During Januscon this year we had the first edition of “VoIP On the Beach” an initiative to discuss issues in VoIP-land and try to come up with ideas or solutions together. This time the topic that was chosen to be discussed was the federation of WebRTC. People attending were split into two groups, one talking about the benefits and solutions to doing WebRTC federation and one group that was making an argument against this. I made some notes for the ‘pro’ side and will share them with you here to try and continue the discussion here. @giavac has the notes for the ‘against’ side and will share those with you here as well.
During our discussions we noted a few areas of interest in which we can see challenges that need to be overcome in some way to create a truly federated WebRTC system.
Trust & identity
One issue would be trust. How can you know for sure that the person calling you/being called is actually who they say they are? And how do you stop spammers from annoying people?
Options could include a ‘web of trust’ like GPG, an AdBlocker type system where public lists are kept of corporations/individuals that spam the network, or potentially a karma based system.
We’ve also looked at previous attempts by protocols & programs to federate to take some learnings from them. XMPP for example is a good example of a protocol that has federation as one of the main benefits, though it can be tricky to setup correctly and end-to-end encryption is the dependent on the client and not enforced by the protocol itself.
Another example is Matrix, which does federation in a different way by creating interoperability between different networks (Slack/IRC/XMPP etc.) using what they call ‘bridges’.
I’m sure I missed some points from the discussions so please add where possible!