Re: BoF item 7/14: dead modules

[Andre Klapper <ak-47 gmx net>]

On Fri, 2012-08-03 at 16:46 +0300, Rūdolfs Mazurs wrote:
> I don't think it would be good idea to base module death only on past
> commits. Take galeon — it is rather actively translated, but Wikipedia
> states it has been discontinued 3 years ago [1]. Even if we only
> consider non-translation commits, there might be only bug fixes, that no
> one wants to release.

25 commits in total in Galeon for the last two years:
  git log --after=2010-08-03 --pretty=oneline | wc -l
25 translation commits in Galeon in the last two years:
  git log --after=2010-08-03 --pretty=oneline -- po/ | wc -l
=> 25 - 25 = 0. No single code commit.

> So we should ask maintainer, if [s]he believes, if the software
> is to see next release, when there are suspicions of dying software.

Yes, but this will require a lot of work, plus a careful wording, as
“there's no point at which he [a maintainer] consciously realizes that
he can no longer fulfill the duties of the role” (Karl Fogel: Producing
Open Source Software, page 216).

> On the other hand, there might be other software, that is useful and
> being used, but not actively developed, which should see translation
> update releases, even if there is no maintainer.

Indeed a module can be small, well-defined, considered feature-complete,
and very mature, so code changes are not needed. 
It does not necessarily have to be obsolete/deprecated/legacy just
because there is no activity.

andre
-- 
mailto:ak-47 gmx net | failed
http://blogs.gnome.org/aklapper