On Fri, Mar 05, 2004 at 02:32:45PM +0000, Martin Michlmayr wrote:
> * Anthony Towns <aj@azure.humbug.org.au> [2004-03-05 15:25]:
> > > I disagree with this. I think that maintainers who neglect their
> > > duties and don't follow documented procedures (orphan their
> > > packages, inform the keyring maintainer that they are leaving the
> > > project [1]) should not be treated the same as maintainers who
> > > leave the project properly.
> > Then how should they be treated, exactly?
> They should be treated like people who don't follow their duties,
> which is what they did. In practice, this means that someone who left
> Debian properly by resigning can easily come back by mailing the
> keyring maintainer. Those who did not retire properly, on the other
> hand, will have to go through New Maintainer in order to ensure they
> understand their duties and procedures in Debian.
So, for example, I should be put through n-m again immediately because I
haven't been doing regular maintenance of cruft or ifupdown? Or if those
packages haven't had severe enough bugs for you, perhaps Branden or the
entire Progeny staff should be put through n-m again for abandoning pgi
which has had an RC bug in unstable for well over a year now?
If not, what makes my or their other contributions to the project valuable
enough to override that policy, that aren't equally well provided by,
say, working on other free software projects, caring for sick family
members, serving your country with compulsory military service, making
millions by exploiting peasants in third world countries, or veging out
in front of the TV?
The constitution states as its *first* general rule that "Nothing in
this constitution imposes an obligation on anyone to do any work for
the Project." Debian's policy of treating everyone as a volunteer,
and thanking them for what they can do, and not criticising, rebuking,
or sanctioning them for what they can't do, or simply chose not to do
is valuable and admirable, IMO, and shouldn't be thrown out so casually.
As a for instance, IMO one of the problems with expecting people to
maintain their packages consistently and well, is that every time there's
an NMU then that can only be interpreted as a direct statement that
they're not meeting Debian's expectations, and hence that they're a below
average maintainer, or worse simply irresponsible. But there're plenty of
reasons why people can't maintain their packages at the level we'd like,
even if they are competent, and even if they are completely responsible.
Life can intervene, your priorities can change, or whatever. There's no
reason to imagine maintainer's in that position aren't up to scratch,
and the consequence that NMUs have to be looked at as an attack on the
maintainer's commitment to the project, rather than a useful contribution
from a co-developer and something to be thankful for, is quite a nuisance.
I dunno, it just seems that the same people who're wanting Debian to be
more encouraging and accessible to newbies, to chicks, or to others,
are at the same time being pretty unwelcoming and unencouraging of
other contributions, whether that be, eg, James work, which can hardly
be mentioned without adding copious criticism, or non-free package
maintainers' contributions which all the candidates seem to want dropped
from the project, or, evidently, to the contributions of folks who don't
offer Debian the same level of dedication our DPL does.
(And, obviously, I'm not saying everyone should be let do anything
without any oversight at all -- if people don't know what they're doing,
we shouldn't be inflicting their shoddy work on our users, and if they're
working against our goals, we shouldn't be shy about making sure they
can't do us or our users any harm; but if people are slacking off, we're
better off making sure we pick that slack up, not make sure there are
apropriate punitive procedures in place.)
Cheers,
aj
--
Anthony Towns <aj@humbug.org.au> <http://azure.humbug.org.au/~aj/>
I don't speak for anyone save myself. GPG signed mail preferred.
Linux.conf.au 2004 -- Because we could.
http://conf.linux.org.au/ -- Jan 12-17, 2004
Attachment:
signature.asc
Description: Digital signature