[Biopython-dev] [Biopython] Switching to PEP440 compliant versioning for our development code

[Peter Cock]

On Thu, Jun 23, 2016 at 8:44 PM, Peter Cock <p.j.a.cock at googlemail.com> wrote:
> On Thu, Jun 23, 2016 at 7:59 PM, Eric Talevich <eric.talevich at gmail.com> wrote:
>>
>> +1
>>
>> I agree with this change. But in my experience, the suffix that plays best
>> with Python versioning- and packaging-related tools, and follows PEP440 most
>> literally, would be "<next_version>.dev0".
>>
>> -Eric
>
> Yes, by my reading of PEP440 etc, 1.68.dev would be normalised
> to 1.68.dev0 for comparison - but personally that looks silly and
> will just lead people to ask what the trailing zero means.
>
> On a related note, we could try to do something clever like
> 1.68.devN where N is the number of commits as in
> https://github.com/astropy/astropy/blob/master/setup.py
> but that seems overly complicated.
>
> Perhaps 1.68.devYYYYMMDD would be a better idea if we
> really cared about telling apart intermediate development
> releases (important the trailing number strictly increases),
> but thus far it hasn't been a problem.
>
> Peter

Change to use the .dev suffix made, i.e." <next_version>.dev"

https://github.com/biopython/biopython/commit/1dc54a492e4e40337586d5b8d1006acaded7440a
https://github.com/biopython/biopython.github.io/commit/60147f86f0a34387f05724d81a14debd8476fe8f

This solves the PEP440 warning from "python setup.py bdist_wheel",
and avoids the ambiguous plus to underscore for the in-development
wheel files:

http://lists.open-bio.org/pipermail/biopython-dev/2016-June/021446.html

Any further thoughts on the .dev0 alternative, or variants .devN?

Peter