[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: NEW processing

* Lucas Nussbaum <lucas@lucas-nussbaum.net> [081203 18:19]:
> > I completely disagree. It's a welcome benefit if packages of inferior
> > quality are prevented from entering the archive in the first place imo.
> We currently have a long reviewing process before packages get into the
> archive. But once they are in, maintainers are free to do whatever they
> want with their packages, without any review happening.
> I'm not advocating that we just stop doing reviews. But IMHO, NEW
> processing should be about the legal problems, not about the random
> lintian warning/errors, and the various other packaging malpractices.
> What we should do instead is encourage reviews (either on mentors, or
> inside teams), because that's also a much more scalable solution.

I think package general reviews are a good idea. But checking packages
upon entering is still the best point. Removing packages that users
already user is not something most people take lightly. Rejecting
packages not having some minimal quality to begin with is easy, though.

> Of course, if ftpmasters come across random problems in the package
> while reviewing legal stuff and mention that to the maintainer, it's
> great. But having other things checked by ftpmasters just reduces their
> bandwidth and causes longer queues.

I think splitting that in two teams and having packages only leave NEW
when both ftp-masters have agreed on the legal aspects and an other
group of people have checked for other problems will not improve the
time packages are in NEW in my eyes.

And suggesting that things are not checked first time will only make
things having less quality. In my eyes there are rather too few checks.
Having things like packages with unneedingly repackaged upstream sources
and things like that entering which are not possible to undo even by
some later maintainer before the next upstream version and things like
that do not benefit the project at all...

	Bernhard R. Link

Reply to: