[standards-jig] software version

Matt Tucker matt at jivesoftware.com
Sun Jul 27 20:54:53 UTC 2003


Just wanted to add that I'm in agreement with your position. If you add 
elements then it's a "new version" of the protocol and in general it's 
more useful to explicitly handle the new protocol versions through 
namespace changes.

The informational JEP's have always confused me completely. I don't 
understand why we even have them. It is indeed good to document old 
protocol usage, but if it's never going to become "standard", then 
what's the point of calling it a JEP (it could be documented somewhere 
else)? :) So yeah, a standards-track version JEP would be cool. If it's 
standards-track, that probably means we want to bite the bullet and 
switch over to:

  1) Using the new style namespaces that are URL's.
  2) Not use an IQ "query" sub-packet but something more useful like:

      <version xmlns="http://www.jabber.org/version"/>

I have a feeling that it's always going to be tricky moving the 
informational JEP's into standards-track ones due to the new conventions 
that are in place today, but it seems worth it.


Evan Prodromou wrote:

>>>>>>"PS" == Peter Saint-Andre <stpeter at jabber.org> writes:
>     PS> So I'd like to propose that we add a <creator/> element to
>     PS> jabber:iq:version, which existing implementations could ignore
>     PS> if they don't understand it.
>     PS> Objections?
> Dang. Annoyingly, I have another objection.
> JEP-0092 is an informational JEP, which "defines an existing protocol
> in use within the Jabber community". The existing protocol doesn't
> have this field or others. Therefore, an informational JEP probably
> shouldn't include them.
> Would it be at all useful for me to write up a new standards-track JEP
> that incorporates the new fields?
> ~ESP

More information about the Standards mailing list