David Zhao

Jumping into this late. Been thinking about this one for a bit. I support having an official "version policy" around Go version support. My reasons are: * set clarity and...

Far from being an expert on this, but I think the current behavior is actually correct. [RTPTransceiver.Stop()](https://github.com/pion/webrtc/blob/ea7964c12c1a1e51fe9d86c8cc9cae8cbca22e58/rtptransceiver.go#L151) in Pion doesn't work exactly the same as the browser. The current implementation...

@k0nserv I believe you are right. I got a bit confused because the behavior is different depending on if it's sending or receiving. When sending, Pion handles transceiver re-use correctly....

> Which `Stop` are you referring to? Is it the same as the one @k0nserv originally referred to? Yes, I'm referring to the same Stop > https://github.com/pion/webrtc/blob/ea7964c12c1a1e51fe9d86c8cc9cae8cbca22e58/peerconnection.go#L1058-L1061 > > This...

Yeah, I think that should be sufficient.. but this chain of events is sort of complex.. wdyt @Sean-Der ?

> Another question on my mind is: How should it be detected when a transceiver transitions from .e.g. `recvonly` to `inactive`? In an SFU case this event must lead to...

> The spec says that RTCP should continue to be sent but RTP media should stop. You are right. though I'm not sure what type of RTCP it would still...

@jech we are using this (and TCPMux) with LiveKit and only see packet loss when UDP buffer gets overwhelmed. (increasing it gets rid of the loss for us). https://github.com/livekit/livekit-server/blob/master/pkg/rtc/config.go#L70 Do...

what is considered double-stack? having both ipv4/6?

With livekit we are using UDP4 with the mux and that could explain the difference. The challenge with dual-stack is ensuring what's advertised to match the dest addr that we...