[python-committers] Pace of change for Python 3.x [was: My cavalier and aggressive manner, API] change and bugs introduced for basically zero benefit

[Neil Schemenauer]

On 2017-01-21, Brett Cannon wrote:
> So the common theme here regardless of whether you agree with Raymond or
> Victor's approach to development is that we are not getting enough code
> reviews to go around. To me that's what the systemic issue is that this
> email is bringing up.

I think there is another issue.  What pace of evolution is
appropriate for Python 3.x?  Python 2.7.x represents an extreme
position: bug fixes only.  Evolution for 3.x is too fast for some
users.  It is natural that as Python matures the average user is
more interested in a stable langage than in bleeding-edge features.
Python 3.x should be compelling to those users if we want to
smoothly end-of-life 2.x.  If 3.x looks like a "change treadmill"
then 2.7 will have a long life.

I don't know how to solve that.  Having experimental and stable
forks of 3.x seems bad.  Incompatible language versions are a drain
on the ecosystem.  However, a total block of language evolution is
also no good.

Maybe we could emulate the Linux kernel releases.  I.e. have
relatively fast moving development but also choose releases to give
long maintenance cycles.  Ideally the long term releases would be
synchronized with OS distribitions (e.g. Red Hat, Debian, Ubuntu).

The Linux kernel has a strict policy of not breaking user space
programs.  For Python, a similar rule would be not breaking Python
programs.  Until 2.x has finally died out, I think we should have an
extra conservative policy on incompatible language changes.

Lack of patch reviewers is a major problem.  Linux seems to have an
advantage of more contributors and that many are paid to work on
Linux.  As far as I'm aware, there is no core Python developer who
works mostly on Python and gets paid for it.  So, we have to work
with what we have or figure out some way to get people funded.

BTW, thanks for working on the GitHub migration Brett.  I think it
will help.