Re: [sipcore] composition or just indirection

["James M. Polk" <[e-mail]>]

At 06:50 PM 7/20/2010, Thomson, Martin wrote:
> > From: Paul Kyzivat [mailto:[e-mail]]:
> > Thomson, Martin wrote:
> > > The Geolocation header identifies a PIDF document by value or
> > > reference.  A Geolocation header indicates that a) this particular
> > > PIDF
> > > document contains location information and b) that this location
> > > information in particular may be relevant to the processing of the
> > > SIP request.
> >
> > IMO this is *too* vague.
> >
> > For instance, I might want request that a call be routed *to* the phone
> > closest to some geolocation. That location is also relevant to the
> > processing of the request, but in an entirely different way.
>
>Damn.  You would have to point out my big blind 
>spot.  I've been concentrating on the 
>routing-based-on-caller-location use case too long.
>
>If we also consider that the location needs to 
>be attributed to a particular entity, you can 
>either add syntax that attempts to identify the 
>entity that you are talking about (a "from" or 
>"to" parameter might do this), or you can make 
>the entity constant.  In the interest of 
>simplicity, the latter seems more fashionable:
>
>"
>The Geolocation header identifies a source of 
>location information.  This might be a PIDF-LO 
>document that is identified by value or 
>reference, or a geo: URI that includes location information directly.
>
>A Geolocation header indicates that a) that this 
>location information is somehow applicable to 
>the entity that initiated the request and b) 
>that this location information might be relevant 
>to the processing of the SIP request.
>"

I'm reading this believing someone here believes 
part of -03 needs to replace text with the above, 
but I'm not sure which text needs replacing from 
this discussion. Can someone point to the paragraph in question?

James


>This is almost the "bold" interpretation that 
>says that this Target’s location should be 
>considered the source of the request.  It does 
>not imply that the Target and source are the 
>same identity.  Nor does it expressly state that 
>the source is actually located anywhere in 
>particular, though we're walking a fairly fine line there.
>
>--Martin