firefox 1-1 session about jitsi-meet

[fippo]

it seems that firefox has DTLS problems when being the focus. @emcho / @lyubomir can you take a look? Testing seems easy, just go into a room with FF and make an empty conference.

This is weird because IIRC we now behave the same for the focus as for participants, no?

[fippo]

interesting... so FF nightly has an about:webrtc page which shows that the host candidates with the remote candidate failed but peer reflexive candidates are succeeded.

wireshark only sees a few STUN packets. No attempt to make a DTLS handshake from the focus.

That is weird...

[fippo]

http://colibri.hancke.name/test.html -- I see two binding requests which are answered by the bridge, nothing beyond that. No ice-controlling/ice-controlled... not sure why it works in the nonfocus case.

[fippo]

This was caused by the lack of generation which caused the bridge to generate a parse exception. I can get relayed A/V via the bridge now which is a step in the right direction.