- From: Bernard Aboba <Bernard.Aboba@microsoft.com>
- Date: Fri, 13 Jul 2018 01:02:51 +0000
- To: Iñaki Baz Castillo <ibc@aliax.net>
- CC: Harald Alvestrand <harald@alvestrand.no>, WebRTC WG <public-webrtc@w3.org>
- Message-ID: <BL0PR00MB03235F8CEFDD1A9A98D7BB5DEC580@BL0PR00MB0323.namprd00.prod.outlook.com>
Inaki said: "Please, don't forget RTX/FLEXFEC stuff in simulcast. If RID is used, RRID must be used. If just SSRC is used then ssrc-group is needed. Etc etc." [BA] JSEP does not mention RRID, nor do any of the examples include the repaired-rtp-stream-id in the SDP. However, RRID is mentioned in draft-ietf-mmusic-sdp-simulcast and SDP examples such as Figure 7 do include it. With respect to the object model, there was a discussion of RRID two years ago: https://github.com/w3c/ortc/issues/406 Conclusion seemed to be that it was not necessary, but perhaps the conclusion was too hasty. [https://avatars3.githubusercontent.com/u/808811?s=400&v=4]<https://github.com/w3c/ortc/issues/406> RRID support · Issue #406 · w3c/ortc · GitHub<https://github.com/w3c/ortc/issues/406> GitHub is where people build software. More than 28 million people use GitHub to discover, fork, and contribute to over 85 million projects. github.com ________________________________ From: Iñaki Baz Castillo <ibc@aliax.net> Sent: Thursday, July 12, 2018 2:03 PM To: Bernard Aboba Cc: Harald Alvestrand; WebRTC WG Subject: Re: Some ideas on SVC support in WebRTC 1.0 (Take 2) Please, don't forget RTX/FLEXFEC stuff in simulcast. If RID is used, RRID must be used. If just SSRC is used then ssrc-group is needed. Etc etc. Too many ways to express the same does not usually end well. El jue., 12 jul. 2018 22:59, Bernard Aboba <Bernard.Aboba@microsoft.com<mailto:Bernard.Aboba@microsoft.com>> escribió: Harald said: “It seems that -jsep took away the requirement to generate a=ssrc in version -17. Should we attempt to add it back, or should we just depend on the good graces of the implementors?” [BA] I don’t think we need to change JSEP. Implementers appear to understand the backward compatibility issues. If SFUs eventually support MID and RID header extensions, then “a=ssrc” line support will be less necessary, and then having a mandate in the specification would become a “ball and chain”.
Received on Friday, 13 July 2018 01:03:18 UTC