maintenance coverage

Oswald Buddenhagen ossi at
Sun May 20 12:22:52 BST 2007


the basic idea is to gather some information about how well particular
parts of kde are maintained. the primary goal would be to give hints
where help - or in the worst case a new maintainer - is needed.

the primary tool would be bugzilla, obviously.
based on how long reports for a component stay in particular states,
"activity percentiles" for "maintenance" (bugs) and "development"
(wishes) could be calculated. additionally, if many reports are dealt
with by somebody else than the registered maintainer, a "de-facto
maintainer hint" could be issued (yes, i'm aware that mailing lists and
dedicated bug triagers have to be taken into account).

in a next step, svn directories could be mapped to bugzilla components.
potential benefits would be more accurate statistics - taking into
account the actual size of the component and non-bugzilla activity.
additionally, the MAINTAINERS files would be obsoleted by this more
formal structure which hopefully would be easier to keep up to date with
the help of automated hints.

now, it's clear to me that many people will choke on the idea of some
data mining apparatus watching their productivity. however, while much
of FOSS is about egos, one should not forget the benefit for the
community. and it's not about taking away power from people, it's about
providing them with hard data to help making responsible decisions.
and if it works out, it will gain us some additional developers.

of course, all this talk is worth nothing if nobody feels like
implementing it - i most probably won't. just throwing in ideas.
now give it a roasting. :)

Hi! I'm a .signature virus! Copy me into your ~/.signature, please!
Chaos, panic, and disorder - my work here is done.

More information about the kde-core-devel mailing list