[standards-jig] thoughts on a new idle protocol

Nathan Walp faceprint at faceprint.com
Sat Jan 17 20:21:46 UTC 2004


I'm thinking of authoring a JEP for a new idle protocol, something along
the lines of:

<presence>
	...
	<idle xmlns='http://jabber.org/protocol/idle'
stamp='2004-01-17T15:06:35-04:00' />
	or
	<idle xmlns='http://jabber.org/protocol/idle' seconds='600' />
	...
</presence>

above, stamp is a timestamp according to the JEP-0082, and seconds is
self-explanatory.  The seconds option would require the server to add
jabber:x:delay tags in order to provide an accurate measure, otherwise
you wouldn't be able to compute the idle time for someone already idle
when you signed on and received their initial presence from the server. 
For the stamp option, there is the potential for skew between the clocks
of the clients.  The seconds option has this same problem, but only when
receiving presence with a jabber:x:delay tag.

jabber:iq:last already provides a mechanism for querying a user's idle
time.  However, that's a pull method, and this is a push method.  For
clients (like Gaim) that display idle time in the buddy list (either
directly, or indirectly, such as by dimming the entry), polling all the
users in the buddy list isn't an acceptable solution.

I don't see this as a huge traffic generator, as many clients already
change their status when they go idle, and this would just piggy-back on
that presence packet.  

Any thoughts?

Nathan, holding out his asbestos hat waiting for someone's $0.02.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://mail.jabber.org/pipermail/standards/attachments/20040117/411fecde/attachment.sig>


More information about the Standards mailing list