[SciPy-Dev] Required Python Version

[David Cournapeau]

On Sun, Jul 18, 2010 at 4:40 PM, Ralf Gommers
<ralf.gommers at googlemail.com> wrote:
>
>
> On Sun, Jul 18, 2010 at 2:21 PM, Matthieu Brucher
> <matthieu.brucher at gmail.com> wrote:
>>
>> 2010/7/17 David Cournapeau <cournape at gmail.com>:
>> > On Sat, Jul 17, 2010 at 2:25 PM, Ralf Gommers
>> > <ralf.gommers at googlemail.com> wrote:
>> >
>> >> We now support python 2.4-2.6, for the next versions we'll add 2.7, 3.1
>> >> and
>> >> 3.2. Supporting more versions has a cost. And it's clear that the
>> >> amount of
>> >> people running 2.4 from svn is at or close to zero, because recent
>> >> syntax
>> >> errors for 2.4 have gone unnoticed for a long period.
>> >
>> > People running svn are a small proportion of our userbase. And the 2.4
>> > support cost is fairly minimal IMO.
>
> I am not the best judge of how much the cost is, but this thread started
> because of a patch that couldn't go in, and Charles question on Py_ssize_t
> is as yet unanswered.

I don't see why we could not add a support function to parse those
values. We need similar functions anyway to deal with size_t and co
correctly.

> That's debatable. 2.4 users have perfectly good releases right now (plus
> numpy 1/5 / scipy 0.8 coming up), 3.x users have nothing.

I meant that we have more 2.4 users than 3.x. It is really a
cost/benefit analysis: the advantages of 2.5 seem rather trivial to me
compared to losing support for quite common platforms.

David