[jdev] Re: anonymous login

[Peter Saint-Andre]

Ah, true, that's the right place to do it. Sounds like an Internet-Draft 
may be in order... ;-)

/psa

In article <82777bea05020722322a491f29 at mail.gmail.com>,
 Joe Hildebrand <hildjj at gmail.com> wrote:

> There are also cases where you just want the server to pick a full JID
> for you, and ANONYMOUS is overkill.  Customers coming in to a customer
> service site is a good example.
> 
> It may make sense to come up with a UNIQUE SASL mechanism that tells
> the server to create a new, unique JID, good for one shot.
> 
> On Tue, 01 Feb 2005 10:49:57 -0700, Peter Saint-Andre
> <stpeter at jabber.org> wrote:
> > In some applications, it might be helpful for an XMPP server to assign a
> > JID to an entity that connects with a stream that is qualified by the
> > 'jabber:client' namespace (or more precisely assign the node identifier
> > portion of the entity's JID, just as the server can assign the resource
> > identifier of the entity's JID using the resource binding protocol).
> > Chatting with someone via IM just now, we worked out the following:
> > 
> > 1. entity authenticates via SASL ANONYMOUS
> > 
> > 2. server tells entity that resource binding is required (via stream
> > feature)
> > 
> > 3. entity requests that server create a resource for it
> > 
> > 4. server replies with full JID:
> > 
> > <iq type='result' id='bind_2'>
> >   <bind xmlns='urn:ietf:params:xml:ns:xmpp-bind'>
> >     <jid>somenode at example.com/someresource</jid>
> >   </bind>
> > </iq>
> > 
> > Thoughts?