[Bioperl-l] Fate of Bio::Perl newbie module?

[Carnë Draug]

On Wed, 27 Nov 2019 at 15:24, Andreas Tille <tille at debian.org> wrote:
>
> Hi,
>
> On Wed, Nov 27, 2019 at 12:08:08PM +0000, Carnė Draug wrote:
> > On Wed, 27 Nov 2019 at 10:52, Andreas Tille <andreas at an3as.eu> wrote:
> > > > The Bio::Perl module is on the Bio-Procedural repository [1].  No one
> > > > has picked up its maintenance and remains unreleased so you won't find
> > > > it on CPAN.
> > >
> > > Would you volunteer to take over this one?
> >
> > I'm sorry but I can't.
> >
> > > If not could you give some
> > > short intro what is needed to do to get that module back on CPAN?
> >
> > Sure.  Someone will need to show interest on take over maintenance of
> > it.  To do so, I would open an issue on the repo [2] saying that I was
> > interested.  This person will need a PAUSE account [3].  The admin of
> > the BIOPERLML on PAUSE (that's cjfields) will need to give permissions
> > to that person.
> >
> > The distribution is already setup to be released with bioperl's dzil
> > plugin bundle [4].  The instructions are there.
> >
> > I think a maintainer should at least address issues #1 [5] and #2 [6]
> > before making the release.
>
> Thanks for the introduction.  For me as a non-Perl programmer that's way
> to complex since even if I make it through the procedure I could not
> take over any maintenance of a Perl module.

I didn't realise you were asking instructions for you.  I wrote them
thinking of someone that could maintain a Perl module.

> > Also, beware of the following optional dependencies which are not part
> > of the core BioPerl distribution (not sure if already packaged in
> > Debian):
> >
> >   * Bio::DB::EMBL
> >   * Bio::DB::GenBank
> >   * Bio::DB::GenPept
> >   * Bio::DB::RefSeq
> >   * Bio::Tools::Run::RemoteBlast
> >
> > As well as these also optional dependencies which are also not on CPAN
> > yet:
> >
> >   * Bio::DB::RefSeq [7]
> >   * Bio::DB::SwissProt [8]
>
> Adding some other lib#AvailableOnCPAN#-perl package seems to be a
> feasible thing to do, thought.  Also if there is some maintained
> upstream source outside CPAN should be fine.
>
> So the question is:
>
>     Do we have some Perl programmer in the Debian Med team who is
>     willing to follow the procedure described above?
>
> If the answer is no, what do we do to salvage packages like prokka and
> possibly others since it seems the old bioperl is broken appart that
> strongly that other packages relying on it will be broken as well. Is
> the answer that we revert the version bump any rely on the old
> monolithic bioperl or should we talk about the problem with upstreams of
> prokka (and may be others) to change their code to only use maintained
> bioperl modules?

On Perl, a distribution will depend on a set of modules not on other
distributions.  So something like roary does not list the BioPerl
distribution as dependency, it lists modules like Bio::SeqIO.  So as
long as roary keeps using maintained modules, meaning released, it
will work, no matter which way bioperl gets split apart.

I think speaking with upstream of prokka, roary, and others may be
best approach since this is not a Debian only problem.  All their
users will face the same issue.  As perl programmers and users of
those modules they are in a good position to take over maintenance of
the bioperl modules they need which them someone on Debian Med team
could package.  Or they can replace their use of the unmaintained
bioperl modules.

I'm sorry I can't offer more help than comments, opinion, and
direction if you want.

David