Re: [rtcweb] Use of offer / answer semantics

Paul Kyzivat <> Thu, 08 September 2011 15:50 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id 08D6421F8557 for <>; Thu, 8 Sep 2011 08:50:16 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -2.521
X-Spam-Status: No, score=-2.521 tagged_above=-999 required=5 tests=[AWL=0.078, BAYES_00=-2.599]
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id U3y+oRyEedpo for <>; Thu, 8 Sep 2011 08:50:15 -0700 (PDT)
Received: from ( []) by (Postfix) with ESMTP id 4C3D121F84DA for <>; Thu, 8 Sep 2011 08:50:15 -0700 (PDT)
Received: from ([]) by with comcast id WF8j1h0030mv7h05AFs8js; Thu, 08 Sep 2011 15:52:08 +0000
Received: from Paul-Kyzivats-MacBook-Pro.local ([]) by with comcast id WFs71h0190tdiYw3XFs7bm; Thu, 08 Sep 2011 15:52:08 +0000
Message-ID: <>
Date: Thu, 08 Sep 2011 11:52:42 -0400
From: Paul Kyzivat <>
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.7; rv:5.0) Gecko/20110624 Thunderbird/5.0
MIME-Version: 1.0
References: <>, <> <BLU152-W507A8040FD123508451E51931E0@phx.gbl>
In-Reply-To: <BLU152-W507A8040FD123508451E51931E0@phx.gbl>
Content-Type: text/plain; charset="ISO-8859-1"; format="flowed"
Content-Transfer-Encoding: 7bit
Subject: Re: [rtcweb] Use of offer / answer semantics
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Real-Time Communication in WEB-browsers working group list <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Thu, 08 Sep 2011 15:50:16 -0000

On 9/8/11 12:45 AM, Bernard Aboba wrote:
>  > > 1) The media negotiations will be done using the same SDP
> offer/answer semantics that are used in SIP.
>  > To be precise - you're suggesting that we use RFC 3264 offer/answer
>  > semantics. (That RFC is 25 pages long. RFC 3261, the core SIP document,
>  > is 269 pages, and is NOT a normative reference from 3264. I am anxious
>  > to avoid having a normative dependency on 3261.)
>  >
>  > I agree with this.
> [BA] I do *not* agree that RTCWEB should have to support every aspect of
> SDP offer/answer. Basic offer/answer, sure. All potential corner cases?
> Not necessarily.

I'm not sure where you are going here?
Are you suggesting that all the mandatory to implement O/A semantics of 
3264 might not be supported? Or are you saying that O/A support may not 
work for all defined SDP extensions?

I think that all mandatory O/A should be supported. If you have 
something specific in mind that is problematic, then maybe we should 
investigate why you think it needs to be excluded. My guess is that 
maybe it is something broken in the spec that ought to be fixed there.

>  > > 2) It will be possible to gateway between legacy SIP devices that
> support ICE and appropriate RTP / SDP mechanisms and codecs without
> using a media gateway. A signaling gateway to convert between the
> signaling on the web side to the SIP signaling may be needed.
>  > I agree with this - I think the "may be needed" will turn out to be
>  > "will be needed", but some portion of that gateway can be implemented by
>  > Javascript running in the browser, if desirable.
> [BA] This seems like a good principle, but I'm not clear that it will
> work with all use cases. For example, what happens in the E911 use cases
> when an RTCWEB implementation attempts to make a call to a PSAP
> implementing NENA i3 Stage 3? If you don't have a media gateway, then
> the browser will need to implement one of the mandated codecs on the
> PSAP side. So in those use cases, eliminating the media gateway implies
> making G.711 and H.264 mandatory-to-implement.

AFAIK, not every rtcweb application will be obligated to support E911.
(In particular, any application that doesn't identify callees by phone 
number is a good candidate to be exempt from E911.

Certainly a server without a media gateway will be limited in what it 
can call based on codec compatibility. That may or may not be a 
limitation, depending on the application. Those that find it an 
unacceptable limitation will probably find a way to incorporate a 
transcoder when needed.