Planning releases

[Havoc Pennington <hp redhat com>]

Joe Shaw <joe@helixcode.com> writes: 
> I think the reasoning behind it is to dictate a release schedule for the
> major releases of GNOME. What is included in GNOME? That's what the board
> will decide. For example, what is GNOME 1.4? Is it gnome-libs
> 1.4? gnome-libs and gnome-core? All of the gnome-* packages? All of the
> gnome-* packages, Evolution and Nautilus? All of the above plus gnumeric,
> X-chat, gnutella, gdm, et cetera?
>

I don't know if the board would decide release plans. This is almost
certainly _the_ big decision that GNOME ever has to make; historically
it's a big gnome-hackers flamewar, and it's a point of much contention
in the steering committee too.

This is also an area where companies are strongly biased.

My impression is that in many free software projects, a release is
simply never made, or is made at ridiculously long intervals. cf. Gimp
or Debian or even the Linux kernel. I wouldn't mind having some way to
avoid spending 2 years between releases, and that means someone or
some group with the power to say "sorry, your package or feature ain't
in the release". Technical requirements are not going to delay a
release 2 years; GNOME isn't that complex. But politics might.


Anyway, releases are the real hot potato. I think working code can
always decide a technical issue; projects can always ignore each other
even. And no one cares that much if we buy a new drive for the CVS
machine. But for releases, everyone _must_ work toward the same goal,
and everyone has quite a strong opinion about what the release will
contain and when they'd like it to happen.

So, we should have a clear statement in our bylaws or whatever about
how we make release decisions. (And it would be nice if the
decision-making process was a good one.)


I imagine the board would be much less controversial if it only did
the things I mentioned in my previous mail (contact with companies,
advocacy, money, etc.) and ignored releases. Maybe we should consider
that. I'm not sure.

Havoc