wg-pic - Re: [wg-pic] PA vs. P2P
Subject: Presence and IntComm WG
List archive
- From: Candace Holman <>
- To:
- Subject: Re: [wg-pic] PA vs. P2P
- Date: Tue, 17 May 2005 17:09:04 -0400
For now let's assume that the PA and pidf-diff (partial pa update) runs on a central server, although clearly there are advantages to running these operations on the client.
Candace
At 03:59 PM 5/17/2005, you wrote:
In fact, there can be a mixed model. Instead of having a completely
centralized PA for all users, every user can have his/her own PA to handle
presence composition for his/her multiple PUAs, but the PAs can be in a
P2P network.
BTW, to correct my answer on partial event notification. If we put the
burden of status composition on watchers, using pidf-diff does not require
a centralized model.
-Xiaotao
===========================================================
Name : Xiaotao Wu
Email :
Homepage : http://www.cs.columbia.edu/~xiaotaow
Phone : (212)939-7054, Fax: (212)666-0140
Phone-PC : (212)939-7133
SIP :
sip:
Office : Room 506, Computer Science building, West 120th
===========================================================
On Tue, 17 May 2005, Ben Teitelbaum wrote:
> Steve Blair
<>
writes:
>
> >>Candace,
> >>
> >>This is a great list. I might add:
> >>
> >>7. Off-line presence (i.e in the absence of any PUAs); for example, a
> >> PA might report when a user was last seen;
> >>
> > Really? This would seem to pull PIC back into the social context
> > conversation. I would think the key is user settable presence. For
> > example if a users wants such off-line presence they could enable it
> > but it would be off by default. Am I missing the big picture on this
> > one?
>
> When I've heard Henning talk about this, the model is still very much
> that the user sets the policy and composition rules and uploads them
> to a PA via XCAP. I'm not sure where the specification of policy
> language stands in SIMPLE.
>
> Note that RPID has "since" and "until" attributes, so some of the
> off-line presence that a PA might furnish could have been explicitly
> published by a user's presentity. For example, a local calendar app
> on a user's laptop might publish "on plane until 18:20 UTC" just
> before disconnecting. Authorized watchers should be able to retrieve
> this presence from the user's PA while he/she is off-line on the
> plane.
>
> -- ben
>
> --
> Ben Teitelbaum http://people.internet2.edu/~ben/
>
- PA vs. P2P, Candace Holman, 05/17/2005
- Re: [wg-pic] PA vs. P2P, Ben Teitelbaum, 05/17/2005
- Re: [wg-pic] PA vs. P2P, Steve Blair, 05/17/2005
- Re: [wg-pic] PA vs. P2P, Ben Teitelbaum, 05/17/2005
- Re: [wg-pic] PA vs. P2P, Xiaotao Wu, 05/17/2005
- Re: [wg-pic] PA vs. P2P, Candace Holman, 05/17/2005
- Re: [wg-pic] PA vs. P2P, Steve Blair, 05/17/2005
- Re: [wg-pic] PA vs. P2P, Xiaotao Wu, 05/17/2005
- Re: [wg-pic] PA vs. P2P, Ben Teitelbaum, 05/17/2005
- Re: [wg-pic] PA vs. P2P, Steve Blair, 05/17/2005
- Re: [wg-pic] PA vs. P2P, Xiaotao Wu, 05/17/2005
- Re: [wg-pic] PA vs. P2P, Candace Holman, 05/17/2005
- Re: [wg-pic] PA vs. P2P, Xiaotao Wu, 05/17/2005
- Re: [wg-pic] PA vs. P2P, Candace Holman, 05/17/2005
- Re: [wg-pic] PA vs. P2P, Ben Teitelbaum, 05/17/2005
Archive powered by MHonArc 2.6.16.