Non-responsive maintainer process for kurzawa / Krzysztof Kurzawski

[Michael Schwendt]

On Fri, 18 Sep 2009 12:23:13 -0400, Jon wrote:

> > Monitoring
> > how many bug reports go unanswered by maintainers for example.
> 
> That would be a good start. One could produce a report (sounds like
> I'm volunteering, but I have no time this weekend because of my
> brother's wedding) for something like "packages with no response from
> the maintainer for ~3 months".
> 
> Actually writing this may be harder than it sounds, though - I could
> certainly find unmodified bugs, but what if they were modified by
> someone/something other than the maintainer?

Well, first of all, we're interested in entirely "inactive tickets".

That are tickets were nobody other than the reporter has posted (and
possibly even the reporter has stopped adding comments). It gets
interesting when such tickets pile up or are still without a reply after a
growing period of time. It gets more interesting when someone/something
closes such tickets and this "someone/something" is not assigned to the
package in pkgdb.

If "something" is some automated bug-triaging script, it becomes
dangerous, since you cannot rely on the bug reporter to do the grunt work
and reopen a ticket. Such tickets may count as "ignored tickets".
A worst-case scenario. Only after months somebody may find out that a
package maintainer has been non-responsive and inactive for months and
that packages are out-of-date or even broken.