[Standards] Re: resource identifiers: a summary

JD Conley jd.conley at coversant.net
Fri Jun 1 11:44:49 CDT 2007


> Ian Paterson schrieb:
> > A very good summary, thanks. :-)
> >
> > +1 for the action points.
> >
> > - Ian
> 
> also +1 from me
> 
> Alex

Ditto on that. I do have one addition though. Servers such as ours and
Google's (and maybe others) use a key contained in the server generated
resource identifier for intra-cluster routing, so it starts out as
something random/unique/opaque, but then gets some real meaning. That
way when anything comes into one node addressed to a full JID it can be
routed to another node by hitting a routing table the size of the number
of cluster nodes instead of the number of resources. I doubt that
belongs in the spec, but just an FYI on some real world usage...

-JD


More information about the Standards mailing list