To be honest I would prefer to not put extra code in Firefox. I'm sure it will result in tons of new edge cases where things can break. I understand that 3264 permits this. My guess is that there are use cases for offer answer in general (SIP or other protocols) where establishing an empty session makes sense. But I fail to see the relevant use for WebRTC. The only use case I'm aware of are the WebRTC WPT trying to utilize this. But I don't think that warrants the extra effort and risk. So even though JSEP allows this from an IETF perspective I would prefer if the W3C would explicitly forbid this, because of lack of use for WebRTC. If this group decides that it is needed for WebRTC then Firefox will implement it (I think that is what @jan-ivar was trying to say). -- GitHub Notification of comment by nils-ohlmeier Please view or discuss this issue at https://github.com/w3c/webrtc-pc/issues/1745#issuecomment-382868693 using your GitHub accountReceived on Thursday, 19 April 2018 20:22:37 UTC
This archive was generated by hypermail 2.4.0 : Saturday, 6 May 2023 21:19:44 UTC