Versioned Namespaces (was Re: [standards-jig] Avatars)

Ralph Meijer jabber.org at ralphm.ik.nu
Wed May 7 21:04:04 UTC 2003


On Wed, May 07, 2003 at 01:31:53PM -0600, Peter Millard wrote:
> Justin Karneges wrote:
> [Much snippage]
> > I don't think JEP authors should have to deal with their own baggage.
> > Seriously, if your JEP hasn't been approved yet, don't be afraid to throw
> > your entire spec out the window.  Tough luck for those that have implemented
> > an earlier revision.
> 
> /me cheers.
> This is exactly the point. If you don't like changing protocols, then don't
> implement experimental protocols. I have Endeavored to do this with Exodus (and
> I have mostly succeeded).
> 
> +1 on everything Justin said here. Kudos :)

I agree too. I am against versioning in namespaces before final. Let me make
it clear again what my original suggestion was: to have experimental
implementations of experimental protocols use another (possibly looking
the same but with the addition of the product name, or whatever, as long
as it doesn't start with jabber: or http://jabber.org) namespace to not have
it interfere with the final version later on.

Of course this isn't at all necessary if people would keep the experimental
stuff out of the 'release quality' branch of their products. I just don't see
why experimental stuff should become available for lots of users. In my view
early implementations are good for proof of concept, finding faults in the
specification and generally make discussions more than theoretical. I repeat my
suggestion to make JEPs visably marked as experimental, be it a watermark image
or a big fat red banner at the top screaming NOT FOR WIDESPREAD USE.

Also, let's keep it clear that I'm not trying to bash people. What's done
is done. Let's not do it again.

Ralphm



More information about the Standards mailing list