[Standards-JIG] httpbind example issue

Ian Paterson ian.paterson at clientside.co.uk
Thu Feb 2 10:55:25 UTC 2006


> > Perhaps this could have been slightly more elegant, but does that
> > matter enough to change it at this stage?
> 
> Isn't draft status mean that it's subject to change?

Yes, "some changes are possible". But since "implementations are
encouraged", IMHO, once a few implementations have been developed and
put into production, it is both the author's and the council's
responsability to ensure only important changes (or changes that won't
affect existing installations) are introduced.

> If the goal is that "it is done", perhaps we need to put this to the
> council with the motion for last call, and move it along to Standard?
>
> That said, we have made similar changes in the recent past,
> which the council approved.  In my opinion, this change is on 
> the same level as those, so I see no reason not to change it. 
>  However, it sounds as if more people are implementing this, 
> so I also think it's perhaps time to push it forward to the 
> next stage.

Yes quite a few implementations have been springing up.

However, my feeling is that this JEP is still a long way from becoming
'Final'. The nice thing about the Draft stage is that it allows us to
make changes that are deemed important - even if they affect existing
installations. I hope only a few changes will be necessary, but I'm sure
some will be - e.g., today's post from Peter Millard about limiting the
size of RIDs.

There are only five Final JEPs (arguably a good thing for now at least).
JEP-0124 is not XMPP, and it is far more complex to implement (as
recommended) than other JEPs. It documents a truly innovative approach
to HTTP tunneling, so implementors are breaking totally new
technological ground. The JEP can be expected to take longer than
average to become Final.

AFAIK, we have no successful interoperability experiences between
applications created by different developers. However, I expect we will
gain that experience during 2006.

- Ian




More information about the Standards mailing list