Jingle IQ with "source-add" action

Hi all,

I’m trying to understand how participants communicate in sending/receiving data.

For instance in 2 participants case, if there is a second participant joining the room (with default value = 2) that actually create/start the conference from jvb. Then the channel config with other contents are sent to jicofo, then session-inititate is created and sent to 2 participants. Then 2 participants from clients sends back session-accept to jicofo, then each participant receive a jingle iq with “source-add” action from jicofo with ssrc/ssrc-group (to update each other’s information and config to client sides, based on my understanding)… Then now they can see each other on each client side.

However, when there is a third person joining the conference, everything is pretty much the same (session-initiate and session-accept between client and jicofo). Then current participant 1 and 2 each receives a “source-add” action jingle iq to update the information about the third person joining, right? But why the third person doesn’t receive any “source-add” action jingle iq to update the current participants on its client side? Am I misunderstanding the use of “source-add” here?

This question is pretty theoretical because I’m just very new to XMPP and WebRTC. I’ll be very appreciated if anyone can explain this to me…

Also, it’s also vague to me about ssrc/ssrc-group and how they are actually used?

Cause it receives the session-initiate with all needed information. So in jingle session-initiate is to create the session and later you can modify the session with messages like source-add, source-remove, source-modify …

This is what ssrc to be used for the media and in the case of simulcast streams are grouped for the different layers and their ssrc.

1 Like

Thank you Damencho, I have looked at the session-initiate again and seen the information of jvb and current participants as source tag in there. Thanks man.