[Biopython-dev] Updating Biopython requirements wording to recommend Python 3

Tiago Rodrigues Antao tra at popgen.net
Sun Oct 4 18:35:31 UTC 2015


On Sun, 4 Oct 2015 20:10:29 +0200
Markus Piotrowski <Markus.Piotrowski at ruhr-uni-bochum.de> wrote:

> Actually, it's not our job to nudge people towards Python 3, our 
> recommendations should be given with regard to Biopython stability.


It is our job, if we decide, as a community, that we prefer to do this.
There is no reason at all to be "neutral" on this. Actually if we are
"neutral" then we really should support Python 1 maybe? And Python 2.1,
2.2 and 2.3. We have never been "neutral" in the past with 2.x
deprecations.

If we do not do this, then we actually are breaking apart from Python's
policy, which is very actively to nudge people to 3. One just has to
read Guido to see this.

Biopython has a profound smell of "Legacy" on itself. It was created in
a time where Python was a minority language, where one could not rely
on the availability of good external Python libraries. I suspect that
if it were not for Python success in itself, this would have been a
dead project long ago.

What we can do, at the very least, is to follow the lead of Python
itself.

If it were on me, the architecture would be completely redesigned (see
Bow's proposal), and dependencies on well-established libraries would be
introduced (scikit-learn, pandas, matplotlib).

Tiago


More information about the Biopython-dev mailing list