On Thursday 09 April 2009 10:26:10 Mridul Muralidharan wrote: > I was trying to understand how current components and clients behave ... > particularly since psi and others would have already faced and worked > around/solved this issue.
Psi doesn't try to detect the type of contact to know if presence without a resource could happen. I'm willing to be no client does that, actually.. Besides, a transport contact could actually send a resource, and probably some of them do. I don't think it's fair to assume a transport would always send presence without a resource. So, a client has to be prepared to accept both. I don't see much point in having special handling depending on contact type. I agree with Robin, that the RFC should at least clarify what it means to have presence from no resource. Probably it should just be treated as a resource of 0-length, that does not overshadow other resources from the same bare jid. I think this would describe current practice. The alternative approach of changing all existing implementations to never send nor support presence from no resource is not practical. That ship has sailed. -Justin
