On Sat, Feb 6, 2010 at 11:04 AM, Peter Saint-Andre <[email protected]> wrote: > Also while travelling yesterday I updated the old distributed MUC proposal: > http://xmpp.org/extensions/inbox/distributedmuc.html
My promised comments: There's another approach to this, discussed by Dave Cridland and myself offlist, which is interesting. The basic premise being that instead of pre-arranging with conference.jabber.org to shadow mucs there, it could do it on the fly (subject to c.j.o supporting this - but there's no need for a peering agreement or the like, because there are only benefits to this). So, if I run example.com, I may decide that it's worth mirroring cjo for my users. I tell my server to mirror c.j.o. It will then start treating incoming c2s packets to cjo as being handled locally. When a user joins a muc on cjo, it goes to cjo and says "hey, I'm going to start mirrorring you on behalf of my users, and this user has joined". To cjo, it will be appearing as e.g. [email protected] or the like, and both ends will fold and unfold stanzas over the link, not dissimilarly from Peter's current version. In the case where cjo themselves want to have available several available peers, they can set up SRV records for the service to point to pre-agreed peers. Then you've got the master/slave split, where some messages are broadcast locally (message/presence) as well as broadcast up the chain, while others are passed straight up (config changes) and don't take effect when you're in a netsplit. Dave will probably now explain how we discussed something entirely different and better. Thoughts? /K
