Unpushing broken or obsolete updates from testing

[Matthias Saou]

Dominik 'Rathann' Mierzejewski wrote :

> On Friday, 06 July 2007 at 03:37, Kevin Kofler wrote:
> > Why are updates lingering in updates-testing for weeks when:
> > * they are known broken and/or
> > * they have already been replaced by a new version, which is already in
> >   stable?
> > IMHO, that's really what the "unpush" option is for. For example, do we
> > really need to have that obsolete Thunderbird RC sitting in testing
> > forever when the actual release is already out in the stable updates?
> > This also makes it harder to track down the legitimate updates which have
> > been sitting in testing for way too long.
> 
> This is exactly the kind of thing that should be automated, i.e. run a
> script once a week that'd check if there are any obsolete updates in
> testing and delete them (notifying the owner perhaps).

Removing packages from updates-testing which have newer releases in
updates-stable makes sense, but removing "broken" updates from
updates-testing might be more of a problem, since it then removes the
most recent version of a package, thus theoretically introduces the
possibility of breaking update paths for users of the repo, or
(possibly worse?) of leaving those users with a broken piece of software
for a long time...

Pushing fixed packages to testing-updates seems like the right thing to
do...

Matthias

-- 
Clean custom Red Hat Linux rpm packages : http://freshrpms.net/
Fedora release 7 (Moonshine) - Linux kernel 2.6.21-1.3228.fc7
Load : 0.68 1.53 1.84