[standards-jig] Administration Namespace/specs

Iain Shigeoka iainshigeoka at yahoo.com
Sat Mar 9 18:38:07 UTC 2002


On 3/8/02 11:54 AM, "Casey Crabb" <debug at nafai.dyndns.org> wrote:

>> I also think this is a great idea.  However, I don't know if the thread
>> belongs in standards-jig.  I am definitely NOT trying to be list mom here
>> but this is an implementation specific discussion about jabberd.
> 
> Yes and no. It may be worthwhile to have the administration features a
> part of the protocol so that there is a standard way to configure
> 'Jabber' servers, and these methods can be implemented in jabber
> clients. From that point of view it is a standards-jig topic in that we
> want the additions to the protocol to make sense.

Agreed.

> Now I may agree that at this point it might not be time yet to post such
> discussion here. What I really need now is feedback from jabber server
> operators about said list. Once I have something resembling protocol
> chunks would probably be a better time to air it here.

Sounds fair.  I think that administration tends to be very implementation
dependent though so I was thinking that it may be best to start the
conversation on jabberd specific lists first...

> Also, the discussion of whether or not to include adminstration of
> jabber servers in the protocol at all is a discussion for standards-jig.

Definitely.  One question though is, are all jabberd operators reading this
list?

> I personally would like administration of the server to be
> server-implementation independant so that the server-implementation
> becomes a swapable component.

I tend to agree, although using Jabber to do the administration is a
questionable approach IMO.  I can understand if you're a do-it-yourself
person and you like administrating servers through whatever is provided for
each server.  However most big shops tend to require standard administration
interfaces like SNMP, administer directory resources through directory
services, etc...  Unless we're talking about those types of interfaces, I'm
wondering about the applicability beyond jabberd.  Due to this, I would tend
to favor having jabberd implement something and allow it to become a de
facto standard if it ever catches on.

Thoughts?

-iain


_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com




More information about the Standards mailing list