Re: [sipcore] Questions on location conveyance and dereferencing

["Tschofenig, Hannes (NSN - FI/Espoo)" <[e-mail]>]

Hi John, 
 

> > In your enterprise LIS for emergency services you will most 
> > likely want
> > to figure out who is going to consume that reference. Quite 
> likely the
> > PSAP will consume the reference. If so, you better talk to the PSAP
> > responsible for that specific area. 
> [JRE] Quite so, except is it likely my VoIP service provider 
> might also want to inspect the location, to confirm correct 
> routing? Then depending on which SP I choose to route that 
> particular call through, they might support different URI 
> schemes? It seems to be a mess not having a MUST implement scheme.

I understand the theoretical concern. 
I have no objections against a mandatory to implement HTTP-based
dereferencing mechanism. 


> > XMPP does not provide equivalent functionality of SIP location
> > conveyance. 
> > So, another non-issue. 
> [JRE] That was not the point I was making. I was simply 
> saying that even applications that support presence might 
> support XMPP rather than SIMPLE, and therefore would not want 
> to implement RFC 3856. Instead they might prefer to use HTTP.
I am not quite sure which scenario you have in mind. 
If I have an end host that supports XMPP and then gets a location URI
then there is the question of who is going to dereference it. If it is
the end point then it would have to implement the protocol indicated by
the URI scheme. Now, if this happens to be SIP then this end point would
have to implement SIP. The end host could, however, request location by
value instead instead of going through the dance of utilizing a LbyR. 
If the end point does not dereference the LbyR but instead some other
entity then there is the question of how this reference then actually
gets there. 

Ciao
Hannes