[Standards-JIG] JEP-126 (Invisibility)

Vinod Panicker vinod.p at gmail.com
Tue Apr 11 06:07:25 UTC 2006

On 4/11/06, Justin Karneges <justin-keyword-jabber.093179 at affinix.com> wrote:
> On Monday 10 April 2006 15:19, Peter Saint-Andre wrote:
> > But not all servers (and even fewer clients) implemented visible. So we
> > never really had consensus on how invisibility is supposed to work via
> > the presence stanza. That lack of consensus concerned me then, and it
> > concerns me now.
> One thing that always bothered me about the old way is how it would play with
> directed presence.  Even with privacy lists, I'm still unsure how to handle
> an "away-to" or "invisible-to" scenario.  Perhaps this has less to do with
> invisibility and more to do with how statuses play with directed presence.

I had a similar question [1] when implementing it on the server side.

> I've resisted adding a per-contact directed presence feature into Psi, because
> it could potentially be misleading.  What if I send a directed presence to
> someone indicating that I'm offline:
>   <presence type='unavailable' to='otherUser at example.com'/>
> Then sometime later, I change my global status:
>   <presence>
>     <show>away</show>
>   </presence>
> Does otherUser get my presence?  I guess I ask this question both about RFC
> 3921 and also about what servers in the wild actually do.

I'd say that otherUser should not get the presence, since a directed
presence has already been sent.  This directed presence should be
valid for the duration of the user's session.  The RFC says otherwise.

> If the answer is that he does get the presence, and privacy lists must be used
> in order to block out the later presence packets, how would it fit with the
> "away-to" scenario?
>   <presence to='otherUser at example.com'>
>     <show>away</show>
>   </presence>
>   <iq ... privacy list stuff here to block otherUser />
>   <presence/>
> OtherUser gets the away packet, and does not get the future online packet.
> Now, what happens if otherUser signs off and then back on?  I assume he will
> see me as offline, when the correct behavior would be for him to see me as
> away.
> If we want these features, then we need an explicit protocol, otherwise the
> user will have flaky experiences.

Right.  Privacy lists don't seem to fit here.  IMO, if the users
session is still active, the server should be returning the "last
directed presence" for otherUser in reply to a presence probe.


[1] - http://mail.jabber.org/pipermail/jdev/2006-February/023082.html

More information about the Standards mailing list