Re: ICE-SIP-SDP updates

Hi Suhas,


On Mon, Sep 18, 2017 at 12:15 AM, Suhas Nandakumar <[email protected]>
wrote:

> 1. Handling of ICE mismatch
>
> Where is ICE mismatch currently defined? I cannot find this in either
> rfc5245bis or ICE-SIP-SDP.
>
> I think ICE mismatch should occur if address in c= line and port in m=
> line do not match any of the candidates and <new>address in c= line is not
> dummy address IN IP4 0.0.0.0 and port in m= line is not dummy port 9
> </new>. This is needed to match https://tools.ietf.org/html/
> draft-ietf-rtcweb-jsep-21#section-5.2.1 and trickle ICE
>
> [Suhas] ICE mismatch is discussed in 4.1.2.3,  4.1.3.1 and 11.2.2
>

[roman] Section 4.1.2.3 should be modified to allow c= line with IN IP4
0.0.0.0 and port in m= line with port 9 not to be treated as ice-mismatch
even if it does not match any of the candidates. Without this, trickle ICE
and JSEP will cause ice-mismatch. Furthermore, rtcp attribute is optional,
so the text needs to be modified to allow for this.

2. Move general requirements from https://tools.ietf.org/html/
> draft-ietf-mmusic-sctp-sdp-26#section-12.2 into ICE-SIP-SDP.
>
> [Suhas] Added the supporting text in the above PR. Also the repository has
> HTML file for easy reading.
>

[roman] This looks good, but I will double check and provide comments if I
find anything.

>
> 3. ICE transport tag
>
> [Suhas] IIRC, this topic was brought to discussion several times in the WG
> and there was no consensus on defining a new ICE specific transport tag.
> Personally, I feel this will be a large undertaking and we should avoid
> doing this as part of this draft work. I am open to suggestions.
>
>
 [roman]  I agree that this can be left out and handled in the future draft
if there is enough support for it.

Regards,
_____________
Roman Shpount

_______________________________________________
mmusic mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/mmusic