openjdk team mailing list archive

["Adam D. Barratt" <adam@xxxxxxxxxxxxxxxxxxx>]

On Sat, 2015-07-04 at 15:11 +0100, Steven Chamberlain wrote:
> Hi,
> 
> Adam D. Barratt wrote:
> > I assume this is due to the presence of
> > 
> > openjdk-7-doc          | 7u75-2.5.4-3  | unstable   | all
> > 
> > which appears to be being kept around due to the old kfreebsd-* binary
> > packages.
> 
> I didn't expect that kfreebsd FTBFS could still affect testing
> migrations.  Is this a (rare, edge-case) bug in britney?

Well, it's arguably not really an anything in britney, as explained in
parts of the message that you didn't include in your reply.

britney doesn't determine which RC bugs are relevant directly. The BTS
produces a list of the RC bugs affecting testing and unstable and
britney merely takes the relevant disjunctions of those lists.

The issue specifically is that the kfreebsd FTBFS means that dak is
keeping the old (and RC-buggy) version of the arch:all package around in
unstable, leading to the BTS indicating that the bug still affects the
unstable version of openjdk-7-doc, as there is an unfixed version in
unstable.

fwiw, it's possible for a hurd FTBFS to keep old buggy arch:all around
in such a fashion, but I don't remember it being an issue in practice
recently at least; possibly because we've historically been happier
asking for removals of things that break hurd.

> We do have a patch to fix the kfreebsd FTBFS however:
> https://bugs.debian.org/788982

Getting that applied would seem the best fix all around.

Regards,

Adam