On 02/13/2014 08:32 PM, Martin Thomson wrote: > On 13 February 2014 10:22, Stefan Håkansson LK > <stefan.lk.hakansson@ericsson.com> wrote: >> What happens if, at a later stage, the app does addTrack to the >> MediaStream that was earlier done AddStreamId for? Or removeTrack? >> >> (Of course I have nothing against what is proposed in principle, I just >> fear we adding a lot of complexity for no practical gain) > Add X can be made idempotent, I see no issue there. > I think Stefan is saying there's a difference between pc.AddStreamID(stream1) .... stream1.AddTrack(track1) (should not cause signalling, should not cause track1 to be sent) pc.AddTrack(track1) (causes track1 to be sent without a stream) pc.AddStreamID(stream1) (causes nothing to happen - stream1 is empty) stream1.AddTrack(track1) (should this cause signalling that track1 is now a member of stream1?) All sequences of *.Add* shoud have a deterministic result, even though the result of nonsensical sequences doesn't have to look very sensible :-) -- Surveillance is pervasive. Go Dark.Received on Thursday, 13 February 2014 19:50:14 UTC
This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:17:54 UTC