Fwd: closing out old bugs of unmaintained releases

[Nils Philippsen]

On Sun, 2008-01-06 at 20:29 -0500, Paul W. Frields wrote:
> On Sat, 2008-01-05 at 22:20 +1100, David Timms wrote:
> > If the bug is still present, there may also be a newer development 
> > package that you could test, available in the development repository. 
> > {see wiki for how to yum --enablerepo=devel update this_package}. Again, 
> > entering a bugzilla comment to say you tested with that rawhide version 
> > and the result of your test, and including a change in bug resolution to 
> > CLOSED - NEXT_RELEASE helps developers to make progress on the issue you 
> > found."
> 
> Hmm.  I wonder how many will take us up on that offer, have a problem
> due to instability of other bleeding-edge stuff brought in as
> dependencies, and then write us off.  I'm not saying it's a bad idea to
> ask reporters to do it, but we should be up front about the
> ramifications.

I think it's a bad idea to ask that from reporters automatically. My
guess is that most packages shouldn't be dealt with that way because
they likely would need a huge dep-chain upgraded. That'd leave the
reporter with a half stable/half Rawhide chimera which would be even
less stable (or supportable) than say updates from a test release to a
final version (which we expressly discourage people to do). The only
situation where I would suggest this course is with noarch packages like
one of the python-only config tools, if you (the maintainer) are sure
that no new deps have been added in the meantime. Checking other
(binary) Rawhide packages is mostly only feasible on testing systems
(where they could install Rawhide wholesale). If a maintainer thinks
it's a good idea to test the Rawhide version(*), he can always build a
scratch package against the stable Fedora version the user runs, but
that isn't something we should even mention in that message in order to
keep it short, to the point.

(*): Some packages are in sync between Rawhide and stable updates. A
user would be pretty upset to test something which is essentially a
rebuild of what he already has ;-).

Nils
-- 
     Nils Philippsen    /    Red Hat    /    nphilipp at redhat.com
"Those who would give up Essential Liberty to purchase a little Temporary
 Safety, deserve neither Liberty nor Safety."  --  B. Franklin, 1759
 PGP fingerprint:  C4A8 9474 5C4C ADE3 2B8F  656D 47D8 9B65 6951 3011