|
The connection between the watcher (the real end user) to
the PS of the presentity can be comprised of two logical
hops.
The access hop (between the watcher and its PS) is using
SUBSCRIBE as is but the PS actually terminates the SUBSCRIBE
dialog.
The interdomain hop (between the PSs) needs to use
extended/modified SIMPLE mechanisms (pending requirements
validity).
I hope
it clarifies one of the possible "architectures" to achieve the
optimizations.
Orit.
From: Alex Audu [mailto:alex.audu@alcatel.com] Sent: Thursday, February 26, 2004 8:22 AM To: Orit Levin Cc: Robert Sparks; IETF SIMPLE WG; Avshalom Houri Subject: Re: [Simple] Inter-domain Requirements for SIMPLE This seems like a mechanism to optimize information trasnfer between a watcher's PS and the remote PS containing the presesnce information of interest. But the requirement is specifying the possibility of "requesting" access to information "without subscribing " to it. The common understanding in SIMPLE is you can't request for presence information without subscribing to the presentity. If you request for it, you must have subscribed to it. But you don't want to be transfering information to an endpoint that didn't subscribe to that information just because you want to optimize bandwidth utilization or something like that. Or am I missing something? Regards, Alex. Orit Levin wrote: This requirement attempts to say: A potential watcher asks a presentity to grant access to presentity's presence information, but the watcher (or more precisely, its PS) is not interested in the presence information being pushed on individual basis from this point on. In other words, it can be implemented by just adding the watcher to presentity's ACL. Or in case of groups, adding the requesting watcher to a specific group. Orit. -----Original Message----- From: Robert Sparks [mailto:rsparks@dynamicsoft.com] Sent: Wednesday, February 25, 2004 8:48 AM To: Orit Levin Cc: IETF SIMPLE WG; Avshalom Houri Subject: Re: [Simple] Inter-domain Requirements for SIMPLE Orit, Avshalom - Thanks for submitting this draft. You've captured some important thoughts around deploying presence systems that deserve attention. I particularly look forward to the discussions we'll have around section 7.3 (Grouping of Watchers for SHARING of Presence Info). One quick question: I don't understand what you're looking for with the first requirement of 7.1: o Presence access: It MUST be possible to request continuous access to the status of a remote presentity without "subscribing" to it. Could you explain this a little more? RjS On Mon, 2004-02-23 at 17:33, Orit Levin wrote: |