[Standards-JIG] 'Reliable' XMPP

Jean-Louis Seguineau/EXC/ENG jean-louis.seguineau at antepo.com
Fri Jul 30 18:44:50 UTC 2004



I understand the concern here. That said, we have designed a few approach in
the past that brought different 'atomic' pieces together under the umbrella
of a 'framework'. What I was trying to say, is that we would probably gained
defining what 'reliable XMPP' mean and how one can design binding profile to
express certain part of it. 
The ideal case would be to be able to re-incorporate the existing JEP into
this framework. So we need an informational JEP and an application JEP to
set the stage.

Comments ?

Jean-Louis


-----Original Message-----
Message: 3
Date: Fri, 30 Jul 2004 09:46:30 -0600
From: Peter Saint-Andre <stpeter at jabber.org>
Subject: [Standards-JIG] Re: Standards-JIG Digest, Vol 5, Issue 29
To: standards-jig at jabber.org
Message-ID: <stpeter-D8958C.09463030072004 at sea.gmane.org>

In article <20040630100536.75C7464589 at hades.jabber.org>,
 "Jean-Louis Seguineau/EXC/ENG" <jean-louis.seguineau at antepo.com> 
 wrote:

> Peter,
> 
> IMHO I believe the issue here is to make XMPP a 'reliable' protocol. We
are
> facing entrenched players (IBM, TIBCO, etc...) that are playing the '...
> XMPP is not and industrial grade reliable protocol compared to (insert
some
> well known brand here)' game.
> 
> There have been separate attempts at bringing some kind of control to the
> way stanzas can be send to multiple recipients (JEP33) can be acted upon
> (JEP79) and how the can be augmented with metadata (JEP131). All this
effort
> is going in the right direction, but seen from the outside, does not
present
> a united front against the incumbents. 
> 
> I have the feeling that XMPP would gain a lot if these were regrouped
under
> a larger umbrella geared to make XMPP as 'reliable' as the market
> incumbents. I also believe while reading the thread about Pub/Sub that
this
> is also a must have.

Sorry for the delayed reply -- I'm just catching up on list traffic.

Perhaps it makes sense to define a "reliable XMPP profile" (via an 
Informational JEP) that shows how all these pieces fit together. While 
some people seem to want everything to be in one big document, I don't 
think that's necessary as long as we clearly spell out exactly how one 
would go about using the "atomic" pieces to roll out reliable XMPP, 
extended presence, or other such "molecular" system behaviors.

/psa





More information about the Standards mailing list