- From: Graham Klyne <GK@ninebynine.org>
- Date: Mon, 22 Nov 2004 20:22:09 +0000
- To: Christopher Schmidt <crschmidt@crschmidt.net>, Danny Ayers <danny.ayers@gmail.com>
- Cc: www-rdf-interest@w3.org
At 08:48 22/11/04 -0500, Christopher Schmidt wrote:
>On Mon, Nov 22, 2004 at 01:32:04PM +0100, Danny Ayers wrote:
> >=20
> > Before long I think we're likely to see an even more rapidly changing
> > 'presence' data set ("currently listening to..."), maybe having a main
> > delivery protocol other than HTTP (i.e. Jabber).
>
>I'm not sure about alternate delivery protocol, but creating a rapidly=20
>changing presence data set is exactly what myself and others are working=20
>on with the MeNow project: http://crschmidt.net/semweb/menow/ , soon to=20
>move to http://www.menow.org/ .
RDF-encoded presence information sounds interesting.
FWIW, I think Jabber would be a good choice for information distributed
using a publish/subscribe type model. (HTTP tends to require polling,
though I suppose if one had an HTTP server at the client one could PUT or
POST to it. But here be messy problems with firewalls, NATs, etc.)
Pure asynchronous event pushing (without the active subscription) would be
more challenging, though one could consider using SMTP. I think there are
some efforts related to event notification bubbling below the surface in
the IETF community. But it's a tricky protocol to get right an Internet scale.
#g
------------
Graham Klyne
For email:
http://www.ninebynine.org/#Contact
Received on Monday, 22 November 2004 20:41:42 UTC