[Biopython-dev] Preparing for Biopython 1.48

[Peter]

Dear all,

Is there anything we need to address before starting to prepare Biopython 1.48?

This is likely to be the final Numeric only release of Biopython, with
the following release hopefully supporting both Numeric and numpy in
some way.

(As an aside, once we support numpy, having scipy as an optional
dependency for the statistics Tiago wanted to use in Bio.PopGen does
seem less onerous.)

Regarding potentially deprecating Bio.Mindy and Martel, I propose we
do this in the following release (i.e. after Biopython 1.48).  We can
then drop the mxTextTools dependency completely.

While I would like to address some of the enhancements (esp Bug 2530)
these can wait.  Ignoring the enhancements, there are several "small"
issues on bugzilla that could be dealt with, but nothing that I think
warrants delaying the release.

One question: Currently Bio.SeqIO in CVS has partial support for
writing GenBank files (basically the sequence and minimal annotation -
no references, no features).  I don't want to rush something out
without proper testing, so do people think it would be better to ship
with this partial support, or temporarily disable it (a one line
change in Bio/SeqIO/__init__.py to the _FormatToWriter dictionary, and
probably refreshing the expected unit test output).

Comments and suggestions welcome!

Thanks,

Peter