[Standards] geoloc: PEP=user, <iq>=resource

Stephen Pendleton pendleto at movsoftware.com
Fri Jun 1 18:12:21 UTC 2007

Ah, I was confused because it looked like the examples were for pubsub
(XEP-0060), not PEP, because the section header says "4.1 Entity publishes
location via pubsub". Even though PEP is a specific form on pubsub I was
thinking of XEP-0060. I just reread it and saw where in parenthesis there is
a reference to the examples being based on PEP (XEP-0163). My mistake.

There are no examples then on how this can be used with pubsub (XEP-0060) in
the XEP. In the pubsub case the <addresses> info is needed, unless you gave
some semantic meaning to the pubsub node name. I hear people talking about
pubsub being deprecated in this case in favor of PEP. If that is true, and
presumably no one has actually implemented a pubsub version off geoloc
(other than me), can we just drop references to using geoloc with XEP-0060
out of the next draft? It would be great for interoperability if there was
only "one true way" of doing geoloc. I did my implementation before PEP was
around, or I would have chosen the PEP path myself.

In PEP, your bare JID simply *is* a virtual pubsub server, so the 'from' 
address is your bare JID, so there is no need for the <addresses> info 
(that was necessary if the node was hosted on a pubsub component with 
its own separate JID, but not in PEP since the pubsub service is your 
bare JID).


Peter Saint-Andre
XMPP Standards Foundation http://www.xmpp.org/xsf/people/stpeter.shtml

More information about the Standards mailing list