Re: Bugfix release

["Elijah Newren" <newren gmail com>]

On 3/22/07, Alex Jones <alex weej com> wrote:
> I remember a while ago some people were talking about how it might be a
> good idea to dedicate a release cycle to fixing bugs, with a feature
> freeze in effect.
>
> Is there anyone still in support of this idea, or has it been generally
> discarded?

I don't think the idea would work well; in fact, we've rejected
similar ideas that were far less imposing in the past[1].  The reasons
are perhaps easiest to understand if we changed the wording of your
suggestion just slightly:  Let's say we declared that only bugfixing
was allowed for a cycle.  This declaration would mean that no new
features could be developed, no translations could be added, no more
user documentation could be written, and work on the webpages would
need to stop.  You might get a few more bug fixers due to manually
altering priorities of those that have multiple interests and
abilities but you'll have far more people who just quit working for a
cycle.  Overall, you'd get much less work going into GNOME.

In the past, this reasoning was used to reject even making an official
pre-declared _focus_ for a cycle, let alone restricting work to _only_
be done in a given focus area.  (See e.g.
http://mail.gnome.org/archives/desktop-devel-list/2004-March/msg00415.html)
That's not to say that focusing on a particular activity is bad, just
that getting people to focus on a pet topic needs to be done in a
different way[2].

Cheers,
Elijah

[1] http://mail.gnome.org/archives/desktop-devel-list/2004-March/msg00415.html

[2] I think the better way to get increased efforts on a given topic
is to start working on them, convincing others to join you, and
letting others know of results you are getting.  At one end, this can
be like the performance work which seemed to be galvanized into higher
gear by Federico's blogs on the topic (specifically starting with the
file chooser).  At the other, it can be as simple as gathering a
couple people on IRC and working on fixing particular bugs (e.g.
Kjartan has started showstopper-fixing hackfests on IRC just by
grabbing people he saw in channels he was in and talking them into
working on some bugs with him; I remember one day where a few of us
(including Kjartan) fixed a small handful of high profile bugs at his
prompting).
Another alternative for increasing focus on a particular topic is to
have individual maintainers choose to follow them for their project.
I have basically imposed feature freezes on myself for an entire
development cycle for certain modules that I maintained, simply
because I felt it was the right focus for my project at the time.