Re: getting on a longer release cycled

["Elijah Newren" <newren gmail com>]

On 9/7/06, Pat Suwalski <pat suwalski net> wrote:
> David Trowbridge wrote:
> > What in particular isn't possible with the 6-month cycle?
>
> Nothing's impossible, but a longer cycle every so often would encourage
> larger and better thought-out changes. I always get the feeling that
> GNOME contributors hold back on a lot of excellent ideas because they
> feel it will take longer to implement properly and test than the time
> they have before the next release.

This argument has never made any sense to me; I don't see why there's
any truth in it.  I'm not saying there isn't any, just that it does
not match my experience[1] and my feelings have been quite the
opposite on the matter.

I've seen various arguments for longer release cycles, and perhaps we
may want to move towards them at some point.  The argument I
personally found most convincing for a change was from Andrew Cowie.
Not sure if I'll paraphrase him correctly, but basically his argument
was that:

 We should have a stable version that is still actively supported by the time
 users have adopted it, and most ("mainstream") users won't adopt
 new versions until they've been stable for more than 6 months.  Because of
 our 6 month release cycle and lack of ability/desire to maintain more than
 two releases (our current handling with 2.odd.x and 2.even.x), we effectively
 don't support and miss out on a lot of the feedback of many "mainstream"
 users.

I used to be firmly in favor of the 6-month cycle, but I found
Andrew's argument quite convincing and it has turned me into more of a
fence sitter for now.  It isn't yet clear to me that a change would be
a definite improvement, let alone enough of a benefit to merit the
change in the process, but that may well change.

Anyway, that's my $0.02; I hope it provides a couple useful viewpoints
for the discussion.

Elijah


[1] I've often worked on big changes that couldn't possibly make it in
by the next release, including during code freezes.  Sure, I can't
commit it to HEAD when I'm doing so, but I can keep working on it even
during hard code freeze (in branches, of course), planning it out for
some later release.  How many years has work on compositing gone on?
I don't think a 9-12 month (or even 18 or 24) would have helped it
actually get in before the "next" release, nor that the overall
quality of it when it finally does get in would be any higher.