[Biopython] Planning Biopython 1.72

[Michiel de Hoon]

 Hi Peter,
I don't think we need to reintroduce the old API; the benefit of doing so for any users who may have used this function does not weigh up against the added complexity. Anyway the version() function was more for internal use and debugging, and it's easy to fix for users who use this function.

Best,-Michiel

    On Saturday, June 23, 2018, 1:11:48 AM GMT+9, Peter Cock <p.j.a.cock at googlemail.com> wrote:  
 
 On Fri, Jun 22, 2018 at 4:49 PM, Peter Cock <p.j.a.cock at googlemail.com> wrote:
> I'm part way though the release now, and for future reference
> I think I would now be able to simplify the build instructions
> as well...

I'd normally ask the developer-only list, but since we merged
the mailing lists, if anyone wants to do any last minute local
testing before things get up to PyPI:

http://biopython.org/DIST/biopython-1.72.tar.gz
http://biopython.org/DIST/biopython-1.72.zip

(There may still be some strange locale specific encoding issues
with the test suite which our current developers are not
running into on their own machines, there are some open issues
on GitHub about this - we need help here.)

> The first hurdle is with the wheel building, it looks like I need
> to debug this cryptic error:
>
> Traceback (most recent call last):
>  File "<string>", line 1, in <module>
> AttributeError: 'module' object has no attribute 'version'
>
> https://travis-ci.org/biopython/biopython-wheels/builds/395521793

That seems easy to avoid, but suggests an accidental API
change to Bio.Cluster was introduced:

https://github.com/biopython/biopython-wheels/commit/4018334b038278e8da3522039584e57dd361f135

Michiel, is it worth re-creating the old API now, something
like this in Bio/Cluster/__init__.py perhaps?

def version():
    return __version__

Peter
  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.open-bio.org/pipermail/biopython/attachments/20180623/d37ffcc5/attachment.html>