Juliusz Chroboczek

Until recently, Pion has been using a system that apparently works for everyone: * people would submit improvements to Pion, without going out of their way to ensure it works...

I'm sorry, but I don't have either the time or the inclination to have this discussion. In the meantime, however, compilation under gccgo is still broken. So please either apply...

> This discussion started, I believe, as a consequence of Galene users... I'm pretty sure that's not the case. You started this discussion single-handed, thus creating a problem where there...

> Most users don't care if a packet is a RTX, they just want to process packets as they arrive In my understanding, the whole point of having a separate...

The same mechanism is used for RED. Unilke RTX, which is nice to have but not difficult to work around, RED is what allows you to do FEC with audio,...

Let me explain some more why a separate track is essential for FEC but not strictly required for retransmissions. Suppose you are doing retransmissions on the same track. You receive...

> Do you find any advantage on using FEC with RED? I haven't done any experiments yet. > On my statics gathered, when network gets congested packets gets dropped in...

I believe that's correct.

I think that pion/webrtc#1803 is fairly important, it regularly appears in my race detector logs. What about https://github.com/pion/webrtc/issues/1675? This causes real issues with non-browser clients. I'm also taking the opportunity...

No. My current hypothesis is that it's the same issue as https://github.com/pion/webrtc/issues/1356, which is apparently due to having multiple local addresses on a single local socket; that's going to happen...