[Biopython-dev] Biopython 1.56 release plans

[Peter]

On Thu, Nov 4, 2010 at 5:13 PM, Peter <biopython at maubp.freeserve.co.uk> wrote:
> Hi all,
>
> I've mentioned in recent threads that I think we should try and
> release Biopython 1.56 this month (November 2010).
>
> I think the NEWS file is pretty up to date, and covers important
> new functionality like Andrea Pierleoni's UniProt XML parser
> and the IMGT support (with Uri Laserson).
>
> Is there any other functionality which is ready for merging?
>
> For example, Tiago - you've been doing lots of work on your
> branch with the PopGen code. Is that code ready? I'm willing
> to do the git merge/rebase.
>
> Is there any reason to bother with a beta release this time?
>
> If there are no pressing additions, I may be able to do the
> release tomorrow - otherwise how about aiming for Thursday
> or Friday next week (11 or 12 November)?

As people will have noticed, the release didn't happen this week.

Tiago has been doing some excellent work with the prototype
buildbot server (see http://events.open-bio.org:8010/grid for
the current temporary home), and as part of this we've set
up a few machines as buildslaves. See this thread:
http://lists.open-bio.org/pipermail/biopython-dev/2010-November/008376.html

Running under Jython on the Mac showed a few problems
which appear to now be sorted, other than an apparent
problem with the GenePop tool.

Unfortunately running under Jython on Windows XP has
revealed several new problems, e.g.
http://lists.open-bio.org/pipermail/biopython-dev/2010-November/008431.html

As things stand all the tests (*) are fine on "C" Python on
Linux, Mac, and Windows. They are also fine on Jython
on Linux, give some warnings on Jython on Mac, and 3
errors on Windows.

Hopefully we can address these three test failures (or
at least understand them) and do Biopython 1.56 at
the end of next week instead.

Peter

(*) We haven't audited all the slave test output to check
which tests are being skipped due to missing optional
dependencies yet. e.g. command line tools, or Python
modules like ReportLab or NetworkX.