W3C home > Mailing lists > Public > public-webrtc@w3.org > July 2017

Re: Mangling a=extmap and a=mid before setLocalDescription

From: Iñaki Baz Castillo <ibc@aliax.net>
Date: Wed, 19 Jul 2017 14:12:44 +0200
Message-ID: <CALiegfmHMr-cDMckAELTkOvaMYY4X4_a0ZBsYoZ7VK3cMJTpgw@mail.gmail.com>
To: Adam Roach <adam@nostrum.com>
Cc: "public-webrtc@w3.org" <public-webrtc@w3.org>
2017-07-19 14:07 GMT+02:00 Adam Roach <adam@nostrum.com>:
> On 7/19/17 13:58, Iñaki Baz Castillo wrote:
>>
>> I know that the idea is to not have to mangle the SDP to get some cool
>> things, but before that happens in the real world, is it allowed to
>> mangle the ID of a=extmap and the value of a=mid between createOffer()
>> and setLocalDescription()?
>>
>> Are current browsers supposed to honor those new values given in
>> setLocalDescription() and use them within the RTP packets?
>
>
> JSEP makes this illegal; see
> https://tools.ietf.org/html/draft-ietf-rtcweb-jsep-21#section-5.4
>
> I'm sure Firefox won't honor such changes.

Nice to know that.


> I suspect Chrome doesn't; and, if it does, it shouldn't.

Yes, but currently a=ssrc, payload numbers (among others) can be
mangled and it works. Anyhow I agree that the app should not do this.

Once this is clear (thanks for the reference) I wonder why do we need
the "createOffer" + "setLocalDescription" ceremony at all? I mean, if
the SDP provided by createOffer() should not be changed in any way and
setLocalDescription() should be called with it, why does it even
exist? It provides nothing.



-- 
Iñaki Baz Castillo
<ibc@aliax.net>
Received on Wednesday, 19 July 2017 12:13:33 UTC

This archive was generated by hypermail 2.3.1 : Monday, 23 October 2017 15:19:51 UTC