Re: Two theses regarding packages
On Sun, Jun 10, 2001 at 01:51:35AM +0200, Adrian Bunk wrote:
> - We have > 700 develers that maintain 4727 source packages with
> 7275 binary packages that have currently 383 RC bugs.
> - Debian has a good reputation for making a high quality distribution.
> - We should sometimes manage to release a new stable Debian.
>
> We are very maintainer orientated - a maintainer can e.g. decide alone
> whether he wants to apply a patch or fix a bug.
>
> We must find ways to ensure the quality of our packages.
We only release packages we must release or whic have no release critical
bugs. So the quality of our release is maintained, and we can release at
"any time", read: a random broken game doesn't hold up the release.
This is a good approach: It enables those who are interested in releases
make a release without waiting for maintainers who don't care much about
releases and don't fix important bugs timely. Stepping out of each others
way is a good approach in an area where > 700 people run around.
OTOH, if you care about a specific program, usually there is no problem in
helping out. Yes, sometimes there are disagreements about NMUs without
approval etc, but they are quite rare. With so many people doing so many
different things with so little communication, we do well. This is because
communication and synchronization is not a requirement: If we were to
introduce more communication, more timing requirements, more schedules and
more rules, things would get worse.
> I don't want to start to punish people and it's clear that most of us do
> our Debian work in our spare time but every member of Debian should be
> aware that he should try to do good work - and we need ways to deal with
> maintainers who don't fulfill this
We have ways to deal with lack of activity, and they work well. You do not
give any arguments to convince me otherwise. BTW, I find remarks like
"every member of Debian should be aware that he should try to do good work"
rather inappropriate.
> "rm -rf /*" to the postinst of one of my packages to see whether this is
> true) - again: My aim isn't to exclude hundreds of people and there aren't
> specific people I want to be excluded but we need to find ways to ensure
> the quality of our packages.
You *can* *not* ensure the quality other than by reaching a helping hand.
A package gets not buggy because someone stands in the way and forbids you
to fix it. A package usually gets buggy because there is nobody
interested and capable of fixing it, and if this is the case, then pointing a
finger and demanding better and more contribution doesn't help you an inch.
A fundamental error is to assume that just because there is a maintainer
registered for a package, there will be someone who fixes all problems with
it. If there is a particular problem that you'd like to have fixed, and the
maintainer doesn't care, you can start the NMU process. If a conflict
arises, the maintainer is assumed to care, and he will have to defend his
position, just as you have to defend your (planned) NMU.
Thanks,
Marcus
--
`Rhubarb is no Egyptian god.' Debian http://www.debian.org brinkmd@debian.org
Marcus Brinkmann GNU http://www.gnu.org marcus@gnu.org
Marcus.Brinkmann@ruhr-uni-bochum.de
http://www.marcus-brinkmann.de
Reply to: