On Thu, 2011-08-25 at 15:00 +0200, Andreas Monitzer wrote: > On Thursday, August 25 2011 at 00:19, Peter Saint-Andre wrote: > > Right, there are use cases (e.g., vCards in chatrooms) that make PEP > > unworkable. I'm willing to have the discussion again. :) > > Then why not fix it at the source (PEP), maybe by creating a separate > XEP for MUC-PEP? Is it better to have a workaround in every XEP that > would otherwise use PEP instead?
Agreed. We've talked about his on several instances of the XMPP Summit, lastly in the Cisco offices near Brussels, sparked by the very same specification for vCard. As far as I can remember we covered a lot of ground, but didn't reach a conclusion. I'm not sure if anyone made notes there, but it would be valuable to have a summary of aspects and issues around PubSub inside MUC. Kevin, Matt, Joe, do you have such notes? > My major issue is that requiring specific server support results in a > huge delay in deployment, and some servers will never support it at > all (which means that client developers will still have to implement > vcard-temp in 2020), since many server developers and admins are of > the kind "if it ain't broke, don't touch it". I'm not sure if this is also about using PEP for transporting vCard4, but I seem to remember some of the possible solutions did impose cooperation of the MUC room for doing PEP 'properly'. -- ralphm
