[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Question for all candidates: Release process

On Mon, Mar 15, 2010 at 4:09 AM, Lucas Nussbaum
<lucas@lucas-nussbaum.net> wrote:

> During the last debconf, the freeze of squeeze was first announced to
> take place in December, then this decision was cancelled, and now we are
> in March.
> - How do you analyze what happened during last summer? What went wrong?

What went wrong: a decision was taken without consulting the affected teams.

This is a situation that has happened over and over again in Debian.
The Debian community does not take it well when they are informed of a
decision already taken without having a say on it. We should learn
from this to never make this mistake again: always keep involved
participants in the loop and consult with other people before
announcing something as a taken decision.

What went right: after the many messages stating that the December
freeze was NOT a good idea, specially after Mark Shuttleworth
clarified what he meant by "freeze" (stating which versions to include
in the release), the Release Team sent a mail saying that freezing in
December was not a good idea and that the freeze would most probably
happen on early 2010.

> - What is your opinion on the motivations for the proposal to freeze in
>  December? Specifically, in the future, should we try to coordinate our
>  release process with Ubuntu's?

I think it's generally a good idea, *BUT* I think we should freeze
about the time when Ubuntu releases the LTS, *NOT* 4 or 5 months
before.  Because, as we all know, Ubuntu doesn't really freeze until
really close to the release. So, freezing before that is the same as
not coordinating at all.  This would mean that Ubuntu would release
the LTS in April and Debian would release maybe in July or August. I'm
perfectly ok with that.

I also think that releasing roughly every two years, with a
kernel/X/etc upgrade in the middle (like was done with etchnhalf), is
good. I value stability far more than "bleeding edge", and I know that
releasing more often than that is going to have quite an impact in
stability.  Where I work, most users (about 150) are still using Etch,
because they are used to it and have no real need of anything newer.

All in all, this is my opinion as just a DD.  The release process is
the work of the Release Team, and it's their call in the end.
However, to avoid the turmoil caused last year after the DebConf
keynote / press release, the RT should always keep involved teams in
the loop when setting a release (or freeze) date.

> - So, we are now in March. What is your opinion with the release process
>  so far? When do you see the release happening?

I'm very saddened by Luk's resignation from yesterday, I think it's
mainly the result of having to do too much work by too few people.
This leads to stress and frustration, and those tend to lead to
resignations.  Now we have to give the team some time to re-arrange
themselves and find a way to continue.

I won't make any predictions, since they'd make very little sense in
the current context.  I do plan to help the Release Team releasing
squeeze, in whatever my capacity, and I'm hoping that it'll happen
this year, so that the whole LTS syncing thing can happen.  However,
if in order to have a stable, robust and clean release we ended up
releasing in 2011, I'd still consider it a good trade off, and I'd
still back up the Release Team.


Reply to: