On Fri, Jan 24, 2014 at 01:00:29AM +0100, Zbigniew Jędrzejewski-Szmek wrote:
On Thu, Jan 23, 2014 at 04:53:47PM -0700, Kevin Fenzi wrote:
> On Thu, 23 Jan 2014 15:26:24 -0800
> Adam Williamson <awilliam(a)redhat.com> wrote:
> I think ideally any process around this should have at least two parts:
>
> a) an automated/scriptable part.
>
> In this part the script uses cold hard facts to look for possible
> packages that are unloved or package maintainers that are not active.
> There's tons of data we have now with fedmsg. Sadly, we don't have
> bugzilla in fedmsg, but we could scrape it directly.
> it generates a list that feeds to the next part.
>
> b) The generated list is examined by humans and action taken.
>
> Some things that are the list will be false positives. Try and adjust
> the script to not generate them.
>
> As a bonus, the script could also possibly try and figure out components
> that 'need help'...ie, lots of unanswered bugs or something.
Even a simple list of packages ordered by the time from last
non-mass-rebuild release multiplied by the number of currently open
bugs would be quite useful. Packages with bug-years above 50 or so
would be good candidates for inspection.
I looked into the build history of our package collection recently, more
specially trying to find the date of the last successful build of all our
packages using datagrepper.
I presented the output in:
http://blog.pingoured.fr/index.php?post/2013/12/17/Fedora-build-history
There is a rather small list that might be something to look into first:
66 packages have not been sucessfully re-built for 200 days or more
Pierre