[Debian-med-packaging] Please fix more bugs to make sure we can keep the usual quality - ${drink} offered

[Andreas Tille]

Hi folks,

I've just created the monthly team statistics and stumbled upon our bugs
closing statistics[1].  Despite I'm fighting very hard the number of our
open bugs (also RC critical ones) is growing (slowly but anyway).  So we
need to do more and I specifically like to encourage people who are
listed as Uploaders to some packages to feel dedicated to those packages
(if somebody can't really remember I can do an UDD query to generate a
list).  Despite I have quite some one bug per day rate and basically
collected packages from all those maintainers that became MIA the help
that is given by some active bug fixers is not sufficient to bear the
workload.

To give some motivation for those who have no idea what to do I try to
give a list of issues that urgently needs fixing.  I'd offer a ${drink}
at next meeting for anybody who contributes to the solution of the named
bugs until 1st of October.

#906811 blasr
  needs fixing pbbam_0.18.0+dfsg-1 first.  I asked for help (and I
  really need it) here:
   https://alioth-lists.debian.net/pipermail/debian-med-packaging/2018-August/064944.html

#902132 falcon
  I commited the latest upstream version that looks promising to fix that
  issue but it does not build.  The issue looks not *that* complicated
  (build system needs adapting to make sure the test suite will write
   to some writable location) but it just needs to be done.

#907819 jellyfish
  The bug is actually fixed in Git.  However I also tried to get the latest
  upstream version build which fails due some issues with the Python module
  and I have currently no idea how to fix this.

#906371 jmodeltest
  Here is a link by Olivier given (who is always very helpful - thanks a lot
  Olivier) but may be some Java educated team member want to pick up this
    https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=906371#27
  and finalises the package

#897795 librcsb-core-wrapper
  My comment in the changelog in Git says:
    While the gcc-8 bug is fixed the build system is somehow messed
    up.  No idea how to fix this for the moment
  Anybody volunteering for this? 

#896531 & #901455 libsbml
  Bug #896531 was easy and is fixed in Git.  The other bug has hints on
  this thread on this mailing list
    https://alioth-lists.debian.net/pipermail/debian-med-packaging/2018-May/062662.html
  which is basically: Remove the Java docs.
  I've started with this but failed.  This package is complex and
  definitely needs way more love than I could spent.  We should also
  check for a new upstream and we also should provide a Python3 package
  (see #847013). 

#842815 libsis-base-java / libsis-jhdf5-java
  Upstream has finally solved the issue of libsis-jhdf5-java but the
  new upstream version needs libsis-base-java.  Both have changed the
  build system to gradle.  I would really LOVE if somebody would spent
  some time into this.  We have interesting dependencies from these
  packages (inside the archive and pending upload to new)

#906156 python-skbio
  I have injected the last upstream version and packaged its new dependency
  python3-hdmedians but the package has several errors in the test suite.
  Pretty please investigate.

#897867 spades
  I've injectes a new upstream version which fixes the bug.  Unfortunately
  something has changed in the build system throwing exception:
    IOError: [Errno 2] No such file or directory: '/build/spades-3.12.0+dfsg/install_spades/bin/VERSION'
  My gut feeling is that this is a pretty simple and stupid error but
  I did not manage to solve this small riddle

#897877 trinityrnaseq
  New upstream fixes the FTBFS bug but depends from salmon > 0.9
  Any volunteer to upgrade salmon which is not very simple since it
  needs heavy (but more clever! - just ask me if you want to give it
  a try) editing of cmake files

There are *lots* of other probably more simple bugs.  Closing *any* bug
helps to enhance our quality.  I have heard that people are concerned
about the rise of bioconda cutting from our "market share".  I
personally think that there is only a small subset of applications we
share with bioconda and both projects have good chances to evolve in
parallel and healthy.  However, I'm afraid way more that we will loose
in terms of quality and thus the acceptance of Debian Med will become
smaller for internal reasons.  So everybody who is interested in keeping
Debian Med healthy and who has profitted from the project so far is
kindly invited to give back something and fix at least one bug per
month.  If anybody manages to contribute to the solution of one of the
named bugs above until 1st of October I offer a ${drink} at the next
meeting (Sprint, DebConf, or ???).

If you have no idea how to start I have another statistics for you:  My
"contributions" to the debian-mentors at lists.debian.org[2] are basically
**asking** questions about things I just don't know.  I'm also asking a
lot on debian-java at lists.debian.org and debian-python at lists.debian.org
(as well as debian-perl but the bugs above do not seem to be connected
to any perl issue).  So don't be shy, use the Debian channels, there is
also IRC (which I'm rarely using).  Try to investigate the issues and
help to make Debian Med the first choice for free software in life
sciences.

If you have further questions its possibly best to answer in the
according bug repoort.

Kind regards

        Andreas.

[1] http://blends.debian.net/liststats/bugs_debian-med.png
[2] http://blends.debian.net/liststats/authorstat_debian-mentors.png


-- 
http://fam-tille.de