Subject: Expectations for the 0.9 cycle
Date: Fri Apr 13 2001 - 02:42:04 CDT
>>>>> "Hubert" == Hubert Figuiere <email@example.com> writes:
Hubert> Sam TH écrit:
>> Well, since the CVS tree will be closed to fancy new features
>> during the freeze, I expect that there will significant incentive
>> for other people to help as well. I think bug squashing will
>> definitely be the order of the day.
Hubert> I think we should make a release branch to put fixes in (an
Hubert> incorporate them into the trunk). Then the branch will lead to
Hubert> a new release, then closed when the final is here. The trunk
Hubert> will be the next release base.
Hubert> This is the only way I see to be able to fix and release
Hubert> something stable will not preventing hackers to put new
Hubert> feature into the next release. Trust me, preventing them to
Hubert> put new features in the tree is a good way to stop them.
I both agree and disagree. True, in Linux land there's a lot of
frustration when the tree closes because people want to add the latest
new cool things.
But if you step back and take a serious look at AbiWord development,
you'll see something that makes it very different from the Linux
kernel: there are no more than 10 or so active developers.
If there should be _any_ chance at all that we reach 1.0, we need to
focus 100% on feature closing and bug fixing. No fancy stuff in this
period of time, no new features from (new) developers to destabilize
Anyone feeling like adding features should do so in private and keep
their trap shut about it. Bless the mailing list with gifts of feature
surprises post 1.0 but for the love of your god, don't make the 0.9
more painful by ranting about this new cool feature you're adding
that'll make AbiWord brew coffee while it does your laundry.
People should probably expect the 0.9 cycle to be long and slow to get
through - adding new features is more sexy than fixing bugs, so assume
that people will be less enthusiastic about it and spend less time
than they might otherwise have. Anyone who has dreams of a painfree
and energetic 0.9 cycle should see their doctor and ask to get on some
This will be a time where time invested on AbiWord Weekly News (kernel
cousin!) will be important, where good bug tracking data sent to the
user list will be important, where really good triaging will be
important and where we _must_ get more people than usual to help us
both file and fix bugs. Otherwise I fear some of the current
developers may well get fed up with the project.
Do I sound negative? Probably do. Put it down to me being realistic :)
This archive was generated by hypermail 2b25 : Fri Apr 13 2001 - 02:43:17 CDT