New suites vs. common release cycle [Was: Productivity suite]

[Jeff Waugh <jdub perkypants org>]

<quote who="Christian Rose">

> > But this is just a long way of saying, "Welcome aboard, we hope you
> > enjoy your flight. Please keep your seatbelt fastened at all times, as
> > we will be experiencing high turbulence during this flight. Vodka and
> > orange, sir?"
> 
> I for one would think it would be a sad day if we've given up the goal of
> having everything in the same CVS, so that cross-module efforts (such as
> translation, documentation, and so on) can have a decent chance to work on
> all the modules that are supposed to be in a release.
> 
> Yes, I know that Fifth-Toe didn't have that requirement either (and with
> the expected results thereof in some cases). However, I think that if we
> want to make the process better by coordinating some aggregate releases
> with the core release, and market those aggregate releases as extra added
> value that are reasonably coherent and integrated with the rest of the
> desktop (which is the whole point of having the aggregate releases in the
> first place, I think) which is released at the same time, then we better
> also add some reasonable requirements for the modules in those aggregated
> releases.

So we're in a confusing spot here. I'll try and summarise where we're at, so
we can figure out where to go next:

 * We currently have three release suites, with differing requirements, for
   three separate, fairly well-defined purposes. They have served us well in
   terms of understanding and expressing GNOME as a coherent product, and as
   a method of simplifying the release management process.
   
 * Unfortunately, our three release suites are viciously exclusive, socially
   and technically. If your software doesn't fit for whatever reason, you're
   out in the cold, and there is a *lot* of fantastic software for GNOME out
   there (note: for GNOME, not necessarily using GNOME). We need to include
   all of those hackers and contributors, and let them "be GNOME" too!

 * One solution we've considered (and tested!) is the addition of new suites
   to the official GNOME release. The bindings suite was a new addition, and
   it has worked pretty well. Only three other suites have ever come under
   serious consideration: GNOME Office (proposed by Martin as productivity),
   GPE (which has never really got off the ground) and multimedia. All would
   have similar requirements to our current suites, being in GNOME CVS, FTP,
   bugzilla, etc.

 * A more extreme solution, desireable because it's maximally inclusive, is
   allowing *any* project to join the GNOME release process, and doing a lot
   of work to figure out how we can make that happen sanely. I think it may
   just come down to tracking regular releases, encouraging the projects to
   stick with freezes, and getting really good status reports at the end so
   we can have awesome release notes covering everything we've done. This is
   more inclusive because it doesn't *force* projects to host with GNOME. We
   can encourage it, but we shouldn't force J. Random Hacker who's happy
   with sourceforge to shift, just so he or she can be "in GNOME". Sure, it
   introduces all kinds of questions we need to find answers for, but if it
   didn't, it wouldn't be fun, right? ;-)

So, in a way I'm tempted to suggest that we *not* kick off the productivity
suite, and we use abiword as a test case for EXTREME CO-OPERATIVE RELEASE
PROCESS ACTION! But, y'know, that could just be me. ;-)

- Jeff

-- 
OSCON 2005: August 1st-5th         http://conferences.oreillynet.com/os2005/
 
   "Consensus is whatever the developers remember or agree with." - Paul
                            Vixie, Open Sources
_______________________________________________
gnome-hackers mailing list
gnome-hackers gnome org
http://mail.gnome.org/mailman/listinfo/gnome-hackers