For your consideration: Secondary Architectures in Fedora

[David Woodhouse]

On Thu, 2007-05-31 at 17:03 +0200, Oliver Falk wrote:
> >  1. A spurious build or test failure which happens on all arches
> >     but intermittently.
> >  2. A simple error introduced in the package update.
> >  3. Something 'hard' which the arch team need to look into.
> >  4. A compiler bug.
> 
> OK, for is possible sure, but doesn't happen quite often hopefully.

You mean that #4 is possible but shouldn't happen often? That's true at
the moment but once we start pulling in new architectures it could
happen more often.

We should make sure that it's relatively easy for a package maintainer
to see a compiler failure, and just say "Don't Care" and leave it for
the arch maintainer to deal with (although the more conscientious
package maintainers would actually file the compiler bug with test case
for themselves). Being able to file a bug and then push a button for
'release the build anyway' seems not to be too much of an imposition.

In fact, _other_ than compiler bugs, I suspect that the _majority_ of
bugs which just happen to kill one build but not others may well be
generic bugs which are sensitive to timing or other criteria, and not
really arch-specific at all. Which is why it's so irresponsible to let
partially-failed builds make it through to the repo without _any_
interaction from the maintainer at all.

-- 
dwmw2