[Council] proto-JEP: Message Archiving

Peter Saint-Andre stpeter at jabber.org
Mon May 10 16:32:00 CDT 2004


On Fri, Apr 30, 2004 at 11:50:35AM -0600, Peter Millard wrote:
> Thomas Muldowney wrote:
> > The last piece I would like to suggest is that we have
> > a partial guideline to what receives a number.  So far the simple
> > suggestion seems to be, at least some backing from the community and it
> > has been technically thought out.  What this causes is young or single
> > idea JEPs to be pushed back, and it also causes messy JEPs to be pushed
> > back.  Both of those are what are generally the dead JEPs right now.
> > So it seems like that would work well.  Others?
> 
> I also agree w/ the comments previously expressed by temas & LW.. The process of
> mandating some "backing" and technical thought seems to be fine requirements for
> getting a JEP number. I agree, that these JEP's that come out of (seemingly)
> no-where often are suspect and need some time to be fleshed out. I also agree
> that this is what the wiki is for.
> 
> I'm wondering if these vague requirements are sufficient, or is the community
> going to need/want something more quantitative?

Mine is seemingly a minority opinion, but personally I don't see why 
"dead-end JEPs" are such a frightening thing. The open-source way is
that lots of people start writing code or developing protocols and they
never go anywhere. How many "dead-end projects" are there on SourceForge
or JabberStudio? Does it matter? Out of that "chaos" there always emerge
some projects or protocols that have strong community interest. Sure,
it's a Darwinian process, but that's what leads to a healthy ecosystem
in the long run. Personally I did not think that we needed to add this
layer to the process. Adding more processes (voting by the membership
or somesuch) on top of processes that are not clearly working strikes me
as a Ptolemaic approach to the organization (more epicycles, anyone) and
is not the answer, I think. In any case, the intent was to improve 
communication between the Council and JEP authors, which it seems has 
not happened as a result of the previous reforms. Personally I am in
favor of letting the JEP Editor publish JEPs as they come in, have the
Standards-JIG list rip things to shreds, and have the Council (1) provide
more direction on that list and (2) push quality protocols through the
process faster. It may not seem pretty or controlled, but I think it
worked just fine in the past.

Peter

-- 
Peter Saint-Andre
Jabber Software Foundation
http://www.jabber.org/people/stpeter.php




More information about the Council mailing list