[Debichem-devel] Bug#1042858: Bug#1042858: promod3: unable to install on arm64

[Sebastian Ramacher]

On 2025-04-09 14:25:54 +0300, Adrian Bunk wrote:
> On Tue, Aug 01, 2023 at 11:23:39PM +0200, Sebastian Ramacher wrote:
> >...
> > The following packages have unmet dependencies:
> >  promod3 : Depends: python3-ost but it is not installable
> >            Depends: python3-promod3 but it is not installable
> > E: Unable to correct problems, you have held broken packages.
> >...
> 
> On Fri, Jan 26, 2024 at 05:31:04PM +0200, Graham Inggs wrote:
> >...
> > On Fri, 26 Jan 2024 at 09:21, Andrius Merkys <merkys at debian.org> wrote:
> >...
> > > Is it necessary to limit the architecture list for promod3?
> > > I am reluctant to do so as there will be the need to manually update the
> > > list every time a new architecture becomes supported by its arch:any
> > > dependencies.
> > 
> > You could make promod3 arch:any instead of arch:all, but instead of
> > trying to maintain a list of architectures, simply add python3-ost as
> > a Build-Dependency (even if it is not really needed).  This will
> > ensure that promod3 is only built on architectures where python3-ost
> > is available.
> 
> For trixie, could you just hint promod3 back into testing?
> 
> The britney dependency check for binary-all packages is only done on 
> amd64 and arm64 exactly for the reason that it is quite normal that
> binary-all packages are not installable on all architectures.
> 
> If the release team wants all such packages in the archive converted 
> from binary-all to binary-any that's doable, but it's a bit late for
> an MBF of RC bugs for trixie.

No, we don't. But the sensibility of an arch: all package that is only
installable on amd64 is beyond me.

Anyway, I will hint it back into testing.

Cheers
-- 
Sebastian Ramacher