6.0pt;margin-left:4.8pt;margin-t=
op:5.0pt;margin-right:0in;margin-bottom:5.0pt"><div><div><p class=3D"MsoNor=
mal">

<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#44546a"></span></p></div></div></blockquote></div></div>=
</div></div></div></blockquote><div><br></div><div>Will add. Thanks.</div>

<div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8=
ex;border-left:1px #ccc solid;padding-left:1ex"><div lang=3D"EN-US" link=3D=
"blue" vlink=3D"purple"><div><div class=3D"h5"><div><div><blockquote style=
=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0in 0in 0in 6.0pt;m=
argin-left:4.8pt;margin-top:5.0pt;margin-right:0in;margin-bottom:5.0pt">

<div><div><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-famil=
y:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a">When </span><sp=
an style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-ser=
if&quot;;color:#1f497d">the </span><span style=3D"font-size:11.0pt;font-fam=
ily:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a">implementatio=
n produces compressed </span><span style=3D"font-size:11.0pt;font-family:&q=
uot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">fragments,</span><s=
pan style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-se=
rif&quot;;color:#44546a"> it periodically produces frames with 0x00 0x00 0x=
FF 0xFF at the end of the frame</span><span style=3D"font-size:11.0pt;font-=
family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"> due to fl=
ushing DEFLATE</span><span style=3D"font-size:11.0pt;font-family:&quot;Cali=
bri&quot;,&quot;sans-serif&quot;;color:#44546a">. But </span><span style=3D=
"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;co=
lor:#1f497d">because of the</span><span style=3D"font-size:11.0pt;font-fami=
ly:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a"> following req=
uirements:</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#44546a">=C2=A0</span><u></u><u></=
u></p><p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&q=
uot;Courier New&quot;">=C2=A0=C2=A0 2. If the resulting data does not end w=
ith an empty DEFLATE block</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Co=
urier New&quot;">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 with no compression (=
the &quot;BTYPE&quot; bits is set to 00), append an</span><u></u><u></u></p=
><p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;C=
ourier New&quot;">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 empty DEFLATE block =
with no compression to the tail end.</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Co=
urier New&quot;">=C2=A0</span><u></u><u></u></p><p class=3D"MsoNormal"><spa=
n style=3D"font-size:10.0pt;font-family:&quot;Courier New&quot;">=C2=A0=C2=
=A0 3.=C2=A0 Remove 4 octets (that are 0x00 0x00 0xff 0xff) from the tail e=
nd.</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Co=
urier New&quot;">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 After this step, the =
last octet of the compressed data contains</span><u></u><u></u></p><p class=
=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Courier Ne=
w&quot;">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 (possibly part of) the DEFLAT=
E header bits with the &quot;BTYPE&quot; bits</span><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Co=
urier New&quot;">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 set to 00.</span><u><=
/u><u></u></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-f=
amily:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a">=C2=A0</spa=
n><u></u><u></u></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">the </span><span style=3D=
"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;co=
lor:#44546a">implementation must </span><span style=3D"font-size:11.0pt;fon=
t-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">e</span>=
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#44546a">nsure that the message ends empty block</span><s=
pan style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-se=
rif&quot;;color:#1f497d">s</span><span style=3D"font-size:11.0pt;font-famil=
y:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a"> without </span=
><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans=
-serif&quot;;color:#1f497d">a trailing</span><span style=3D"font-size:11.0p=
t;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a">=C2=
=A00x0000FFFF, so </span><span style=3D"font-size:11.0pt;font-family:&quot;=
Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">the </span><span style=
=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;=
;color:#44546a">implementation must keep track on what type of block was at=
 the end of the frame that was sent last. If the last frame sent ends with =
0x0000FFFF</span><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&=
quot;,&quot;sans-serif&quot;;color:#1f497d"> then</span><span style=3D"font=
-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#=
44546a"> </span><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&q=
uot;,&quot;sans-serif&quot;;color:#1f497d">the </span><span style=3D"font-s=
ize:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44=
546a">implementation cannot remove those 4 bytes from the wire</span><span =
style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&=
quot;;color:#1f497d">, but </span><span style=3D"font-size:11.0pt;font-fami=
ly:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#44546a">it must artifi=
cially produce an empty DEFLATE block to send </span><span style=3D"font-si=
ze:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f4=
97d">as the</span><span style=3D"font-size:11.0pt;font-family:&quot;Calibri=
&quot;,&quot;sans-serif&quot;;color:#44546a"> final frame. </span><span sty=
le=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-serif&quo=
t;;color:#1f497d">This is only due to the requirement to remove </span><spa=
n style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-seri=
f&quot;;color:#44546a">0x0000FFFF.</span><u></u><u></u></p>

</div></div></blockquote><div><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></=
p></div><div><p class=3D"MsoNormal">Yes. It&#39;s just two byte long. DEFLA=
TE block header (3bit) + Fixed Huffman code for end of block symbol (7bit) =
+ padding (6bit).=C2=A0</p>

</div></div></div></div></div></div></blockquote><div><br></div><div>Sorry.=
 This comment is (partially) wrong. An artificial uncompressed DEFLATE bloc=
k must be used instead. It&#39;ll be 1 octet long. Please check a new examp=
le in new draft when published.<br>

</div><div><br></div><div>Thanks</div></div></div></div>

--001a11c3841a12452f04f086dd46--

From internet-drafts@ietf.org  Tue Jan 21 19:47:04 2014
Return-Path: <internet-drafts@ietf.org>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id ECEB51A0017; Tue, 21 Jan 2014 19:47:03 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Q9CVq7bu-eEO; Tue, 21 Jan 2014 19:47:02 -0800 (PST)
Received: from ietfa.amsl.com (localhost [IPv6:::1]) by ietfa.amsl.com (Postfix) with ESMTP id DBD121A01A8; Tue, 21 Jan 2014 19:47:00 -0800 (PST)
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable
From: internet-drafts@ietf.org
To: i-d-announce@ietf.org
X-Test-IDTracker: no
X-IETF-IDTracker: 4.90.p2
Auto-Submitted: auto-generated
Precedence: bulk
Message-ID: <20140122034700.9780.55724.idtracker@ietfa.amsl.com>
Date: Tue, 21 Jan 2014 19:47:00 -0800
Cc: hybi@ietf.org
Subject: [hybi] I-D Action: draft-ietf-hybi-permessage-compression-17.txt
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 22 Jan 2014 03:47:04 -0000

A New Internet-Draft is available from the on-line Internet-Drafts director=
ies.
 This draft is a work item of the BiDirectional or Server-Initiated HTTP Wo=
rking Group of the IETF.

        Title           : Compression Extensions for WebSocket
        Author          : Takeshi Yoshino
	Filename        : draft-ietf-hybi-permessage-compression-17.txt
	Pages           : 32
	Date            : 2014-01-21

Abstract:
   This document specifies a framework for creating WebSocket extensions
   that add compression functionality to the WebSocket Protocol.  An
   extension based on this framework compresses the payload data portion
   of non-control WebSocket messages on a per-message basis using
   parameters negotiated during the opening handshake.  This framework
   provides a general method to apply a compression algorithm to the
   contents of WebSocket messages.  For each compression algorithm, an
   extension is defined by specifying parameter negotiation and
   compression algorithm in detail.  This document also specifies one
   specific compression extension using the DEFLATE algorithm.

   Please send feedback to the hybi@ietf.org mailing list.


The IETF datatracker status page for this draft is:
https://datatracker.ietf.org/doc/draft-ietf-hybi-permessage-compression/

There's also a htmlized version available at:
http://tools.ietf.org/html/draft-ietf-hybi-permessage-compression-17

A diff from the previous version is available at:
http://www.ietf.org/rfcdiff?url2=3Ddraft-ietf-hybi-permessage-compression-17


Please note that it may take a couple of minutes from the time of submission
until the htmlized version and diff are available at tools.ietf.org.

Internet-Drafts are also available by anonymous FTP at:
ftp://ftp.ietf.org/internet-drafts/


From tyoshino@google.com  Tue Jan 21 19:49:05 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id F04B31A01BF for <hybi@ietfa.amsl.com>; Tue, 21 Jan 2014 19:49:04 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.913
X-Spam-Level: 
X-Spam-Status: No, score=-1.913 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Y5xnXtywCP2a for <hybi@ietfa.amsl.com>; Tue, 21 Jan 2014 19:49:03 -0800 (PST)
Received: from mail-wi0-x22c.google.com (mail-wi0-x22c.google.com [IPv6:2a00:1450:400c:c05::22c]) by ietfa.amsl.com (Postfix) with ESMTP id 1AD131A0017 for <hybi@ietf.org>; Tue, 21 Jan 2014 19:49:02 -0800 (PST)
Received: by mail-wi0-f172.google.com with SMTP id ex4so5174910wid.17 for <hybi@ietf.org>; Tue, 21 Jan 2014 19:49:02 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :content-type; bh=GYa/AzMyTcN1hAbnmr2ceo7FELCIOdOZUc9VK7sLAVo=; b=QbMBuIXJ5aJd5H8VEXssHbmwo19qwwJb4MaFra9BziaZ6Qmgbpcjko6ELcuS848b+h fTtQw6BE+6blV7r/MgoTC9peVOc9CTnWVg3KLj/mFb1xMcotDYcZT85K7C3qe9yC3+Kv QLK5L3iJOHkskHH9MCCWIH9PYguXsyQh/gULxYnViBE+NXwS3AWbubG4hSXTCLWkZrcb uWrvYiGmQZ/K2cd1b1VGKGwvNraDcBd9yPdleO9CG6glI3D6h1zQJ9ETRT+fcsD7/4Bd nb6BInWShN4hEsgKcKi8R+DIvUhQGRxmcw37WwKigDUHZXSKniKGewR67JT4rZiYOvJl bzzA==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:content-type; bh=GYa/AzMyTcN1hAbnmr2ceo7FELCIOdOZUc9VK7sLAVo=; b=QCKolfLGia/Yv5RJi7VySCb0zo1WUPnjo4RhyxPYVSVbQnQvWxiX4JbIhCgVDbc7hh DSJoy8FbdPAIxilbt+5HV2ko8dYJ3yf6F5i6tNsn7zbokXzYFN3TzoxS/qpZXonVh8W6 4W/xKLsUvpqdke+EOcECYIkQQdqVcUfeXKDkZlJuvK0/hBc0uQTa+mS3xr/YucOOWHgv xNIGMDnkhxkuUhK0+QW+rdPlLfJ3NoARxKAkQBK7/TQSYFnxUs9Xxnx+/2Ert6v9Incb V4RVYiB+Uqpg2FBzWCpJDad9pgqdYHBtOg2PalRtfanqpEdTGNRzVtAKbW9qht+ds7h9 ou2w==
X-Gm-Message-State: ALoCoQn1sZ8Q1vFoVMVSkijFCbrIWMWrDlWZBeVHokt7JOFQcO3SsBTfSaV7U7y6kgq9XTlJ6QJZOhjtJ3HZplk5HXF9b/+WAsvO5TKBafYr2f/AtVETjTr6oLd3QnFAR947UV/lJhX1t/E09G90qZCFOdPSrczbFbbDVa28bIg25hyYCIhMTh8XHzVE9OBwaywvBwwxFP1b
X-Received: by 10.195.13.164 with SMTP id ez4mr22188877wjd.11.1390362542269; Tue, 21 Jan 2014 19:49:02 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Tue, 21 Jan 2014 19:48:41 -0800 (PST)
In-Reply-To: <20140122034700.9780.55724.idtracker@ietfa.amsl.com>
References: <20140122034700.9780.55724.idtracker@ietfa.amsl.com>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Wed, 22 Jan 2014 12:48:41 +0900
Message-ID: <CAH9hSJbnU_rF1=drgWTUMosxCntyk7A9PhXcEFRuB3HcZTpgQw@mail.gmail.com>
To: "hybi@ietf.org" <hybi@ietf.org>
Content-Type: multipart/alternative; boundary=047d7bb04f6871081004f08701f2
Subject: Re: [hybi] I-D Action: draft-ietf-hybi-permessage-compression-17.txt
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 22 Jan 2014 03:49:05 -0000

--047d7bb04f6871081004f08701f2
Content-Type: text/plain; charset=UTF-8

Added an example of how to generate an empty uncompressed DEFLATE block
manually.

Based on Arman's suggestion here.
http://www.ietf.org/mail-archive/web/hybi/current/msg10342.html

--047d7bb04f6871081004f08701f2
Content-Type: text/html; charset=UTF-8

<div dir="ltr"><div class="gmail_extra">Added an example of how to generate an empty uncompressed DEFLATE block manually.</div><div class="gmail_extra"><br></div><div class="gmail_extra">Based on Arman&#39;s suggestion here.</div>

<div class="gmail_extra"><a href="http://www.ietf.org/mail-archive/web/hybi/current/msg10342.html">http://www.ietf.org/mail-archive/web/hybi/current/msg10342.html</a><br></div></div>

--047d7bb04f6871081004f08701f2--

From arman@noemax.com  Wed Jan 22 04:19:26 2014
Return-Path: <arman@noemax.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 6CA9D1A00B1 for <hybi@ietfa.amsl.com>; Wed, 22 Jan 2014 04:19:26 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.535
X-Spam-Level: 
X-Spam-Status: No, score=-2.535 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id PuBjslpE6IBF for <hybi@ietfa.amsl.com>; Wed, 22 Jan 2014 04:19:24 -0800 (PST)
Received: from mail.noemax.com (mail.noemax.com [74.208.113.37]) by ietfa.amsl.com (Postfix) with ESMTP id B4B771A0438 for <hybi@ietf.org>; Wed, 22 Jan 2014 04:19:24 -0800 (PST)
DKIM-Signature: a=rsa-sha1; t=1390393116; x=1390997916; s=m2048; d=noemax.com; c=relaxed/relaxed; v=1; bh=346lGQ8jq0bOhRrvh5UD7BZeM9A=; h=From:Subject:Date:Message-ID:To:MIME-Version:Content-Type:In-Reply-To:References; b=a1rx21nxbt80bTWVGBS7ZbggO7NfMpTxQUn40Vldl1qqGks9dYF1XgIHcko5O49ejcbSmNXq2/K1ewJYlzEm/vbPqqup//ZtIlyFkuwhKqcSkqXWqY3gEFrtqK5zmFWomBkgDPx0RBEpkdgvjEKcUvtV8jpVoGP3EkAZz++DghGJD5MKnvy10An7dZVkZKwMEuOPWQ9dFHuTyQcn3Iq5qjjPVLuhCGLD2L9XfxTg7W+PFjdu1pCpeK8GpRxCiITt+ucihbEL3tRaBmfNR4sS+Y0El+u/iU1zyrmnLQLh8FZg4aq0SdR/3s9qdeL19yHiNg78KZ5iFSn6CxEfRS0LIA==
Received: from mail.noemax.com by mail.noemax.com (Noemax Mail Server) with ASMTP (SSL) id 201401221418351357; Wed, 22 Jan 2014 14:18:35 +0200
From: "Arman Djusupov" <arman@noemax.com>
To: "'Takeshi Yoshino'" <tyoshino@google.com>, <hybi@ietf.org>
References: <20140122034700.9780.55724.idtracker@ietfa.amsl.com> <CAH9hSJbnU_rF1=drgWTUMosxCntyk7A9PhXcEFRuB3HcZTpgQw@mail.gmail.com>
In-Reply-To: <CAH9hSJbnU_rF1=drgWTUMosxCntyk7A9PhXcEFRuB3HcZTpgQw@mail.gmail.com>
Date: Wed, 22 Jan 2014 14:19:18 +0200
Message-ID: <004701cf176c$2ebfeaf0$8c3fc0d0$@noemax.com>
MIME-Version: 1.0
Content-Type: multipart/alternative; boundary="----=_NextPart_000_0048_01CF177C.F249A550"
X-Mailer: Microsoft Outlook 14.0
Thread-Index: AQHu5m4OcX8dmY8xcQVyQnyH0n7lhAIp2YUjmj+8OhA=
Content-Language: en-us
Subject: Re: [hybi] I-D Action: draft-ietf-hybi-permessage-compression-17.txt
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 22 Jan 2014 12:19:26 -0000

This is a multipart message in MIME format.

------=_NextPart_000_0048_01CF177C.F249A550
Content-Type: text/plain;
	charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Thanks Takeshi, that=E2=80=99s great.

=20

With best regards,

Arman

=20

=20

From: hybi [mailto:hybi-bounces@ietf.org] On Behalf Of Takeshi Yoshino
Sent: Wednesday, January 22, 2014 5:49 AM
To: hybi@ietf.org
Subject: Re: [hybi] I-D Action: =
draft-ietf-hybi-permessage-compression-17.txt

=20

Added an example of how to generate an empty uncompressed DEFLATE block =
manually.

=20

Based on Arman's suggestion here.

http://www.ietf.org/mail-archive/web/hybi/current/msg10342.html


------=_NextPart_000_0048_01CF177C.F249A550
Content-Type: text/html;
	charset="UTF-8"
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:m=3D"http://schemas.microsoft.com/office/2004/12/omml" =
xmlns=3D"http://www.w3.org/TR/REC-html40"><head><meta =
http-equiv=3DContent-Type content=3D"text/html; charset=3Dutf-8"><meta =
name=3DGenerator content=3D"Microsoft Word 14 (filtered =
medium)"><style><!--
/* Font Definitions */
@font-face
	{font-family:"Cambria Math";
	panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
	{font-family:Calibri;
	panose-1:2 15 5 2 2 2 4 3 2 4;}
@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:0in;
	margin-bottom:.0001pt;
	font-size:12.0pt;
	font-family:"Times New Roman","serif";}
a:link, span.MsoHyperlink
	{mso-style-priority:99;
	color:blue;
	text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
	{mso-style-priority:99;
	color:purple;
	text-decoration:underline;}
span.EmailStyle17
	{mso-style-type:personal-reply;
	font-family:"Calibri","sans-serif";
	color:#44546A;}
.MsoChpDefault
	{mso-style-type:export-only;
	font-family:"Calibri","sans-serif";
	mso-ligatures:standard;}
@page WordSection1
	{size:8.5in 11.0in;
	margin:1.0in 1.0in 1.0in 1.0in;}
div.WordSection1
	{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext=3D"edit" spidmax=3D"1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext=3D"edit">
<o:idmap v:ext=3D"edit" data=3D"1" />
</o:shapelayout></xml><![endif]--></head><body lang=3DEN-US link=3Dblue =
vlink=3Dpurple><div class=3DWordSection1><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>Thanks </span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'>Takeshi</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>, that=E2=80=99s</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497=
D'> great</span><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>.<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>With best regards,<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'>Arman<o:p></o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span =
style=3D'font-size:11.0pt;font-family:"Calibri","sans-serif";color:#44546=
A'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><b><span =
style=3D'font-size:10.0pt;font-family:"Tahoma","sans-serif"'>From:</span>=
</b><span style=3D'font-size:10.0pt;font-family:"Tahoma","sans-serif"'> =
hybi [mailto:hybi-bounces@ietf.org] <b>On Behalf Of </b>Takeshi =
Yoshino<br><b>Sent:</b> Wednesday, January 22, 2014 5:49 =
AM<br><b>To:</b> hybi@ietf.org<br><b>Subject:</b> Re: [hybi] I-D Action: =
draft-ietf-hybi-permessage-compression-17.txt<o:p></o:p></span></p><p =
class=3DMsoNormal><o:p>&nbsp;</o:p></p><div><div><p =
class=3DMsoNormal>Added an example of how to generate an empty =
uncompressed DEFLATE block manually.<o:p></o:p></p></div><div><p =
class=3DMsoNormal><o:p>&nbsp;</o:p></p></div><div><p =
class=3DMsoNormal>Based on Arman's suggestion =
here.<o:p></o:p></p></div><div><p class=3DMsoNormal><a =
href=3D"http://www.ietf.org/mail-archive/web/hybi/current/msg10342.html">=
http://www.ietf.org/mail-archive/web/hybi/current/msg10342.html</a><o:p><=
/o:p></p></div></div></div></body></html>
------=_NextPart_000_0048_01CF177C.F249A550--


From tyoshino@google.com  Wed Jan 22 20:50:16 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id EF7621A01AD for <hybi@ietfa.amsl.com>; Wed, 22 Jan 2014 20:50:15 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.713
X-Spam-Level: 
X-Spam-Status: No, score=-0.713 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, J_CHICKENPOX_34=0.6, J_CHICKENPOX_37=0.6, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ldG7DklXmT8Z for <hybi@ietfa.amsl.com>; Wed, 22 Jan 2014 20:50:13 -0800 (PST)
Received: from mail-we0-x22d.google.com (mail-we0-x22d.google.com [IPv6:2a00:1450:400c:c03::22d]) by ietfa.amsl.com (Postfix) with ESMTP id 1FD881A018D for <hybi@ietf.org>; Wed, 22 Jan 2014 20:50:12 -0800 (PST)
Received: by mail-we0-f173.google.com with SMTP id t60so723839wes.18 for <hybi@ietf.org>; Wed, 22 Jan 2014 20:50:11 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=n5PA5i6lgzdrIQeIH1NgQVBV8v5jDpub62e4O/8futI=; b=Dyfxkog0wqr5xxMXqsmozN8mPEDUbu9Wk8opGGq/B2mb86Ud3KYxRN5ETKmj5rsSU2 SmOKNBECknUZqejKzYXdFFC4evq5emhrUhlJ2nqCH3EDnZCoWIkHT48PvhBThVP7cRJw cLEtpzyefU+Vnx9XpHIstrzYCbrGoGqJ1BJ/OorWbsdGnD+3N+TCsFVbbmufl4fIcjeC jPuI9P1hYgt7Ou65QYSw5mEJXGQm3sndj0HWvqX55k/NzzSB2VuDpZKQ1Vt3sTVKUP/p 7SLnH2B/rDEH7palC0qzW7i01F5kikAXN0/Io6tMN9pbKX2ysSdJPRlNsSOBW88pYQaq 2wPQ==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=n5PA5i6lgzdrIQeIH1NgQVBV8v5jDpub62e4O/8futI=; b=aRMUKJv4S40NkApDq1tFuW1p0yfmTaak844rQ/dFz7MnfoXusvHiA8g3bFnJ6m3+oM 66yBw2+fZ3sFG5Vs8ssNyI43bEST7guSuTWsBclMm7JkOMrBA8xK+qFjLzrZPRhO8uWN ThMOvN0PjEhD+nacDmvdHkDZirUXtJSSlXiZYWFkbRq3cIwNks346JCsi0uLort0jyi3 Jil83QvR/HzJfo6ORileJF0erUPcFnPtJ41w/8lwR6BnkIE5i1y4stiXx4EGguwkBSlX CgVRAH5ZFdyvZE9uTIVkinHgjKJZyI/QPze/67r3Xcggkg1s7H1nAPtWKSgF6/B0Et6C EXAQ==
X-Gm-Message-State: ALoCoQnFUIF3qvPT046CbfDNY9LOIqcYUoe68sIlaSGg1sJ4NQ2psByeC7O6H9bT/n9a71K1HR7No3laetCRLWJdHHzTOv5STCen29usVxIiYml2dPzAAb8FKz8gySRyRzjep9Q6NGGgbSCPspbWlw1tFlhLR5DoZhhfXFswgnHQkao0FhYymCMTeUQVRKFFskDGj9PYMKSH
X-Received: by 10.194.78.141 with SMTP id b13mr4859690wjx.32.1390452611665; Wed, 22 Jan 2014 20:50:11 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Wed, 22 Jan 2014 20:49:51 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446C082A0D@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF9948F@EXVMBX020-12.exch020.serverdata.net> <52D90F99.6080205@gmx.de> <634914A010D0B943A035D226786325D4446BF9949D@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJZQ3NxVW36PnZ0TpMF4tPPaJLr12M8NtbPb6pUejf1wcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99977@EXVMBX020-12.exch020.serverdata.net> <B8ED6F18-B710-44AE-829B-EDE5859B2C5B@zaphoyd.com> <52DE3B1B.1010804@it.aoyama.ac.jp> <634914A010D0B943A035D226786325D4446C082A0D@EXVMBX020-12.exch020.serverdata.net>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Thu, 23 Jan 2014 13:49:51 +0900
Message-ID: <CAH9hSJZwf4ZuOnOwt7k+YjqCd5zxYYWXdp=krPjZKE+Tq=9LXw@mail.gmail.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=047d7bfcfc82fefe7304f09bf94e
Cc: Julian Reschke <julian.reschke@gmx.de>, "hybi@ietf.org" <hybi@ietf.org>, Peter Thorson <webmaster@zaphoyd.com>
Subject: Re: [hybi] WebSocket subprotocol parameters
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 23 Jan 2014 04:50:16 -0000

--047d7bfcfc82fefe7304f09bf94e
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

On Tue, Jan 21, 2014 at 7:12 PM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> > > My impression was that subprotocol primarily changed the interpretati=
on
> > of messages payloads by the end applications themselves. WebSocket
> > messages allow essentially arbitrary data so you can define whatever
> > parameter setting mechanism you like with whatever grammar makes sense
> > for your subprotocol and send the parameters in the opening =E2=80=9Che=
llo=E2=80=9D
> > message. Defining a grammar in the socket spec to allow application
> > protocols to configure themselves seems limiting and out of scope.
> >
> > Yes indeed. It has always been my impression that even subprotocols
> > themselves are rather unnecessary. On the server side, just use a
> different
> > URI/IRI as endpoint. On the client side, it's all JS anyway.
>
> No quite, since using a different URL does not allow the server to
> communicate back
> it's subprotocol selection, e.g.
>
>
The server can immediately send back a frame following the opening
handshake response. So, there's not much difference between subprotocol
selection reply in form of opening handshake and in form of the first
message. No additional RTT.


> Sec-WebSocket-Protocol: foo.msgpack, foo.json
>
> allows the client to announce the "foo" protocol in 2 variants (MsgPack
> and JSON serialization)
> and the server to _select_ one of those, and communicate back it's
> selection to the client.
>
> An URL
>
> ws://server.com?serializers=3Dmsgpack;json
>
> does not allow the server to communicate back it's selection.
>

So, you'll define a thin extension layer to your application level protocol
which uses the URI and the first message from the server to negotiate
subprotocol parameter.


>
> Further, as an example, imagine a subprotocol that allows message batchin=
g
> with length prefixes, that wants to negotiate the number of bytes used fo=
r
> the length prefix.
>
> Sec-WebSocket-Protocol: foo; max_len_prefix=3D4
>
> A server that wants to use "len_prefix=3D3" cannot communicate that -
> neither with URL (since there is no "URL response"), and not with
> Sec-WebSocket-Protocol,
> since the server _must_ choose one subprotocol announced by the client
> _exactly_.
>
> It cannot answer
>
> Sec-WebSocket-Protocol: foo; len_prefix=3D3
>
> Since the number of bytes per length prefix in batched messages is needed
> to _parse_ already the first WebSocket message at application level,
> negotiating at the WS application message level does not work either (or
> gets complicated).
>

I understand it'll be simpler if we could complete that with in subprotocol
negotiation framework. But does it really get so complicated if we do that
using URI and bootstrap message? The client just need to pass one phase
where it parses len_prefix from the server and then transit to real app
mode where len_prefix will be used.

I don't object to raising a request to relax subprotocol constraint from
the next version. But I think it doesn't make much difference technically.

--047d7bfcfc82fefe7304f09bf94e
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On T=
ue, Jan 21, 2014 at 7:12 PM, Tobias Oberstein <span dir=3D"ltr">&lt;<a href=
=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstein@=
tavendo.de</a>&gt;</span> wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-=
left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;p=
adding-left:1ex"><div class=3D"im">&gt; &gt; My impression was that subprot=
ocol primarily changed the interpretation<br>


&gt; of messages payloads by the end applications themselves. WebSocket<br>
&gt; messages allow essentially arbitrary data so you can define whatever<b=
r>
&gt; parameter setting mechanism you like with whatever grammar makes sense=
<br>
&gt; for your subprotocol and send the parameters in the opening =E2=80=9Ch=
ello=E2=80=9D<br>
&gt; message. Defining a grammar in the socket spec to allow application<br=
>
&gt; protocols to configure themselves seems limiting and out of scope.<br>
&gt;<br>
&gt; Yes indeed. It has always been my impression that even subprotocols<br=
>
&gt; themselves are rather unnecessary. On the server side, just use a diff=
erent<br>
&gt; URI/IRI as endpoint. On the client side, it&#39;s all JS anyway.<br>
<br>
</div>No quite, since using a different URL does not allow the server to co=
mmunicate back<br>
it&#39;s subprotocol selection, e.g.<br>
<br></blockquote><div><br></div><div>The server can immediately send back a=
 frame following the opening handshake response. So, there&#39;s not much d=
ifference between subprotocol selection reply in form of opening handshake =
and in form of the first message. No additional RTT.</div>

<div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px =
0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-l=
eft-style:solid;padding-left:1ex">
Sec-WebSocket-Protocol: foo.msgpack, foo.json<br>
<br>
allows the client to announce the &quot;foo&quot; protocol in 2 variants (M=
sgPack and JSON serialization)<br>
and the server to _select_ one of those, and communicate back it&#39;s sele=
ction to the client.<br>
<br>
An URL<br>
<br>
ws://<a href=3D"http://server.com?serializers=3Dmsgpack;json" target=3D"_bl=
ank">server.com?serializers=3Dmsgpack;json</a><br>
<br>
does not allow the server to communicate back it&#39;s selection.<br></bloc=
kquote><div><br></div><div>So, you&#39;ll define a thin extension layer to =
your application level protocol which uses the URI and the first message fr=
om the server to negotiate subprotocol parameter.</div>

<div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px =
0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-l=
eft-style:solid;padding-left:1ex">
<br>
Further, as an example, imagine a subprotocol that allows message batching =
with length prefixes, that wants to negotiate the number of bytes used for =
the length prefix.<br>
<br>
Sec-WebSocket-Protocol: foo; max_len_prefix=3D4<br>
<br>
A server that wants to use &quot;len_prefix=3D3&quot; cannot communicate th=
at - neither with URL (since there is no &quot;URL response&quot;), and not=
 with Sec-WebSocket-Protocol,<br>
since the server _must_ choose one subprotocol announced by the client _exa=
ctly_.<br>
<br>
It cannot answer<br>
<br>
Sec-WebSocket-Protocol: foo; len_prefix=3D3<br>
<br>
Since the number of bytes per length prefix in batched messages is needed t=
o _parse_ already the first WebSocket message at application level,<br>
negotiating at the WS application message level does not work either (or ge=
ts complicated).<br></blockquote><div><br></div><div>I understand it&#39;ll=
 be simpler if we could complete that with in subprotocol negotiation frame=
work. But does it really get so complicated if we do that using URI and boo=
tstrap message? The client just need to pass one phase where it parses len_=
prefix from the server and then transit to real app mode where len_prefix w=
ill be used.</div>

<div><br></div><div>I don&#39;t object to raising a request to relax subpro=
tocol constraint from the next version. But I think it doesn&#39;t make muc=
h difference technically.</div><div>=C2=A0</div></div></div></div>

--047d7bfcfc82fefe7304f09bf94e--

From ricea@google.com  Thu Jan 23 00:32:36 2014
Return-Path: <ricea@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C19591A0236 for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 00:32:36 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.913
X-Spam-Level: 
X-Spam-Status: No, score=-1.913 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id DTCvMPzllgyv for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 00:32:33 -0800 (PST)
Received: from mail-oa0-x22c.google.com (mail-oa0-x22c.google.com [IPv6:2607:f8b0:4003:c02::22c]) by ietfa.amsl.com (Postfix) with ESMTP id 9B4B91A0230 for <hybi@ietf.org>; Thu, 23 Jan 2014 00:32:33 -0800 (PST)
Received: by mail-oa0-f44.google.com with SMTP id g12so1800917oah.3 for <hybi@ietf.org>; Thu, 23 Jan 2014 00:32:32 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type; bh=AOYOlhVFCm2ekWpau5sThd5T+Xz5qpJp7lDj+SJPwaI=; b=WM7oQKFdc75uZo7P6iEw/u0wwAZoK6qe7F/OsYMFTiXzv+Boysva0JvErBSQ8Z93tj DYw3qHJQHW1YRo5Otrruyhtl7P8PTpH+WxgO7eUz/B5jwurccqDn08RkCz+D8qakXkPY yb3rBN0oshqt3gw88vPQtx7j615ou27Mb1lRp7enMRGips37sf8WlAzTDm1YeujQS7uf K8PZyJ8/W5BKMPZZHU+XogSN/sRFfULu0uLrFe0vfIZyBysMx7cv67ZWqglxjxAeUOvX mBmP7l0tcYKz8SBiLXzO2MqialqMNURx1UzeDxvrcXOmwBIArVAaooiyuH46CZZ45P1/ QOVA==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=AOYOlhVFCm2ekWpau5sThd5T+Xz5qpJp7lDj+SJPwaI=; b=cuHd7EkQcxh1FUA80EA7AjI1vd32pSd12PcfDhxoXz1mVRQIGiTRUDh1MzKf0uIIVb 3C0e+GKph9mhwUhWMRVkTRwGz/gUDw4jm+neaE7910Hb2mMgKIXZh0dekjCZ1h/66Ulh wfB63sPWJPGWgF3wntV6+elxEVMcdMUjm9S7gTOc+GPJBJBzJsGQ482ht+S4hsR9+SAf +UO3bxXBnwTJxPwClQdg6au5y5nLHMz2j0p8gn8NN7xUCc9+b1oTYLWYUXQk61iUYfe4 nM4W14f9jcVABRjhZ4pmfzGpxq1xQhIZSGB3swk8DTbOXyFnnt40xYhZ45cVoLKWnqR1 GMLA==
X-Gm-Message-State: ALoCoQnqfRRHT8C6EZkoq6+RnXG+LKHduHqZ/rApLQeVXipzWWo9ET7g3U5rliFJCutweve9FJ4C3lP8UvkTmQcZuCNQKuoM/QCQz+VX6ZlwjFwe4LlLOPcpUok2pMiF2Qg5iBtIacxXxhoBvkzdzQD0xle7DXLQ2XtvCW7U5iBulyRPQkx2tkRwlkI7FEyBuZkgHBaTrc35
MIME-Version: 1.0
X-Received: by 10.182.40.201 with SMTP id z9mr5547937obk.45.1390465952568; Thu, 23 Jan 2014 00:32:32 -0800 (PST)
Received: by 10.182.118.41 with HTTP; Thu, 23 Jan 2014 00:32:32 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446BF9987B@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF991A9@EXVMBX020-12.exch020.serverdata.net> <634914A010D0B943A035D226786325D4446BF9987B@EXVMBX020-12.exch020.serverdata.net>
Date: Thu, 23 Jan 2014 17:32:32 +0900
Message-ID: <CAHixhFr6wPQQRFrV72eXp9cUA=m1Qvxd2ZhYpqPXFa9pRhnZcQ@mail.gmail.com>
From: Adam Rice <ricea@google.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=001a11c33b022cf00204f09f158e
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] permessage-priority extension draft
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 23 Jan 2014 08:32:36 -0000

--001a11c33b022cf00204f09f158e
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

This part contains a contradiction:

Intermediaries unaware of the permessage-priority extension

      1. MUST NOT change the message fragmentation (further fragment or
      coalesce non-control frames)

      2. MUST preserve strict frame order while processing and
      forwarding non-control frames

      3. MUST ignore WebSocket base protocol rules as they apply to
      continuation frames and the FIN bit

   or

      1. MUST remove the "permessage-priority" element from the client's
      "Sec-WebSocket-Extensions" header in the opening handshake


If the intermediary is unaware of the extension, then there is no way for
it to follow the rules in extension specification.

I think we can extrapolate a more general "all or nothing" rule from this.
A WebSocket intermediary must either treat all data as opaque and pass it
through unchanged, or completely implement the specification and all
extensions in use.

It follows that a WebSocket intermediary that understands the protocol must
strip any unrecognised extensions from the client handshake.

Thinking about compatibility with permessage-deflate, I came to the
conclusion that RFC6455 is too optimistic about extension compatibility.
The default expectation should be that extensions are incompatible, and the
WebSocket Extension Name Registery should have a list of "Known Compatible
Extensions" rather than "Known Incompatible Extensions".

As an implementer, my personal preference would be to not implement
permessage-priority. It requires changes in almost every layer of the
stack, adding complexity and overhead. I see my job as to get bytes from
the Javascript send() call into the kernel send buffer as efficiently as
reasonably possible; ideally I would like to avoid any additional queuing
altogether, never mind re-ordering of queued messages.

Thanks,
Adam Rice


On 18 January 2014 08:10, Tobias Oberstein <tobias.oberstein@tavendo.de>wro=
te:

> There is a collection of open issues here:
> https://github.com/oberstet/permessage-priority/issues
>
> One of those is how to make permessage-priority compatible with
> permessage-deflate.
>
> I guess it can be done by maintaining a separate compression context per
> priority class.
> This allows intermediaries to prioritize and reorder frames without
> problems. However,
> it multiplies memory requirements.
>
> Was there any discussion/plan about how permessage-deflate would have
> worked with MUX
> when compression was applied to the physical WebSocket connection?
>
> I mean, when compression was applied to the logical MUXed WS connections,
> there
> would have been a separate compression context for each (logical) WS
> connection also,
> and intermediaries can reMUX without recompressing.
>
> But when applied to the physical WS?
>
> /Tobias
>
> > -----Urspr=C3=BCngliche Nachricht-----
> > Von: hybi [mailto:hybi-bounces@ietf.org] Im Auftrag von Tobias Oberstei=
n
> > Gesendet: Donnerstag, 16. Januar 2014 17:39
> > An: hybi@ietf.org
> > Betreff: [hybi] permessage-priority extension draft
> >
> > Alright, I have written up a draft for a "permessage-priority" WebSocke=
t
> > extension:
> >
> >
> https://github.com/oberstet/permessage-priority/blob/master/draft-oberste=
in-
> > hybi-permessage-priority.txt
> >
> > In writing this, I came to the conclusion that "message prioritization"
> on a
> > single WebSocket connection in fact is orthogonal to multiplexing
> multiple
> > (logical) WebSocket connections over a single physical one (be it via
> WS-MUX
> > or a variant of WS/SPDY).
> >
> > So this proposal should not be seen as a competitive proposal to WS-MUX
> or
> > WS/SPDY, but as ... something different;)
> >
> > The motivation and use case is outlined in the introduction.
> >
> > I'd appreciate any feedback, suggestions, corrections or even PRs (the
> draft is
> > on GitHub in a dedicated repo).
> >
> > If the draft doesn't conform to RFC requirements or formatting rules,
> please
> > forgive - this is my first RFC.
> >
> > Cheers,
> > /Tobias
> >
> > PS: For one thing I don't know, how do I submit this draft with the IET=
F
> so it can
> > follow the proper way?
> >
> > _______________________________________________
> > hybi mailing list
> > hybi@ietf.org
> > https://www.ietf.org/mailman/listinfo/hybi
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>

--001a11c33b022cf00204f09f158e
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">This part contains a contradiction:<div><br></div><div><pr=
e style=3D"margin-top:0px;margin-bottom:0px"><div class=3D"" id=3D"LC650" s=
tyle=3D"color:rgb(51,51,51);font-family:Consolas,&#39;Liberation Mono&#39;,=
Courier,monospace;font-size:12px;line-height:18px;padding-left:10px">
Intermediaries unaware of the permessage-priority extension</div><div class=
=3D"" id=3D"LC651" style=3D"color:rgb(51,51,51);font-family:Consolas,&#39;L=
iberation Mono&#39;,Courier,monospace;font-size:12px;line-height:18px;paddi=
ng-left:10px">
<br style></div><div class=3D"" id=3D"LC652" style=3D"color:rgb(51,51,51);f=
ont-family:Consolas,&#39;Liberation Mono&#39;,Courier,monospace;font-size:1=
2px;line-height:18px;padding-left:10px">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A01. MUST NOT change the message fragmentation (further fragment or</div>
<div class=3D"" id=3D"LC653" style=3D"color:rgb(51,51,51);font-family:Conso=
las,&#39;Liberation Mono&#39;,Courier,monospace;font-size:12px;line-height:=
18px;padding-left:10px">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0coalesce non-co=
ntrol frames)</div><div class=3D"" id=3D"LC654" style=3D"color:rgb(51,51,51=
);font-family:Consolas,&#39;Liberation Mono&#39;,Courier,monospace;font-siz=
e:12px;line-height:18px;padding-left:10px">
<br style></div><div class=3D"" id=3D"LC655" style=3D"color:rgb(51,51,51);f=
ont-family:Consolas,&#39;Liberation Mono&#39;,Courier,monospace;font-size:1=
2px;line-height:18px;padding-left:10px">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A02. MUST preserve strict frame order while processing and</div>
<div class=3D"" id=3D"LC656" style=3D"color:rgb(51,51,51);font-family:Conso=
las,&#39;Liberation Mono&#39;,Courier,monospace;font-size:12px;line-height:=
18px;padding-left:10px">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0forwarding non-=
control frames</div><div class=3D"" id=3D"LC657" style=3D"color:rgb(51,51,5=
1);font-family:Consolas,&#39;Liberation Mono&#39;,Courier,monospace;font-si=
ze:12px;line-height:18px;padding-left:10px">
<br style></div><div class=3D"" id=3D"LC658" style=3D"color:rgb(51,51,51);f=
ont-family:Consolas,&#39;Liberation Mono&#39;,Courier,monospace;font-size:1=
2px;line-height:18px;padding-left:10px">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A03. MUST ignore WebSocket base protocol rules as they apply to</div>
<div class=3D"" id=3D"LC659" style=3D"color:rgb(51,51,51);font-family:Conso=
las,&#39;Liberation Mono&#39;,Courier,monospace;font-size:12px;line-height:=
18px;padding-left:10px">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0continuation fr=
ames and the FIN bit</div><div class=3D"" id=3D"LC660" style=3D"color:rgb(5=
1,51,51);font-family:Consolas,&#39;Liberation Mono&#39;,Courier,monospace;f=
ont-size:12px;line-height:18px;padding-left:10px">
<br style></div><div class=3D"" id=3D"LC661" style=3D"color:rgb(51,51,51);f=
ont-family:Consolas,&#39;Liberation Mono&#39;,Courier,monospace;font-size:1=
2px;line-height:18px;padding-left:10px">=C2=A0=C2=A0=C2=A0or</div><div clas=
s=3D"" id=3D"LC662" style=3D"color:rgb(51,51,51);font-family:Consolas,&#39;=
Liberation Mono&#39;,Courier,monospace;font-size:12px;line-height:18px;padd=
ing-left:10px">
<br style></div><div class=3D"" id=3D"LC663" style=3D"color:rgb(51,51,51);f=
ont-family:Consolas,&#39;Liberation Mono&#39;,Courier,monospace;font-size:1=
2px;line-height:18px;padding-left:10px">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A01. MUST remove the &quot;permessage-priority&quot; element from the clie=
nt&#39;s</div>
<div class=3D"" id=3D"LC664" style=3D"color:rgb(51,51,51);font-family:Conso=
las,&#39;Liberation Mono&#39;,Courier,monospace;font-size:12px;line-height:=
18px;padding-left:10px">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0&quot;Sec-WebSo=
cket-Extensions&quot; header in the opening handshake</div>
</pre></div><br>If the intermediary is unaware of the extension, then there=
 is no way for it to follow the rules in extension specification.<div><br><=
/div><div>I think we can extrapolate a more general &quot;all or nothing&qu=
ot; rule from this. A WebSocket intermediary must either treat all data as =
opaque and pass it through unchanged, or completely implement the specifica=
tion and all extensions in use.</div>
<div><br></div><div>It follows that a WebSocket intermediary that understan=
ds the protocol must strip any unrecognised extensions from the client hand=
shake.</div><div><br></div><div>Thinking about compatibility with permessag=
e-deflate, I came to the conclusion that RFC6455 is too optimistic about ex=
tension compatibility. The default expectation should be that extensions ar=
e incompatible, and the WebSocket Extension Name Registery should have a li=
st of &quot;<span style=3D"color:rgb(0,0,0);font-size:1em">Known Compatible=
 Extensions</span>&quot; rather than &quot;<span style=3D"color:rgb(0,0,0);=
font-size:1em">Known Incompatible Extensions</span>&quot;.</div>
<div><br></div><div>As an implementer, my personal preference would be to n=
ot implement permessage-priority. It requires changes in almost every layer=
 of the stack, adding complexity and overhead. I see my job as to get bytes=
 from the Javascript send() call into the kernel send buffer as efficiently=
 as reasonably possible; ideally I would like to avoid any additional queui=
ng altogether, never mind re-ordering of queued messages.</div>
<div><br></div><div>Thanks,</div><div>Adam Rice</div></div><div class=3D"gm=
ail_extra"><br><br><div class=3D"gmail_quote">On 18 January 2014 08:10, Tob=
ias Oberstein <span dir=3D"ltr">&lt;<a href=3D"mailto:tobias.oberstein@tave=
ndo.de" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt;</span> wrote:=
<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">There is a collection of open issues here:<b=
r>
<a href=3D"https://github.com/oberstet/permessage-priority/issues" target=
=3D"_blank">https://github.com/oberstet/permessage-priority/issues</a><br>
<br>
One of those is how to make permessage-priority compatible with permessage-=
deflate.<br>
<br>
I guess it can be done by maintaining a separate compression context per pr=
iority class.<br>
This allows intermediaries to prioritize and reorder frames without problem=
s. However,<br>
it multiplies memory requirements.<br>
<br>
Was there any discussion/plan about how permessage-deflate would have worke=
d with MUX<br>
when compression was applied to the physical WebSocket connection?<br>
<br>
I mean, when compression was applied to the logical MUXed WS connections, t=
here<br>
would have been a separate compression context for each (logical) WS connec=
tion also,<br>
and intermediaries can reMUX without recompressing.<br>
<br>
But when applied to the physical WS?<br>
<br>
/Tobias<br>
<br>
&gt; -----Urspr=C3=BCngliche Nachricht-----<br>
&gt; Von: hybi [mailto:<a href=3D"mailto:hybi-bounces@ietf.org">hybi-bounce=
s@ietf.org</a>] Im Auftrag von Tobias Oberstein<br>
&gt; Gesendet: Donnerstag, 16. Januar 2014 17:39<br>
&gt; An: <a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br>
&gt; Betreff: [hybi] permessage-priority extension draft<br>
<div class=3D"HOEnZb"><div class=3D"h5">&gt;<br>
&gt; Alright, I have written up a draft for a &quot;permessage-priority&quo=
t; WebSocket<br>
&gt; extension:<br>
&gt;<br>
&gt; <a href=3D"https://github.com/oberstet/permessage-priority/blob/master=
/draft-oberstein-" target=3D"_blank">https://github.com/oberstet/permessage=
-priority/blob/master/draft-oberstein-</a><br>
&gt; hybi-permessage-priority.txt<br>
&gt;<br>
&gt; In writing this, I came to the conclusion that &quot;message prioritiz=
ation&quot; on a<br>
&gt; single WebSocket connection in fact is orthogonal to multiplexing mult=
iple<br>
&gt; (logical) WebSocket connections over a single physical one (be it via =
WS-MUX<br>
&gt; or a variant of WS/SPDY).<br>
&gt;<br>
&gt; So this proposal should not be seen as a competitive proposal to WS-MU=
X or<br>
&gt; WS/SPDY, but as ... something different;)<br>
&gt;<br>
&gt; The motivation and use case is outlined in the introduction.<br>
&gt;<br>
&gt; I&#39;d appreciate any feedback, suggestions, corrections or even PRs =
(the draft is<br>
&gt; on GitHub in a dedicated repo).<br>
&gt;<br>
&gt; If the draft doesn&#39;t conform to RFC requirements or formatting rul=
es, please<br>
&gt; forgive - this is my first RFC.<br>
&gt;<br>
&gt; Cheers,<br>
&gt; /Tobias<br>
&gt;<br>
&gt; PS: For one thing I don&#39;t know, how do I submit this draft with th=
e IETF so it can<br>
&gt; follow the proper way?<br>
&gt;<br>
&gt; _______________________________________________<br>
&gt; hybi mailing list<br>
&gt; <a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br>
&gt; <a href=3D"https://www.ietf.org/mailman/listinfo/hybi" target=3D"_blan=
k">https://www.ietf.org/mailman/listinfo/hybi</a><br>
_______________________________________________<br>
hybi mailing list<br>
<a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br>
<a href=3D"https://www.ietf.org/mailman/listinfo/hybi" target=3D"_blank">ht=
tps://www.ietf.org/mailman/listinfo/hybi</a><br>
</div></div></blockquote></div><br></div>

--001a11c33b022cf00204f09f158e--

From tobias.oberstein@tavendo.de  Thu Jan 23 02:37:10 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 57B611A03EC for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 02:37:10 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.899
X-Spam-Level: 
X-Spam-Status: No, score=-1.899 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id VSG6Ckwa4UUE for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 02:37:06 -0800 (PST)
Received: from EXHUB020-5.exch020.serverdata.net (exhub020-5.exch020.serverdata.net [206.225.164.32]) by ietfa.amsl.com (Postfix) with ESMTP id 1C8C51A03A9 for <hybi@ietf.org>; Thu, 23 Jan 2014 02:37:06 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-5.exch020.serverdata.net ([206.225.164.32]) with mapi; Thu, 23 Jan 2014 02:37:05 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Adam Rice <ricea@google.com>
Date: Thu, 23 Jan 2014 02:37:00 -0800
Thread-Topic: [hybi] permessage-priority extension draft
Thread-Index: Ac8YFaxq/pMbEFjfTUynoguZX974iwADG2cA
Message-ID: <634914A010D0B943A035D226786325D4446C083400@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF991A9@EXVMBX020-12.exch020.serverdata.net> <634914A010D0B943A035D226786325D4446BF9987B@EXVMBX020-12.exch020.serverdata.net> <CAHixhFr6wPQQRFrV72eXp9cUA=m1Qvxd2ZhYpqPXFa9pRhnZcQ@mail.gmail.com>
In-Reply-To: <CAHixhFr6wPQQRFrV72eXp9cUA=m1Qvxd2ZhYpqPXFa9pRhnZcQ@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: multipart/alternative; boundary="_000_634914A010D0B943A035D226786325D4446C083400EXVMBX02012ex_"
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] permessage-priority extension draft
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 23 Jan 2014 10:37:10 -0000

--_000_634914A010D0B943A035D226786325D4446C083400EXVMBX02012ex_
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
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==

--_000_634914A010D0B943A035D226786325D4446C083400EXVMBX02012ex_
Content-Type: text/html; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_634914A010D0B943A035D226786325D4446C083400EXVMBX02012ex_--


From yhirano@google.com  Thu Jan 23 03:16:10 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 34EC31A034E for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 03:16:10 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.913
X-Spam-Level: 
X-Spam-Status: No, score=-1.913 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id csHdclm8I3sw for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 03:16:07 -0800 (PST)
Received: from mail-qa0-x231.google.com (mail-qa0-x231.google.com [IPv6:2607:f8b0:400d:c00::231]) by ietfa.amsl.com (Postfix) with ESMTP id 676531A039D for <hybi@ietf.org>; Thu, 23 Jan 2014 03:16:07 -0800 (PST)
Received: by mail-qa0-f49.google.com with SMTP id w8so1963121qac.22 for <hybi@ietf.org>; Thu, 23 Jan 2014 03:16:06 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=yaX7mFedSu9nvmDswj+aRafmK8I5expyEvpzrZxv5BE=; b=Tg3eXlqwDtOB7ofrLRw/ffs3RD69Wma+5/SiYd5yYYGKXpQ2hoSkKqhWaWgphTtU2o SWNJ6d7IFfvHwl5pBxr0Kk6iPUXymGZsHExjAhCdCI817oiLFtgGAF4cSNGN31ZQKzsf 6cUKaZcmUN5N5gF14DDcixZi3tlOG2RPKQKd5015wy8Y7m6irfSSUwe99eikwcqxreXP A9YQK6YIjHv1KvJb3CKBhU6TxtZJ0ZOTOMmfnrmpNed47F+j2R4kSzuEBo+VGchLaFtm ZWVp26+UzQcvKjA2pR1IBiGk7rHg9/cVc2AasmqhvPWvBC69IIaZFvVlZPYIFoOH6UKT vNQA==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=yaX7mFedSu9nvmDswj+aRafmK8I5expyEvpzrZxv5BE=; b=HZXCLeCjl9tDrVWzo/Wlu6++gFL5MQ385q1KTq1muTKI6Fa26tkSQwpVtEvTMLt7wf IbMNyUC1GibPnwlBD1YCnmwRgVqesYE+O74au0VnGPxmWYu4vOzrjVr8Vk9wm3o+TA5x 8lzlu/RJ95ynJYccZDTdco2TgiXefLyDJvxD8=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=yaX7mFedSu9nvmDswj+aRafmK8I5expyEvpzrZxv5BE=; b=apsP7rEwVQBiWWQe1kkQO6hgkNxCd3t2NcKJO+iujqdcvtRZ+c0jJOog7dRzynMgXD 2FtVjvLvpAQWxA3WgjrTzfyuvO5fXX/bPZmrKeZr6Ylk6M1WGiCOKn63XNloXbkD8o/3 9ffmrpaSLNUUJu2EqSSi6wdOpyrDiud42O21/fO2RP5i1NhE6u5geO4iATZPthrYjndw +bneQyrlyEpoitGjF0ugo9yJr8i7sTUBDHnynaGIvgRFXnMopCuZagzGGv0ITXB65Qgz gXHQ/iTqyIezTINKW7CUt4eTdrEqM2yjdgFuDdA4wT4HipVpgLgrzVWQxQCtzuiVIBb2 11xA==
X-Gm-Message-State: ALoCoQntsiND2kRurBE+mor7cZCiRgme2fT5mpn41yLigUL+2OVqp/ojv1yMAYibyZkvDuHLBcylASu8W4NxjbLQgP4KNKMFQUDb63ljQ1KSvYtMPPYJ5axGuA9GmgSOkzhJjW1+ZXGcTgMbyCEOPP1Zq89GMJaDv+rYXanoj3xzgZSGvgmk2Ktyk4rx6iydlieSoZVjn1A1
MIME-Version: 1.0
X-Received: by 10.140.22.145 with SMTP id 17mr10105973qgn.0.1390475766414; Thu, 23 Jan 2014 03:16:06 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.140.109.202 with HTTP; Thu, 23 Jan 2014 03:16:06 -0800 (PST)
In-Reply-To: <6652CC7E-7E92-40EF-BD50-2DDD39E15533@zaphoyd.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net> <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com> <634914A010D0B943A035D226786325D4446C0826EC@EXVMBX020-12.exch020.serverdata.net> <6652CC7E-7E92-40EF-BD50-2DDD39E15533@zaphoyd.com>
Date: Thu, 23 Jan 2014 20:16:06 +0900
X-Google-Sender-Auth: MNLaEmquG53MdXPYk5Z24p8ngnk
Message-ID: <CABihn6GSrU8oJa4nbMJiDDo2Z+-c4fT69roVFxcHYM4XYFToRQ@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: Peter Thorson <webmaster@zaphoyd.com>
Content-Type: multipart/alternative; boundary=001a11c130102050b304f0a15edc
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 23 Jan 2014 11:16:10 -0000

--001a11c130102050b304f0a15edc
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Thanks for comments and sorry for the late response.

I now agree with your opinion that having both SPDY frame and WS frame is
redundant.
I would like extensions and having RSVs seems a good idea.
As a result, I think it is OK not to support extensions that works on
individual WS frames not WS messages. (i.e. We should support
permessage-deflate.)

- An HTTP 2.0 WEBSOCKET frame type could be used to deliver WebSocket
> messages. You=E2=80=99d need a FIN flag for the message boundaries, frame
> boundaries would be just standard HTTP 2.0 frames. No need to repackage o=
r
> reframe or make a distinction between =E2=80=9Cwebsocket=E2=80=9D frames =
and HTTP 2.0
> frames.
> - Depending on how closely we want HTTP 2.0 WebSockets to match RFC6455
> WebSockets additional support could be added.. for example, define the
> WEBSOCKET frame type=E2=80=99s flag bits as follows (full compatibility f=
lag
> configuration):

I don't think it is a good idea, for various reasons.
- Using many flags field reduces extensiveness of the spec (there are only
8 bits).
- Using many flags and adding a WebSocket-specific frame type will not make
httpbis people happy.
- Robustness against SPDY data reframing is important but this framing plan
doesn't have the property.

So the minimum appears to be 2 bits in SPDY/DATA frame headers:
> 1 bit for MSG_DONE on the last SPDY/DATA frame and
> 1 bit for MSG_TYPE (UTF8/binary) on the first SPDY/DATA frame that starts
> a new msg

Hmm, then, I think it is good to have
- MSG_DONE in flags for each frame
- MSG_TYPE at the head of a message
- RSVs at the head of a message
.

By the way, I still think it is preferable to decouple metadata from
payload.
Is it acceptable for you to represent a WS data Message by HEADERS + DATAs?
Specifically,
- A WebSocket data message is mapped to a HEADERS frame and multiple DATA
frames.
 - The last frame in a message turns MSG_DONE flag on.
 - The HEADERS frame consists of several headers
  - opcode
  - rsv1
  - rsv2
  - rsv3
- A WebSocket CLOSE message is mapped to a HEADERS frame.
 - The HEADERS frame consists of several headers
  - opcode
  - code
  - reason
 - The HEADERS frame turns END_STREAM on
- A WebSocket PING message is mapped to SPDY PING message
- A WebSocket PONG message is mapped to SPDY PONG message

Thanks,



On Tue, Jan 21, 2014 at 2:52 AM, Peter Thorson <webmaster@zaphoyd.com>wrote=
:

>
> On Jan 20, 2014, at 11:35, Tobias Oberstein <tobias.oberstein@tavendo.de>
> wrote:
>
> >> Re: minimum bits
> >>
> >> I definitely agree that as written today only message done, stream
> done, and
> >> utf8/binary are necessary. Stream done is present already in the DATA
> frame
> >> type. This does seem to ignore the fact that RFC6455 defines 8 data
> message
> >> types though and might limit compatibility between the two in the
> future.
> >
> > Yep, that's the idea. Ignore other WS opcodes, since those are not
> exposed in the W3C API
> > anyway, and compatibility with the latter only was the supposed goal.
> >
> > There is 1 more API thing: W3C API lets me close a WS with "code" and
> "reason".
> >
> > How do would this be mapped?
>
> I=E2=80=99d probably map it exactly the way it is now in RFC6455. The pay=
load of
> an HTTP 2.0 frame with stream fin bit set would be the close code + reaso=
n.
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>

--001a11c130102050b304f0a15edc
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div>Thanks for comments and sorry for the late response.<=
/div><div><br></div>I now agree with your opinion that having both SPDY fra=
me and WS frame is redundant.<div>I would like extensions and having RSVs s=
eems a good idea.</div>
<div>As a result, I think it is OK not to support extensions that works on =
individual WS frames not WS messages. (i.e. We should support permessage-de=
flate.)</div><div><br></div><blockquote class=3D"gmail_quote" style=3D"marg=
in:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,20=
4);border-left-style:solid;padding-left:1ex">
- An HTTP 2.0 WEBSOCKET frame type could be used to deliver WebSocket messa=
ges. You=E2=80=99d need a FIN flag for the message boundaries, frame bounda=
ries would be just standard HTTP 2.0 frames. No need to repackage or refram=
e or make a distinction between =E2=80=9Cwebsocket=E2=80=9D frames and HTTP=
 2.0 frames.<br>
- Depending on how closely we want HTTP 2.0 WebSockets to match RFC6455 Web=
Sockets additional support could be added.. for example, define the WEBSOCK=
ET frame type=E2=80=99s flag bits as follows (full compatibility flag confi=
guration):</blockquote>
<div>I don&#39;t think it is a good idea, for various reasons.</div><div>- =
Using many flags field reduces extensiveness of the spec (there are only 8 =
bits).</div><div>- Using many flags and adding a WebSocket-specific frame t=
ype will not make httpbis people happy.</div>
<div>- Robustness against SPDY data reframing is important but this framing=
 plan doesn&#39;t have the property.</div><div><br></div><blockquote class=
=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;bo=
rder-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">
<span style=3D"font-family:arial,sans-serif;font-size:13px">So the minimum =
appears to be 2 bits in SPDY/DATA frame headers:</span><br style=3D"font-fa=
mily:arial,sans-serif;font-size:13px"><span style=3D"font-family:arial,sans=
-serif;font-size:13px">1 bit for MSG_DONE on the last SPDY/DATA frame and<b=
r>
</span><span style=3D"font-family:arial,sans-serif;font-size:13px">1 bit fo=
r MSG_TYPE (UTF8/binary) on the first SPDY/DATA frame that starts a new msg=
</span></blockquote><div>Hmm, then, I think it is good to have</div><div>
- MSG_DONE in flags for each frame</div><div>- MSG_TYPE at the head of a me=
ssage</div><div>- RSVs at the head of a message</div><div>.</div><div><br><=
/div><div>By the way, I still think it is preferable to decouple metadata f=
rom payload.<br>
</div><div>Is it acceptable for you to represent a WS data Message by HEADE=
RS + DATAs?</div><div>Specifically,=C2=A0</div><div>- A WebSocket data mess=
age is mapped to a HEADERS frame and multiple DATA frames.</div><div>=C2=A0=
- The last frame in a message turns MSG_DONE flag on.</div>
<div>=C2=A0- The HEADERS frame consists of several headers</div><div>=C2=A0=
 - opcode</div><div>=C2=A0 - rsv1</div><div>=C2=A0 - rsv2</div><div>=C2=A0 =
- rsv3</div><div>- A WebSocket CLOSE message is mapped to a HEADERS frame.<=
/div><div>=C2=A0- The HEADERS frame consists of several headers</div>
<div>=C2=A0 - opcode</div><div>=C2=A0 - code</div><div>=C2=A0 - reason</div=
><div>=C2=A0- The HEADERS frame turns END_STREAM on</div><div><div>- A WebS=
ocket PING message is mapped to SPDY PING message</div></div><div><div>- A =
WebSocket PONG message is mapped to SPDY PONG message</div>
</div><div><br></div><div>Thanks,</div><div><br></div></div><div class=3D"g=
mail_extra"><br><br><div class=3D"gmail_quote">On Tue, Jan 21, 2014 at 2:52=
 AM, Peter Thorson <span dir=3D"ltr">&lt;<a href=3D"mailto:webmaster@zaphoy=
d.com" target=3D"_blank">webmaster@zaphoyd.com</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div class=3D"im"><br>
On Jan 20, 2014, at 11:35, Tobias Oberstein &lt;<a href=3D"mailto:tobias.ob=
erstein@tavendo.de">tobias.oberstein@tavendo.de</a>&gt; wrote:<br>
<br>
&gt;&gt; Re: minimum bits<br>
&gt;&gt;<br>
&gt;&gt; I definitely agree that as written today only message done, stream=
 done, and<br>
&gt;&gt; utf8/binary are necessary. Stream done is present already in the D=
ATA frame<br>
&gt;&gt; type. This does seem to ignore the fact that RFC6455 defines 8 dat=
a message<br>
&gt;&gt; types though and might limit compatibility between the two in the =
future.<br>
&gt;<br>
&gt; Yep, that&#39;s the idea. Ignore other WS opcodes, since those are not=
 exposed in the W3C API<br>
&gt; anyway, and compatibility with the latter only was the supposed goal.<=
br>
&gt;<br>
&gt; There is 1 more API thing: W3C API lets me close a WS with &quot;code&=
quot; and &quot;reason&quot;.<br>
&gt;<br>
&gt; How do would this be mapped?<br>
<br>
</div>I=E2=80=99d probably map it exactly the way it is now in RFC6455. The=
 payload of an HTTP 2.0 frame with stream fin bit set would be the close co=
de + reason.<br>
<div class=3D"HOEnZb"><div class=3D"h5">___________________________________=
____________<br>
hybi mailing list<br>
<a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br>
<a href=3D"https://www.ietf.org/mailman/listinfo/hybi" target=3D"_blank">ht=
tps://www.ietf.org/mailman/listinfo/hybi</a><br>
</div></div></blockquote></div><br></div>

--001a11c130102050b304f0a15edc--

From jat@jaet.org  Thu Jan 23 11:23:09 2014
Return-Path: <jat@jaet.org>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 0C01D1A00F9 for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 11:23:09 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.977
X-Spam-Level: 
X-Spam-Status: No, score=-1.977 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_LOW=-0.7] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id tVvWT_4Btji6 for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 11:23:04 -0800 (PST)
Received: from mail-pb0-f49.google.com (mail-pb0-f49.google.com [209.85.160.49]) by ietfa.amsl.com (Postfix) with ESMTP id 4C9011A0061 for <hybi@ietf.org>; Thu, 23 Jan 2014 11:22:57 -0800 (PST)
Received: by mail-pb0-f49.google.com with SMTP id up15so2229840pbc.36 for <hybi@ietf.org>; Thu, 23 Jan 2014 11:22:56 -0800 (PST)
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=Y5CFW/1vYBferpE53XunfIunVKSISqmKjaTV1F9H+cQ=; b=kRPrkFu4j5amgsUJr/YnlzBXJd8k1WFCvdCAsjQOt7y9zk4vgWFrqjVnlSxjPibSz+ DIXQNzfnc1x3V3tACcYLpT6vSCxiV+O7bSeRSc/5YkUdqh/FRD3AFIgtG9XWKPNWNJ2/ 4Z/pieLJNpXqpwMWVpzlM5hcgw3KWxki0t/yxyYqbfdJ4nVEqseRz2et2LyAhToIDH2J I1ylQFDkbyFEm/3SRWDMVkT4VAzC5hTa/b+fvcrxFcfKFwa03L4xhUIiQOG3LfQ1+QjA omXKVyJ0cBjsuNKlOTkhmDRCQmrXSNemttJTeBQMiZsxbGAJ5xbynXpfp1DZL9uXNOPF 2BSw==
X-Gm-Message-State: ALoCoQny72y1oTPkgI1h1MNwPSZv2m5bPTdb+gQ4tN1rG4YZANq8XfQkFfFe+ElnAWhDfkzEjJPf
X-Received: by 10.67.4.169 with SMTP id cf9mr9816681pad.45.1390504976533; Thu, 23 Jan 2014 11:22:56 -0800 (PST)
MIME-Version: 1.0
Received: by 10.68.200.37 with HTTP; Thu, 23 Jan 2014 11:22:36 -0800 (PST)
X-Originating-IP: [198.49.103.3]
In-Reply-To: <634914A010D0B943A035D226786325D4446C083400@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF991A9@EXVMBX020-12.exch020.serverdata.net> <634914A010D0B943A035D226786325D4446BF9987B@EXVMBX020-12.exch020.serverdata.net> <CAHixhFr6wPQQRFrV72eXp9cUA=m1Qvxd2ZhYpqPXFa9pRhnZcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446C083400@EXVMBX020-12.exch020.serverdata.net>
From: "John A. Tamplin" <jat@jaet.org>
Date: Thu, 23 Jan 2014 14:22:36 -0500
Message-ID: <CAM5k6X_gjYoS8Ah64MO84H_r0TDWKh_Eh9_BNHzh9T7mquHhxQ@mail.gmail.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=047d7b1601572f64a904f0a82b30
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] permessage-priority extension draft
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 23 Jan 2014 19:23:09 -0000

--047d7b1601572f64a904f0a82b30
Content-Type: text/plain; charset=UTF-8

On Thu, Jan 23, 2014 at 5:37 AM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> What is probably missing is:
>
>
>
> "An intermediary that is not aware of the semantics of negotiated
> extensions MUST ignore the rules that
>
> normally apply to continuation frames (opcode 0) and FIN bit"
>
>
>
> Since without such rule, a (verifying) intermediary unaware of the
> extension semantics might comply
>
> by not changing fragmentation, but still continue checking and (falsely)
> failing the WS connection if
>
> it believes the continuation frame / FIN rules haven't been followed. And
> those rules might have
>
> been changed by the extension.
>

I disagree - I don't think an extension should be allowed to change those
rules.  If you don't even have the framing, then really there is no
standard WebSocket protocol.

-- 
John A. Tamplin

--047d7b1601572f64a904f0a82b30
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">On Thu, Jan 23, 2014 at 5:37 AM, Tobias Oberstein <span di=
r=3D"ltr">&lt;<a href=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_bla=
nk" class=3D"cremed">tobias.oberstein@tavendo.de</a>&gt;</span> wrote:<br><=
div class=3D"gmail_extra">

<div class=3D"gmail_quote"><blockquote class=3D"gmail_quote" style=3D"margi=
n:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div lang=3D"DE" =
link=3D"blue" vlink=3D"purple"><p class=3D"MsoNormal"><span style=3D"color:=
rgb(31,73,125);font-family:Calibri,sans-serif;font-size:11pt">What is proba=
bly missing is:</span><br>

</p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quo=
t;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u><=
/span></p><p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-famil=
y:&quot;Calibri&quot;,&quot;sans-serif&quot;;color:#1f497d">&quot;An interm=
ediary that is not aware of the semantics of negotiated extensions MUST ign=
ore the rules that<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">normally apply to continu=
ation frames (opcode 0) and FIN bit&quot;<u></u><u></u></span></p><p class=
=3D"MsoNormal">

<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-=
serif&quot;;color:#1f497d"><u></u>=C2=A0<u></u></span></p><p class=3D"MsoNo=
rmal"><span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot=
;sans-serif&quot;;color:#1f497d">Since without such rule, a (verifying) int=
ermediary unaware of the extension semantics might comply<u></u><u></u></sp=
an></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">by not changing fragmenta=
tion, but still continue checking and (falsely) failing the WS connection i=
f<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">it believes the continuat=
ion frame / FIN rules haven&#39;t been followed. And those rules might have=
<u></u><u></u></span></p>

<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:#1f497d">been changed by the exten=
sion.</span></p></div></blockquote></div><br>I disagree - I don&#39;t think=
 an extension should be allowed to change those rules. =C2=A0If you don&#39=
;t even have the framing, then really there is no standard WebSocket protoc=
ol.<br clear=3D"all">

<div><br></div>-- <br>John A. Tamplin
</div></div>

--047d7b1601572f64a904f0a82b30--

From jat@jaet.org  Thu Jan 23 11:27:46 2014
Return-Path: <jat@jaet.org>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 6DCB81A0061 for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 11:27:46 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.977
X-Spam-Level: 
X-Spam-Status: No, score=-1.977 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_LOW=-0.7] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Glk3iyHBhX2Y for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 11:27:44 -0800 (PST)
Received: from mail-pa0-f50.google.com (mail-pa0-f50.google.com [209.85.220.50]) by ietfa.amsl.com (Postfix) with ESMTP id BFB1B1A004C for <hybi@ietf.org>; Thu, 23 Jan 2014 11:27:44 -0800 (PST)
Received: by mail-pa0-f50.google.com with SMTP id kp14so2256075pab.37 for <hybi@ietf.org>; Thu, 23 Jan 2014 11:27:44 -0800 (PST)
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=Mu2vQM++YIlh3eG9MrKir6C5xP14RN2vhUvMRIdRL4E=; b=iJ6UCOby7QRU1k9nqJh+mS84XkTKkCmEtZI+U7+hrK+2WcyKzhYOFk45YbHMgItpJ9 6BvQGNF+Pxc0AIDc1ADmiSoBrY/dbiXR1icItbBaTEi6HpGPk3Cb47PpXt/kLJ5ZCzYs VOA4ho1ObPhoD86ruvrwp1/W2IoFZnZu0Ttj/ubhRmS/41NYV0/8H3NDLDkvKSyZPHLV wNfZlguYg2sy1vDRqew7lpvpIFbm/hKqwDbVlMQ1Yi2hxmcJIds9EKRaYO52cI4r3ylT QBb2VZ3WAHUFvYFTsL8sJsUza6eYA1Fdbuooa/K3eLPDKudPuEakNIkFWCVsUnmjt/op XMUw==
X-Gm-Message-State: ALoCoQmpawT7v/d0bOyf36VTFnVA/KAL3NFl3dC/a6bZLyoJ7Yg4KlkWvD5/fjHBzldjuYBgjYBV
X-Received: by 10.66.14.41 with SMTP id m9mr9898637pac.123.1390505264007; Thu, 23 Jan 2014 11:27:44 -0800 (PST)
MIME-Version: 1.0
Received: by 10.68.200.37 with HTTP; Thu, 23 Jan 2014 11:27:23 -0800 (PST)
X-Originating-IP: [198.49.103.3]
In-Reply-To: <634914A010D0B943A035D226786325D4446C083400@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF991A9@EXVMBX020-12.exch020.serverdata.net> <634914A010D0B943A035D226786325D4446BF9987B@EXVMBX020-12.exch020.serverdata.net> <CAHixhFr6wPQQRFrV72eXp9cUA=m1Qvxd2ZhYpqPXFa9pRhnZcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446C083400@EXVMBX020-12.exch020.serverdata.net>
From: "John A. Tamplin" <jat@jaet.org>
Date: Thu, 23 Jan 2014 14:27:23 -0500
Message-ID: <CAM5k6X8cRGtUHp=jTSW5=h05ZTFSnm0Ow-Bak+XYC_qCT8LjXA@mail.gmail.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=bcaec520f51751ec8d04f0a83c47
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] permessage-priority extension draft
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 23 Jan 2014 19:27:46 -0000

--bcaec520f51751ec8d04f0a83c47
Content-Type: text/plain; charset=UTF-8

On Thu, Jan 23, 2014 at 5:37 AM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> And in the end: why do we have frames at all in WS if we don't make good
> use of them?  ;)
>

The primary reason is so we can write streams of data we don't know the
size of up front without having to buffer the whole thing.  Aside from
that, there are benefits for small-memory senders to limit the buffers
required.

-- 
John A. Tamplin

--bcaec520f51751ec8d04f0a83c47
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">On Thu, Jan 23, 2014 at 5:37 AM, Tobias Oberstein <span di=
r=3D"ltr">&lt;<a href=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_bla=
nk" class=3D"cremed">tobias.oberstein@tavendo.de</a>&gt;</span> wrote:<br><=
div class=3D"gmail_extra">

<div class=3D"gmail_quote"><blockquote class=3D"gmail_quote" style=3D"margi=
n:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div lang=3D"DE" =
link=3D"blue" vlink=3D"purple"><p class=3D"MsoNormal"><span style=3D"color:=
rgb(31,73,125);font-family:Calibri,sans-serif;font-size:11pt">And in the en=
d: why do we have frames at all in WS if we don&#39;t make good use of them=
? =C2=A0;)</span></p>

</div></blockquote></div><br>The primary reason is so we can write streams =
of data we don&#39;t know the size of up front without having to buffer the=
 whole thing. =C2=A0Aside from that, there are benefits for small-memory se=
nders to limit the buffers required.<br clear=3D"all">

<div><br></div>-- <br>John A. Tamplin
</div></div>

--bcaec520f51751ec8d04f0a83c47--

From tobias.oberstein@tavendo.de  Thu Jan 23 12:29:28 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 48E5D1A00ED for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 12:29:28 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id KU9DbQ9LAq4z for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 12:29:26 -0800 (PST)
Received: from EXHUB020-3.exch020.serverdata.net (exhub020-3.exch020.serverdata.net [206.225.164.30]) by ietfa.amsl.com (Postfix) with ESMTP id 4F95F1A00C5 for <hybi@ietf.org>; Thu, 23 Jan 2014 12:29:26 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-3.exch020.serverdata.net ([206.225.164.30]) with mapi; Thu, 23 Jan 2014 12:29:25 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: "John A. Tamplin" <jat@jaet.org>
Date: Thu, 23 Jan 2014 12:29:20 -0800
Thread-Topic: [hybi] permessage-priority extension draft
Thread-Index: Ac8YcIYvuQY/GTrLSWaIZi27aV5qqgAB/Y/Q
Message-ID: <634914A010D0B943A035D226786325D4446C13ADC6@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF991A9@EXVMBX020-12.exch020.serverdata.net> <634914A010D0B943A035D226786325D4446BF9987B@EXVMBX020-12.exch020.serverdata.net> <CAHixhFr6wPQQRFrV72eXp9cUA=m1Qvxd2ZhYpqPXFa9pRhnZcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446C083400@EXVMBX020-12.exch020.serverdata.net> <CAM5k6X_gjYoS8Ah64MO84H_r0TDWKh_Eh9_BNHzh9T7mquHhxQ@mail.gmail.com>
In-Reply-To: <CAM5k6X_gjYoS8Ah64MO84H_r0TDWKh_Eh9_BNHzh9T7mquHhxQ@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] permessage-priority extension draft
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 23 Jan 2014 20:29:28 -0000
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From tobias.oberstein@tavendo.de  Thu Jan 23 12:33:45 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id AE0FA1A013C for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 12:33:45 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 5Nt55suEwGbw for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 12:33:42 -0800 (PST)
Received: from EXHUB020-1.exch020.serverdata.net (exhub020-1.exch020.serverdata.net [206.225.164.28]) by ietfa.amsl.com (Postfix) with ESMTP id F357B1A0138 for <hybi@ietf.org>; Thu, 23 Jan 2014 12:33:41 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-1.exch020.serverdata.net ([206.225.164.28]) with mapi; Thu, 23 Jan 2014 12:33:40 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: "John A. Tamplin" <jat@jaet.org>
Date: Thu, 23 Jan 2014 12:33:34 -0800
Thread-Topic: [hybi] permessage-priority extension draft
Thread-Index: Ac8YcTFhXtiUkh2ETluPyTOHFI/OlwACKM5Q
Message-ID: <634914A010D0B943A035D226786325D4446C13ADD5@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF991A9@EXVMBX020-12.exch020.serverdata.net> <634914A010D0B943A035D226786325D4446BF9987B@EXVMBX020-12.exch020.serverdata.net> <CAHixhFr6wPQQRFrV72eXp9cUA=m1Qvxd2ZhYpqPXFa9pRhnZcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446C083400@EXVMBX020-12.exch020.serverdata.net> <CAM5k6X8cRGtUHp=jTSW5=h05ZTFSnm0Ow-Bak+XYC_qCT8LjXA@mail.gmail.com>
In-Reply-To: <CAM5k6X8cRGtUHp=jTSW5=h05ZTFSnm0Ow-Bak+XYC_qCT8LjXA@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] permessage-priority extension draft
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 23 Jan 2014 20:33:45 -0000
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From ricea@google.com  Thu Jan 23 22:41:12 2014
Return-Path: <ricea@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id BB0791A013A for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 22:41:12 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.913
X-Spam-Level: 
X-Spam-Status: No, score=-1.913 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id tUjpWGBEzUFt for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 22:41:11 -0800 (PST)
Received: from mail-oa0-x22d.google.com (mail-oa0-x22d.google.com [IPv6:2607:f8b0:4003:c02::22d]) by ietfa.amsl.com (Postfix) with ESMTP id 4264B1A00EA for <hybi@ietf.org>; Thu, 23 Jan 2014 22:41:11 -0800 (PST)
Received: by mail-oa0-f45.google.com with SMTP id i11so3374260oag.32 for <hybi@ietf.org>; Thu, 23 Jan 2014 22:41:09 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type; bh=S66K4flaYATMXPYDHVoRBqnoNnPKeqplT3raNKtiQ3Y=; b=LMhhnO3HTpC1SDjw9jZR8CYvWqJFDDPklgq0wxeF9YTSt7EnWBh5XzjR4SXDUolzyG wddDFz8gdSow9HIMvcss+3ARQPn+vs4BEHMl8a/SGwRVnS6zT8JE5el16wIGP5FZIEgM CXmwtbh7LqjbnxKQXsFWFCR/IpY/1PPFBUDJJJm9nL9X6OleKYXE+8RyzQ6QHPSBa7Mx YjCQGRpLNYV8Ynz3LXmhWNGwFqB2qpq9HEtp6HaMEMT3WvNwb1SBr6C94JAk0pp3TU4B gJtAg2SDaPKKQRlACO/5Fs85vK4ekTj2hMNs+DnCNr77kwZAI6IUd2IObPxZ3AVX2xNP D0kg==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=S66K4flaYATMXPYDHVoRBqnoNnPKeqplT3raNKtiQ3Y=; b=CQz04Kqc+MBXwuwMvpyJvK3Weq8mUoNsiRmlCsFe35UQ76nQ73BioERd6VLmQ2iuK6 Hv2XTtgR5A4Z45hssgh6xElluL+MK+ysfonLHyUZM/1Hc+Rn6ycw2+9BP9Y+TeqpHG5e mktCM8leCDWu2DMKVblWvie02kCLFjt0F2AGOPPVwkPiMLZyLiERtrOQzBHQTlDm+ytI KE5HoS7396c8hT2pIBqvQDqbuKhvFVJXWyeLQ6zS4rtshynBkKgy9mkGyn/SuYdTcp56 bq+hUL1gKW4bXHzT/Q/ohx4aZUgWdAFSx1UKO4G4JuH48qfiE3kQMTidhHfEEkJuZMh9 oUpA==
X-Gm-Message-State: ALoCoQkLyEEj4JvUt6V4iKUpn+/V3v/awR9HEnOmpVc/i+AM8TwFem7XqjzjRsbVvrShS6tpeAGpRx2+jT/ewXWfmHYIQxER730jY7fC1ba5jlmkwG6UN9/FnGEGiYE5YQdie53gCunamn6gyazBLOcnNHByQPU21VnlkO1xQVcwY8hh1AsnO0a8PWopngqFnFOcrUYNhkVp
MIME-Version: 1.0
X-Received: by 10.60.138.228 with SMTP id qt4mr1330634oeb.38.1390545668981; Thu, 23 Jan 2014 22:41:08 -0800 (PST)
Received: by 10.182.118.41 with HTTP; Thu, 23 Jan 2014 22:41:08 -0800 (PST)
In-Reply-To: <CAM5k6X8cRGtUHp=jTSW5=h05ZTFSnm0Ow-Bak+XYC_qCT8LjXA@mail.gmail.com>
References: <634914A010D0B943A035D226786325D4446BF991A9@EXVMBX020-12.exch020.serverdata.net> <634914A010D0B943A035D226786325D4446BF9987B@EXVMBX020-12.exch020.serverdata.net> <CAHixhFr6wPQQRFrV72eXp9cUA=m1Qvxd2ZhYpqPXFa9pRhnZcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446C083400@EXVMBX020-12.exch020.serverdata.net> <CAM5k6X8cRGtUHp=jTSW5=h05ZTFSnm0Ow-Bak+XYC_qCT8LjXA@mail.gmail.com>
Date: Fri, 24 Jan 2014 15:41:08 +0900
Message-ID: <CAHixhFpZc8_=Rb7cj-iai-2atbu6fr1jPPUCjfo-VdRMA0p7iw@mail.gmail.com>
From: Adam Rice <ricea@google.com>
To: "John A. Tamplin" <jat@jaet.org>
Content-Type: multipart/alternative; boundary=047d7b342c96a4de3804f0b1a4a5
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] permessage-priority extension draft
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 24 Jan 2014 06:41:12 -0000

--047d7b342c96a4de3804f0b1a4a5
Content-Type: text/plain; charset=UTF-8

On 24 January 2014 04:27, John A. Tamplin <jat@jaet.org> wrote:
>
> The primary reason is so we can write streams of data we don't know the
> size of up front without having to buffer the whole thing.  Aside from
> that, there are benefits for small-memory senders to limit the buffers
> required.
>

Another legitimate use is to allow the timely sending of control frames.

For example, a server sending a large message down a slow link might want
to interrupt the message with a "Ping" every few minutes to elicit a "Pong"
from the client and keep the NAT mapping alive.

Adam Rice

--047d7b342c96a4de3804f0b1a4a5
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On 2=
4 January 2014 04:27, John A. Tamplin <span dir=3D"ltr">&lt;<a href=3D"mail=
to:jat@jaet.org" target=3D"_blank" class=3D"cremed">jat@jaet.org</a>&gt;</s=
pan> wrote:<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;bor=
der-left:1px #ccc solid;padding-left:1ex">
<div dir=3D"ltr"><div class=3D"gmail_extra">The primary reason is so we can=
 write streams of data we don&#39;t know the size of up front without havin=
g to buffer the whole thing. =C2=A0Aside from that, there are benefits for =
small-memory senders to limit the buffers required.</div>
</div></blockquote><div><br></div><div>Another legitimate use is to allow t=
he timely sending of control frames.</div><div><br></div><div>For example, =
a server sending a large message down a slow link might want to interrupt t=
he message with a &quot;Ping&quot; every few minutes to elicit a &quot;Pong=
&quot; from the client and keep the NAT mapping alive.</div>
<div><br></div><div>Adam Rice</div></div></div></div>

--047d7b342c96a4de3804f0b1a4a5--

From duerst@it.aoyama.ac.jp  Thu Jan 23 23:33:12 2014
Return-Path: <duerst@it.aoyama.ac.jp>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9561E1A015F for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 23:33:12 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: 0.374
X-Spam-Level: 
X-Spam-Status: No, score=0.374 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HELO_EQ_JP=1.244, HOST_EQ_JP=1.265, MIME_8BIT_HEADER=0.3, RP_MATCHES_RCVD=-0.535] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ZyObL6MFAhbl for <hybi@ietfa.amsl.com>; Thu, 23 Jan 2014 23:33:11 -0800 (PST)
Received: from scintmta01.scbb.aoyama.ac.jp (scintmta01.scbb.aoyama.ac.jp [133.2.253.33]) by ietfa.amsl.com (Postfix) with ESMTP id DB8771A010E for <hybi@ietf.org>; Thu, 23 Jan 2014 23:33:10 -0800 (PST)
Received: from scmse02.scbb.aoyama.ac.jp ([133.2.253.231]) by scintmta01.scbb.aoyama.ac.jp (secret/secret) with SMTP id s0O7X5cK014876; Fri, 24 Jan 2014 16:33:05 +0900
Received: from (unknown [133.2.206.134]) by scmse02.scbb.aoyama.ac.jp with smtp id 61c3_59b5_c3973352_84c9_11e3_8346_001e6722eec2; Fri, 24 Jan 2014 16:33:04 +0900
Received: from [IPv6:::1] (unknown [133.2.210.1]) by itmail2.it.aoyama.ac.jp (Postfix) with ESMTP id EC636BF54C; Fri, 24 Jan 2014 16:33:04 +0900 (JST)
Message-ID: <52E2171F.4030602@it.aoyama.ac.jp>
Date: Fri, 24 Jan 2014 16:32:47 +0900
From: =?UTF-8?B?Ik1hcnRpbiBKLiBEw7xyc3Qi?= <duerst@it.aoyama.ac.jp>
Organization: Aoyama Gakuin University
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.1.9) Gecko/20100722 Eudora/3.0.4
MIME-Version: 1.0
To: Adam Rice <ricea@google.com>
References: <634914A010D0B943A035D226786325D4446BF991A9@EXVMBX020-12.exch020.serverdata.net> <634914A010D0B943A035D226786325D4446BF9987B@EXVMBX020-12.exch020.serverdata.net> <CAHixhFr6wPQQRFrV72eXp9cUA=m1Qvxd2ZhYpqPXFa9pRhnZcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446C083400@EXVMBX020-12.exch020.serverdata.net> <CAM5k6X8cRGtUHp=jTSW5=h05ZTFSnm0Ow-Bak+XYC_qCT8LjXA@mail.gmail.com> <CAHixhFpZc8_=Rb7cj-iai-2atbu6fr1jPPUCjfo-VdRMA0p7iw@mail.gmail.com>
In-Reply-To: <CAHixhFpZc8_=Rb7cj-iai-2atbu6fr1jPPUCjfo-VdRMA0p7iw@mail.gmail.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] permessage-priority extension draft
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 24 Jan 2014 07:33:12 -0000

On 2014/01/24 15:41, Adam Rice wrote:

> Another legitimate use is to allow the timely sending of control frames.
>
> For example, a server sending a large message down a slow link might want
> to interrupt the message with a "Ping" every few minutes to elicit a "Pong"
> from the client and keep the NAT mapping alive.

Wouldn't the NAT mapping be kept alive anyway as long as data is flowing?

Regards,   Martin.

From webmaster@zaphoyd.com  Sat Jan 25 08:17:42 2014
Return-Path: <webmaster@zaphoyd.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 837FA1A03D9 for <hybi@ietfa.amsl.com>; Sat, 25 Jan 2014 08:17:42 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.899
X-Spam-Level: 
X-Spam-Status: No, score=-1.899 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id iS-G2J7D2QQT for <hybi@ietfa.amsl.com>; Sat, 25 Jan 2014 08:17:38 -0800 (PST)
Received: from authsmtp01.uchicago.edu (authsmtp01.uchicago.edu [128.135.12.121]) by ietfa.amsl.com (Postfix) with ESMTP id A55C81A03CA for <hybi@ietf.org>; Sat, 25 Jan 2014 08:17:38 -0800 (PST)
Received: from [10.0.1.3] (c-68-51-76-178.hsd1.il.comcast.net [68.51.76.178]) (Authenticated sender: zaphoyd) by authsmtp01.uchicago.edu (Postfix) with ESMTP id 61E43498039; Sat, 25 Jan 2014 10:17:36 -0600 (CST)
Content-Type: multipart/alternative; boundary="Apple-Mail=_AD79EEFA-161C-47BD-8AC9-9EBC7025FA89"
Mime-Version: 1.0 (Mac OS X Mail 7.0 \(1822\))
From: Peter Thorson <webmaster@zaphoyd.com>
In-Reply-To: <CABihn6GSrU8oJa4nbMJiDDo2Z+-c4fT69roVFxcHYM4XYFToRQ@mail.gmail.com>
Date: Sat, 25 Jan 2014 10:24:09 -0600
Message-Id: <F1F115D4-29EB-457B-B02C-71E46D3E1C81@zaphoyd.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net> <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com> <634914A010D0B943A035D226786325D4446C0826EC@EXVMBX020-12. exch020.serverdata.net> <6652CC7E-7E92-40EF-BD50-2DDD39E15533@zaphoyd.com> <CABihn6GSrU8oJa4nbMJiDDo2Z+-c4fT69roVFxcHYM4XYFToRQ@mail.gmail.com>
To: Yutaka Hirano <yhirano@chromium.org>
X-Mailer: Apple Mail (2.1822)
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sat, 25 Jan 2014 16:17:42 -0000

--Apple-Mail=_AD79EEFA-161C-47BD-8AC9-9EBC7025FA89
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=windows-1252

Critical first question here: Do we have the cooperation of the HTTP 2.0 =
working group or is the WS over HTTP 2.0 effort something being done =
unilaterally by the WebSocket working group?

If we are working unilaterally I think the best option is to tunnel raw =
RFC6455 over a standard HTTP 2.0 data channel and leave it at that. By =
doing this..
- We retain full control over the standard and protocol and only have to =
deal with a single protocol when considering future versions, =
extensions, and other errata fixes.
- We can re-use existing RFC6455 parsing code. Such code is readily =
available in libraries and already present in many pieces of software =
that will likely implement HTTP 2.0.
- Tunneling over HTTP 2.0 is more efficient in terms of latency and byte =
overhead than proposals that try and use HTTP 2.0 framing without =
cooperation from the HTTP 2.0 standard itself.
- Tunneling makes it fairly easy for intermediaries to be built that =
package or extract a WebSocket stream from the HTTP 2.0 stream and route =
it elsewhere. Other sorts of intermediaries are no more difficult to =
construct than they are today where one must be able to read both HTTP =
1.1 and RFC6455.

The only reason I see to do something else would be if we can reduce =
complexity and improve performance. The HTTP 2.0 spec has goals stated =
as: "HTTP/2.0 enables a more efficient use of network resources and a =
reduced perception of latency.". Any proposal for WS over HTTP 2.0 that =
makes things slower doesn't make sense. HTTP 2.0's single TCP connection =
introduces some improvements but its internal mechanisms (framing, flow =
control, header compression, etc) introduce significant (as compared to =
RFC6455) additional processing, latency, and framing overhead.

Compared to a separate HTTP and WS connection, the performance of =
tunneling WS over HTTP likely to be better for certain things (initial =
startup time, TCP packet coalescing) but worse at others (message =
latency, message overhead after HTTP resources are done loading). Any =
proposal for more direct framing needs to improve on the performance of =
tunneling to justify the complexity of maintaining two separate specs =
and code paths.

I believe the proposal that uses a HEADERS frame + DATA frames is worse =
than just tunneling raw RFC6455. A headers frame alone is 12 bytes not =
including the payload. Assume we leave out headers with default values. =
Assume we get the 88% compression ratio Google saw in its tests of full =
HTTP header compression. Even then we are talking 15+ bytes on the wire =
to send 7 bits of data, of which 4-6 are "reserved for future use". Add =
another 8 bytes per data frame. We are talking invoking string header =
processing systems and the fixed costs of spinning up the header =
compression engine and enforcing its strict ordering guarantees for 1-2 =
headers on every message. This is nearly a full blown HTTP request.

My initial proposal that uses a dedicated WEBSOCKET frame type was =
specifically set up to improve on the performance of the tunneling =
strategy as an ordered bidirectional messaging channel while maintaining =
100% compatibility with all of the extensibility built into RFC6455. It =
does that by making full use of the extensibility features of HTTP 2.0 =
itself. I do not think that using all of the flags for this type reduces =
the extensibility of the spec because it matches RFC6455 exactly which =
itself leaves half of those bits "reserved for future use".

It may *look* like there are no more flags available for future use, but =
in reality most are still reserved for future use just in an order that =
is compatible with RFC6455. The use of a WebSocket specific HTTP 2.0 =
frame type means that using all of the flags here doesn't affect the =
extensibility of the other types of frames used for HTTP data and the =
separate processing rules are cleanly delineated.

If we are willing to break perfect RFC6455 compatibility we should get =
rid of the opcode as well. This wouldn't change the wire overhead but =
would get rid of more redundant logic, free up more flag bits, and make =
the whole thing more cohesive with the rest of HTTP 2.0. We'd be giving =
up the ability to transparently support extensions and future protocol =
versions that define new opcodes (we could still deploy the saved bits =
explicitly as needed to support them). It would allow using regular DATA =
frames as well (though I still think a dedicated messaging frame type is =
beneficial).

Regarding robustness against reframing: There is nothing that would =
prevent dedicated WEBSOCKET frames from being split or coalesced. The =
intermediary would just need to understand whatever mechanism is used to =
determine message boundaries (either opcode or MSG_DONE bit). This is =
identical to the methods of reframing regular DATA frames.

On Jan 23, 2014, at 5:16, Yutaka Hirano <yhirano@chromium.org> wrote:

> Thanks for comments and sorry for the late response.
>=20
> I now agree with your opinion that having both SPDY frame and WS frame =
is redundant.
> I would like extensions and having RSVs seems a good idea.
> As a result, I think it is OK not to support extensions that works on =
individual WS frames not WS messages. (i.e. We should support =
permessage-deflate.)
>=20
> - An HTTP 2.0 WEBSOCKET frame type could be used to deliver WebSocket =
messages. You=92d need a FIN flag for the message boundaries, frame =
boundaries would be just standard HTTP 2.0 frames. No need to repackage =
or reframe or make a distinction between =93websocket=94 frames and HTTP =
2.0 frames.
> - Depending on how closely we want HTTP 2.0 WebSockets to match =
RFC6455 WebSockets additional support could be added.. for example, =
define the WEBSOCKET frame type=92s flag bits as follows (full =
compatibility flag configuration):
> I don't think it is a good idea, for various reasons.
> - Using many flags field reduces extensiveness of the spec (there are =
only 8 bits).
> - Using many flags and adding a WebSocket-specific frame type will not =
make httpbis people happy.
> - Robustness against SPDY data reframing is important but this framing =
plan doesn't have the property.
>=20
> So the minimum appears to be 2 bits in SPDY/DATA frame headers:
> 1 bit for MSG_DONE on the last SPDY/DATA frame and
> 1 bit for MSG_TYPE (UTF8/binary) on the first SPDY/DATA frame that =
starts a new msg
> Hmm, then, I think it is good to have
> - MSG_DONE in flags for each frame
> - MSG_TYPE at the head of a message
> - RSVs at the head of a message
> .
>=20
> By the way, I still think it is preferable to decouple metadata from =
payload.
> Is it acceptable for you to represent a WS data Message by HEADERS + =
DATAs?
> Specifically,=20
> - A WebSocket data message is mapped to a HEADERS frame and multiple =
DATA frames.
>  - The last frame in a message turns MSG_DONE flag on.
>  - The HEADERS frame consists of several headers
>   - opcode
>   - rsv1
>   - rsv2
>   - rsv3
> - A WebSocket CLOSE message is mapped to a HEADERS frame.
>  - The HEADERS frame consists of several headers
>   - opcode
>   - code
>   - reason
>  - The HEADERS frame turns END_STREAM on
> - A WebSocket PING message is mapped to SPDY PING message
> - A WebSocket PONG message is mapped to SPDY PONG message
>=20
> Thanks,
>=20
>=20
>=20
> On Tue, Jan 21, 2014 at 2:52 AM, Peter Thorson <webmaster@zaphoyd.com> =
wrote:
>=20
> On Jan 20, 2014, at 11:35, Tobias Oberstein =
<tobias.oberstein@tavendo.de> wrote:
>=20
> >> Re: minimum bits
> >>
> >> I definitely agree that as written today only message done, stream =
done, and
> >> utf8/binary are necessary. Stream done is present already in the =
DATA frame
> >> type. This does seem to ignore the fact that RFC6455 defines 8 data =
message
> >> types though and might limit compatibility between the two in the =
future.
> >
> > Yep, that's the idea. Ignore other WS opcodes, since those are not =
exposed in the W3C API
> > anyway, and compatibility with the latter only was the supposed =
goal.
> >
> > There is 1 more API thing: W3C API lets me close a WS with "code" =
and "reason".
> >
> > How do would this be mapped?
>=20
> I=92d probably map it exactly the way it is now in RFC6455. The =
payload of an HTTP 2.0 frame with stream fin bit set would be the close =
code + reason.
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>=20


--Apple-Mail=_AD79EEFA-161C-47BD-8AC9-9EBC7025FA89
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
	charset=windows-1252

<html><head><meta http-equiv=3D"Content-Type" content=3D"text/html =
charset=3Dwindows-1252"><meta http-equiv=3D"Content-Type" =
content=3D"text/html charset=3Dwindows-1252"></head><body =
style=3D"word-wrap: break-word; -webkit-nbsp-mode: space; =
-webkit-line-break: after-white-space;"><div>Critical first question =
here: Do we have the cooperation of the HTTP 2.0 working group or is the =
WS over HTTP 2.0 effort something being done unilaterally by the =
WebSocket working group?</div><div><br></div><div>If we are working =
unilaterally I think the best option is to tunnel raw RFC6455 over a =
standard HTTP 2.0 data channel and leave it at that. By doing =
this..</div><div>- We retain full control over the standard and protocol =
and only have to deal with a single protocol when considering future =
versions, extensions, and other errata fixes.</div><div>- We can re-use =
existing RFC6455 parsing code. Such code is readily available in =
libraries and already present in many pieces of software that will =
likely implement HTTP 2.0.</div><div>- Tunneling over HTTP 2.0 is more =
efficient in terms of latency and byte overhead than proposals that try =
and use HTTP 2.0 framing without cooperation from the HTTP 2.0 standard =
itself.</div><div>- Tunneling makes it fairly easy for intermediaries to =
be built that package or extract a WebSocket stream from the HTTP 2.0 =
stream and route it elsewhere. Other sorts of intermediaries are no more =
difficult to construct than they are today where one must be able to =
read both HTTP 1.1 and RFC6455.</div><div><br></div><div>The only reason =
I see to do something else would be if we can reduce complexity and =
improve performance. The HTTP 2.0 spec has goals stated as: "HTTP/2.0 =
enables a more efficient use of network resources and a reduced =
perception of latency.". Any proposal for WS over HTTP 2.0 that makes =
things slower doesn't make sense. HTTP 2.0's single TCP connection =
introduces some improvements but its internal mechanisms (framing, flow =
control, header compression, etc) introduce significant (as compared to =
RFC6455) additional processing, latency, and framing =
overhead.</div><div><br></div><div>Compared to a separate HTTP and WS =
connection, the performance of tunneling WS over HTTP likely to be =
better for certain things (initial startup time, TCP packet coalescing) =
but worse at others (message latency, message overhead after HTTP =
resources are done loading). Any proposal for more direct framing needs =
to improve on the performance of tunneling to justify the complexity of =
maintaining two separate specs and code =
paths.</div><div><br></div><div>I believe the proposal that uses a =
HEADERS frame + DATA frames is worse than just tunneling raw RFC6455. A =
headers frame alone is 12 bytes not including the payload. Assume we =
leave out headers with default values. Assume we get the 88% compression =
ratio Google saw in its tests of full HTTP header compression. Even then =
we are talking 15+ bytes on the wire to send 7 bits of data, of which =
4-6 are "reserved for future use". Add another 8 bytes per data frame. =
We are talking invoking string header processing systems and the fixed =
costs of spinning up the header compression engine and enforcing its =
strict ordering guarantees for 1-2 headers on every message. This is =
nearly a full blown HTTP request.</div><div><br></div><div>My initial =
proposal that uses a dedicated WEBSOCKET frame type was specifically set =
up to improve on the performance of the tunneling strategy as an ordered =
bidirectional messaging channel while maintaining 100% compatibility =
with all of the extensibility built into RFC6455. It does that by making =
full use of the extensibility features of HTTP 2.0 itself. I do not =
think that using all of the flags for this type reduces the =
extensibility of the spec because it matches RFC6455 exactly which =
itself leaves half of those bits "reserved for future =
use".</div><div><br></div><div>It may *look* like there are no more =
flags available for future use, but in reality most are still reserved =
for future use just in an order that is compatible with RFC6455. The use =
of a WebSocket specific HTTP 2.0 frame type means that using all of the =
flags here doesn't affect the extensibility of the other types of frames =
used for HTTP data and the separate processing rules are cleanly =
delineated.</div><div><br></div><div>If we are willing to break perfect =
RFC6455 compatibility we should get rid of the opcode as well. This =
wouldn't change the wire overhead but would get rid of more redundant =
logic, free up more flag bits, and make the whole thing more cohesive =
with the rest of HTTP 2.0. We'd be giving up the ability to =
transparently support extensions and future protocol versions that =
define new opcodes (we could still deploy the saved bits explicitly as =
needed to support them). It would allow using regular DATA frames as =
well (though I still think a dedicated messaging frame type is =
beneficial).</div><div><br></div><div>Regarding robustness against =
reframing: There is nothing that would prevent dedicated WEBSOCKET =
frames from being split or coalesced. The intermediary would just need =
to understand whatever mechanism is used to determine message boundaries =
(either opcode or MSG_DONE bit). This is identical to the methods of =
reframing regular DATA frames.</div><div><br></div><div><div>On Jan 23, =
2014, at 5:16, Yutaka Hirano &lt;<a =
href=3D"mailto:yhirano@chromium.org">yhirano@chromium.org</a>&gt; =
wrote:</div><br class=3D"Apple-interchange-newline"><blockquote =
type=3D"cite"><div dir=3D"ltr"><div>Thanks for comments and sorry for =
the late response.</div><div><br></div>I now agree with your opinion =
that having both SPDY frame and WS frame is redundant.<div>I would like =
extensions and having RSVs seems a good idea.</div>
<div>As a result, I think it is OK not to support extensions that works =
on individual WS frames not WS messages. (i.e. We should support =
permessage-deflate.)</div><div><br></div><blockquote class=3D"gmail_quote"=
 style=3D"margin:0px 0px 0px =
0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left=
-style:solid;padding-left:1ex">
- An HTTP 2.0 WEBSOCKET frame type could be used to deliver WebSocket =
messages. You=92d need a FIN flag for the message boundaries, frame =
boundaries would be just standard HTTP 2.0 frames. No need to repackage =
or reframe or make a distinction between =93websocket=94 frames and HTTP =
2.0 frames.<br>
- Depending on how closely we want HTTP 2.0 WebSockets to match RFC6455 =
WebSockets additional support could be added.. for example, define the =
WEBSOCKET frame type=92s flag bits as follows (full compatibility flag =
configuration):</blockquote>
<div>I don't think it is a good idea, for various reasons.</div><div>- =
Using many flags field reduces extensiveness of the spec (there are only =
8 bits).</div><div>- Using many flags and adding a WebSocket-specific =
frame type will not make httpbis people happy.</div>
<div>- Robustness against SPDY data reframing is important but this =
framing plan doesn't have the property.</div><div><br></div><blockquote =
class=3D"gmail_quote" style=3D"margin:0px 0px 0px =
0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left=
-style:solid;padding-left:1ex">
<span style=3D"font-family:arial,sans-serif;font-size:13px">So the =
minimum appears to be 2 bits in SPDY/DATA frame headers:</span><br =
style=3D"font-family:arial,sans-serif;font-size:13px"><span =
style=3D"font-family:arial,sans-serif;font-size:13px">1 bit for MSG_DONE =
on the last SPDY/DATA frame and<br>
</span><span style=3D"font-family:arial,sans-serif;font-size:13px">1 bit =
for MSG_TYPE (UTF8/binary) on the first SPDY/DATA frame that starts a =
new msg</span></blockquote><div>Hmm, then, I think it is good to =
have</div><div>
- MSG_DONE in flags for each frame</div><div>- MSG_TYPE at the head of a =
message</div><div>- RSVs at the head of a =
message</div><div>.</div><div><br></div><div>By the way, I still think =
it is preferable to decouple metadata from payload.<br>
</div><div>Is it acceptable for you to represent a WS data Message by =
HEADERS + DATAs?</div><div>Specifically,&nbsp;</div><div>- A WebSocket =
data message is mapped to a HEADERS frame and multiple DATA =
frames.</div><div>&nbsp;- The last frame in a message turns MSG_DONE =
flag on.</div>
<div>&nbsp;- The HEADERS frame consists of several =
headers</div><div>&nbsp; - opcode</div><div>&nbsp; - =
rsv1</div><div>&nbsp; - rsv2</div><div>&nbsp; - rsv3</div><div>- A =
WebSocket CLOSE message is mapped to a HEADERS frame.</div><div>&nbsp;- =
The HEADERS frame consists of several headers</div>
<div>&nbsp; - opcode</div><div>&nbsp; - code</div><div>&nbsp; - =
reason</div><div>&nbsp;- The HEADERS frame turns END_STREAM =
on</div><div>- A WebSocket PING message is mapped to SPDY PING =
message</div><div><div>- A WebSocket PONG message is mapped to SPDY PONG =
message</div>
</div><div><br></div><div>Thanks,</div><div><br></div></div><div =
class=3D"gmail_extra"><br><br><div class=3D"gmail_quote">On Tue, Jan 21, =
2014 at 2:52 AM, Peter Thorson <span dir=3D"ltr">&lt;<a =
href=3D"mailto:webmaster@zaphoyd.com" =
target=3D"_blank">webmaster@zaphoyd.com</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 =
.8ex;border-left:1px #ccc solid;padding-left:1ex"><div class=3D"im"><br>
On Jan 20, 2014, at 11:35, Tobias Oberstein &lt;<a =
href=3D"mailto:tobias.oberstein@tavendo.de">tobias.oberstein@tavendo.de</a=
>&gt; wrote:<br>
<br>
&gt;&gt; Re: minimum bits<br>
&gt;&gt;<br>
&gt;&gt; I definitely agree that as written today only message done, =
stream done, and<br>
&gt;&gt; utf8/binary are necessary. Stream done is present already in =
the DATA frame<br>
&gt;&gt; type. This does seem to ignore the fact that RFC6455 defines 8 =
data message<br>
&gt;&gt; types though and might limit compatibility between the two in =
the future.<br>
&gt;<br>
&gt; Yep, that's the idea. Ignore other WS opcodes, since those are not =
exposed in the W3C API<br>
&gt; anyway, and compatibility with the latter only was the supposed =
goal.<br>
&gt;<br>
&gt; There is 1 more API thing: W3C API lets me close a WS with "code" =
and "reason".<br>
&gt;<br>
&gt; How do would this be mapped?<br>
<br>
</div>I=92d probably map it exactly the way it is now in RFC6455. The =
payload of an HTTP 2.0 frame with stream fin bit set would be the close =
code + reason.<br>
<div class=3D"HOEnZb"><div =
class=3D"h5">_______________________________________________<br>
hybi mailing list<br>
<a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br>
<a href=3D"https://www.ietf.org/mailman/listinfo/hybi" =
target=3D"_blank">https://www.ietf.org/mailman/listinfo/hybi</a><br>
</div></div></blockquote></div><br></div>
</blockquote></div><br></body></html>=

--Apple-Mail=_AD79EEFA-161C-47BD-8AC9-9EBC7025FA89--

From tyoshino@google.com  Sun Jan 26 22:04:19 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C7CBA1A01C7 for <hybi@ietfa.amsl.com>; Sun, 26 Jan 2014 22:04:19 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.913
X-Spam-Level: 
X-Spam-Status: No, score=-1.913 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id fcvjFOxbxMnC for <hybi@ietfa.amsl.com>; Sun, 26 Jan 2014 22:04:17 -0800 (PST)
Received: from mail-we0-x234.google.com (mail-we0-x234.google.com [IPv6:2a00:1450:400c:c03::234]) by ietfa.amsl.com (Postfix) with ESMTP id 58DAF1A01C6 for <hybi@ietf.org>; Sun, 26 Jan 2014 22:04:17 -0800 (PST)
Received: by mail-we0-f180.google.com with SMTP id q59so4687392wes.25 for <hybi@ietf.org>; Sun, 26 Jan 2014 22:04:14 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=XiC55NnF+80J8+sPNK7sS9IbzwfyLqQ4iV+A4nvTLFo=; b=K7zD10TGi1ljasj8IxeKskw6Xr7Oq+UuU3frzvDkLHz+5YedQDCWOanGh+QCniXYkl 8o1UK2nPxgtLpUXFysU4mvkd3yaiKgza2Zlv8m5jeldzwQEZdxF6UUMrWprqo70cYatD BY2LNeiWDdMx77qEkT2BcmSY5GsTBQlw0HhodWWG5ISn+5KJJTy8lmfBZgRiOMVzzvBT Y0/z6G8kmkv6BrTQ/MWj3XMjB/1BtCzvLVd9hQkb8p4YHES1hAmMirT8S6pez5S7Mj2T Gca7geOE2VXRIX+24dU7iRdFcfzQCs7RcUt5Ija9RodkNTPnwj9fdfgdAdAdCdUqw10V nQxw==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=XiC55NnF+80J8+sPNK7sS9IbzwfyLqQ4iV+A4nvTLFo=; b=X7kSkknlXOhJLh38m457jTXmK67UOuuvB6NS4sr8a8TTaYSEDXZOxAxeGkboH4vVKl /yXre2eHqHrJHrf1dJFTW4lJjdR/VTxEcnUUpF2yNCju4nFGlEkvyZaJb+DQiNGbsYD5 QzVWbFcxxQ0D+ED0K9dIHpfBoYGphJXCEXO4TDqrNHSPsF0aaK6Ob+YgXkB8BZCoI/4k MY7kx13vJyo7v1JCpfu8u6ck/6fQFqA79b7C3lh/6CyCD16lPqdSjxwV4RMU4lbTjMGN AyoRpKGA67LTqsoWCSXWHYZYkSYnCSYi895l/BlvatLL4m0hRKmcLEQl16MbHev65DNY mOUA==
X-Gm-Message-State: ALoCoQnYv1vP0vQDQP9hFklIUUPa+Kb/toNHWVTs3CSEW8SFUhZ/HgANmev4XFhS65qwEklHt1aLZ842xW0j+QnE8f2ytag4Lw7A1kxPI1VTjJ8cG++AfBKPlJvl3Ga57cHbb5vdV8UOK78kyjjgGqKxvosoCXjHq6KoHHCDyp1qr43tzICm/cYI2tDUFbUri5PDJGaXyPom
X-Received: by 10.194.76.134 with SMTP id k6mr500310wjw.32.1390802654734; Sun, 26 Jan 2014 22:04:14 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Sun, 26 Jan 2014 22:03:54 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446C13ADD5@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF991A9@EXVMBX020-12.exch020.serverdata.net> <634914A010D0B943A035D226786325D4446BF9987B@EXVMBX020-12.exch020.serverdata.net> <CAHixhFr6wPQQRFrV72eXp9cUA=m1Qvxd2ZhYpqPXFa9pRhnZcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446C083400@EXVMBX020-12.exch020.serverdata.net> <CAM5k6X8cRGtUHp=jTSW5=h05ZTFSnm0Ow-Bak+XYC_qCT8LjXA@mail.gmail.com> <634914A010D0B943A035D226786325D4446C13ADD5@EXVMBX020-12.exch020.serverdata.net>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Mon, 27 Jan 2014 15:03:54 +0900
Message-ID: <CAH9hSJbnQ5q_wgK4U0TMk2gmy9urmyFKeatUGwMKq+_ya2=xpw@mail.gmail.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=047d7bfcf016304a1104f0ed7a83
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] permessage-priority extension draft
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 27 Jan 2014 06:04:20 -0000

--047d7bfcf016304a1104f0ed7a83
Content-Type: text/plain; charset=UTF-8

On Fri, Jan 24, 2014 at 5:33 AM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> >>And in the end: why do we have frames at all in WS if we don't make good
> use of them?  ;)
>
> >The primary reason is so we can write streams of data we don't know the
> size of up front without having to buffer the whole thing.  Aside from
> that, there are benefits for small-memory senders to limit the >buffers
> required.
>
> Wasn't fragmentation also indented to limit head-of-line blocking /
> channel occupation with extensions like WebSocket MUX?
>
>
Right. When looking for the way to interleave frames but keeping message
semantics, we said "yeah, there's frame!"

But in evolution of the mux spec, I
- stopped using frame-wise information (extension data)
- stopped use of inconsistent opcode/FIN combination

As a result, in this spec, everything are encapsulated into normal sequence
of fragments, and intermediaries which don't understand mux are allowed to
fragment/coalesce them.

I'd recommend encapsulation of frames into an encapsulating message
considering integration with other extensions and operability with old
intermediaries as John said.


> From http://tools.ietf.org/html/draft-ietf-hybi-websocket-multiplexing-11
>
> "
>   o  Fragment a large message into smaller frames to prevent a large
>       message in a logical channel occupying the physical connection and
>       thus delaying messages in other logical channels.
> "
>
> [Sender algorithm]
>
> This is similar to permessage-priority's use of fragmentation.
>
> /Tobias
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>

--047d7bfcf016304a1104f0ed7a83
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On F=
ri, Jan 24, 2014 at 5:33 AM, Tobias Oberstein <span dir=3D"ltr">&lt;<a href=
=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.oberstein@=
tavendo.de</a>&gt;</span> wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div class=3D"im">&gt;&gt;And in the end: wh=
y do we have frames at all in WS if we don&#39;t make good use of them? =C2=
=A0;)<br>


<br>
&gt;The primary reason is so we can write streams of data we don&#39;t know=
 the size of up front without having to buffer the whole thing. =C2=A0Aside=
 from that, there are benefits for small-memory senders to limit the &gt;bu=
ffers required.<br>


<br>
</div>Wasn&#39;t fragmentation also indented to limit head-of-line blocking=
 / channel occupation with extensions like WebSocket MUX?<br>
<br></blockquote><div><br></div><div>Right. When looking for the way to int=
erleave frames but keeping message semantics, we said &quot;yeah, there&#39=
;s frame!&quot;</div><div><br></div><div>But in evolution of the mux spec, =
I</div>

<div>- stopped using frame-wise information (extension data)</div><div>- st=
opped use of inconsistent opcode/FIN combination</div><div><br></div><div>A=
s a result, in this spec, everything are encapsulated into normal sequence =
of fragments, and intermediaries which don&#39;t understand mux are allowed=
 to fragment/coalesce them.</div>

<div><br></div><div>I&#39;d recommend encapsulation of frames into an encap=
sulating message considering integration with other extensions and operabil=
ity with old intermediaries as John said.</div><div>=C2=A0</div><blockquote=
 class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc soli=
d;padding-left:1ex">


>From <a href=3D"http://tools.ietf.org/html/draft-ietf-hybi-websocket-multip=
lexing-11" target=3D"_blank">http://tools.ietf.org/html/draft-ietf-hybi-web=
socket-multiplexing-11</a><br>
<br>
&quot;<br>
=C2=A0 o =C2=A0Fragment a large message into smaller frames to prevent a la=
rge<br>
=C2=A0 =C2=A0 =C2=A0 message in a logical channel occupying the physical co=
nnection and<br>
=C2=A0 =C2=A0 =C2=A0 thus delaying messages in other logical channels.<br>
&quot;<br>
<br>
[Sender algorithm]<br>
<br>
This is similar to permessage-priority&#39;s use of fragmentation.<br>
<span class=3D"HOEnZb"><font color=3D"#888888"><br>
/Tobias<br>
</font></span><div class=3D"HOEnZb"><div class=3D"h5">_____________________=
__________________________<br>
hybi mailing list<br>
<a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br>
<a href=3D"https://www.ietf.org/mailman/listinfo/hybi" target=3D"_blank">ht=
tps://www.ietf.org/mailman/listinfo/hybi</a><br>
</div></div></blockquote></div><br></div></div>

--047d7bfcf016304a1104f0ed7a83--

From tyoshino@google.com  Sun Jan 26 22:13:26 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 2D3A91A01CD for <hybi@ietfa.amsl.com>; Sun, 26 Jan 2014 22:13:26 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.613
X-Spam-Level: 
X-Spam-Status: No, score=-1.613 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, MIME_8BIT_HEADER=0.3, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 116YrcGNLIqn for <hybi@ietfa.amsl.com>; Sun, 26 Jan 2014 22:13:24 -0800 (PST)
Received: from mail-we0-x22d.google.com (mail-we0-x22d.google.com [IPv6:2a00:1450:400c:c03::22d]) by ietfa.amsl.com (Postfix) with ESMTP id 7B83E1A01C6 for <hybi@ietf.org>; Sun, 26 Jan 2014 22:13:24 -0800 (PST)
Received: by mail-we0-f173.google.com with SMTP id t60so4728156wes.4 for <hybi@ietf.org>; Sun, 26 Jan 2014 22:13:22 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=qs54ALGErlcPlseQs8F1TvsiLzTzeVCC9en+prQZVRw=; b=LaUsEoaWGafpMjS08A+h7vMU+QvknAPLdxfRuibbSCYEpgGlP9P90EqoahwJrEQzQH mpBd+X7kI89Rmyi98XlGz2oRx66gYq78pTMHZydO9DeOi048NSYQFHKty+dapx/1ccnQ rvWDqIcfJ1zq0hu1GijlFTzAeePXMpXdnsyJchqtVTqv5kYNCHCpTQzaYRGOJLBN7j5l 6U1eRyGoZmjOwbl0uzm1aFvSOksHDHDaemDjpUroED9rx1WW8UCG+2dWD6ggGQnUhawh YBx0lcQhFd7LTknOhhQ4pydHFQ0d6856A4Yppng0R94R98YWhMPBRsnqF0WnrlRF1e7z SunA==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=qs54ALGErlcPlseQs8F1TvsiLzTzeVCC9en+prQZVRw=; b=Rv+/FKulOoVDxfRdRoPk5w1UAmKqqv/mhNh+NdEVTZRbiHT7JUzNkyTy8AUNjfxMvV bD3I78ayBttDHvaOLozyXCwUWyUu7xOx0ijqhUto/LSV7GY8iZSi/dVNROcgCS6Gi421 xOeAjqnqgTsOwFiqWcvw56JZLV1rCcEXv2Q2MwZq/JcXKfOeZb2hQVdU8khc+cpqgkxk ZXiHYKusfoWZkzHSlI2K2cDGZpBmhxRYdULxIhfs1Bufv+BOGS2wu3Wl6T9Moz+qpBFl OzMF8hiMpdG9Fx+u7bWZwYqU2g/KxfTDS0qpEgh4dE2VgYra9AO+YwXDsBs+5KU4nkJf itaA==
X-Gm-Message-State: ALoCoQnKdsKOSw+Bz51hBrR714DBCsTe4j7hkK5B3LUhc6XIIsRCIj2Wf+s93PQICJLlnwsnAUld/pMaeckKd0Y3Vf4usr736ORBlGEw0H2FpYhYr05dgLDu3+k8nUhP6GNB2JjNwdZlEzKIywpCPnmFvKy3lSb8ieCS5SxXGVrmntxnjOtzGo609b079pnCnkNTsJCZg3er
X-Received: by 10.180.188.197 with SMTP id gc5mr10480495wic.30.1390803201855;  Sun, 26 Jan 2014 22:13:21 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Sun, 26 Jan 2014 22:13:01 -0800 (PST)
In-Reply-To: <52E2171F.4030602@it.aoyama.ac.jp>
References: <634914A010D0B943A035D226786325D4446BF991A9@EXVMBX020-12.exch020.serverdata.net> <634914A010D0B943A035D226786325D4446BF9987B@EXVMBX020-12.exch020.serverdata.net> <CAHixhFr6wPQQRFrV72eXp9cUA=m1Qvxd2ZhYpqPXFa9pRhnZcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446C083400@EXVMBX020-12.exch020.serverdata.net> <CAM5k6X8cRGtUHp=jTSW5=h05ZTFSnm0Ow-Bak+XYC_qCT8LjXA@mail.gmail.com> <CAHixhFpZc8_=Rb7cj-iai-2atbu6fr1jPPUCjfo-VdRMA0p7iw@mail.gmail.com> <52E2171F.4030602@it.aoyama.ac.jp>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Mon, 27 Jan 2014 15:13:01 +0900
Message-ID: <CAH9hSJYOcpQh8xFLYBPVm4HdQsMGSLYZV-NU95GQCpFBNQJdoQ@mail.gmail.com>
To: =?UTF-8?Q?Martin_J=2E_D=C3=BCrst?= <duerst@it.aoyama.ac.jp>
Content-Type: multipart/alternative; boundary=001a11c3841accb1ee04f0ed9a2a
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] permessage-priority extension draft
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 27 Jan 2014 06:13:26 -0000

--001a11c3841accb1ee04f0ed9a2a
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

On Fri, Jan 24, 2014 at 4:32 PM, "Martin J. D=C3=BCrst"
<duerst@it.aoyama.ac.jp>wrote:

> On 2014/01/24 15:41, Adam Rice wrote:
>
>  Another legitimate use is to allow the timely sending of control frames.
>>
>> For example, a server sending a large message down a slow link might wan=
t
>> to interrupt the message with a "Ping" every few minutes to elicit a
>> "Pong"
>> from the client and keep the NAT mapping alive.
>>
>
> Wouldn't the NAT mapping be kept alive anyway as long as data is flowing?
>

Not so important, but a close frame may be sent in the middle of big
message to notify the reason of closure to the peer than just closing the
underlying connection.

I don't remember, but maybe close frames were made interjectable not based
on concrete requirement but considering future extensibility and belief on
this kind of feature (signaling should be able to be sent independent of
main flow. or something...).

--001a11c3841accb1ee04f0ed9a2a
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On F=
ri, Jan 24, 2014 at 4:32 PM, &quot;Martin J. D=C3=BCrst&quot; <span dir=3D"=
ltr">&lt;<a href=3D"mailto:duerst@it.aoyama.ac.jp" target=3D"_blank">duerst=
@it.aoyama.ac.jp</a>&gt;</span> wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div class=3D"im">On 2014/01/24 15:41, Adam =
Rice wrote:<br>
<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">
Another legitimate use is to allow the timely sending of control frames.<br=
>
<br>
For example, a server sending a large message down a slow link might want<b=
r>
to interrupt the message with a &quot;Ping&quot; every few minutes to elici=
t a &quot;Pong&quot;<br>
from the client and keep the NAT mapping alive.<br>
</blockquote>
<br></div>
Wouldn&#39;t the NAT mapping be kept alive anyway as long as data is flowin=
g?<br></blockquote><div><br></div><div>Not so important, but a close frame =
may be sent in the middle of big message to notify the reason of closure to=
 the peer than just closing the underlying connection.</div>

<div><br></div><div>I don&#39;t remember, but maybe close frames were made =
interjectable not based on concrete requirement but considering future exte=
nsibility and belief on this kind of feature (signaling should be able to b=
e sent independent of main flow. or something...).</div>

<div>=C2=A0</div></div></div></div>

--001a11c3841accb1ee04f0ed9a2a--

From tyoshino@google.com  Sun Jan 26 23:21:44 2014
Return-Path: <tyoshino@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 77F2B1A01D0 for <hybi@ietfa.amsl.com>; Sun, 26 Jan 2014 23:21:44 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.913
X-Spam-Level: 
X-Spam-Status: No, score=-1.913 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id hzHvR9xBFbSU for <hybi@ietfa.amsl.com>; Sun, 26 Jan 2014 23:21:43 -0800 (PST)
Received: from mail-we0-x22d.google.com (mail-we0-x22d.google.com [IPv6:2a00:1450:400c:c03::22d]) by ietfa.amsl.com (Postfix) with ESMTP id A5B551A01CF for <hybi@ietf.org>; Sun, 26 Jan 2014 23:21:42 -0800 (PST)
Received: by mail-we0-f173.google.com with SMTP id t60so4784286wes.18 for <hybi@ietf.org>; Sun, 26 Jan 2014 23:21:40 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type; bh=k5dnL0ji3tZ970UQ8RSRHvjGrNpvL+PCTmVLsVoP5IY=; b=m7w7GIW1gl+m81dStMYr1SYGilRJI9OqsU127mtyT72SYXxXTsGbjJK6B2G6a6z3rl J/TaFKvTDFZ2XwHsvi4uJUGVzKXbgWr20MlDRRwIWkEgzYSvWr1j5v4mJ+lrMUwZgfcb Ft9AA0+J8x1hwE5PGKyXczD30p1+d6DAyf16kltnCj1E+oSZCYX6KWNaraHeu1SiGI9Q eHheaV9esOG94syBpycsnP5L07RiXg2xOyVTIsbrV5sbeY5aNem3ehwke3xtQAOqXlZy arH2/MNbIbUYuj3eJyrZHOYVddmD60W7ILlaZNreaI8BdfiJwSXbP4NYryUTpUTc6gWe fkHw==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; bh=k5dnL0ji3tZ970UQ8RSRHvjGrNpvL+PCTmVLsVoP5IY=; b=HSC9C0GNkFKbY58vy0U+A3VRix9I0cqD8nwD0ssus7/sreEg+odcXw57bdFLQft+/i cvg+TVPh3NF9LFU+VZaojMUhzyzX8mG+PVwfMAiLxUzd4Hhu6b/G3/WO61KpP1lAMymX ZDJoCloP6rRn2mudy33LOhuUi1H6Mhb+Zzap/v8q0QE0x9vgR8P2owciqLQW638CC5HX dFwfWBGA92h6Q4Gt7jUIMLm5bMmzP/0M7/oW5aUBNX6dsdl0nmx2BXexiWvhlgyxe9nW +nsCT6712P8pE8q7DBvmgSb4721npA+5mJZpEdZSsYkxpebVQkcHM9g+UyoPVUZwY8pq EUZQ==
X-Gm-Message-State: ALoCoQlEQp9N67k92W4dGvQfYmYe0NTg8KjbaEmUzzyrniLSHmbpI8uJlPOQ+VEGSLPxPrFrGd0CQul/Yq4OB7/Y4WSQE/jHRkw92h9lMqlyU+nGTmfjckOsh7Fi7FaYsiTmR5m03gcefA/9yXOovRn0/A4PinMLfVzBVSoMQrC4H4cIsjxVUtt6Cs0C8LU+KTctIiKoKwEB
X-Received: by 10.194.85.75 with SMTP id f11mr255032wjz.47.1390807300033; Sun, 26 Jan 2014 23:21:40 -0800 (PST)
MIME-Version: 1.0
Received: by 10.194.8.231 with HTTP; Sun, 26 Jan 2014 23:21:19 -0800 (PST)
In-Reply-To: <F1F115D4-29EB-457B-B02C-71E46D3E1C81@zaphoyd.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net> <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com> <6652CC7E-7E92-40EF-BD50-2DDD39E15533@zaphoyd.com> <CABihn6GSrU8oJa4nbMJiDDo2Z+-c4fT69roVFxcHYM4XYFToRQ@mail.gmail.com> <F1F115D4-29EB-457B-B02C-71E46D3E1C81@zaphoyd.com>
From: Takeshi Yoshino <tyoshino@google.com>
Date: Mon, 27 Jan 2014 16:21:19 +0900
Message-ID: <CAH9hSJY3+ZJ05i0z9Hy39Ro4C9xEp9nw7-u1bhLA-sKdn=R7Ew@mail.gmail.com>
To: Peter Thorson <webmaster@zaphoyd.com>
Content-Type: multipart/alternative; boundary=089e016347a411f81204f0ee8fd3
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 27 Jan 2014 07:21:44 -0000

--089e016347a411f81204f0ee8fd3
Content-Type: text/plain; charset=UTF-8

On Sun, Jan 26, 2014 at 1:24 AM, Peter Thorson <webmaster@zaphoyd.com>wrote:

> Critical first question here: Do we have the cooperation of the HTTP 2.0
> working group or is the WS over HTTP 2.0 effort something being done
> unilaterally by the WebSocket working group?
>

We want to make HTTP/2.0 good enough for transporting WebSocket API
traffic. MSG_DONE was proposed by Roberto in response to our discussion and
we also want to provide him and httpbis how it can success based on our
experience from RFC 6455 standardization/implementation/deployment.

If it turns out the best things to do is leaving HTTP/2.0 untouched and
layer RFC 6455 onto HTTP/2.0 streams, we can choose it.

The question is what's the best layering. Whether we work with httpbis or
not depends on that question.

--089e016347a411f81204f0ee8fd3
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On S=
un, Jan 26, 2014 at 1:24 AM, Peter Thorson <span dir=3D"ltr">&lt;<a href=3D=
"mailto:webmaster@zaphoyd.com" target=3D"_blank">webmaster@zaphoyd.com</a>&=
gt;</span> wrote:<br>


<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div style=3D"word-wrap:break-word"><div>Cri=
tical first question here: Do we have the cooperation of the HTTP 2.0 worki=
ng group or is the WS over HTTP 2.0 effort something being done unilaterall=
y by the WebSocket working group?</div>


</div></blockquote><div><br></div><div>We want to make HTTP/2.0 good enough=
 for transporting WebSocket API traffic. MSG_DONE was proposed by Roberto i=
n response to our discussion and we also want to provide him and httpbis ho=
w it can success based on our experience from RFC 6455 standardization/impl=
ementation/deployment.</div>


<div><br></div><div>If it turns out the best things to do is leaving HTTP/2=
.0 untouched and layer RFC 6455 onto HTTP/2.0 streams, we can choose it.</d=
iv><div><br></div><div>The question is what&#39;s the best layering. Whethe=
r we work with httpbis or not depends on that question.</div>

<div>=C2=A0</div></div></div></div>

--089e016347a411f81204f0ee8fd3--

From ricea@google.com  Mon Jan 27 02:08:28 2014
Return-Path: <ricea@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id EC5031A01EC for <hybi@ietfa.amsl.com>; Mon, 27 Jan 2014 02:08:27 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.613
X-Spam-Level: 
X-Spam-Status: No, score=-1.613 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, MIME_8BIT_HEADER=0.3, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id S6RzvvTpMJnn for <hybi@ietfa.amsl.com>; Mon, 27 Jan 2014 02:08:27 -0800 (PST)
Received: from mail-ob0-x22d.google.com (mail-ob0-x22d.google.com [IPv6:2607:f8b0:4003:c01::22d]) by ietfa.amsl.com (Postfix) with ESMTP id 51C291A01E1 for <hybi@ietf.org>; Mon, 27 Jan 2014 02:08:24 -0800 (PST)
Received: by mail-ob0-f173.google.com with SMTP id vb8so6193043obc.18 for <hybi@ietf.org>; Mon, 27 Jan 2014 02:08:22 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type; bh=Bfs09/6FtuaHXvBOQ8jqrdCDUOkdt7OKkOhlvp1fgeQ=; b=Zz2oL/w3JSt/2eSA9GmYU2VuQ914gPt0yIbTVrl7dbsPutydXYNRDgctyx0M9ZD+Z+ +2+hK4pzF5eh+hpfaY/5tiORzCwVUoRgZZbzyv7Lnq6bcB003TMjC4EknV1UiDVaVCNP pQxxvXkiGC5dE4XS+XpdXrU/iW9wz1yRCohYMLR4MeFr3FxrTNvPqrx7aw1TiDo0UwBB Uh7YC66qsb9QWXsVdYEZpM6LuI6zQZBiswf+tcZIajXgWvUQ9+2xmZ54oO55GS8YW2y+ bjRqqwVXXZMIvJqYKT5MtAiprLku2BdMFpaTfxElYWsbUzcE/1/Kd9ZsS9H+npsOSy50 Sf5Q==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=Bfs09/6FtuaHXvBOQ8jqrdCDUOkdt7OKkOhlvp1fgeQ=; b=KRkFeJA+HT9dN7PPTYAVnmMUPasuBjQUslf+uhtyMjCedVELWwyTVUPG2wUW5SoOIz RklABRaL6APXPLLWu5aAlwIoeAn/YJsnkxcLLoAjPbaxsqflYLxQCkDdg1EeVVSilrwD 50QygArFs3PEPyAOjhhofC1DCm5OUdBTJsrQOK1KtcQvqNPzO2zKOL7eOYY9JoF5jzok qlAkIoTAueZo6bALJziaQAv9PNVlbuNeiIDgEQaeOgEEWkjDr7hcIaUJtZADWweGugL1 CJ2VSVyGoj7GYZViZgrvF1WkYrJDKVAtPvou+HtEB4H7IXU7GxwuR1CKl0Z6OE+7z1K7 ObEw==
X-Gm-Message-State: ALoCoQl78rXKVh+VOMfPHhbVgnCr7FBMSqr7rjtRufBb+obFdq+ayz5rYzZCeoNZ+JI44tQm/519Pyy1EqtSz58SpEu4zUMRlw7TLTrRpdkeUW8MtUVgzrl+CKEXs0NQy+MKMO0WX8dfTur+d9a6oVLeJuJFF7ukvBZWRD7FZvE4nh8E9eIIyycQD/OH0qwo0xtWT7Ywl7/g
MIME-Version: 1.0
X-Received: by 10.182.48.130 with SMTP id l2mr1016091obn.44.1390817302049; Mon, 27 Jan 2014 02:08:22 -0800 (PST)
Received: by 10.182.118.41 with HTTP; Mon, 27 Jan 2014 02:08:21 -0800 (PST)
In-Reply-To: <52E2171F.4030602@it.aoyama.ac.jp>
References: <634914A010D0B943A035D226786325D4446BF991A9@EXVMBX020-12.exch020.serverdata.net> <634914A010D0B943A035D226786325D4446BF9987B@EXVMBX020-12.exch020.serverdata.net> <CAHixhFr6wPQQRFrV72eXp9cUA=m1Qvxd2ZhYpqPXFa9pRhnZcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446C083400@EXVMBX020-12.exch020.serverdata.net> <CAM5k6X8cRGtUHp=jTSW5=h05ZTFSnm0Ow-Bak+XYC_qCT8LjXA@mail.gmail.com> <CAHixhFpZc8_=Rb7cj-iai-2atbu6fr1jPPUCjfo-VdRMA0p7iw@mail.gmail.com> <52E2171F.4030602@it.aoyama.ac.jp>
Date: Mon, 27 Jan 2014 19:08:21 +0900
Message-ID: <CAHixhFq+RfQUyMut_UQEo1RP8=RPV2JGd6UoQ-qLjUpqj5E4Uw@mail.gmail.com>
From: Adam Rice <ricea@google.com>
To: =?UTF-8?Q?Martin_J=2E_D=C3=BCrst?= <duerst@it.aoyama.ac.jp>
Content-Type: multipart/alternative; boundary=089e0160b3dc3c927804f0f0e363
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] permessage-priority extension draft
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 27 Jan 2014 10:08:28 -0000

--089e0160b3dc3c927804f0f0e363
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

On 24 January 2014 16:32, "Martin J. D=C3=BCrst" <duerst@it.aoyama.ac.jp> w=
rote:
>
>
> Wouldn't the NAT mapping be kept alive anyway as long as data is flowing?
>
>
I was thinking of NAT devices that only refresh the mapping when they see a
packet in the "outgoing" direction.

However, I had forgotten that the ACKs from the client of the data being
sent from the server would be sufficient to keep the NAT mapping alive.

So yes, there should be no need for Pings as long as data is in transit.

--089e0160b3dc3c927804f0f0e363
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On 2=
4 January 2014 16:32, &quot;Martin J. D=C3=BCrst&quot; <span dir=3D"ltr">&l=
t;<a href=3D"mailto:duerst@it.aoyama.ac.jp" target=3D"_blank" class=3D"crem=
ed">duerst@it.aoyama.ac.jp</a>&gt;</span> wrote:<blockquote class=3D"gmail_=
quote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1=
ex">
<div class=3D"im">
<br></div>
Wouldn&#39;t the NAT mapping be kept alive anyway as long as data is flowin=
g?<br><br></blockquote><div><br></div><div>I was thinking of NAT devices th=
at only refresh the mapping when they see a packet in the &quot;outgoing&qu=
ot; direction.</div>
<div><br></div><div>However, I had forgotten that the ACKs from the client =
of the data being sent from the server would be sufficient to keep the NAT =
mapping alive.</div><div><br></div><div>So yes, there should be no need for=
 Pings as long as data is in transit.=C2=A0</div>
</div></div></div>

--089e0160b3dc3c927804f0f0e363--

From yhirano@google.com  Tue Jan 28 03:52:48 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 5C5CF1A0122 for <hybi@ietfa.amsl.com>; Tue, 28 Jan 2014 03:52:48 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.913
X-Spam-Level: 
X-Spam-Status: No, score=-1.913 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id maIrtHtxUh4z for <hybi@ietfa.amsl.com>; Tue, 28 Jan 2014 03:52:44 -0800 (PST)
Received: from mail-qa0-x232.google.com (mail-qa0-x232.google.com [IPv6:2607:f8b0:400d:c00::232]) by ietfa.amsl.com (Postfix) with ESMTP id 51E3F1A0018 for <hybi@ietf.org>; Tue, 28 Jan 2014 03:52:44 -0800 (PST)
Received: by mail-qa0-f50.google.com with SMTP id cm18so305632qab.9 for <hybi@ietf.org>; Tue, 28 Jan 2014 03:52:41 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=Lx3Ygx2SllTDw2I+5KDFsst4HDZFBc/gGdUSGN00vuY=; b=ATyzEjfmML0liRPuCsYPdBDwH5GtLqfNTirg+jqCc0h+rxg1/YTkRpkjfS3YcxsmYE O6DoS0oX3PHIJhLQaroedMiSG33ttkOwqOnELCO5nOHbPK1X0qnFQ91pyla2lLzeFPIN Vd0qkKkP16+av47Yl/FJlxAUr8j/anYqD8hCFw3imr5hoXKm8eeu7Lf13zPwd8ndbyk7 DzTw0LK6pp2aQWULbxt4RZW9rHkJHj389gA53k/C8+ubIS7ZQwKTCoSNkjkdl0rg9dL5 Dcjpp6hkff0I/utXIp3d4HKalFn2AQ1i34QJbjSKNxvwa7qPpHKkC2SYygmNRmGuE4oi oghQ==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=Lx3Ygx2SllTDw2I+5KDFsst4HDZFBc/gGdUSGN00vuY=; b=MvJ80QBfZG85IPO307l26IBv42RT4f/MmVjOR4lvDncOmORWEP05OLyPSQB3IIyE9Q TfDU4EbJNwzBigWBziw5qIpBC5M8Dvh3DdkqJj1OWsaZEcTkX54jQTiSJRB5jWkvK77c XljRey0lJhDHYSRLF3ay3yz1DiYQufCf9AQ8U=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=Lx3Ygx2SllTDw2I+5KDFsst4HDZFBc/gGdUSGN00vuY=; b=FdpGmG71upIARqASRvKFjfCYvN+IyTxjtqG8aVxe+NTXVqIgrwr1tWln0t9bu3oIv0 AaXnfHIky/Mqumzp8/J2pBhfj+7JXZSkunBSS02JMO3PjJaMhd7kj3yRr46sOTDruCfF RqR3K1eZaaAPLeD8siC5t8WJefay1TBsTJd5ioYC6L8n9maJHwUpwhz3u88WlAebj8WI zaaGS3nIwj2nGCe7n/oyG9q0hpoZQh3ZfkVjnuqvfJeeKlv4J+m8sDFBPVAPUG9B1Yvy 4RFPTW8Lnsij9VeC2PR27SXfoZAEw+mx4oT7avsXRcY+lix5hupn2fTPv1Aetaw7NUB+ Lcvg==
X-Gm-Message-State: ALoCoQlqUHPqAV4InKrBgHx63f/ZckOYQOGV3xTwM+IvktjmAe2xtWdz/pFWt/XzmmLJq1mULwDrVfomYl4HbLPhGMHxqiRFS/3N4wouuDlGnF3wbEaMoA3F+dqyfzlUEh8/btp6CY8/SayTAjgqv9WldGLtlk5Lt+G4OWyNFFM71xNGTt407/C4wvX1P0LfS84GjjG4b5sB
MIME-Version: 1.0
X-Received: by 10.140.47.101 with SMTP id l92mr1397458qga.9.1390909961451; Tue, 28 Jan 2014 03:52:41 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.140.109.202 with HTTP; Tue, 28 Jan 2014 03:52:41 -0800 (PST)
In-Reply-To: <F1F115D4-29EB-457B-B02C-71E46D3E1C81@zaphoyd.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net> <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com> <6652CC7E-7E92-40EF-BD50-2DDD39E15533@zaphoyd.com> <CABihn6GSrU8oJa4nbMJiDDo2Z+-c4fT69roVFxcHYM4XYFToRQ@mail.gmail.com> <F1F115D4-29EB-457B-B02C-71E46D3E1C81@zaphoyd.com>
Date: Tue, 28 Jan 2014 20:52:41 +0900
X-Google-Sender-Auth: uCmbce5CRqdx_Bb8vtkTvafjEsg
Message-ID: <CABihn6GXxUwHwLd+926VPEkZxHRZUUi-3dw0PoNaO6aK8X4sEQ@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: Peter Thorson <webmaster@zaphoyd.com>
Content-Type: multipart/alternative; boundary=001a11c1640e2aca9c04f10676c2
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 28 Jan 2014 11:52:48 -0000

--001a11c1640e2aca9c04f10676c2
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

>
> Critical first question here: Do we have the cooperation of the HTTP 2.0
> working group or is the WS over HTTP 2.0 effort something being done
> unilaterally by the WebSocket working group?

If we truly need, we should not hesitate to propose introducing something
to the HTTP 2.0 spec.

My initial proposal that uses a dedicated WEBSOCKET frame type was
> specifically set up to improve on the performance of the tunneling strate=
gy
> as an ordered bidirectional messaging channel while maintaining 100%
> compatibility with all of the extensibility built into RFC6455.


Regarding robustness against reframing: There is nothing that would prevent
> dedicated WEBSOCKET frames from being split or coalesced. The intermediar=
y
> would just need to understand whatever mechanism is used to determine
> message boundaries (either opcode or MSG_DONE bit). This is identical to
> the methods of reframing regular DATA frames.

Sorry, I'm still not sure.
When we are saying "maintaining 100% compatibility with all of the
extensibility built into RFC6455", we should care about the case WS frames
have semantic meanings, as Tobias rightly pointed out. Since the proposal
maps one WS Frame to one HTTP2.0 Frame, HTTP2.0 reframing incurs WS
reframing that can change semantics.

Or, flags except FIN are shared among all frames in a WS Message? If so, it
makes sense.

 I believe the proposal that uses a HEADERS frame + DATA frames is worse
> than just tunneling raw RFC6455. A headers frame alone is 12 bytes not
> including the payload. Assume we leave out headers with default values.
> Assume we get the 88% compression ratio Google saw in its tests of full
> HTTP header compression. Even then we are talking 15+ bytes on the wire t=
o
> send 7 bits of data, of which 4-6 are "reserved for future use". Add
> another 8 bytes per data frame. We are talking invoking string header
> processing systems and the fixed costs of spinning up the header
> compression engine and enforcing its strict ordering guarantees for 1-2
> headers on every message. This is nearly a full blown HTTP request.

The proposal is mapping one WS Message to one HEADERS + multiple DATAs.
For large messages, it can achieve less overhead than WS-TUNNEL at the cost
of some RFC6455 compatibility.

Thanks,



On Sun, Jan 26, 2014 at 1:24 AM, Peter Thorson <webmaster@zaphoyd.com>wrote=
:

> Critical first question here: Do we have the cooperation of the HTTP 2.0
> working group or is the WS over HTTP 2.0 effort something being done
> unilaterally by the WebSocket working group?
>
> If we are working unilaterally I think the best option is to tunnel raw
> RFC6455 over a standard HTTP 2.0 data channel and leave it at that. By
> doing this..
> - We retain full control over the standard and protocol and only have to
> deal with a single protocol when considering future versions, extensions,
> and other errata fixes.
> - We can re-use existing RFC6455 parsing code. Such code is readily
> available in libraries and already present in many pieces of software tha=
t
> will likely implement HTTP 2.0.
> - Tunneling over HTTP 2.0 is more efficient in terms of latency and byte
> overhead than proposals that try and use HTTP 2.0 framing without
> cooperation from the HTTP 2.0 standard itself.
> - Tunneling makes it fairly easy for intermediaries to be built that
> package or extract a WebSocket stream from the HTTP 2.0 stream and route =
it
> elsewhere. Other sorts of intermediaries are no more difficult to constru=
ct
> than they are today where one must be able to read both HTTP 1.1 and
> RFC6455.
>
> The only reason I see to do something else would be if we can reduce
> complexity and improve performance. The HTTP 2.0 spec has goals stated as=
:
> "HTTP/2.0 enables a more efficient use of network resources and a reduced
> perception of latency.". Any proposal for WS over HTTP 2.0 that makes
> things slower doesn't make sense. HTTP 2.0's single TCP connection
> introduces some improvements but its internal mechanisms (framing, flow
> control, header compression, etc) introduce significant (as compared to
> RFC6455) additional processing, latency, and framing overhead.
>
> Compared to a separate HTTP and WS connection, the performance of
> tunneling WS over HTTP likely to be better for certain things (initial
> startup time, TCP packet coalescing) but worse at others (message latency=
,
> message overhead after HTTP resources are done loading). Any proposal for
> more direct framing needs to improve on the performance of tunneling to
> justify the complexity of maintaining two separate specs and code paths.
>
> I believe the proposal that uses a HEADERS frame + DATA frames is worse
> than just tunneling raw RFC6455. A headers frame alone is 12 bytes not
> including the payload. Assume we leave out headers with default values.
> Assume we get the 88% compression ratio Google saw in its tests of full
> HTTP header compression. Even then we are talking 15+ bytes on the wire t=
o
> send 7 bits of data, of which 4-6 are "reserved for future use". Add
> another 8 bytes per data frame. We are talking invoking string header
> processing systems and the fixed costs of spinning up the header
> compression engine and enforcing its strict ordering guarantees for 1-2
> headers on every message. This is nearly a full blown HTTP request.
>
> My initial proposal that uses a dedicated WEBSOCKET frame type was
> specifically set up to improve on the performance of the tunneling strate=
gy
> as an ordered bidirectional messaging channel while maintaining 100%
> compatibility with all of the extensibility built into RFC6455. It does
> that by making full use of the extensibility features of HTTP 2.0 itself.=
 I
> do not think that using all of the flags for this type reduces the
> extensibility of the spec because it matches RFC6455 exactly which itself
> leaves half of those bits "reserved for future use".
>
> It may *look* like there are no more flags available for future use, but
> in reality most are still reserved for future use just in an order that i=
s
> compatible with RFC6455. The use of a WebSocket specific HTTP 2.0 frame
> type means that using all of the flags here doesn't affect the
> extensibility of the other types of frames used for HTTP data and the
> separate processing rules are cleanly delineated.
>
> If we are willing to break perfect RFC6455 compatibility we should get ri=
d
> of the opcode as well. This wouldn't change the wire overhead but would g=
et
> rid of more redundant logic, free up more flag bits, and make the whole
> thing more cohesive with the rest of HTTP 2.0. We'd be giving up the
> ability to transparently support extensions and future protocol versions
> that define new opcodes (we could still deploy the saved bits explicitly =
as
> needed to support them). It would allow using regular DATA frames as well
> (though I still think a dedicated messaging frame type is beneficial).
>
> Regarding robustness against reframing: There is nothing that would
> prevent dedicated WEBSOCKET frames from being split or coalesced. The
> intermediary would just need to understand whatever mechanism is used to
> determine message boundaries (either opcode or MSG_DONE bit). This is
> identical to the methods of reframing regular DATA frames.
>
> On Jan 23, 2014, at 5:16, Yutaka Hirano <yhirano@chromium.org> wrote:
>
> Thanks for comments and sorry for the late response.
>
> I now agree with your opinion that having both SPDY frame and WS frame is
> redundant.
> I would like extensions and having RSVs seems a good idea.
> As a result, I think it is OK not to support extensions that works on
> individual WS frames not WS messages. (i.e. We should support
> permessage-deflate.)
>
> - An HTTP 2.0 WEBSOCKET frame type could be used to deliver WebSocket
>> messages. You=E2=80=99d need a FIN flag for the message boundaries, fram=
e
>> boundaries would be just standard HTTP 2.0 frames. No need to repackage =
or
>> reframe or make a distinction between =E2=80=9Cwebsocket=E2=80=9D frames=
 and HTTP 2.0
>> frames.
>> - Depending on how closely we want HTTP 2.0 WebSockets to match RFC6455
>> WebSockets additional support could be added.. for example, define the
>> WEBSOCKET frame type=E2=80=99s flag bits as follows (full compatibility =
flag
>> configuration):
>
> I don't think it is a good idea, for various reasons.
> - Using many flags field reduces extensiveness of the spec (there are onl=
y
> 8 bits).
> - Using many flags and adding a WebSocket-specific frame type will not
> make httpbis people happy.
> - Robustness against SPDY data reframing is important but this framing
> plan doesn't have the property.
>
> So the minimum appears to be 2 bits in SPDY/DATA frame headers:
>> 1 bit for MSG_DONE on the last SPDY/DATA frame and
>> 1 bit for MSG_TYPE (UTF8/binary) on the first SPDY/DATA frame that start=
s
>> a new msg
>
> Hmm, then, I think it is good to have
> - MSG_DONE in flags for each frame
> - MSG_TYPE at the head of a message
> - RSVs at the head of a message
> .
>
> By the way, I still think it is preferable to decouple metadata from
> payload.
> Is it acceptable for you to represent a WS data Message by HEADERS + DATA=
s?
> Specifically,
> - A WebSocket data message is mapped to a HEADERS frame and multiple DATA
> frames.
>  - The last frame in a message turns MSG_DONE flag on.
>  - The HEADERS frame consists of several headers
>   - opcode
>   - rsv1
>   - rsv2
>   - rsv3
> - A WebSocket CLOSE message is mapped to a HEADERS frame.
>  - The HEADERS frame consists of several headers
>   - opcode
>   - code
>   - reason
>  - The HEADERS frame turns END_STREAM on
> - A WebSocket PING message is mapped to SPDY PING message
> - A WebSocket PONG message is mapped to SPDY PONG message
>
> Thanks,
>
>
>
> On Tue, Jan 21, 2014 at 2:52 AM, Peter Thorson <webmaster@zaphoyd.com>wro=
te:
>
>>
>> On Jan 20, 2014, at 11:35, Tobias Oberstein <tobias.oberstein@tavendo.de=
>
>> wrote:
>>
>> >> Re: minimum bits
>> >>
>> >> I definitely agree that as written today only message done, stream
>> done, and
>> >> utf8/binary are necessary. Stream done is present already in the DATA
>> frame
>> >> type. This does seem to ignore the fact that RFC6455 defines 8 data
>> message
>> >> types though and might limit compatibility between the two in the
>> future.
>> >
>> > Yep, that's the idea. Ignore other WS opcodes, since those are not
>> exposed in the W3C API
>> > anyway, and compatibility with the latter only was the supposed goal.
>> >
>> > There is 1 more API thing: W3C API lets me close a WS with "code" and
>> "reason".
>> >
>> > How do would this be mapped?
>>
>> I=E2=80=99d probably map it exactly the way it is now in RFC6455. The pa=
yload of
>> an HTTP 2.0 frame with stream fin bit set would be the close code + reas=
on.
>> _______________________________________________
>> hybi mailing list
>> hybi@ietf.org
>> https://www.ietf.org/mailman/listinfo/hybi
>>
>
>
>

--001a11c1640e2aca9c04f10676c2
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div><blockquote class=3D"gmail_quote" style=3D"margin:0px=
 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);bor=
der-left-style:solid;padding-left:1ex">Critical first question here: Do we =
have the cooperation of the HTTP 2.0 working group or is the WS over HTTP 2=
.0 effort something being done unilaterally by the WebSocket working group?=
</blockquote>


</div><div>If we truly need, we should not hesitate to propose introducing =
something to the HTTP 2.0 spec.<br></div><div><br></div><blockquote class=
=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;bo=
rder-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">


<span style=3D"font-family:arial,sans-serif;font-size:13px">My initial prop=
osal that uses a dedicated WEBSOCKET frame type was specifically set up to =
improve on the performance of the tunneling strategy as an ordered bidirect=
ional messaging channel while maintaining=C2=A0</span><span style=3D"font-f=
amily:arial,sans-serif;font-size:13px">100% compatibility with all of the e=
xtensibility built into RFC6455.</span></blockquote>


<div><br></div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0p=
x 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-lef=
t-style:solid;padding-left:1ex"><span style=3D"font-family:arial,sans-serif=
;font-size:13px">Regarding robustness against reframing: There is nothing t=
hat would prevent dedicated WEBSOCKET frames from being split or coalesced.=
 The intermediary would just need to understand whatever mechanism is used =
to determine message boundaries (either opcode or MSG_DONE bit). This is id=
entical to the methods of reframing regular DATA frames.</span>=C2=A0</bloc=
kquote>


<div>Sorry, I&#39;m still not sure.</div><div>When we are saying &quot;<spa=
n style=3D"font-family:arial,sans-serif;font-size:13px">maintaining=C2=A0</=
span><span style=3D"font-family:arial,sans-serif;font-size:13px">100% compa=
tibility with all of the extensibility built into RFC6455&quot;, we should =
care about the case WS frames have semantic meanings, as Tobias rightly poi=
nted out.=C2=A0</span><span style=3D"font-family:arial,sans-serif;font-size=
:13px">Since the proposal maps one WS Frame to one HTTP2.0 Frame, HTTP2.0 r=
eframing incurs WS reframing that can change semantics.</span></div>
<div><span style=3D"font-family:arial,sans-serif;font-size:13px"><br></span=
></div><div><span style=3D"font-family:arial,sans-serif;font-size:13px">Or,=
 flags except FIN are shared among all frames in a WS Message? If so, it ma=
kes sense.</span></div>

<div><span style=3D"font-family:arial,sans-serif;font-size:13px"><br>
</span></div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px =
0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-=
style:solid;padding-left:1ex">=C2=A0<span style=3D"font-family:arial,sans-s=
erif;font-size:13px">I believe the proposal that uses a HEADERS frame + DAT=
A frames is worse than just tunneling raw RFC6455. A headers frame alone is=
 12 bytes not including the payload. Assume we leave out headers with defau=
lt values. Assume we get the 88% compression ratio Google saw in its tests =
of full HTTP header compression. Even then we are talking 15+ bytes on the =
wire to send 7 bits of data, of which 4-6 are &quot;reserved for future use=
&quot;. Add another 8 bytes per data frame. We are talking invoking string =
header processing systems and the fixed costs of spinning up the header com=
pression engine and enforcing its strict ordering guarantees for 1-2 header=
s on every message. This is nearly a full blown HTTP request.</span></block=
quote>


<div style=3D"font-family:arial,sans-serif;font-size:13px">The proposal is =
mapping one WS Message to one HEADERS + multiple DATAs.</div><div style=3D"=
font-family:arial,sans-serif;font-size:13px">For large messages, it can ach=
ieve less overhead than WS-TUNNEL at the cost of some RFC6455 compatibility=
.</div>


<div style=3D"font-family:arial,sans-serif;font-size:13px"><br></div><div s=
tyle=3D"font-family:arial,sans-serif;font-size:13px">Thanks,</div><div styl=
e=3D"font-family:arial,sans-serif;font-size:13px"><br></div></div><div clas=
s=3D"gmail_extra">
<br><br><div class=3D"gmail_quote">On Sun, Jan 26, 2014 at 1:24 AM, Peter T=
horson <span dir=3D"ltr">&lt;<a href=3D"mailto:webmaster@zaphoyd.com" targe=
t=3D"_blank">webmaster@zaphoyd.com</a>&gt;</span> wrote:<br><blockquote cla=
ss=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;pa=
dding-left:1ex">
<div style=3D"word-wrap:break-word"><div>Critical first question here: Do w=
e have the cooperation of the HTTP 2.0 working group or is the WS over HTTP=
 2.0 effort something being done unilaterally by the WebSocket working grou=
p?</div>
<div><br></div><div>If we are working unilaterally I think the best option =
is to tunnel raw RFC6455 over a standard HTTP 2.0 data channel and leave it=
 at that. By doing this..</div><div>- We retain full control over the stand=
ard and protocol and only have to deal with a single protocol when consider=
ing future versions, extensions, and other errata fixes.</div>
<div>- We can re-use existing RFC6455 parsing code. Such code is readily av=
ailable in libraries and already present in many pieces of software that wi=
ll likely implement HTTP 2.0.</div><div>- Tunneling over HTTP 2.0 is more e=
fficient in terms of latency and byte overhead than proposals that try and =
use HTTP 2.0 framing without cooperation from the HTTP 2.0 standard itself.=
</div>
<div>- Tunneling makes it fairly easy for intermediaries to be built that p=
ackage or extract a WebSocket stream from the HTTP 2.0 stream and route it =
elsewhere. Other sorts of intermediaries are no more difficult to construct=
 than they are today where one must be able to read both HTTP 1.1 and RFC64=
55.</div>
<div><br></div><div>The only reason I see to do something else would be if =
we can reduce complexity and improve performance. The HTTP 2.0 spec has goa=
ls stated as: &quot;HTTP/2.0 enables a more efficient use of network resour=
ces and a reduced perception of latency.&quot;. Any proposal for WS over HT=
TP 2.0 that makes things slower doesn&#39;t make sense. HTTP 2.0&#39;s sing=
le TCP connection introduces some improvements but its internal mechanisms =
(framing, flow control, header compression, etc) introduce significant (as =
compared to RFC6455) additional processing, latency, and framing overhead.<=
/div>
<div><br></div><div>Compared to a separate HTTP and WS connection, the perf=
ormance of tunneling WS over HTTP likely to be better for certain things (i=
nitial startup time, TCP packet coalescing) but worse at others (message la=
tency, message overhead after HTTP resources are done loading). Any proposa=
l for more direct framing needs to improve on the performance of tunneling =
to justify the complexity of maintaining two separate specs and code paths.=
</div>
<div><br></div><div>I believe the proposal that uses a HEADERS frame + DATA=
 frames is worse than just tunneling raw RFC6455. A headers frame alone is =
12 bytes not including the payload. Assume we leave out headers with defaul=
t values. Assume we get the 88% compression ratio Google saw in its tests o=
f full HTTP header compression. Even then we are talking 15+ bytes on the w=
ire to send 7 bits of data, of which 4-6 are &quot;reserved for future use&=
quot;. Add another 8 bytes per data frame. We are talking invoking string h=
eader processing systems and the fixed costs of spinning up the header comp=
ression engine and enforcing its strict ordering guarantees for 1-2 headers=
 on every message. This is nearly a full blown HTTP request.</div>
<div><br></div><div>My initial proposal that uses a dedicated WEBSOCKET fra=
me type was specifically set up to improve on the performance of the tunnel=
ing strategy as an ordered bidirectional messaging channel while maintainin=
g 100% compatibility with all of the extensibility built into RFC6455. It d=
oes that by making full use of the extensibility features of HTTP 2.0 itsel=
f. I do not think that using all of the flags for this type reduces the ext=
ensibility of the spec because it matches RFC6455 exactly which itself leav=
es half of those bits &quot;reserved for future use&quot;.</div>
<div><br></div><div>It may *look* like there are no more flags available fo=
r future use, but in reality most are still reserved for future use just in=
 an order that is compatible with RFC6455. The use of a WebSocket specific =
HTTP 2.0 frame type means that using all of the flags here doesn&#39;t affe=
ct the extensibility of the other types of frames used for HTTP data and th=
e separate processing rules are cleanly delineated.</div>
<div><br></div><div>If we are willing to break perfect RFC6455 compatibilit=
y we should get rid of the opcode as well. This wouldn&#39;t change the wir=
e overhead but would get rid of more redundant logic, free up more flag bit=
s, and make the whole thing more cohesive with the rest of HTTP 2.0. We&#39=
;d be giving up the ability to transparently support extensions and future =
protocol versions that define new opcodes (we could still deploy the saved =
bits explicitly as needed to support them). It would allow using regular DA=
TA frames as well (though I still think a dedicated messaging frame type is=
 beneficial).</div>
<div><br></div><div>Regarding robustness against reframing: There is nothin=
g that would prevent dedicated WEBSOCKET frames from being split or coalesc=
ed. The intermediary would just need to understand whatever mechanism is us=
ed to determine message boundaries (either opcode or MSG_DONE bit). This is=
 identical to the methods of reframing regular DATA frames.</div>
<div><div class=3D"h5"><div><br></div><div><div>On Jan 23, 2014, at 5:16, Y=
utaka Hirano &lt;<a href=3D"mailto:yhirano@chromium.org" target=3D"_blank">=
yhirano@chromium.org</a>&gt; wrote:</div><br><blockquote type=3D"cite"><div=
 dir=3D"ltr">
<div>Thanks for comments and sorry for the late response.</div><div><br></d=
iv>I now agree with your opinion that having both SPDY frame and WS frame i=
s redundant.<div>I would like extensions and having RSVs seems a good idea.=
</div>

<div>As a result, I think it is OK not to support extensions that works on =
individual WS frames not WS messages. (i.e. We should support permessage-de=
flate.)</div><div><br></div><blockquote class=3D"gmail_quote" style=3D"marg=
in:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,20=
4);border-left-style:solid;padding-left:1ex">

- An HTTP 2.0 WEBSOCKET frame type could be used to deliver WebSocket messa=
ges. You=E2=80=99d need a FIN flag for the message boundaries, frame bounda=
ries would be just standard HTTP 2.0 frames. No need to repackage or refram=
e or make a distinction between =E2=80=9Cwebsocket=E2=80=9D frames and HTTP=
 2.0 frames.<br>

- Depending on how closely we want HTTP 2.0 WebSockets to match RFC6455 Web=
Sockets additional support could be added.. for example, define the WEBSOCK=
ET frame type=E2=80=99s flag bits as follows (full compatibility flag confi=
guration):</blockquote>

<div>I don&#39;t think it is a good idea, for various reasons.</div><div>- =
Using many flags field reduces extensiveness of the spec (there are only 8 =
bits).</div><div>- Using many flags and adding a WebSocket-specific frame t=
ype will not make httpbis people happy.</div>

<div>- Robustness against SPDY data reframing is important but this framing=
 plan doesn&#39;t have the property.</div><div><br></div><blockquote class=
=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;bo=
rder-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">

<span style=3D"font-family:arial,sans-serif;font-size:13px">So the minimum =
appears to be 2 bits in SPDY/DATA frame headers:</span><br style=3D"font-fa=
mily:arial,sans-serif;font-size:13px"><span style=3D"font-family:arial,sans=
-serif;font-size:13px">1 bit for MSG_DONE on the last SPDY/DATA frame and<b=
r>

</span><span style=3D"font-family:arial,sans-serif;font-size:13px">1 bit fo=
r MSG_TYPE (UTF8/binary) on the first SPDY/DATA frame that starts a new msg=
</span></blockquote><div>Hmm, then, I think it is good to have</div><div>

- MSG_DONE in flags for each frame</div><div>- MSG_TYPE at the head of a me=
ssage</div><div>- RSVs at the head of a message</div><div>.</div><div><br><=
/div><div>By the way, I still think it is preferable to decouple metadata f=
rom payload.<br>

</div><div>Is it acceptable for you to represent a WS data Message by HEADE=
RS + DATAs?</div><div>Specifically,=C2=A0</div><div>- A WebSocket data mess=
age is mapped to a HEADERS frame and multiple DATA frames.</div><div>=C2=A0=
- The last frame in a message turns MSG_DONE flag on.</div>

<div>=C2=A0- The HEADERS frame consists of several headers</div><div>=C2=A0=
 - opcode</div><div>=C2=A0 - rsv1</div><div>=C2=A0 - rsv2</div><div>=C2=A0 =
- rsv3</div><div>- A WebSocket CLOSE message is mapped to a HEADERS frame.<=
/div><div>=C2=A0- The HEADERS frame consists of several headers</div>

<div>=C2=A0 - opcode</div><div>=C2=A0 - code</div><div>=C2=A0 - reason</div=
><div>=C2=A0- The HEADERS frame turns END_STREAM on</div><div>- A WebSocket=
 PING message is mapped to SPDY PING message</div><div><div>- A WebSocket P=
ONG message is mapped to SPDY PONG message</div>

</div><div><br></div><div>Thanks,</div><div><br></div></div><div class=3D"g=
mail_extra"><br><br><div class=3D"gmail_quote">On Tue, Jan 21, 2014 at 2:52=
 AM, Peter Thorson <span dir=3D"ltr">&lt;<a href=3D"mailto:webmaster@zaphoy=
d.com" target=3D"_blank">webmaster@zaphoyd.com</a>&gt;</span> wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div><br>
On Jan 20, 2014, at 11:35, Tobias Oberstein &lt;<a href=3D"mailto:tobias.ob=
erstein@tavendo.de" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt; w=
rote:<br>
<br>
&gt;&gt; Re: minimum bits<br>
&gt;&gt;<br>
&gt;&gt; I definitely agree that as written today only message done, stream=
 done, and<br>
&gt;&gt; utf8/binary are necessary. Stream done is present already in the D=
ATA frame<br>
&gt;&gt; type. This does seem to ignore the fact that RFC6455 defines 8 dat=
a message<br>
&gt;&gt; types though and might limit compatibility between the two in the =
future.<br>
&gt;<br>
&gt; Yep, that&#39;s the idea. Ignore other WS opcodes, since those are not=
 exposed in the W3C API<br>
&gt; anyway, and compatibility with the latter only was the supposed goal.<=
br>
&gt;<br>
&gt; There is 1 more API thing: W3C API lets me close a WS with &quot;code&=
quot; and &quot;reason&quot;.<br>
&gt;<br>
&gt; How do would this be mapped?<br>
<br>
</div>I=E2=80=99d probably map it exactly the way it is now in RFC6455. The=
 payload of an HTTP 2.0 frame with stream fin bit set would be the close co=
de + reason.<br>
<div><div>_______________________________________________<br>
hybi mailing list<br>
<a href=3D"mailto:hybi@ietf.org" target=3D"_blank">hybi@ietf.org</a><br>
<a href=3D"https://www.ietf.org/mailman/listinfo/hybi" target=3D"_blank">ht=
tps://www.ietf.org/mailman/listinfo/hybi</a><br>
</div></div></blockquote></div><br></div>
</blockquote></div><br></div></div></div></blockquote></div><br></div>

--001a11c1640e2aca9c04f10676c2--

From yhirano@google.com  Wed Jan 29 05:10:27 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id BC5F01A01FE for <hybi@ietfa.amsl.com>; Wed, 29 Jan 2014 05:10:27 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.913
X-Spam-Level: 
X-Spam-Status: No, score=-1.913 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Pzwe7J-vnYBr for <hybi@ietfa.amsl.com>; Wed, 29 Jan 2014 05:10:21 -0800 (PST)
Received: from mail-yk0-x230.google.com (mail-yk0-x230.google.com [IPv6:2607:f8b0:4002:c07::230]) by ietfa.amsl.com (Postfix) with ESMTP id 61DA61A01C1 for <hybi@ietf.org>; Wed, 29 Jan 2014 05:10:21 -0800 (PST)
Received: by mail-yk0-f176.google.com with SMTP id 131so8027856ykp.7 for <hybi@ietf.org>; Wed, 29 Jan 2014 05:10:18 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=U+5MEEMfPA09c0Ka3tewFusetkOeUavmo5jJ4AEqFjE=; b=FziYcj4lZXvg+zmtu8Q57vgeA2IYUG9lGupWEHcWDx3njCA/IpnPCXAU2DuYym6DPY rVUYWFdtloCJTY2GlhgT5OslZ6HIxAmBOId8TW4Cc6qSevsRd/AvFTIkkdm17UP9I/5M R0nOvRSUXs5ZwUKmtHzEYYBrSKgZavWafV8aTprWUPmS4Oo1GpSwCLCqFYSnobK7wplO 0zgx6rIbxhyIufjvkMdVkPIs4oHYKece/87tx8Pt26I1wE8fBEx+eaOHCI5c9+vetNJv uTxMd5I1U50otdCepSY40fKEPPiGstDde0IXi4/LP3kV0Jo5BNXvuCqXCR1b2gLEjizW /cag==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=U+5MEEMfPA09c0Ka3tewFusetkOeUavmo5jJ4AEqFjE=; b=JBqzynsAcTf2IumH163kKEX65adRedRE3y4KDZ3PrrZh+CpWBnG+o27Hh/1EiH0x7R R+qhQNzavW0kj6gV3G/AtkGbiDMbegY3KmaiiU5S8OdT2cWef/1kP5j1cwELU2ZfLMXN q6zskAm3GZAHYeiKYutWdj+rDCrImVgoq8jFA=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=U+5MEEMfPA09c0Ka3tewFusetkOeUavmo5jJ4AEqFjE=; b=kU7LWfNwN/482lxR+JSdJSyqf/gio90Ga88R9ok4gViAsMJDfC53rGfeSSmFvfc1TN gp28xtfIHv7Ob1vABvOcmtZJl3UkozqbvPZuzuaGeCIMtqJhXMh3yGSiCMD/iMOvOZQi KRc4EL9LoV8+qEIQUAFDVXJhs1N8j+C1NgjK/nds+NjuEOhgQI0CbfgooIYhRsbNYj6b 21GbKuZzL7NSfRb6Bl7dZZeSzvTHfHYSzBDYUonFxXD+tiENX4seImWZjqwzcK6mhE8c 3yQz2Gh87h05nCWkGlTeBO0/9xCYhoyP//B9cg46GpPR773M5Jz8fFfDyiiy+xl8RfrC J86A==
X-Gm-Message-State: ALoCoQnFXX+usmTA+KEnLPfUvbSHQfe+n7WIvrNnC1/8SfhoHjn+FugajoaiwHO5AY4p4b+noQdaO1V851VsNUhbkWnrjMkJhbwXaT4Rv6/jcSWjnm+ZLvpjCw/a20e3gsjkJb8bpskTJy6rbpN/2qq4jVj0ytuZYk2GRvkpqTmDQMh5Nvc6PNwhQA+Bm3drLnsxlfrcGiJ2
MIME-Version: 1.0
X-Received: by 10.236.38.74 with SMTP id z50mr252319yha.134.1391001017901; Wed, 29 Jan 2014 05:10:17 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.170.135.22 with HTTP; Wed, 29 Jan 2014 05:10:17 -0800 (PST)
In-Reply-To: <CABihn6GXxUwHwLd+926VPEkZxHRZUUi-3dw0PoNaO6aK8X4sEQ@mail.gmail.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net> <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com> <6652CC7E-7E92-40EF-BD50-2DDD39E15533@zaphoyd.com> <CABihn6GSrU8oJa4nbMJiDDo2Z+-c4fT69roVFxcHYM4XYFToRQ@mail.gmail.com> <F1F115D4-29EB-457B-B02C-71E46D3E1C81@zaphoyd.com> <CABihn6GXxUwHwLd+926VPEkZxHRZUUi-3dw0PoNaO6aK8X4sEQ@mail.gmail.com>
Date: Wed, 29 Jan 2014 22:10:17 +0900
X-Google-Sender-Auth: YpV_pzAxV7-zlCV5vQSterzVGgI
Message-ID: <CABihn6E5h6Xpu2kvZdAAjfp6nKhAqhEwQE2binyai=VC6HkWsw@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: Peter Thorson <webmaster@zaphoyd.com>
Content-Type: multipart/alternative; boundary=047d7bea41a48df44c04f11ba9d6
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 29 Jan 2014 13:10:27 -0000

--047d7bea41a48df44c04f11ba9d6
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Hi all,

I have updated the draft. The new URL is:
https://docs.google.com/a/google.com/document/d/10dKnaojn_d5xkxOanUsdCsRo4n=
AKfNodmWArWYDAClg/edit#

I added framing proposals proposed in this thread. Please see
"Bi-directional Communication" section.
Correct me if I misunderstand something.
And, other comments, especially for the opening handshake, will be
appreciated.

Thanks,


On Tue, Jan 28, 2014 at 8:52 PM, Yutaka Hirano <yhirano@chromium.org> wrote=
:

> Critical first question here: Do we have the cooperation of the HTTP 2.0
>> working group or is the WS over HTTP 2.0 effort something being done
>> unilaterally by the WebSocket working group?
>
> If we truly need, we should not hesitate to propose introducing something
> to the HTTP 2.0 spec.
>
> My initial proposal that uses a dedicated WEBSOCKET frame type was
>> specifically set up to improve on the performance of the tunneling strat=
egy
>> as an ordered bidirectional messaging channel while maintaining 100%
>> compatibility with all of the extensibility built into RFC6455.
>
>
> Regarding robustness against reframing: There is nothing that would
>> prevent dedicated WEBSOCKET frames from being split or coalesced. The
>> intermediary would just need to understand whatever mechanism is used to
>> determine message boundaries (either opcode or MSG_DONE bit). This is
>> identical to the methods of reframing regular DATA frames.
>
> Sorry, I'm still not sure.
> When we are saying "maintaining 100% compatibility with all of the
> extensibility built into RFC6455", we should care about the case WS frame=
s
> have semantic meanings, as Tobias rightly pointed out. Since the proposal
> maps one WS Frame to one HTTP2.0 Frame, HTTP2.0 reframing incurs WS
> reframing that can change semantics.
>
> Or, flags except FIN are shared among all frames in a WS Message? If so,
> it makes sense.
>
>  I believe the proposal that uses a HEADERS frame + DATA frames is worse
>> than just tunneling raw RFC6455. A headers frame alone is 12 bytes not
>> including the payload. Assume we leave out headers with default values.
>> Assume we get the 88% compression ratio Google saw in its tests of full
>> HTTP header compression. Even then we are talking 15+ bytes on the wire =
to
>> send 7 bits of data, of which 4-6 are "reserved for future use". Add
>> another 8 bytes per data frame. We are talking invoking string header
>> processing systems and the fixed costs of spinning up the header
>> compression engine and enforcing its strict ordering guarantees for 1-2
>> headers on every message. This is nearly a full blown HTTP request.
>
> The proposal is mapping one WS Message to one HEADERS + multiple DATAs.
> For large messages, it can achieve less overhead than WS-TUNNEL at the
> cost of some RFC6455 compatibility.
>
> Thanks,
>
>
>
> On Sun, Jan 26, 2014 at 1:24 AM, Peter Thorson <webmaster@zaphoyd.com>wro=
te:
>
>> Critical first question here: Do we have the cooperation of the HTTP 2.0
>> working group or is the WS over HTTP 2.0 effort something being done
>> unilaterally by the WebSocket working group?
>>
>> If we are working unilaterally I think the best option is to tunnel raw
>> RFC6455 over a standard HTTP 2.0 data channel and leave it at that. By
>> doing this..
>> - We retain full control over the standard and protocol and only have to
>> deal with a single protocol when considering future versions, extensions=
,
>> and other errata fixes.
>> - We can re-use existing RFC6455 parsing code. Such code is readily
>> available in libraries and already present in many pieces of software th=
at
>> will likely implement HTTP 2.0.
>> - Tunneling over HTTP 2.0 is more efficient in terms of latency and byte
>> overhead than proposals that try and use HTTP 2.0 framing without
>> cooperation from the HTTP 2.0 standard itself.
>> - Tunneling makes it fairly easy for intermediaries to be built that
>> package or extract a WebSocket stream from the HTTP 2.0 stream and route=
 it
>> elsewhere. Other sorts of intermediaries are no more difficult to constr=
uct
>> than they are today where one must be able to read both HTTP 1.1 and
>> RFC6455.
>>
>> The only reason I see to do something else would be if we can reduce
>> complexity and improve performance. The HTTP 2.0 spec has goals stated a=
s:
>> "HTTP/2.0 enables a more efficient use of network resources and a reduce=
d
>> perception of latency.". Any proposal for WS over HTTP 2.0 that makes
>> things slower doesn't make sense. HTTP 2.0's single TCP connection
>> introduces some improvements but its internal mechanisms (framing, flow
>> control, header compression, etc) introduce significant (as compared to
>> RFC6455) additional processing, latency, and framing overhead.
>>
>> Compared to a separate HTTP and WS connection, the performance of
>> tunneling WS over HTTP likely to be better for certain things (initial
>> startup time, TCP packet coalescing) but worse at others (message latenc=
y,
>> message overhead after HTTP resources are done loading). Any proposal fo=
r
>> more direct framing needs to improve on the performance of tunneling to
>> justify the complexity of maintaining two separate specs and code paths.
>>
>> I believe the proposal that uses a HEADERS frame + DATA frames is worse
>> than just tunneling raw RFC6455. A headers frame alone is 12 bytes not
>> including the payload. Assume we leave out headers with default values.
>> Assume we get the 88% compression ratio Google saw in its tests of full
>> HTTP header compression. Even then we are talking 15+ bytes on the wire =
to
>> send 7 bits of data, of which 4-6 are "reserved for future use". Add
>> another 8 bytes per data frame. We are talking invoking string header
>> processing systems and the fixed costs of spinning up the header
>> compression engine and enforcing its strict ordering guarantees for 1-2
>> headers on every message. This is nearly a full blown HTTP request.
>>
>> My initial proposal that uses a dedicated WEBSOCKET frame type was
>> specifically set up to improve on the performance of the tunneling strat=
egy
>> as an ordered bidirectional messaging channel while maintaining 100%
>> compatibility with all of the extensibility built into RFC6455. It does
>> that by making full use of the extensibility features of HTTP 2.0 itself=
. I
>> do not think that using all of the flags for this type reduces the
>> extensibility of the spec because it matches RFC6455 exactly which itsel=
f
>> leaves half of those bits "reserved for future use".
>>
>> It may *look* like there are no more flags available for future use, but
>> in reality most are still reserved for future use just in an order that =
is
>> compatible with RFC6455. The use of a WebSocket specific HTTP 2.0 frame
>> type means that using all of the flags here doesn't affect the
>> extensibility of the other types of frames used for HTTP data and the
>> separate processing rules are cleanly delineated.
>>
>> If we are willing to break perfect RFC6455 compatibility we should get
>> rid of the opcode as well. This wouldn't change the wire overhead but wo=
uld
>> get rid of more redundant logic, free up more flag bits, and make the wh=
ole
>> thing more cohesive with the rest of HTTP 2.0. We'd be giving up the
>> ability to transparently support extensions and future protocol versions
>> that define new opcodes (we could still deploy the saved bits explicitly=
 as
>> needed to support them). It would allow using regular DATA frames as wel=
l
>> (though I still think a dedicated messaging frame type is beneficial).
>>
>> Regarding robustness against reframing: There is nothing that would
>> prevent dedicated WEBSOCKET frames from being split or coalesced. The
>> intermediary would just need to understand whatever mechanism is used to
>> determine message boundaries (either opcode or MSG_DONE bit). This is
>> identical to the methods of reframing regular DATA frames.
>>
>> On Jan 23, 2014, at 5:16, Yutaka Hirano <yhirano@chromium.org> wrote:
>>
>> Thanks for comments and sorry for the late response.
>>
>> I now agree with your opinion that having both SPDY frame and WS frame i=
s
>> redundant.
>> I would like extensions and having RSVs seems a good idea.
>> As a result, I think it is OK not to support extensions that works on
>> individual WS frames not WS messages. (i.e. We should support
>> permessage-deflate.)
>>
>> - An HTTP 2.0 WEBSOCKET frame type could be used to deliver WebSocket
>>> messages. You=E2=80=99d need a FIN flag for the message boundaries, fra=
me
>>> boundaries would be just standard HTTP 2.0 frames. No need to repackage=
 or
>>> reframe or make a distinction between =E2=80=9Cwebsocket=E2=80=9D frame=
s and HTTP 2.0
>>> frames.
>>> - Depending on how closely we want HTTP 2.0 WebSockets to match RFC6455
>>> WebSockets additional support could be added.. for example, define the
>>> WEBSOCKET frame type=E2=80=99s flag bits as follows (full compatibility=
 flag
>>> configuration):
>>
>> I don't think it is a good idea, for various reasons.
>> - Using many flags field reduces extensiveness of the spec (there are
>> only 8 bits).
>> - Using many flags and adding a WebSocket-specific frame type will not
>> make httpbis people happy.
>> - Robustness against SPDY data reframing is important but this framing
>> plan doesn't have the property.
>>
>> So the minimum appears to be 2 bits in SPDY/DATA frame headers:
>>> 1 bit for MSG_DONE on the last SPDY/DATA frame and
>>> 1 bit for MSG_TYPE (UTF8/binary) on the first SPDY/DATA frame that
>>> starts a new msg
>>
>> Hmm, then, I think it is good to have
>> - MSG_DONE in flags for each frame
>> - MSG_TYPE at the head of a message
>> - RSVs at the head of a message
>> .
>>
>> By the way, I still think it is preferable to decouple metadata from
>> payload.
>> Is it acceptable for you to represent a WS data Message by HEADERS +
>> DATAs?
>> Specifically,
>> - A WebSocket data message is mapped to a HEADERS frame and multiple DAT=
A
>> frames.
>>  - The last frame in a message turns MSG_DONE flag on.
>>  - The HEADERS frame consists of several headers
>>   - opcode
>>   - rsv1
>>   - rsv2
>>   - rsv3
>> - A WebSocket CLOSE message is mapped to a HEADERS frame.
>>  - The HEADERS frame consists of several headers
>>   - opcode
>>   - code
>>   - reason
>>  - The HEADERS frame turns END_STREAM on
>> - A WebSocket PING message is mapped to SPDY PING message
>> - A WebSocket PONG message is mapped to SPDY PONG message
>>
>> Thanks,
>>
>>
>>
>> On Tue, Jan 21, 2014 at 2:52 AM, Peter Thorson <webmaster@zaphoyd.com>wr=
ote:
>>
>>>
>>> On Jan 20, 2014, at 11:35, Tobias Oberstein <tobias.oberstein@tavendo.d=
e>
>>> wrote:
>>>
>>> >> Re: minimum bits
>>> >>
>>> >> I definitely agree that as written today only message done, stream
>>> done, and
>>> >> utf8/binary are necessary. Stream done is present already in the DAT=
A
>>> frame
>>> >> type. This does seem to ignore the fact that RFC6455 defines 8 data
>>> message
>>> >> types though and might limit compatibility between the two in the
>>> future.
>>> >
>>> > Yep, that's the idea. Ignore other WS opcodes, since those are not
>>> exposed in the W3C API
>>> > anyway, and compatibility with the latter only was the supposed goal.
>>> >
>>> > There is 1 more API thing: W3C API lets me close a WS with "code" and
>>> "reason".
>>> >
>>> > How do would this be mapped?
>>>
>>> I=E2=80=99d probably map it exactly the way it is now in RFC6455. The p=
ayload of
>>> an HTTP 2.0 frame with stream fin bit set would be the close code + rea=
son.
>>> _______________________________________________
>>> hybi mailing list
>>> hybi@ietf.org
>>> https://www.ietf.org/mailman/listinfo/hybi
>>>
>>
>>
>>
>

--047d7bea41a48df44c04f11ba9d6
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Hi all,<div><br></div><div>I have updated the draft. The n=
ew URL is:</div><div><a href=3D"https://docs.google.com/a/google.com/docume=
nt/d/10dKnaojn_d5xkxOanUsdCsRo4nAKfNodmWArWYDAClg/edit#">https://docs.googl=
e.com/a/google.com/document/d/10dKnaojn_d5xkxOanUsdCsRo4nAKfNodmWArWYDAClg/=
edit#</a></div>
<div><br></div><div>I added framing proposals proposed in this thread. Plea=
se see &quot;Bi-directional Communication&quot; section.</div><div>Correct =
me if I misunderstand something.</div><div>And, other comments, especially =
for the opening handshake, will be appreciated.</div>
<div><br></div><div>Thanks,</div></div><div class=3D"gmail_extra"><br><br><=
div class=3D"gmail_quote">On Tue, Jan 28, 2014 at 8:52 PM, Yutaka Hirano <s=
pan dir=3D"ltr">&lt;<a href=3D"mailto:yhirano@chromium.org" target=3D"_blan=
k">yhirano@chromium.org</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div dir=3D"ltr"><div class=3D"im"><div><blo=
ckquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left=
-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;paddi=
ng-left:1ex">
Critical first question here: Do we have the cooperation of the HTTP 2.0 wo=
rking group or is the WS over HTTP 2.0 effort something being done unilater=
ally by the WebSocket working group?</blockquote>


</div></div><div>If we truly need, we should not hesitate to propose introd=
ucing something to the HTTP 2.0 spec.<br></div><div class=3D"im"><div><br><=
/div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;bo=
rder-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:so=
lid;padding-left:1ex">



<span style=3D"font-family:arial,sans-serif;font-size:13px">My initial prop=
osal that uses a dedicated WEBSOCKET frame type was specifically set up to =
improve on the performance of the tunneling strategy as an ordered bidirect=
ional messaging channel while maintaining=C2=A0</span><span style=3D"font-f=
amily:arial,sans-serif;font-size:13px">100% compatibility with all of the e=
xtensibility built into RFC6455.</span></blockquote>



<div><br></div></div><div class=3D"im"><blockquote class=3D"gmail_quote" st=
yle=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb=
(204,204,204);border-left-style:solid;padding-left:1ex"><span style=3D"font=
-family:arial,sans-serif;font-size:13px">Regarding robustness against refra=
ming: There is nothing that would prevent dedicated WEBSOCKET frames from b=
eing split or coalesced. The intermediary would just need to understand wha=
tever mechanism is used to determine message boundaries (either opcode or M=
SG_DONE bit). This is identical to the methods of reframing regular DATA fr=
ames.</span>=C2=A0</blockquote>



</div><div>Sorry, I&#39;m still not sure.</div><div>When we are saying &quo=
t;<span style=3D"font-family:arial,sans-serif;font-size:13px">maintaining=
=C2=A0</span><span style=3D"font-family:arial,sans-serif;font-size:13px">10=
0% compatibility with all of the extensibility built into RFC6455&quot;, we=
 should care about the case WS frames have semantic meanings, as Tobias rig=
htly pointed out.=C2=A0</span><span style=3D"font-family:arial,sans-serif;f=
ont-size:13px">Since the proposal maps one WS Frame to one HTTP2.0 Frame, H=
TTP2.0 reframing incurs WS reframing that can change semantics.</span></div=
>

<div><span style=3D"font-family:arial,sans-serif;font-size:13px"><br></span=
></div><div><span style=3D"font-family:arial,sans-serif;font-size:13px">Or,=
 flags except FIN are shared among all frames in a WS Message? If so, it ma=
kes sense.</span></div>
<div class=3D"im">

<div><span style=3D"font-family:arial,sans-serif;font-size:13px"><br>
</span></div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px =
0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-=
style:solid;padding-left:1ex">=C2=A0<span style=3D"font-family:arial,sans-s=
erif;font-size:13px">I believe the proposal that uses a HEADERS frame + DAT=
A frames is worse than just tunneling raw RFC6455. A headers frame alone is=
 12 bytes not including the payload. Assume we leave out headers with defau=
lt values. Assume we get the 88% compression ratio Google saw in its tests =
of full HTTP header compression. Even then we are talking 15+ bytes on the =
wire to send 7 bits of data, of which 4-6 are &quot;reserved for future use=
&quot;. Add another 8 bytes per data frame. We are talking invoking string =
header processing systems and the fixed costs of spinning up the header com=
pression engine and enforcing its strict ordering guarantees for 1-2 header=
s on every message. This is nearly a full blown HTTP request.</span></block=
quote>



</div><div style=3D"font-family:arial,sans-serif;font-size:13px">The propos=
al is mapping one WS Message to one HEADERS + multiple DATAs.</div><div sty=
le=3D"font-family:arial,sans-serif;font-size:13px">For large messages, it c=
an achieve less overhead than WS-TUNNEL at the cost of some RFC6455 compati=
bility.</div>



<div style=3D"font-family:arial,sans-serif;font-size:13px"><br></div><div s=
tyle=3D"font-family:arial,sans-serif;font-size:13px">Thanks,</div><div styl=
e=3D"font-family:arial,sans-serif;font-size:13px"><br></div></div><div clas=
s=3D"HOEnZb">
<div class=3D"h5"><div class=3D"gmail_extra">
<br><br><div class=3D"gmail_quote">On Sun, Jan 26, 2014 at 1:24 AM, Peter T=
horson <span dir=3D"ltr">&lt;<a href=3D"mailto:webmaster@zaphoyd.com" targe=
t=3D"_blank">webmaster@zaphoyd.com</a>&gt;</span> wrote:<br><blockquote cla=
ss=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;pa=
dding-left:1ex">

<div style=3D"word-wrap:break-word"><div>Critical first question here: Do w=
e have the cooperation of the HTTP 2.0 working group or is the WS over HTTP=
 2.0 effort something being done unilaterally by the WebSocket working grou=
p?</div>

<div><br></div><div>If we are working unilaterally I think the best option =
is to tunnel raw RFC6455 over a standard HTTP 2.0 data channel and leave it=
 at that. By doing this..</div><div>- We retain full control over the stand=
ard and protocol and only have to deal with a single protocol when consider=
ing future versions, extensions, and other errata fixes.</div>

<div>- We can re-use existing RFC6455 parsing code. Such code is readily av=
ailable in libraries and already present in many pieces of software that wi=
ll likely implement HTTP 2.0.</div><div>- Tunneling over HTTP 2.0 is more e=
fficient in terms of latency and byte overhead than proposals that try and =
use HTTP 2.0 framing without cooperation from the HTTP 2.0 standard itself.=
</div>

<div>- Tunneling makes it fairly easy for intermediaries to be built that p=
ackage or extract a WebSocket stream from the HTTP 2.0 stream and route it =
elsewhere. Other sorts of intermediaries are no more difficult to construct=
 than they are today where one must be able to read both HTTP 1.1 and RFC64=
55.</div>

<div><br></div><div>The only reason I see to do something else would be if =
we can reduce complexity and improve performance. The HTTP 2.0 spec has goa=
ls stated as: &quot;HTTP/2.0 enables a more efficient use of network resour=
ces and a reduced perception of latency.&quot;. Any proposal for WS over HT=
TP 2.0 that makes things slower doesn&#39;t make sense. HTTP 2.0&#39;s sing=
le TCP connection introduces some improvements but its internal mechanisms =
(framing, flow control, header compression, etc) introduce significant (as =
compared to RFC6455) additional processing, latency, and framing overhead.<=
/div>

<div><br></div><div>Compared to a separate HTTP and WS connection, the perf=
ormance of tunneling WS over HTTP likely to be better for certain things (i=
nitial startup time, TCP packet coalescing) but worse at others (message la=
tency, message overhead after HTTP resources are done loading). Any proposa=
l for more direct framing needs to improve on the performance of tunneling =
to justify the complexity of maintaining two separate specs and code paths.=
</div>

<div><br></div><div>I believe the proposal that uses a HEADERS frame + DATA=
 frames is worse than just tunneling raw RFC6455. A headers frame alone is =
12 bytes not including the payload. Assume we leave out headers with defaul=
t values. Assume we get the 88% compression ratio Google saw in its tests o=
f full HTTP header compression. Even then we are talking 15+ bytes on the w=
ire to send 7 bits of data, of which 4-6 are &quot;reserved for future use&=
quot;. Add another 8 bytes per data frame. We are talking invoking string h=
eader processing systems and the fixed costs of spinning up the header comp=
ression engine and enforcing its strict ordering guarantees for 1-2 headers=
 on every message. This is nearly a full blown HTTP request.</div>

<div><br></div><div>My initial proposal that uses a dedicated WEBSOCKET fra=
me type was specifically set up to improve on the performance of the tunnel=
ing strategy as an ordered bidirectional messaging channel while maintainin=
g 100% compatibility with all of the extensibility built into RFC6455. It d=
oes that by making full use of the extensibility features of HTTP 2.0 itsel=
f. I do not think that using all of the flags for this type reduces the ext=
ensibility of the spec because it matches RFC6455 exactly which itself leav=
es half of those bits &quot;reserved for future use&quot;.</div>

<div><br></div><div>It may *look* like there are no more flags available fo=
r future use, but in reality most are still reserved for future use just in=
 an order that is compatible with RFC6455. The use of a WebSocket specific =
HTTP 2.0 frame type means that using all of the flags here doesn&#39;t affe=
ct the extensibility of the other types of frames used for HTTP data and th=
e separate processing rules are cleanly delineated.</div>

<div><br></div><div>If we are willing to break perfect RFC6455 compatibilit=
y we should get rid of the opcode as well. This wouldn&#39;t change the wir=
e overhead but would get rid of more redundant logic, free up more flag bit=
s, and make the whole thing more cohesive with the rest of HTTP 2.0. We&#39=
;d be giving up the ability to transparently support extensions and future =
protocol versions that define new opcodes (we could still deploy the saved =
bits explicitly as needed to support them). It would allow using regular DA=
TA frames as well (though I still think a dedicated messaging frame type is=
 beneficial).</div>

<div><br></div><div>Regarding robustness against reframing: There is nothin=
g that would prevent dedicated WEBSOCKET frames from being split or coalesc=
ed. The intermediary would just need to understand whatever mechanism is us=
ed to determine message boundaries (either opcode or MSG_DONE bit). This is=
 identical to the methods of reframing regular DATA frames.</div>

<div><div><div><br></div><div><div>On Jan 23, 2014, at 5:16, Yutaka Hirano =
&lt;<a href=3D"mailto:yhirano@chromium.org" target=3D"_blank">yhirano@chrom=
ium.org</a>&gt; wrote:</div><br><blockquote type=3D"cite"><div dir=3D"ltr">
<div>Thanks for comments and sorry for the late response.</div><div><br></d=
iv>I now agree with your opinion that having both SPDY frame and WS frame i=
s redundant.<div>I would like extensions and having RSVs seems a good idea.=
</div>


<div>As a result, I think it is OK not to support extensions that works on =
individual WS frames not WS messages. (i.e. We should support permessage-de=
flate.)</div><div><br></div><blockquote class=3D"gmail_quote" style=3D"marg=
in:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,20=
4);border-left-style:solid;padding-left:1ex">


- An HTTP 2.0 WEBSOCKET frame type could be used to deliver WebSocket messa=
ges. You=E2=80=99d need a FIN flag for the message boundaries, frame bounda=
ries would be just standard HTTP 2.0 frames. No need to repackage or refram=
e or make a distinction between =E2=80=9Cwebsocket=E2=80=9D frames and HTTP=
 2.0 frames.<br>


- Depending on how closely we want HTTP 2.0 WebSockets to match RFC6455 Web=
Sockets additional support could be added.. for example, define the WEBSOCK=
ET frame type=E2=80=99s flag bits as follows (full compatibility flag confi=
guration):</blockquote>


<div>I don&#39;t think it is a good idea, for various reasons.</div><div>- =
Using many flags field reduces extensiveness of the spec (there are only 8 =
bits).</div><div>- Using many flags and adding a WebSocket-specific frame t=
ype will not make httpbis people happy.</div>


<div>- Robustness against SPDY data reframing is important but this framing=
 plan doesn&#39;t have the property.</div><div><br></div><blockquote class=
=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;bo=
rder-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">


<span style=3D"font-family:arial,sans-serif;font-size:13px">So the minimum =
appears to be 2 bits in SPDY/DATA frame headers:</span><br style=3D"font-fa=
mily:arial,sans-serif;font-size:13px"><span style=3D"font-family:arial,sans=
-serif;font-size:13px">1 bit for MSG_DONE on the last SPDY/DATA frame and<b=
r>


</span><span style=3D"font-family:arial,sans-serif;font-size:13px">1 bit fo=
r MSG_TYPE (UTF8/binary) on the first SPDY/DATA frame that starts a new msg=
</span></blockquote><div>Hmm, then, I think it is good to have</div><div>


- MSG_DONE in flags for each frame</div><div>- MSG_TYPE at the head of a me=
ssage</div><div>- RSVs at the head of a message</div><div>.</div><div><br><=
/div><div>By the way, I still think it is preferable to decouple metadata f=
rom payload.<br>


</div><div>Is it acceptable for you to represent a WS data Message by HEADE=
RS + DATAs?</div><div>Specifically,=C2=A0</div><div>- A WebSocket data mess=
age is mapped to a HEADERS frame and multiple DATA frames.</div><div>=C2=A0=
- The last frame in a message turns MSG_DONE flag on.</div>


<div>=C2=A0- The HEADERS frame consists of several headers</div><div>=C2=A0=
 - opcode</div><div>=C2=A0 - rsv1</div><div>=C2=A0 - rsv2</div><div>=C2=A0 =
- rsv3</div><div>- A WebSocket CLOSE message is mapped to a HEADERS frame.<=
/div><div>=C2=A0- The HEADERS frame consists of several headers</div>


<div>=C2=A0 - opcode</div><div>=C2=A0 - code</div><div>=C2=A0 - reason</div=
><div>=C2=A0- The HEADERS frame turns END_STREAM on</div><div>- A WebSocket=
 PING message is mapped to SPDY PING message</div><div><div>- A WebSocket P=
ONG message is mapped to SPDY PONG message</div>


</div><div><br></div><div>Thanks,</div><div><br></div></div><div class=3D"g=
mail_extra"><br><br><div class=3D"gmail_quote">On Tue, Jan 21, 2014 at 2:52=
 AM, Peter Thorson <span dir=3D"ltr">&lt;<a href=3D"mailto:webmaster@zaphoy=
d.com" target=3D"_blank">webmaster@zaphoyd.com</a>&gt;</span> wrote:<br>


<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div><br>
On Jan 20, 2014, at 11:35, Tobias Oberstein &lt;<a href=3D"mailto:tobias.ob=
erstein@tavendo.de" target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt; w=
rote:<br>
<br>
&gt;&gt; Re: minimum bits<br>
&gt;&gt;<br>
&gt;&gt; I definitely agree that as written today only message done, stream=
 done, and<br>
&gt;&gt; utf8/binary are necessary. Stream done is present already in the D=
ATA frame<br>
&gt;&gt; type. This does seem to ignore the fact that RFC6455 defines 8 dat=
a message<br>
&gt;&gt; types though and might limit compatibility between the two in the =
future.<br>
&gt;<br>
&gt; Yep, that&#39;s the idea. Ignore other WS opcodes, since those are not=
 exposed in the W3C API<br>
&gt; anyway, and compatibility with the latter only was the supposed goal.<=
br>
&gt;<br>
&gt; There is 1 more API thing: W3C API lets me close a WS with &quot;code&=
quot; and &quot;reason&quot;.<br>
&gt;<br>
&gt; How do would this be mapped?<br>
<br>
</div>I=E2=80=99d probably map it exactly the way it is now in RFC6455. The=
 payload of an HTTP 2.0 frame with stream fin bit set would be the close co=
de + reason.<br>
<div><div>_______________________________________________<br>
hybi mailing list<br>
<a href=3D"mailto:hybi@ietf.org" target=3D"_blank">hybi@ietf.org</a><br>
<a href=3D"https://www.ietf.org/mailman/listinfo/hybi" target=3D"_blank">ht=
tps://www.ietf.org/mailman/listinfo/hybi</a><br>
</div></div></blockquote></div><br></div>
</blockquote></div><br></div></div></div></blockquote></div><br></div>
</div></div></blockquote></div><br></div>

--047d7bea41a48df44c04f11ba9d6--

From tobias.oberstein@tavendo.de  Wed Jan 29 05:49:38 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id A176B1A02C8 for <hybi@ietfa.amsl.com>; Wed, 29 Jan 2014 05:49:38 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id jChRzvfb_g0l for <hybi@ietfa.amsl.com>; Wed, 29 Jan 2014 05:49:37 -0800 (PST)
Received: from EXHUB020-2.exch020.serverdata.net (exhub020-2.exch020.serverdata.net [206.225.164.29]) by ietfa.amsl.com (Postfix) with ESMTP id 69CB01A01FE for <hybi@ietf.org>; Wed, 29 Jan 2014 05:49:37 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-2.exch020.serverdata.net ([206.225.164.29]) with mapi; Wed, 29 Jan 2014 05:49:34 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Peter Thorson <webmaster@zaphoyd.com>, Yutaka Hirano <yhirano@chromium.org>
Date: Wed, 29 Jan 2014 05:49:29 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8Z6PkC125xMiLGRGqo8frCKewTJgDDycNw
Message-ID: <634914A010D0B943A035D226786325D4446C1E8D24@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net> <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com> <634914A010D0B943A035D226786325D4446C0826EC@EXVMBX020-12. exch020.serverdata.net> <6652CC7E-7E92-40EF-BD50-2DDD39E15533@zaphoyd.com> <CABihn6GSrU8oJa4nbMJiDDo2Z+-c4fT69roVFxcHYM4XYFToRQ@mail.gmail.com> <F1F115D4-29EB-457B-B02C-71E46D3E1C81@zaphoyd.com>
In-Reply-To: <F1F115D4-29EB-457B-B02C-71E46D3E1C81@zaphoyd.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 29 Jan 2014 13:49:38 -0000

>If we are working unilaterally I think the best option is to tunnel raw RF=
C6455 over a standard HTTP 2.0 data channel and leave it at that. By doing =
this..
>- We retain full control over the standard and protocol and only have to d=
eal with a single protocol when considering future versions, extensions, an=
d other errata fixes.
>- We can re-use existing RFC6455 parsing code. Such code is readily availa=
ble in libraries and already present in many pieces of software that will l=
ikely implement HTTP 2.0.
>- Tunneling over HTTP 2.0 is more efficient in terms of latency and byte o=
verhead than proposals that try and use HTTP 2.0 framing without cooperatio=
n from the HTTP 2.0 standard itself.
>- Tunneling makes it fairly easy for intermediaries to be built that packa=
ge or extract a WebSocket stream from the HTTP 2.0 stream and route it else=
where. Other sorts of intermediaries are no more difficult to construct tha=
n they are today where one must be able to read both HTTP 1.1 and RFC6455.

Those are all good reasons, and I agree, tunneling of full RFC6455 over HTT=
P2/DATA frame payloads seems the most desirable option overall to me also.

/Tobias


From tobias.oberstein@tavendo.de  Wed Jan 29 05:56:14 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id A5FA51A0367 for <hybi@ietfa.amsl.com>; Wed, 29 Jan 2014 05:56:14 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.3
X-Spam-Level: 
X-Spam-Status: No, score=-1.3 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, J_CHICKENPOX_65=0.6, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id KsL2y5xK0f5s for <hybi@ietfa.amsl.com>; Wed, 29 Jan 2014 05:56:14 -0800 (PST)
Received: from EXHUB020-3.exch020.serverdata.net (exhub020-3.exch020.serverdata.net [206.225.164.30]) by ietfa.amsl.com (Postfix) with ESMTP id BE8E81A02C8 for <hybi@ietf.org>; Wed, 29 Jan 2014 05:56:13 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-3.exch020.serverdata.net ([206.225.164.30]) with mapi; Wed, 29 Jan 2014 05:56:10 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Yutaka Hirano <yhirano@chromium.org>, Peter Thorson <webmaster@zaphoyd.com>
Date: Wed, 29 Jan 2014 05:56:07 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8cH3RiHTp8ZsoSQ+uLv50mYHe8twA2ZJHg
Message-ID: <634914A010D0B943A035D226786325D4446C1E8D2C@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net> <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com> <6652CC7E-7E92-40EF-BD50-2DDD39E15533@zaphoyd.com> <CABihn6GSrU8oJa4nbMJiDDo2Z+-c4fT69roVFxcHYM4XYFToRQ@mail.gmail.com> <F1F115D4-29EB-457B-B02C-71E46D3E1C81@zaphoyd.com> <CABihn6GXxUwHwLd+926VPEkZxHRZUUi-3dw0PoNaO6aK8X4sEQ@mail.gmail.com>
In-Reply-To: <CABihn6GXxUwHwLd+926VPEkZxHRZUUi-3dw0PoNaO6aK8X4sEQ@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 29 Jan 2014 13:56:14 -0000
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From tobias.oberstein@tavendo.de  Wed Jan 29 06:16:41 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 98DCA1A0218 for <hybi@ietfa.amsl.com>; Wed, 29 Jan 2014 06:16:41 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 8Cs11q4BO5_9 for <hybi@ietfa.amsl.com>; Wed, 29 Jan 2014 06:16:39 -0800 (PST)
Received: from EXHUB020-1.exch020.serverdata.net (exhub020-1.exch020.serverdata.net [206.225.164.28]) by ietfa.amsl.com (Postfix) with ESMTP id C96701A0207 for <hybi@ietf.org>; Wed, 29 Jan 2014 06:16:38 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-1.exch020.serverdata.net ([206.225.164.28]) with mapi; Wed, 29 Jan 2014 06:16:35 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Takeshi Yoshino <tyoshino@google.com>
Date: Wed, 29 Jan 2014 06:16:32 -0800
Thread-Topic: [hybi] permessage-priority extension draft
Thread-Index: Ac8bJZ4HD8aQLBXOQnq+CNbPAya6nwB1ES7w
Message-ID: <634914A010D0B943A035D226786325D4446C1E8D5B@EXVMBX020-12.exch020.serverdata.net>
References: <634914A010D0B943A035D226786325D4446BF991A9@EXVMBX020-12.exch020.serverdata.net> <634914A010D0B943A035D226786325D4446BF9987B@EXVMBX020-12.exch020.serverdata.net> <CAHixhFr6wPQQRFrV72eXp9cUA=m1Qvxd2ZhYpqPXFa9pRhnZcQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446C083400@EXVMBX020-12.exch020.serverdata.net> <CAM5k6X8cRGtUHp=jTSW5=h05ZTFSnm0Ow-Bak+XYC_qCT8LjXA@mail.gmail.com> <634914A010D0B943A035D226786325D4446C13ADD5@EXVMBX020-12.exch020.serverdata.net> <CAH9hSJbnQ5q_wgK4U0TMk2gmy9urmyFKeatUGwMKq+_ya2=xpw@mail.gmail.com>
In-Reply-To: <CAH9hSJbnQ5q_wgK4U0TMk2gmy9urmyFKeatUGwMKq+_ya2=xpw@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] permessage-priority extension draft
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 29 Jan 2014 14:16:41 -0000
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From fenix@google.com  Wed Jan 29 15:40:24 2014
Return-Path: <fenix@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id EF0291A03D6 for <hybi@ietfa.amsl.com>; Wed, 29 Jan 2014 15:40:24 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.313
X-Spam-Level: 
X-Spam-Status: No, score=-1.313 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, J_CHICKENPOX_65=0.6, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id xMFztjaQlAcI for <hybi@ietfa.amsl.com>; Wed, 29 Jan 2014 15:40:23 -0800 (PST)
Received: from mail-qc0-x229.google.com (mail-qc0-x229.google.com [IPv6:2607:f8b0:400d:c01::229]) by ietfa.amsl.com (Postfix) with ESMTP id 2B5AD1A03B7 for <hybi@ietf.org>; Wed, 29 Jan 2014 15:40:22 -0800 (PST)
Received: by mail-qc0-f169.google.com with SMTP id w7so3930911qcr.28 for <hybi@ietf.org>; Wed, 29 Jan 2014 15:40:19 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type; bh=6W26I7jlFbASKW+y6+QNd3xU7av1xWD2XF9XXvb/8T0=; b=nLCj8DBV3kRJ86idLOtU/q6QyHFs7fJBCQjsOfU0Pnb4yNwcVpnyvX5I1cMacGk5NY s6J83IDTUZJufqYHpP0v4RpmcRUR8OZmVdixKLC6uYxqDugGU6uLfXVzfTjIGa7mQcic vOBOcRugHlbSHnPh7BID4Kpyx5gFraHBIHbmZtbrcu85G91qcf8XFDc48pYDNtbxw79p NqVUPUNX27893KzYrFllo5DbOvN7X4a5nGPlHIc8tXfUE2CRUPvnonXglSleoSV5YxCw HjmTjLlEIPeuovNZEuyKf+mpUYcqHMN6uRMhfjJVGtTo3mP+aNROQaBk9EjflVMDXC3Q +wuA==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=6W26I7jlFbASKW+y6+QNd3xU7av1xWD2XF9XXvb/8T0=; b=CkXKfJgkRW5js7Pq+1H+Fcpu2MgmKtyrPfpNfyQKuQKDyv+v88bZKP+UMAptfrgNr4 0LOqHESj8DAtzFhkEhj2pfhl/NGcazcbvWlh0WUguaB+Vaffb5NI4nlFekctIGTTM9bU VHXD7IN+tFTKMYHQgO/neS0KU5b6E17DPbrOiVdh96Y0K3aWTLYPgxaIlOw8CWh23zNj fKp16qxZ2G8mGX6o62BgPFOUuk5qkGTnoJ75qqXobIbWYtJuAURNO02buTFnHDJHCBiF vkSyhVwkipudkc+47vWnlMDLEY828k6vYFUc2cYiA+dSL4lX9lkL8+wWXCxIttd0Kwpm Dcrg==
X-Gm-Message-State: ALoCoQkGFf9pbGIWKQ6zNes+ggHnXBa9Q/qogvN0Mr2gma0xdqxhV8tEnU2nFWmlz6hVT96vDgXL6CgseqWUrKUUPImjGFpeMQjqKbgZiyNo1GPHRi38VDSqCVYAo7S7BYnTa5RAf0v1IZZqVz4j7pzFLBfzPuNmh+ybRGcf+BBuHyzdJI7+8NAyFXDRe35oLY/u+FDwCWG/
MIME-Version: 1.0
X-Received: by 10.140.106.225 with SMTP id e88mr15925272qgf.66.1391038819903;  Wed, 29 Jan 2014 15:40:19 -0800 (PST)
Received: by 10.229.13.7 with HTTP; Wed, 29 Jan 2014 15:40:19 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446C1E8D2C@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net> <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com> <6652CC7E-7E92-40EF-BD50-2DDD39E15533@zaphoyd.com> <CABihn6GSrU8oJa4nbMJiDDo2Z+-c4fT69roVFxcHYM4XYFToRQ@mail.gmail.com> <F1F115D4-29EB-457B-B02C-71E46D3E1C81@zaphoyd.com> <CABihn6GXxUwHwLd+926VPEkZxHRZUUi-3dw0PoNaO6aK8X4sEQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446C1E8D2C@EXVMBX020-12.exch020.serverdata.net>
Date: Wed, 29 Jan 2014 15:40:19 -0800
Message-ID: <CAGzyod7pt_PzjexmnZj8dSMg3BooUmuMb6xB5P0sG1t6S5FXZw@mail.gmail.com>
From: Roberto Peon <fenix@google.com>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=001a113a8626bac18c04f1247695
Cc: "hybi@ietf.org" <hybi@ietf.org>, Peter Thorson <webmaster@zaphoyd.com>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 29 Jan 2014 23:40:25 -0000

--001a113a8626bac18c04f1247695
Content-Type: text/plain; charset=UTF-8

I don't find the frame-size arguments very compelling since I haven't seen
or met anyone with a substantial deployment who cares significantly about
the overhead difference given packet, TLS, etc. overheads.
There is no deployment that we know about that uses frame-based
extensibility- we can (and should) remove it from the standard since it
hasn't proven to be interesting.
I'm not sure I agree with your statement that libraries supporting HTTP/2
will include necessarily also have WS framing. It seems unlikely to me-- I
know of at least one large deployment where this isn't at all true.

The HTTP/2 standard has (or shortly will have, as soon as the pull request
goes in) all support necessary to handle any stream-of-message protocol
given support of END_MESSAGE.
I'd strongly prefer that we map WS to those primitives so it, and any other
stream-of-message protocols can be handled using loadbalancer/routing
infrastructure which is and can remain protocol agnostic.

-=R


On Wed, Jan 29, 2014 at 5:56 AM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> >The proposal is mapping one WS Message to one HEADERS + multiple DATAs.
> >For large messages, it can achieve less overhead than WS-TUNNEL at the
> cost of some RFC6455 compatibility.
>
> How can it achieve less overhead?
>
> A SPDY HEADER with information from first WS frame seems to be always
> bigger than a SPDY DATA header (8 octets) + WS frame header.
>
> The only situation I can think of: a large, _fragmented_ WS message is
> reframed into a SPDY HEADER+DATAs sequence.
>
> Effectively, you do away with the individual WS frame information. And
> this exactly breaks full RFC6455 compatibility.
>
> /Tobias
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>

--001a113a8626bac18c04f1247695
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div>I don&#39;t find the frame-size arguments very compel=
ling since I haven&#39;t seen or met anyone with a substantial deployment w=
ho cares significantly about the overhead difference given packet, TLS, etc=
. overheads.</div>
<div>There is no deployment that we know about that uses frame-based extens=
ibility- we can (and should) remove it from the standard since it hasn&#39;=
t proven to be interesting.</div><div>I&#39;m not sure I agree with your st=
atement that libraries supporting HTTP/2 will include necessarily also have=
 WS framing. It seems unlikely to me-- I know of at least one large deploym=
ent where this isn&#39;t at all true.</div>
<div><br></div><div>The HTTP/2 standard has (or shortly will have, as soon =
as the pull request goes in) all support necessary to handle any stream-of-=
message protocol given support of END_MESSAGE.</div><div>I&#39;d strongly p=
refer that we map WS to those primitives so it, and any other stream-of-mes=
sage protocols can be handled using loadbalancer/routing infrastructure whi=
ch is and can remain protocol agnostic.</div>
<div><br></div><div>-=3DR</div></div><div class=3D"gmail_extra"><br><br><di=
v class=3D"gmail_quote">On Wed, Jan 29, 2014 at 5:56 AM, Tobias Oberstein <=
span dir=3D"ltr">&lt;<a href=3D"mailto:tobias.oberstein@tavendo.de" target=
=3D"_blank">tobias.oberstein@tavendo.de</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div class=3D"im">&gt;The proposal is mappin=
g one WS Message to one HEADERS + multiple DATAs.<br>
&gt;For large messages, it can achieve less overhead than WS-TUNNEL at the =
cost of some RFC6455 compatibility.<br>
<br>
</div>How can it achieve less overhead?<br>
<br>
A SPDY HEADER with information from first WS frame seems to be always bigge=
r than a SPDY DATA header (8 octets) + WS frame header.<br>
<br>
The only situation I can think of: a large, _fragmented_ WS message is refr=
amed into a SPDY HEADER+DATAs sequence.<br>
<br>
Effectively, you do away with the individual WS frame information. And this=
 exactly breaks full RFC6455 compatibility.<br>
<span class=3D"HOEnZb"><font color=3D"#888888"><br>
/Tobias<br>
</font></span><div class=3D"HOEnZb"><div class=3D"h5">_____________________=
__________________________<br>
hybi mailing list<br>
<a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br>
<a href=3D"https://www.ietf.org/mailman/listinfo/hybi" target=3D"_blank">ht=
tps://www.ietf.org/mailman/listinfo/hybi</a><br>
</div></div></blockquote></div><br></div>

--001a113a8626bac18c04f1247695--

From webmaster@zaphoyd.com  Wed Jan 29 20:35:40 2014
Return-Path: <webmaster@zaphoyd.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 451991A03EC for <hybi@ietfa.amsl.com>; Wed, 29 Jan 2014 20:35:40 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.299
X-Spam-Level: 
X-Spam-Status: No, score=-1.299 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, J_CHICKENPOX_65=0.6] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 6kprAWvyHnrZ for <hybi@ietfa.amsl.com>; Wed, 29 Jan 2014 20:35:38 -0800 (PST)
Received: from authsmtp00.uchicago.edu (authsmtp00.uchicago.edu [128.135.12.120]) by ietfa.amsl.com (Postfix) with ESMTP id 33C0A1A03E7 for <hybi@ietf.org>; Wed, 29 Jan 2014 20:35:38 -0800 (PST)
Received: from [10.0.1.89] (c-68-51-76-178.hsd1.il.comcast.net [68.51.76.178]) (Authenticated sender: zaphoyd) by authsmtp00.uchicago.edu (Postfix) with ESMTP id 66FB84980A9; Wed, 29 Jan 2014 22:35:34 -0600 (CST)
Content-Type: multipart/alternative; boundary="Apple-Mail=_F836F48D-F5A1-4F19-95D6-F11E1740C602"
Mime-Version: 1.0 (Mac OS X Mail 7.0 \(1822\))
From: Peter Thorson <webmaster@zaphoyd.com>
In-Reply-To: <CAGzyod7pt_PzjexmnZj8dSMg3BooUmuMb6xB5P0sG1t6S5FXZw@mail.gmail.com>
Date: Wed, 29 Jan 2014 22:45:12 -0600
Message-Id: <89555FBE-7443-4CDF-A624-4274E11563F2@zaphoyd.com>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net> <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com> <6652CC7E-7E92-40EF-BD50-2DDD39E15533@zaphoyd.com> <CABih n6GSrU8oJa4nbMJiDDo2Z+-c4fT69roVFxcHYM4XYFToRQ@mail.gmail.com> <F1F115D4-29EB-457B-B02C-71E46D3E1C81@zaphoyd.com> <CABihn6GXxUwHwLd+926VPEkZxHRZUUi-3dw0PoNaO6aK8X4sEQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446C1E8D2C@EXVMBX020-12.exch020.serverdata.net> <CAGzyod7pt_PzjexmnZj8dSMg3BooUmuMb6xB5P0sG1t6S5FXZw@mail.gmail.com>
To: Roberto Peon <fenix@google.com>
X-Mailer: Apple Mail (2.1822)
Cc: "hybi@ietf.org" <hybi@ietf.org>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 30 Jan 2014 04:35:40 -0000

--Apple-Mail=_F836F48D-F5A1-4F19-95D6-F11E1740C602
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=us-ascii


On Jan 29, 2014, at 17:40, Roberto Peon <fenix@google.com> wrote:

> I don't find the frame-size arguments very compelling since I haven't =
seen or met anyone with a substantial deployment who cares significantly =
about the overhead difference given packet, TLS, etc. overheads.

In general I agree, but given that one of the primary goals of HTTP/2 is =
to improve performance by using compression and multiplexing to more =
efficiently use fixed TCP and TLS overhead it would be a shame to throw =
away any of those gains by introducing significant new unnecessary and =
incompressible framing.

> There is no deployment that we know about that uses frame-based =
extensibility- we can (and should) remove it from the standard since it =
hasn't proven to be interesting.

I have no problem with this.

> I'm not sure I agree with your statement that libraries supporting =
HTTP/2 will include necessarily also have WS framing. It seems unlikely =
to me-- I know of at least one large deployment where this isn't at all =
true.

Libraries, not necessarily. Programs (in particular, browsers) probably, =
unless they never supported WebSocket at all.

> The HTTP/2 standard has (or shortly will have, as soon as the pull =
request goes in) all support necessary to handle any stream-of-message =
protocol given support of END_MESSAGE.
> I'd strongly prefer that we map WS to those primitives so it, and any =
other stream-of-message protocols can be handled using =
loadbalancer/routing infrastructure which is and can remain protocol =
agnostic.

An official END_MESSAGE flag in the HTTP DATA frame spec sounds good for =
this purpose. A message delimiter alone is insufficient to allow =
WebSocket traffic as it exists today however. At minimum an additional =
type bit would be needed to determine binary vs UTF8 and ideally a =
compressed vs not bit. These do not strike me as universally useful for =
HTTP/2 data frames as for HTTP requests this type of information has =
many more values and is typically carried in HTTP headers. Does it make =
sense to have a single DATA frame type given that the flag needs for =
message type frames are different from those of full HTTP request =
frames?

> -=3DR
>=20
>=20
> On Wed, Jan 29, 2014 at 5:56 AM, Tobias Oberstein =
<tobias.oberstein@tavendo.de> wrote:
> >The proposal is mapping one WS Message to one HEADERS + multiple =
DATAs.
> >For large messages, it can achieve less overhead than WS-TUNNEL at =
the cost of some RFC6455 compatibility.
>=20
> How can it achieve less overhead?
>=20
> A SPDY HEADER with information from first WS frame seems to be always =
bigger than a SPDY DATA header (8 octets) + WS frame header.
>=20
> The only situation I can think of: a large, _fragmented_ WS message is =
reframed into a SPDY HEADER+DATAs sequence.
>=20
> Effectively, you do away with the individual WS frame information. And =
this exactly breaks full RFC6455 compatibility.
>=20
> /Tobias
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>=20
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi


--Apple-Mail=_F836F48D-F5A1-4F19-95D6-F11E1740C602
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
	charset=us-ascii

<html><head><meta http-equiv=3D"Content-Type" content=3D"text/html =
charset=3Dus-ascii"><meta http-equiv=3D"Content-Type" content=3D"text/html=
 charset=3Dus-ascii"></head><body style=3D"word-wrap: break-word; =
-webkit-nbsp-mode: space; -webkit-line-break: =
after-white-space;"><br><div><div>On Jan 29, 2014, at 17:40, Roberto =
Peon &lt;<a href=3D"mailto:fenix@google.com">fenix@google.com</a>&gt; =
wrote:</div><br class=3D"Apple-interchange-newline"><blockquote =
type=3D"cite"><div dir=3D"ltr">I don't find the frame-size arguments =
very compelling since I haven't seen or met anyone with a substantial =
deployment who cares significantly about the overhead difference given =
packet, TLS, etc. overheads.</div></blockquote><div><br></div><div>In =
general I agree, but given that one of the primary goals of HTTP/2 is to =
improve performance by using compression and multiplexing to more =
efficiently use fixed TCP and TLS overhead it would be a shame to throw =
away any of those gains by introducing significant new unnecessary and =
incompressible framing.</div><br><blockquote type=3D"cite"><div =
dir=3D"ltr">
<div>There is no deployment that we know about that uses frame-based =
extensibility- we can (and should) remove it from the standard since it =
hasn't proven to be =
interesting.</div></div></blockquote><div><br></div><div>I have no =
problem with this.</div><br><blockquote type=3D"cite"><div dir=3D"ltr">I'm=
 not sure I agree with your statement that libraries supporting HTTP/2 =
will include necessarily also have WS framing. It seems unlikely to me-- =
I know of at least one large deployment where this isn't at all =
true.</div></blockquote><div><br></div><div>Libraries, not necessarily. =
Programs (in particular, browsers) probably, unless they never supported =
WebSocket at all.</div><br><blockquote type=3D"cite"><div =
dir=3D"ltr"><div>The HTTP/2 standard has (or shortly will have, as soon =
as the pull request goes in) all support necessary to handle any =
stream-of-message protocol given support of END_MESSAGE.</div><div>I'd =
strongly prefer that we map WS to those primitives so it, and any other =
stream-of-message protocols can be handled using loadbalancer/routing =
infrastructure which is and can remain protocol =
agnostic.</div></div></blockquote><div><br></div><div>An official =
END_MESSAGE flag in the HTTP DATA frame spec sounds good for this =
purpose. A message delimiter alone is insufficient to allow WebSocket =
traffic as it exists today however. At minimum an additional type bit =
would be needed to determine binary vs UTF8 and ideally a compressed vs =
not bit. These do not strike me as universally useful for HTTP/2 data =
frames as for HTTP requests this type of information has many more =
values and is typically carried in HTTP headers. Does it make sense to =
have a single DATA frame type given that the flag needs for message type =
frames are different from those of full HTTP request frames?</div>
<br><blockquote type=3D"cite"><div dir=3D"ltr">-=3DR</div><div =
class=3D"gmail_extra"><br><br><div class=3D"gmail_quote">On Wed, Jan 29, =
2014 at 5:56 AM, Tobias Oberstein <span dir=3D"ltr">&lt;<a =
href=3D"mailto:tobias.oberstein@tavendo.de" =
target=3D"_blank">tobias.oberstein@tavendo.de</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 =
.8ex;border-left:1px #ccc solid;padding-left:1ex"><div =
class=3D"im">&gt;The proposal is mapping one WS Message to one HEADERS + =
multiple DATAs.<br>
&gt;For large messages, it can achieve less overhead than WS-TUNNEL at =
the cost of some RFC6455 compatibility.<br>
<br>
</div>How can it achieve less overhead?<br>
<br>
A SPDY HEADER with information from first WS frame seems to be always =
bigger than a SPDY DATA header (8 octets) + WS frame header.<br>
<br>
The only situation I can think of: a large, _fragmented_ WS message is =
reframed into a SPDY HEADER+DATAs sequence.<br>
<br>
Effectively, you do away with the individual WS frame information. And =
this exactly breaks full RFC6455 compatibility.<br>
<span class=3D"HOEnZb"><font color=3D"#888888"><br>
/Tobias<br>
</font></span><div class=3D"HOEnZb"><div =
class=3D"h5">_______________________________________________<br>
hybi mailing list<br>
<a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br>
<a href=3D"https://www.ietf.org/mailman/listinfo/hybi" =
target=3D"_blank">https://www.ietf.org/mailman/listinfo/hybi</a><br>
</div></div></blockquote></div><br></div>
_______________________________________________<br>hybi mailing =
list<br><a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br><a =
href=3D"https://www.ietf.org/mailman/listinfo/hybi">https://www.ietf.org/m=
ailman/listinfo/hybi</a><br></blockquote></div><br></body></html>=

--Apple-Mail=_F836F48D-F5A1-4F19-95D6-F11E1740C602--

From yhirano@google.com  Wed Jan 29 22:45:27 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 0B40E1A04EB for <hybi@ietfa.amsl.com>; Wed, 29 Jan 2014 22:45:27 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.913
X-Spam-Level: 
X-Spam-Status: No, score=-1.913 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id YAOjUe04vLLO for <hybi@ietfa.amsl.com>; Wed, 29 Jan 2014 22:45:25 -0800 (PST)
Received: from mail-yk0-x230.google.com (mail-yk0-x230.google.com [IPv6:2607:f8b0:4002:c07::230]) by ietfa.amsl.com (Postfix) with ESMTP id 15DBB1A04CE for <hybi@ietf.org>; Wed, 29 Jan 2014 22:45:24 -0800 (PST)
Received: by mail-yk0-f176.google.com with SMTP id 131so13905110ykp.7 for <hybi@ietf.org>; Wed, 29 Jan 2014 22:45:21 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:date:message-id:subject:from:to:content-type; bh=imJ3aLvvfO+nsrvBCzNaqW5SNRyCdkzSWJNz19iVm9A=; b=Zk/2Pc/8X2PUQtj6Y+x1UaXLLVtWb2lsDHXwNP99mAUuB9A59DUOPxcyR37Zs3/iA5 Br0oJNUv1ZYaeTZYpNh6GMEKxkNls+LTh3IiomBEMcHD0uR3Be+YspxrwJUlnzRs6v7f XdOv4F3tTeXbLsfxU2+zfe6sTknFTPneF4QvOHaVxoTTzJEazrkUAwB4ZEddc8f4fOoN e+EVGAfTvSNPbgUfxQHSB4zOiEg69tcSqNSWI7ea5YW7ynOW4x+HK4SIG8fUKj5aEpyz rdgrCnb/dgvarEogQLz2UF3OoVcY3Hevux/QklbHT/9sRlH/jCjnAWiSoVqhaef3ejia 4+vw==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:date:message-id:subject:from:to:content-type; bh=imJ3aLvvfO+nsrvBCzNaqW5SNRyCdkzSWJNz19iVm9A=; b=AfjN1h5NRcfPSfJddMOUh/+E2fJPjGRI+10EZnCb5uK4lLPrv7w99JtlUSk7WBiWN5 /mf5D8dhiPZb8ys+4nRTh6+RWZzQTtGNlDCJny3tb8+6opDL4A2Q1fLDO964PSzAAUH7 oeTGOXky/Eu+oVMF9Obho23tzJy9UOkFwzIOk=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:date:message-id:subject:from :to:content-type; bh=imJ3aLvvfO+nsrvBCzNaqW5SNRyCdkzSWJNz19iVm9A=; b=fs9UY2hFy/OZL2CgRzGK906x7JzFw6wCLt0CXIKPyQ88wmxbQoVGnRz5JaumDOSaWD MthSbGqT8JUXC7bLq5yxdjjEZjC6Hdoqekn4qoWKDfIh/x0RHOda5krOzla7ZjHadEPq nyB4efLcpG6RGc5bFOA8+xT4q3M8yYADR67xF0hl366TnooOEEZ2QER900Uj7/ckdBiv ez/04i/GaBxezSjPq73U6JfcibGOh9elTqVa1CgTD9rkMuDkvqjwhN8m9kF4PVz9FKVM ws/EKd703j96nd3Dtmzm5oMjKJ/bq9YCsb5uTLPfr9QWDfFVf056GE4Zic2hIxV8WCnz HLEw==
X-Gm-Message-State: ALoCoQmyzFL5PLn3PeCj0nRuD/g93vwsezqsC49TSe4uhOKgFAGiUvfCMHXvBEQmaJ6jHJvyrAiN4MHmeZf/caF/8NWIAkSfcRf7z/N4InmC2CM5bdRdCu5zrs25tev5bwD7YvQNaV0hF5Zgi3cW0pr2gnDos6KtNY5v8YAZxP6k/IybFxb+v+2DdGy596u4sU1eJa78/7jE
MIME-Version: 1.0
X-Received: by 10.236.132.74 with SMTP id n50mr11437629yhi.20.1391064321776; Wed, 29 Jan 2014 22:45:21 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.170.135.22 with HTTP; Wed, 29 Jan 2014 22:45:21 -0800 (PST)
Date: Thu, 30 Jan 2014 15:45:21 +0900
X-Google-Sender-Auth: sKCl3pBs7GG9bPGNrQHVQr-z_8Q
Message-ID: <CABihn6EtrUQvCjRXnjmEYHq_3z4KrPzX5e_JeRev0mAX-Od+fQ@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: "hybi@ietf.org" <hybi@ietf.org>
Content-Type: multipart/alternative; boundary=20cf301afa8bc2722304f12a66fe
Subject: [hybi] WebSocket over HTTP/2.0 protocol negotiation
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 30 Jan 2014 06:45:27 -0000

--20cf301afa8bc2722304f12a66fe
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Hi,

As I announced, I've updated the draft and requested for comments.
Tobias gave a comment on protocol negotiation (thanks!) and I think it is
good to discuss here, not at the draft comment fields.

The draft says:

   1. If the client has a HTTP/2.0 session established by previous requests
   and it doesn=E2=80=99t know that the session does NOT support WebSocket =
layering
   over HTTP/2.0, create a WebSocket stream on the session. Both of WebSock=
et
   (ws) and secure WebSocket (wss) could share the same HTTP/2.0 session.
   2. If the client is going to establish a secure WebSocket connection,
   create a WebSocket connection over TLS.
   3. If the client is going to establish a non-secure WebSocket
   connection, try the WebSocket protocol negotiation.

Tobias doesn't like the first step. He says:
>
> I think native WS should be tried first. Native WS has less overhead than
> any WS-over-SPDY layering, and hence should be preferred.


I have some comments about it.
 - WebSocket over HTTP/2.0 is multiplexed, so it is not so obvious which is
preferable. I think it depends on the popularity of the services:
   - WebSocket over HTTP/2.0
   - Native WebSocket and HTTP/2.0
   - Native WebSocket and WebSocket over HTTP/2.0
   Note that servers which doesn't support HTTP/2.0 don't matter, because
they don't meet the precondition of the first step.
 -  There are two scheme negotiation hints described below.
    If we choose Hint X, before receiving a SETTINGS frame from the server,
the client doesn't know which service is preferable. It is rare, but trying
connecting to the wrong service can occur.
    If we choose Hint Y, whether the server supports WebSocket over
HTTP/2.0 or not is detected by the client prior to HTTP/2.0 session
establishment and hence the client always understands the server's
preference.
    So, the actual negotiation decision process depends on the hint we
choose.

I'm sorry that the draft is not clear at this point, but it will be when
the hint is chosen.

Thanks,

--20cf301afa8bc2722304f12a66fe
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Hi,<div><br></div><div>As I announced, I&#39;ve updated th=
e draft and requested for comments.</div><div>Tobias gave a comment on prot=
ocol negotiation (thanks!) and I think it is good to discuss here, not at t=
he draft comment fields.</div>
<div><br></div><div>The draft says:</div><ol><li>If the client has a HTTP/2=
.0 session established by previous requests and it doesn=E2=80=99t know tha=
t the session does NOT support WebSocket layering over HTTP/2.0, create a W=
ebSocket stream on the session. Both of WebSocket (ws) and secure WebSocket=
 (wss) could share the same HTTP/2.0 session.<br>
</li><li>If the client is going to establish a secure WebSocket connection,=
 create a WebSocket connection over TLS.<br></li><li>If the client is going=
 to establish a non-secure WebSocket connection, try the WebSocket protocol=
 negotiation.<br>
</li></ol>Tobias doesn&#39;t like the first step. He says:<blockquote class=
=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;bo=
rder-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">
I think native WS should be tried first. Native WS has less overhead than a=
ny WS-over-SPDY layering, and hence should be preferred.</blockquote><div><=
br></div><div>I have some comments about it.</div><div>=C2=A0- WebSocket ov=
er HTTP/2.0 is multiplexed, so it is not so obvious which is preferable. I =
think it depends on the popularity of the services:<br>
</div><div>=C2=A0 =C2=A0- WebSocket over HTTP/2.0</div><div>=C2=A0 =C2=A0- =
Native WebSocket and HTTP/2.0</div><div>=C2=A0 =C2=A0- Native WebSocket and=
 WebSocket over HTTP/2.0</div><div>=C2=A0 =C2=A0Note that servers which doe=
sn&#39;t support HTTP/2.0 don&#39;t matter, because they don&#39;t meet the=
 precondition of the first step.</div>
<div>=C2=A0- =C2=A0There are two scheme negotiation hints described below.<=
/div><div>=C2=A0 =C2=A0 If we choose Hint X, before receiving a SETTINGS fr=
ame from the server, the client doesn&#39;t know which service is preferabl=
e. It is rare, but trying connecting to the wrong service can occur.</div>
<div>=C2=A0 =C2=A0 If we choose Hint Y, whether the server supports WebSock=
et over HTTP/2.0 or not is detected by the client prior to HTTP/2.0 session=
 establishment and hence the client always understands the server&#39;s pre=
ference.</div>
<div>=C2=A0 =C2=A0 So, the actual negotiation decision process depends on t=
he hint we choose.<br></div><div><br></div><div>I&#39;m sorry that the draf=
t is not clear at this point, but it will be when the hint is chosen.<br></=
div><div>
<br></div><div>Thanks,</div><div><br></div></div>

--20cf301afa8bc2722304f12a66fe--

From yhirano@google.com  Thu Jan 30 00:58:59 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 0F72F1A0500 for <hybi@ietfa.amsl.com>; Thu, 30 Jan 2014 00:58:59 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.313
X-Spam-Level: 
X-Spam-Status: No, score=-1.313 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, J_CHICKENPOX_65=0.6, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id O-YmdCrswXhx for <hybi@ietfa.amsl.com>; Thu, 30 Jan 2014 00:58:57 -0800 (PST)
Received: from mail-yk0-x231.google.com (mail-yk0-x231.google.com [IPv6:2607:f8b0:4002:c07::231]) by ietfa.amsl.com (Postfix) with ESMTP id 97DA61A04FF for <hybi@ietf.org>; Thu, 30 Jan 2014 00:58:57 -0800 (PST)
Received: by mail-yk0-f177.google.com with SMTP id 19so14493009ykq.8 for <hybi@ietf.org>; Thu, 30 Jan 2014 00:58:54 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=mYLwx424OjKxaLFAZB+QgJWVIsTudMB0yhxTObjELXQ=; b=RuCgp0pMTCNo7uolJL7vM94V+pC97ZaHHGsv5dC9BAzPNFXbjWiu/NfAnttxey3xp4 QARFDYKsCbioJppXdvCmW04s8GaPoABunpk+53Wqta/WrkgDPY3d41IA33ShVFPloQ62 nZpsDKR4FnVTufO3Ohg91IigQQomg+bl12IDe28jtYkuhnU6B0yLd1elKLRrOmbq2pZl 2+WkijFzoc6HnOOpxwiJlFZQi3/mB1ctJjc4jtYo7QzV1fPYvrKoKNMjqLgNnYKJJkqg PZTNF2RRaHFItJW2A7x2YXfN+XnpdreCYjSyJGs3DFgrzr9X4102CcCNi/uZ7aSvgz8T 57aw==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=mYLwx424OjKxaLFAZB+QgJWVIsTudMB0yhxTObjELXQ=; b=eQrwUts3ITLIZW/S5pB9VFiLpbSNqNvGOX0GbANjPWdgABJsrG8ZL2RPEf+fII/gK/ 3lJ0rLc11uG53d3n8uMms/jtcFvmpoZz53bVUylv0czHjjm4XVmXHRvPZ22nO8GZ0aLr 8wiD+W42Z//4P89yPhtAf4F5ndudJmVXZSkSQ=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=mYLwx424OjKxaLFAZB+QgJWVIsTudMB0yhxTObjELXQ=; b=jUnVDrkX41HCmZO7UIU866P3Ob+py9Hoel7WFF1tLHWXikuYfrfVR16zaC3ggF+BEN VQR64uwN0uepuqZIp7IO/Dp3pKj0300qhTCxRTLsApkiQv0naCL7cKVtFGvNsnXBj2Zy XIWoGRF/6VBsDMooLacK3nLJaerOoY/dTMBxXddVvpF0m6tX9Ex21zxkU35qYs+jZbnw VtXo1vAret4t4FhNRMX853X9HqGh/2BUVpTCBNnCjwo4yhtNg9PEysz65DhqIJ/52C5h PK6mI4eyVPk+B0T8CT/NcKQ8PZT3SdE+lQF5kResdg7xQqbxQFy2f/Xv6d4O7ElMiLJb 2mjg==
X-Gm-Message-State: ALoCoQlsk0YiquZf1ci5szppd+xvkE3JWSD3ncKuBtix3zky2+vB/jI6OfqL/ai/vz28yr52qJ+QbBb/BKfYumdy6/pc/rtsqpnAkiaACakdonTVM84Lj10gk3dk0n2ydHDgkHHv4BeeG1hv13N0ONneTCyQZkWjipHidIIbmIw6w8LPlkzFLbSZ7PoGkpgJqtpS6aUMZ0l2
MIME-Version: 1.0
X-Received: by 10.236.100.173 with SMTP id z33mr12244901yhf.9.1391072334122; Thu, 30 Jan 2014 00:58:54 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.170.135.22 with HTTP; Thu, 30 Jan 2014 00:58:53 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446C1E8D2C@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net> <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com> <6652CC7E-7E92-40EF-BD50-2DDD39E15533@zaphoyd.com> <CABihn6GSrU8oJa4nbMJiDDo2Z+-c4fT69roVFxcHYM4XYFToRQ@mail.gmail.com> <F1F115D4-29EB-457B-B02C-71E46D3E1C81@zaphoyd.com> <CABihn6GXxUwHwLd+926VPEkZxHRZUUi-3dw0PoNaO6aK8X4sEQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446C1E8D2C@EXVMBX020-12.exch020.serverdata.net>
Date: Thu, 30 Jan 2014 17:58:53 +0900
X-Google-Sender-Auth: zjuUwy71zBHx_udzJkHpphal-CA
Message-ID: <CABihn6H5R3oBYskgM_HoyAc4rYfyDt6xpRK7fZf3EapKCibGYg@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=20cf3005e018552ed704f12c4425
Cc: "hybi@ietf.org" <hybi@ietf.org>, Peter Thorson <webmaster@zaphoyd.com>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 30 Jan 2014 08:58:59 -0000

--20cf3005e018552ed704f12c4425
Content-Type: text/plain; charset=UTF-8

>
> The only situation I can think of: a large, _fragmented_ WS message is
> reframed into a SPDY HEADER+DATAs sequence.

Yes. A WebSocket message is mapped to one HEADERS + multiple DATAs, even if
it consists of multiple WebSocket frames.

> Effectively, you do away with the individual WS frame information. And
> this exactly breaks full RFC6455 compatibility.

Yes. Most of WebSocket headers are shared among all frames in a WebSocket
message.



On Wed, Jan 29, 2014 at 10:56 PM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> >The proposal is mapping one WS Message to one HEADERS + multiple DATAs.
> >For large messages, it can achieve less overhead than WS-TUNNEL at the
> cost of some RFC6455 compatibility.
>
> How can it achieve less overhead?
>
> A SPDY HEADER with information from first WS frame seems to be always
> bigger than a SPDY DATA header (8 octets) + WS frame header.
>
> The only situation I can think of: a large, _fragmented_ WS message is
> reframed into a SPDY HEADER+DATAs sequence.
>
> Effectively, you do away with the individual WS frame information. And
> this exactly breaks full RFC6455 compatibility.
>
> /Tobias
>

--20cf3005e018552ed704f12c4425
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px =
0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-l=
eft-style:solid;padding-left:1ex"><span style=3D"font-family:arial,sans-ser=
if;font-size:13px">The only situation I can think of: a large, _fragmented_=
 WS message is reframed into a SPDY HEADER+DATAs sequence.</span></blockquo=
te>

<div>Yes. A WebSocket message is mapped to one HEADERS + multiple DATAs, ev=
en if it consists of multiple WebSocket frames.</div><blockquote class=3D"g=
mail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;border-=
left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">
<span style=3D"font-family:arial,sans-serif;font-size:13px">Effectively, yo=
u do away with the individual WS frame information. And this exactly breaks=
 full RFC6455 compatibility.</span></blockquote><div>Yes. Most of WebSocket=
 headers are shared among all frames in a WebSocket message.</div>
<div>=C2=A0</div></div><div class=3D"gmail_extra"><br><br><div class=3D"gma=
il_quote">On Wed, Jan 29, 2014 at 10:56 PM, Tobias Oberstein <span dir=3D"l=
tr">&lt;<a href=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">to=
bias.oberstein@tavendo.de</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div class=3D"im">&gt;The proposal is mappin=
g one WS Message to one HEADERS + multiple DATAs.<br>
&gt;For large messages, it can achieve less overhead than WS-TUNNEL at the =
cost of some RFC6455 compatibility.<br>
<br>
</div>How can it achieve less overhead?<br>
<br>
A SPDY HEADER with information from first WS frame seems to be always bigge=
r than a SPDY DATA header (8 octets) + WS frame header.<br>
<br>
The only situation I can think of: a large, _fragmented_ WS message is refr=
amed into a SPDY HEADER+DATAs sequence.<br>
<br>
Effectively, you do away with the individual WS frame information. And this=
 exactly breaks full RFC6455 compatibility.<br>
<span class=3D"HOEnZb"><font color=3D"#888888"><br>
/Tobias<br>
</font></span></blockquote></div><br></div>

--20cf3005e018552ed704f12c4425--

From tobias.oberstein@tavendo.de  Thu Jan 30 03:55:04 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 7CB8B1A01FB for <hybi@ietfa.amsl.com>; Thu, 30 Jan 2014 03:55:04 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.3
X-Spam-Level: 
X-Spam-Status: No, score=-1.3 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, J_CHICKENPOX_65=0.6, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 8hnEFYFJkb20 for <hybi@ietfa.amsl.com>; Thu, 30 Jan 2014 03:55:03 -0800 (PST)
Received: from EXHUB020-5.exch020.serverdata.net (exhub020-5.exch020.serverdata.net [206.225.164.32]) by ietfa.amsl.com (Postfix) with ESMTP id 4C9821A01ED for <hybi@ietf.org>; Thu, 30 Jan 2014 03:55:03 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-5.exch020.serverdata.net ([206.225.164.32]) with mapi; Thu, 30 Jan 2014 03:54:59 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Roberto Peon <fenix@google.com>
Date: Thu, 30 Jan 2014 03:54:56 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8dS3ky7cMHzuscSumqvO6isoFhlQAZJe/A
Message-ID: <634914A010D0B943A035D226786325D4446C1E925F@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net> <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com> <6652CC7E-7E92-40EF-BD50-2DDD39E15533@zaphoyd.com> <CABihn6GSrU8oJa4nbMJiDDo2Z+-c4fT69roVFxcHYM4XYFToRQ@mail.gmail.com> <F1F115D4-29EB-457B-B02C-71E46D3E1C81@zaphoyd.com> <CABihn6GXxUwHwLd+926VPEkZxHRZUUi-3dw0PoNaO6aK8X4sEQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446C1E8D2C@EXVMBX020-12.exch020.serverdata.net> <CAGzyod7pt_PzjexmnZj8dSMg3BooUmuMb6xB5P0sG1t6S5FXZw@mail.gmail.com>
In-Reply-To: <CAGzyod7pt_PzjexmnZj8dSMg3BooUmuMb6xB5P0sG1t6S5FXZw@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>, Peter Thorson <webmaster@zaphoyd.com>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 30 Jan 2014 11:55:04 -0000
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From tobias.oberstein@tavendo.de  Thu Jan 30 04:24:50 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 964DF1A038E for <hybi@ietfa.amsl.com>; Thu, 30 Jan 2014 04:24:50 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.899
X-Spam-Level: 
X-Spam-Status: No, score=-1.899 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=ham
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id bwh_ZaaEjA4J for <hybi@ietfa.amsl.com>; Thu, 30 Jan 2014 04:24:48 -0800 (PST)
Received: from EXHUB020-3.exch020.serverdata.net (exhub020-3.exch020.serverdata.net [206.225.164.30]) by ietfa.amsl.com (Postfix) with ESMTP id 0E75B1A0207 for <hybi@ietf.org>; Thu, 30 Jan 2014 04:24:48 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-3.exch020.serverdata.net ([206.225.164.30]) with mapi; Thu, 30 Jan 2014 04:24:44 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Yutaka Hirano <yhirano@chromium.org>, "hybi@ietf.org" <hybi@ietf.org>
Date: Thu, 30 Jan 2014 04:24:43 -0800
Thread-Topic: [hybi] WebSocket over HTTP/2.0 protocol negotiation
Thread-Index: Ac8dhtznVBTNM2O1Quyg+vibzRfs/AALGbYA
Message-ID: <634914A010D0B943A035D226786325D4446C1E9261@EXVMBX020-12.exch020.serverdata.net>
References: <CABihn6EtrUQvCjRXnjmEYHq_3z4KrPzX5e_JeRev0mAX-Od+fQ@mail.gmail.com>
In-Reply-To: <CABihn6EtrUQvCjRXnjmEYHq_3z4KrPzX5e_JeRev0mAX-Od+fQ@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: multipart/alternative; boundary="_000_634914A010D0B943A035D226786325D4446C1E9261EXVMBX02012ex_"
MIME-Version: 1.0
Subject: Re: [hybi] WebSocket over HTTP/2.0 protocol negotiation
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 30 Jan 2014 12:24:50 -0000

--_000_634914A010D0B943A035D226786325D4446C1E9261EXVMBX02012ex_
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64

SGkgWXV0YWthLA0KDQp0aGUgcG9pbnQgSSB3YXMgdHJ5aW5nIHRvIHJhaXNlIGlzIHRoaXM6DQoN
CkZvciBhIEphdmFTY3JpcHQgaW4gYSBicm93c2VyIHRoYXQgb3BlbnMgYSAid3M6Ly/igKYiIGNv
bm5lY3Rpb24sIHRoZXJlIHNob3VsZCBuZXZlciBiZSBhbiBhZGRpdGlvbmFsIHJvdW5kdHJpcA0K
aW5jdXJyZWQgYnkgZmlyc3QgcHJvYmluZyBmb3IgYW55IFNQRFkvV1MgYW5kIHRoZW4gZmFsbGlu
ZyBiYWNrIHRvIG5hdGl2ZSBXUywgYnV0IHRoZSBuYXRpdmUgV1Mgc2hvdWxkIGJlDQp0cmllZCBm
aXJzdC4NCg0KSW4gb3RoZXIgd29yZHM6IHRoZSBjdXJyZW50LCBpbW1lZGlhdGUgY29ubmVjdGlv
biBlc3RhYmxpc2htZW50IGZvciBuYXRpdmUgV1Mgc2hvdWxkIG5vdCBiZSBwdXQNCmF0IGFuIGRp
c2FkdmFudGFnZSBqdXN0IGJlY2F1c2Ugb2YgaW50cm9kdWNpbmcgYW55IFNQRFkvV1MuDQoNCkEg
c2VydmVyIHNob3VsZCBiZSBhYmxlIHRvIHByb3ZpZGUgYm90aCBTUERZIGFuZCBuYXRpdmUgV2Vi
U29ja2V0IG9uIHRoZSBfc2FtZV8gb3JpZ2luIHdpdGhvdXQNCmJlaW5nIHJlcXVpcmVkIHRvIGlt
cGxlbWVudCBTUERZL1dTIGFsc28sIGFuZCB3aXRob3V0IGluY3VycmluZyBhbiBhZGRpdGlvbmFs
IHJvdW5kdHJpcCBmb3INCnByb2JpbmcgU1BEWS9XUyB3aGlsZSBlc3RhYmxpc2hpbmcgYSBXUyBj
b25uZWN0aW9uLg0KDQpJbiBvdGhlciB3b3JkczogSSB0aGluayBhIHNlcnZlciBzaG91bGQgYWN0
aXZlbHkgYW5ub3VuY2Ugc3VwcG9ydCBmb3IgYW55IFNQRFkvV1MsIGFuZCBvbmx5IHRoZW4NCnNo
b3VsZCBTUERZL1dTIGJlIHRyaWVkIGZpcnN0IHdoZW4gYSBKUyBvcGVucyBhIFdlYlNvY2tldCBj
b25uZWN0aW9uIHRvIGFuIG9yaWdpbiB0aGF0IGhhcyBhbg0KZXhpc3RpbmcgU1BEWSBjb25uZWN0
aW9uIGFuZCB3aGljaCBhbHJlYWR5IGFubm91bmNlZCBpdCdzIHN1cHBvcnQgYW5kIHdpbGxpbmdu
ZXNzIHRvIHNwZWFrDQphbnkgU1BEWS9XUy4NCg0KUHJvYmFibHkgSSBtaXN1bmRlcnN0b29kIHRo
ZSBkcmFmdC4NCg0KQXJlIGJyb3dzZXJzIHN1cHBvc2VkIHRvIHdvcmsgbGlrZSBhYm92ZSB3aGVu
IHRoZXkgaW1wbGVtZW50IGFueSBTUERZL1dTPw0KDQovVG9iaWFzDQoNCg0KVm9uOiBoeWJpIFtt
YWlsdG86aHliaS1ib3VuY2VzQGlldGYub3JnXSBJbSBBdWZ0cmFnIHZvbiBZdXRha2EgSGlyYW5v
DQpHZXNlbmRldDogRG9ubmVyc3RhZywgMzAuIEphbnVhciAyMDE0IDA3OjQ1DQpBbjogaHliaUBp
ZXRmLm9yZw0KQmV0cmVmZjogW2h5YmldIFdlYlNvY2tldCBvdmVyIEhUVFAvMi4wIHByb3RvY29s
IG5lZ290aWF0aW9uDQoNCkhpLA0KDQpBcyBJIGFubm91bmNlZCwgSSd2ZSB1cGRhdGVkIHRoZSBk
cmFmdCBhbmQgcmVxdWVzdGVkIGZvciBjb21tZW50cy4NClRvYmlhcyBnYXZlIGEgY29tbWVudCBv
biBwcm90b2NvbCBuZWdvdGlhdGlvbiAodGhhbmtzISkgYW5kIEkgdGhpbmsgaXQgaXMgZ29vZCB0
byBkaXNjdXNzIGhlcmUsIG5vdCBhdCB0aGUgZHJhZnQgY29tbWVudCBmaWVsZHMuDQoNClRoZSBk
cmFmdCBzYXlzOg0KDQogMS4gIElmIHRoZSBjbGllbnQgaGFzIGEgSFRUUC8yLjAgc2Vzc2lvbiBl
c3RhYmxpc2hlZCBieSBwcmV2aW91cyByZXF1ZXN0cyBhbmQgaXQgZG9lc27igJl0IGtub3cgdGhh
dCB0aGUgc2Vzc2lvbiBkb2VzIE5PVCBzdXBwb3J0IFdlYlNvY2tldCBsYXllcmluZyBvdmVyIEhU
VFAvMi4wLCBjcmVhdGUgYSBXZWJTb2NrZXQgc3RyZWFtIG9uIHRoZSBzZXNzaW9uLiBCb3RoIG9m
IFdlYlNvY2tldCAod3MpIGFuZCBzZWN1cmUgV2ViU29ja2V0ICh3c3MpIGNvdWxkIHNoYXJlIHRo
ZSBzYW1lIEhUVFAvMi4wIHNlc3Npb24uDQogMi4gIElmIHRoZSBjbGllbnQgaXMgZ29pbmcgdG8g
ZXN0YWJsaXNoIGEgc2VjdXJlIFdlYlNvY2tldCBjb25uZWN0aW9uLCBjcmVhdGUgYSBXZWJTb2Nr
ZXQgY29ubmVjdGlvbiBvdmVyIFRMUy4NCiAzLiAgSWYgdGhlIGNsaWVudCBpcyBnb2luZyB0byBl
c3RhYmxpc2ggYSBub24tc2VjdXJlIFdlYlNvY2tldCBjb25uZWN0aW9uLCB0cnkgdGhlIFdlYlNv
Y2tldCBwcm90b2NvbCBuZWdvdGlhdGlvbi4NClRvYmlhcyBkb2Vzbid0IGxpa2UgdGhlIGZpcnN0
IHN0ZXAuIEhlIHNheXM6DQpJIHRoaW5rIG5hdGl2ZSBXUyBzaG91bGQgYmUgdHJpZWQgZmlyc3Qu
IE5hdGl2ZSBXUyBoYXMgbGVzcyBvdmVyaGVhZCB0aGFuIGFueSBXUy1vdmVyLVNQRFkgbGF5ZXJp
bmcsIGFuZCBoZW5jZSBzaG91bGQgYmUgcHJlZmVycmVkLg0KDQpJIGhhdmUgc29tZSBjb21tZW50
cyBhYm91dCBpdC4NCiAtIFdlYlNvY2tldCBvdmVyIEhUVFAvMi4wIGlzIG11bHRpcGxleGVkLCBz
byBpdCBpcyBub3Qgc28gb2J2aW91cyB3aGljaCBpcyBwcmVmZXJhYmxlLiBJIHRoaW5rIGl0IGRl
cGVuZHMgb24gdGhlIHBvcHVsYXJpdHkgb2YgdGhlIHNlcnZpY2VzOg0KICAgLSBXZWJTb2NrZXQg
b3ZlciBIVFRQLzIuMA0KICAgLSBOYXRpdmUgV2ViU29ja2V0IGFuZCBIVFRQLzIuMA0KICAgLSBO
YXRpdmUgV2ViU29ja2V0IGFuZCBXZWJTb2NrZXQgb3ZlciBIVFRQLzIuMA0KICAgTm90ZSB0aGF0
IHNlcnZlcnMgd2hpY2ggZG9lc24ndCBzdXBwb3J0IEhUVFAvMi4wIGRvbid0IG1hdHRlciwgYmVj
YXVzZSB0aGV5IGRvbid0IG1lZXQgdGhlIHByZWNvbmRpdGlvbiBvZiB0aGUgZmlyc3Qgc3RlcC4N
CiAtICBUaGVyZSBhcmUgdHdvIHNjaGVtZSBuZWdvdGlhdGlvbiBoaW50cyBkZXNjcmliZWQgYmVs
b3cuDQogICAgSWYgd2UgY2hvb3NlIEhpbnQgWCwgYmVmb3JlIHJlY2VpdmluZyBhIFNFVFRJTkdT
IGZyYW1lIGZyb20gdGhlIHNlcnZlciwgdGhlIGNsaWVudCBkb2Vzbid0IGtub3cgd2hpY2ggc2Vy
dmljZSBpcyBwcmVmZXJhYmxlLiBJdCBpcyByYXJlLCBidXQgdHJ5aW5nIGNvbm5lY3RpbmcgdG8g
dGhlIHdyb25nIHNlcnZpY2UgY2FuIG9jY3VyLg0KICAgIElmIHdlIGNob29zZSBIaW50IFksIHdo
ZXRoZXIgdGhlIHNlcnZlciBzdXBwb3J0cyBXZWJTb2NrZXQgb3ZlciBIVFRQLzIuMCBvciBub3Qg
aXMgZGV0ZWN0ZWQgYnkgdGhlIGNsaWVudCBwcmlvciB0byBIVFRQLzIuMCBzZXNzaW9uIGVzdGFi
bGlzaG1lbnQgYW5kIGhlbmNlIHRoZSBjbGllbnQgYWx3YXlzIHVuZGVyc3RhbmRzIHRoZSBzZXJ2
ZXIncyBwcmVmZXJlbmNlLg0KICAgIFNvLCB0aGUgYWN0dWFsIG5lZ290aWF0aW9uIGRlY2lzaW9u
IHByb2Nlc3MgZGVwZW5kcyBvbiB0aGUgaGludCB3ZSBjaG9vc2UuDQoNCkknbSBzb3JyeSB0aGF0
IHRoZSBkcmFmdCBpcyBub3QgY2xlYXIgYXQgdGhpcyBwb2ludCwgYnV0IGl0IHdpbGwgYmUgd2hl
biB0aGUgaGludCBpcyBjaG9zZW4uDQoNClRoYW5rcywNCg0K

--_000_634914A010D0B943A035D226786325D4446C1E9261EXVMBX02012ex_
Content-Type: text/html; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_634914A010D0B943A035D226786325D4446C1E9261EXVMBX02012ex_--

From yhirano@google.com  Fri Jan 31 02:41:02 2014
Return-Path: <yhirano@google.com>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id CE7791A0584 for <hybi@ietfa.amsl.com>; Fri, 31 Jan 2014 02:41:02 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.313
X-Spam-Level: 
X-Spam-Status: No, score=-1.313 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, J_CHICKENPOX_65=0.6, RP_MATCHES_RCVD=-0.535, SPF_PASS=-0.001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id rnPQzMmV55Hh for <hybi@ietfa.amsl.com>; Fri, 31 Jan 2014 02:40:59 -0800 (PST)
Received: from mail-yk0-x22a.google.com (mail-yk0-x22a.google.com [IPv6:2607:f8b0:4002:c07::22a]) by ietfa.amsl.com (Postfix) with ESMTP id 931001A0580 for <hybi@ietf.org>; Fri, 31 Jan 2014 02:40:59 -0800 (PST)
Received: by mail-yk0-f170.google.com with SMTP id 9so22274026ykp.1 for <hybi@ietf.org>; Fri, 31 Jan 2014 02:40:56 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=dUZ2+1bkg9lTt+1vvivnjWQ7WWv3riHvSqOyna9uoCY=; b=iHpC9GjYa8BO+LDAVnZ3c//4r6SvVwUe6zigyNs4cXVP3thGmiXloBAMqsqevFua68 sS/pjpAuNbKIIoNGfUqa6GhiVQOPTUtiJbNXvy09d3OLVB0oaLFG2pbfazhV4uqWoRZ+ v/15m/jbP9HVv6UkAjsNVuWShHBa0G71IFo5mq/Jnr9ACqZQwSwgWzZ+uLP9C08Zwy1j gEcGiAUAsVXVrGVNK4d5Ro0C/VRdZKNT6aNbDhsr0KryuWSfioRdbpKta4qdau+WbmW3 KEPMBPVq3B/XewYU4foyB5DU+lGJTxhdtwdGXtEqPCX6oLWVGcp43Gyv8O2vS+wbFmj5 oY+A==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=dUZ2+1bkg9lTt+1vvivnjWQ7WWv3riHvSqOyna9uoCY=; b=da+y6kTLyJvb2Q9T6wzuJAC3f0OaZC6xyWGibiaqgD/7aRwqEK/wxFW47smPjCoRr9 OrhgzhwVZo3l/LGgsGeqU8/esxrnqRAFHZrLKc3243WEaZ8gou7yU997B3eK/TKOvAIa KVIjS7Y0I8fUf8yMsSp8CERVq5JXKVepYYIiY=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=dUZ2+1bkg9lTt+1vvivnjWQ7WWv3riHvSqOyna9uoCY=; b=MxYTyPIHaZab0ArAnnDVVy+e55T8eOZp70Uq5vUB47AqPKWZNRdRWR4vCXa1Hlx/Fj rwFQAfu5px8z+1ChcfnyqqipXgjt+YwGbkS8AqFWWhll2jcCSX5LByJHhCNw6sV3SFWC R9825O26hFLBskmLAxYZjJ65ySfErr0cBedHY4l65wZ1DLEItH8LehpAimuuBdxT8SKL SVegOyAWW5KqLjNjsbJlILRpIS/5bw5QiIH5BCehmRTnMlhQruj2kWC9/3A3MeaSq4E4 RRum9Jg+Num1SYHRMyy9poc7SqWD+AfzBnC9Agt0hiTfMUwNXm7kg88tGdgzWgc1S+On YiLA==
X-Gm-Message-State: ALoCoQmv8qzDfyZngvup/YISyI/HTnhqoGazH+JtQc0CljEYS0wNc264Kdm+sZZI6S/Mu43vp5ewbscCZ3LZR5FOcl41a3/n3j4e9+IPaJTlPVh+tolkLpgbbQcDmdAiyaA90IrJD78BLzb10WD44edfJIgcqYQG4c5wzVHeW8P4hqUATPaUwmDuaYpwbKXdqJEH5mO7kTa5
MIME-Version: 1.0
X-Received: by 10.236.88.179 with SMTP id a39mr1029721yhf.83.1391164855757; Fri, 31 Jan 2014 02:40:55 -0800 (PST)
Sender: yhirano@google.com
Received: by 10.170.135.22 with HTTP; Fri, 31 Jan 2014 02:40:55 -0800 (PST)
In-Reply-To: <634914A010D0B943A035D226786325D4446C1E925F@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net> <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com> <6652CC7E-7E92-40EF-BD50-2DDD39E15533@zaphoyd.com> <CABihn6GSrU8oJa4nbMJiDDo2Z+-c4fT69roVFxcHYM4XYFToRQ@mail.gmail.com> <F1F115D4-29EB-457B-B02C-71E46D3E1C81@zaphoyd.com> <CABihn6GXxUwHwLd+926VPEkZxHRZUUi-3dw0PoNaO6aK8X4sEQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446C1E8D2C@EXVMBX020-12.exch020.serverdata.net> <CAGzyod7pt_PzjexmnZj8dSMg3BooUmuMb6xB5P0sG1t6S5FXZw@mail.gmail.com> <634914A010D0B943A035D226786325D4446C1E925F@EXVMBX020-12.exch020.serverdata.net>
Date: Fri, 31 Jan 2014 19:40:55 +0900
X-Google-Sender-Auth: e-5U5Z3vBE4zutoo5LiPaHQSI2s
Message-ID: <CABihn6GrDwu=74WEACTrd8RMwDe78G-Q3X+e_OUrYbPcFm8gZQ@mail.gmail.com>
From: Yutaka Hirano <yhirano@chromium.org>
To: Tobias Oberstein <tobias.oberstein@tavendo.de>
Content-Type: multipart/alternative; boundary=14dae9d2fb340d3c5804f141cfb4
Cc: "hybi@ietf.org" <hybi@ietf.org>, Peter Thorson <webmaster@zaphoyd.com>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 31 Jan 2014 10:41:03 -0000

--14dae9d2fb340d3c5804f141cfb4
Content-Type: text/plain; charset=UTF-8

>
> Seriously: _what_ is the goal of this effort?

My motivation is, specifying a multiplexed bidirectional message protocol
conforming to the W3C WebSocket API on HTTP/2.0.

 - The W3C WebSocket API
 - Multiplexing
is mandatory for me.
Since It is built on HTTP/2.0, I prefer strongly the following:
 - A mechanism that enables intermediaries to transfer WS over HTTP/2.0
without fully understanding WS over HTTP/2.0.

The remaining, though they are preferable, are not mandatory for me.
 - RFC6455 conformance
 - small amount of bandwidth efficiency
 - the spec cleanness
 - less-invasiveness

Note: It does not mean I can't accept WS-TUNNEL.

Tobias, Roberto, Peter, can you share your priority?



On Thu, Jan 30, 2014 at 8:54 PM, Tobias Oberstein <
tobias.oberstein@tavendo.de> wrote:

> >I don't find the frame-size arguments very compelling since I haven't
> seen or met anyone with a substantial deployment who cares significantly
> about the overhead difference given packet, TLS, etc. overheads.
> >There is no deployment that we know about that uses frame-based
> extensibility- we can (and should) remove it from the standard since it
> hasn't proven to be interesting.
>
> There is a draft for a WebSocket extension ("permessage-priority") being
> currently developed that makes use of this and we will start an
> experimental implementation in Autobahn. It _is_ interesting.
>
> In any case, the ability for framing, and the ability for WS extensions to
> interleave WS messages is part of RFC6455.
>
> Why remove something from RFC6455 just to fit the limited capabilities
> that SPDY offers?
>
> >I'm not sure I agree with your statement that libraries supporting HTTP/2
> will include necessarily also have WS framing. It seems unlikely to me-- I
> know of at least one large deployment where this isn't at all true.
>
> I'd put it like this: any networking library that wants to support
> "messaging" on the Web will need to support native WebSocket, plus
> optionally any SPDY-based messaging.
>
> Since, WebSocket is deployed _today_ and HTTP2 won't be everywhere anytime
> soon.
>
> E.g. HTTP/2 support on Windows seems to be coming only on Windows 8+
> (since they have HTTP in-kernel). So native WebSocket is the only messaging
> that works on any IE version on Windows <8.
>
> >The HTTP/2 standard has (or shortly will have, as soon as the pull
> request goes in) all support necessary to handle any stream-of-message
> protocol given support of END_MESSAGE.
> >I'd strongly prefer that we map WS to those primitives so it, and any
> other stream-of-message protocols can be handled using loadbalancer/routing
> infrastructure which is and can remain protocol agnostic.
>
> That's fine, let's just not call it WebSocket-over-SPDY then;)
>
> Seriously: _what_ is the goal of this effort?
>
> /Tobias
>
>
> -=R
>
> On Wed, Jan 29, 2014 at 5:56 AM, Tobias Oberstein <
> tobias.oberstein@tavendo.de> wrote:
> >The proposal is mapping one WS Message to one HEADERS + multiple DATAs.
> >For large messages, it can achieve less overhead than WS-TUNNEL at the
> cost of some RFC6455 compatibility.
> How can it achieve less overhead?
>
> A SPDY HEADER with information from first WS frame seems to be always
> bigger than a SPDY DATA header (8 octets) + WS frame header.
>
> The only situation I can think of: a large, _fragmented_ WS message is
> reframed into a SPDY HEADER+DATAs sequence.
>
> Effectively, you do away with the individual WS frame information. And
> this exactly breaks full RFC6455 compatibility.
>
> /Tobias
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>
>

--14dae9d2fb340d3c5804f141cfb4
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px =
0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-l=
eft-style:solid;padding-left:1ex"><span style=3D"font-family:arial,sans-ser=
if;font-size:13px">Seriously: _what_ is the goal of this effort?</span></bl=
ockquote>
<div>My motivation is, specifying a multiplexed bidirectional message proto=
col conforming to the W3C WebSocket API on HTTP/2.0.</div><div><br></div><d=
iv>=C2=A0- The W3C WebSocket API</div><div>=C2=A0- Multiplexing</div><div>i=
s mandatory for me.<br>
</div><div>Since It is built on HTTP/2.0, I prefer strongly the following:<=
/div><div>=C2=A0- A mechanism that enables intermediaries to transfer WS ov=
er HTTP/2.0 without fully understanding WS over HTTP/2.0.<br></div><div><br=
>
</div><div>The remaining, though they are preferable, are not mandatory for=
 me.</div><div>=C2=A0- RFC6455 conformance</div><div>=C2=A0- small amount o=
f bandwidth efficiency</div><div>=C2=A0- the spec cleanness</div><div>=C2=
=A0- less-invasiveness</div>
<div><br></div><div>Note: It does not mean I can&#39;t accept WS-TUNNEL.</d=
iv><div><br></div><div>Tobias, Roberto, Peter, can you share your priority?=
</div><div><br></div></div><div class=3D"gmail_extra"><br><br><div class=3D=
"gmail_quote">
On Thu, Jan 30, 2014 at 8:54 PM, Tobias Oberstein <span dir=3D"ltr">&lt;<a =
href=3D"mailto:tobias.oberstein@tavendo.de" target=3D"_blank">tobias.oberst=
ein@tavendo.de</a>&gt;</span> wrote:<br><blockquote class=3D"gmail_quote" s=
tyle=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class=3D"im">&gt;I don&#39;t find the frame-size arguments very compel=
ling since I haven&#39;t seen or met anyone with a substantial deployment w=
ho cares significantly about the overhead difference given packet, TLS, etc=
. overheads.<br>

&gt;There is no deployment that we know about that uses frame-based extensi=
bility- we can (and should) remove it from the standard since it hasn&#39;t=
 proven to be interesting.<br>
<br>
</div>There is a draft for a WebSocket extension (&quot;permessage-priority=
&quot;) being currently developed that makes use of this and we will start =
an experimental implementation in Autobahn. It _is_ interesting.<br>
<br>
In any case, the ability for framing, and the ability for WS extensions to =
interleave WS messages is part of RFC6455.<br>
<br>
Why remove something from RFC6455 just to fit the limited capabilities that=
 SPDY offers?<br>
<div class=3D"im"><br>
&gt;I&#39;m not sure I agree with your statement that libraries supporting =
HTTP/2 will include necessarily also have WS framing. It seems unlikely to =
me-- I know of at least one large deployment where this isn&#39;t at all tr=
ue.<br>

<br>
</div>I&#39;d put it like this: any networking library that wants to suppor=
t &quot;messaging&quot; on the Web will need to support native WebSocket, p=
lus optionally any SPDY-based messaging.<br>
<br>
Since, WebSocket is deployed _today_ and HTTP2 won&#39;t be everywhere anyt=
ime soon.<br>
<br>
E.g. HTTP/2 support on Windows seems to be coming only on Windows 8+ (since=
 they have HTTP in-kernel). So native WebSocket is the only messaging that =
works on any IE version on Windows &lt;8.<br>
<div class=3D"im"><br>
&gt;The HTTP/2 standard has (or shortly will have, as soon as the pull requ=
est goes in) all support necessary to handle any stream-of-message protocol=
 given support of END_MESSAGE.<br>
&gt;I&#39;d strongly prefer that we map WS to those primitives so it, and a=
ny other stream-of-message protocols can be handled using loadbalancer/rout=
ing infrastructure which is and can remain protocol agnostic.<br>
<br>
</div>That&#39;s fine, let&#39;s just not call it WebSocket-over-SPDY then;=
)<br>
<br>
Seriously: _what_ is the goal of this effort?<br>
<span class=3D"HOEnZb"><font color=3D"#888888"><br>
/Tobias<br>
</font></span><div class=3D"HOEnZb"><div class=3D"h5"><br>
<br>
-=3DR<br>
<br>
On Wed, Jan 29, 2014 at 5:56 AM, Tobias Oberstein &lt;<a href=3D"mailto:tob=
ias.oberstein@tavendo.de">tobias.oberstein@tavendo.de</a>&gt; wrote:<br>
&gt;The proposal is mapping one WS Message to one HEADERS + multiple DATAs.=
<br>
&gt;For large messages, it can achieve less overhead than WS-TUNNEL at the =
cost of some RFC6455 compatibility.<br>
How can it achieve less overhead?<br>
<br>
A SPDY HEADER with information from first WS frame seems to be always bigge=
r than a SPDY DATA header (8 octets) + WS frame header.<br>
<br>
The only situation I can think of: a large, _fragmented_ WS message is refr=
amed into a SPDY HEADER+DATAs sequence.<br>
<br>
Effectively, you do away with the individual WS frame information. And this=
 exactly breaks full RFC6455 compatibility.<br>
<br>
/Tobias<br>
_______________________________________________<br>
hybi mailing list<br>
<a href=3D"mailto:hybi@ietf.org">hybi@ietf.org</a><br>
<a href=3D"https://www.ietf.org/mailman/listinfo/hybi" target=3D"_blank">ht=
tps://www.ietf.org/mailman/listinfo/hybi</a><br>
<br>
</div></div></blockquote></div><br></div>

--14dae9d2fb340d3c5804f141cfb4--

From tobias.oberstein@tavendo.de  Fri Jan 31 04:13:58 2014
Return-Path: <tobias.oberstein@tavendo.de>
X-Original-To: hybi@ietfa.amsl.com
Delivered-To: hybi@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 170271A0598 for <hybi@ietfa.amsl.com>; Fri, 31 Jan 2014 04:13:58 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.299
X-Spam-Level: 
X-Spam-Status: No, score=-1.299 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, J_CHICKENPOX_65=0.6, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id lXAeurJAK-io for <hybi@ietfa.amsl.com>; Fri, 31 Jan 2014 04:13:56 -0800 (PST)
Received: from EXHUB020-1.exch020.serverdata.net (exhub020-1.exch020.serverdata.net [206.225.164.28]) by ietfa.amsl.com (Postfix) with ESMTP id CDCCE1A0588 for <hybi@ietf.org>; Fri, 31 Jan 2014 04:13:55 -0800 (PST)
Received: from EXVMBX020-12.exch020.serverdata.net ([169.254.3.11]) by EXHUB020-1.exch020.serverdata.net ([206.225.164.28]) with mapi; Fri, 31 Jan 2014 04:13:51 -0800
From: Tobias Oberstein <tobias.oberstein@tavendo.de>
To: Yutaka Hirano <yhirano@chromium.org>
Date: Fri, 31 Jan 2014 04:13:48 -0800
Thread-Topic: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
Thread-Index: Ac8ecOzevOgS8iycSh6Vh+HJ2jxp2wAC3NOA
Message-ID: <634914A010D0B943A035D226786325D4446C1E9735@EXVMBX020-12.exch020.serverdata.net>
References: <CAH9hSJah=4Z4-NufVrUrP965epc9eMdEbC+3LmucN4DhR=bWiQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403DF@EXVMBX020-12.exch020.serverdata.net> <CABihn6H5EtDwhXMYhh-BMb5m6yxL9nYv6=Ud4fN5DeQkjDGyPQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446BE403E1@EXVMBX020-12.exch020.serverdata.net> <CABihn6EaE+69q3AJqxEj6vGztBHu4FdOcLCvZ_9Jg-e2cFLwEg@mail.gmail.com> <CAGzyod7ymnpOTiiH7xN3W-aBxD2NNPc=S_Oawdihfhuu84unLw@mail.gmail.com> <634914A010D0B943A035D226786325D4446BED9CDE@EXVMBX020-12.exch020.serverdata.net> <CABihn6EGA3znv_WsJ1gELXyNXgY2fy8weBJ62yD+M2iP7xEnCA@mail.gmail.com> <634914A010D0B943A035D226786325D4446BF99976@EXVMBX020-12.exch020.serverdata.net> <CABihn6GFnGTB9sOZ1=2UnOv-pHPRnRY96DnOb_Q24bNUaxgR=w@mail.gmail.com> <9230B7A2-56AD-4807-A53A-928253364B52@zaphoyd.com> <634914A010D0B943A035D226786325D4446BF99A9B@EXVMBX020-12.exch020.serverdata.net> <F088AE82-E1AF-489B-B220-6E911078C1F3@zaphoyd.com> <6652CC7E-7E92-40EF-BD50-2DDD39E15533@zaphoyd.com> <CABihn6GSrU8oJa4nbMJiDDo2Z+-c4fT69roVFxcHYM4XYFToRQ@mail.gmail.com> <F1F115D4-29EB-457B-B02C-71E46D3E1C81@zaphoyd.com> <CABihn6GXxUwHwLd+926VPEkZxHRZUUi-3dw0PoNaO6aK8X4sEQ@mail.gmail.com> <634914A010D0B943A035D226786325D4446C1E8D2C@EXVMBX020-12.exch020.serverdata.net> <CAGzyod7pt_PzjexmnZj8dSMg3BooUmuMb6xB5P0sG1t6S5FXZw@mail.gmail.com> <634914A010D0B943A035D226786325D4446C1E925F@EXVMBX020-12.exch020.serverdata.net> <CABihn6GrDwu=74WEACTrd8RMwDe78G-Q3X+e_OUrYbPcFm8gZQ@mail.gmail.com>
In-Reply-To: <CABihn6GrDwu=74WEACTrd8RMwDe78G-Q3X+e_OUrYbPcFm8gZQ@mail.gmail.com>
Accept-Language: de-DE, en-US
Content-Language: de-DE
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: de-DE, en-US
Content-Type: multipart/alternative; boundary="_000_634914A010D0B943A035D226786325D4446C1E9735EXVMBX02012ex_"
MIME-Version: 1.0
Cc: "hybi@ietf.org" <hybi@ietf.org>, Peter Thorson <webmaster@zaphoyd.com>
Subject: Re: [hybi] Discontinuation of mux standardizaton in favor of WS/HTTP/2.0
X-BeenThere: hybi@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Server-Initiated HTTP <hybi.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/hybi>, <mailto:hybi-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/hybi/>
List-Post: <mailto:hybi@ietf.org>
List-Help: <mailto:hybi-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/hybi>, <mailto:hybi-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 31 Jan 2014 12:13:58 -0000

--_000_634914A010D0B943A035D226786325D4446C1E9735EXVMBX02012ex_
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_634914A010D0B943A035D226786325D4446C1E9735EXVMBX02012ex_
Content-Type: text/html; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_634914A010D0B943A035D226786325D4446C1E9735EXVMBX02012ex_--
