[standards-jig] jabber:iq:stats

Iain Shigeoka iain.shigeoka at messaginglogic.com
Mon Aug 5 16:58:57 UTC 2002


On 8/2/02 1:07 PM, "Russell Davis" <rkdavis at burninghorse.com> wrote:

> there seem to be two directions that iq:stats can take.
> 
> 1. A "pure" jabber way with translation to industry standards done by a
> transport
> 
> 2. use industry standards from the very beginning and no translation is
> required.
> 
> I see pros and cons with both ideas and would like to get together with
> interested parties at a mutually convienient time to hash out a
> consensious. So if anyone who is interested in the direction iq:stats
> takes please send me an email with the best time for you to participate
> in an online discussion in the next few days.

Live meetings may be difficult (as discussed Re: foundation debates) due to
time zones and such.  I also doubt the efficiency of them for technical
discussions.  Why not discuss this on the list?  That way more people can
see it, the discussion will be archived, and we can start discussion now.
:)

IMO, the direction for iq:stats will be best chosen by deciding who the
target audience is.  If we're looking for smaller shops, Jabber only
installations, etc, I'd say #1 is the ticket.  If we're looking at
enterprise users and larger installations #2 is the way to go.

IMO, the #1 option could just as well be served by creating a standard
server log format similar to that for web server logs.  I would imagine most
smaller shops would be looking at post mortem analysis of server
behavior/performance/etc.  That leaves a dynamic protocol for the enterprise
users with #2 being most appropriate.  I could definitely be wrong though.
:)

-iain




More information about the Standards mailing list