bug #149713

[Panu Matilainen]

On Mon, 2005-02-28 at 17:11 +0100, Michael Schwendt wrote:
> On Mon, 28 Feb 2005 17:40:10 +0200 (EET), Panu Matilainen wrote:
> 
> > > After verifying the compiler's warnings, such reports should be forwarded
> > > to upstream developers and are worth fixing, if the code is executed.
> > 
> > I don't think we can expect an extras packager to fix upstream bugs in the 
> > code, packagers deal with *packaging* bugs. If the packager *can* fix bugs 
> > (and certainly many can, to some extent at least) then fine,
> 
> My definition of "packager" and "package maintainer" is different, and
> I've seen other community people see it like me.  Enough insight into the
> code (or programming language) provided, a _package maintainer_ fixes bugs
> to improve the overall package. This ranges from fixes for C/C++ Standard
> issues (e.g. compiler rejecting older code) to fixes for bugs causing
> segmentation faults or other run-time defects.

If Extras packagers (for Core it's IMHO different since that's being
done by paid professionals) are required to be able to fix bugs in the
software they package I think it should be clearly communicated
somewhere instead of being left up to people's own definitions of what
consists of package maintainership.

> I'd rather increase the array size by one element than compile code which
> writes beyond the array boundary.
> 
> > but me thinks 
> > things like this can safely be closed with "please report upstream 
> > instead" unless the bug is caused by FE-specific changes to the software.
> 
> The reporter expects to reach package maintainers, who keep contact with
> the various upstream project developers. Otherwise he would not spend the
> time on doing these test builds. He would not look up the contact
> addresses or bug tracking databases for dozens (at the worst case
> several hundreds) of projects.
> 
> If you expect him to report the bugs upstream and he doesn't do it, it may
> be that the same old bugs bite you in the future and cause run-time
> misbehaviour.

> This is his way of trying to help.  Yes, I realise that it fans out the
> bug traffic. A single reporter in a single bug database -> many package
> maintainers forwarding the bugs to multiple upstream projects.

Then again I've seen more than a few cases of "please file to upstream
(bugzilla)" comments from RH package maintainers in bug reports. Where
do you draw the line? My line has always been "if it's caused by my
changes it's obviously my responsibility to fix, otherwise complain
upstream." If that's not considered enough I think I'm out here apart
from the couple of projects where I actually contribute upstream.

	- Panu -