bug#70815: [PATCH] ; Enahnce python-tests.el to adapt different python interpreters

[kobarity <kobarity@gmail.com>]

Mattias Engdegård wrote:
> 28 maj 2024 kl. 17.17 skrev kobarity <kobarity@gmail.com>:
> 
> > I think it is a "bug".  Because it is working with Python 2 on Linux
> > (with native and non-native completions), and I believe many would
> > expect it to work.
> 
> Could be, but it's hard to believe that whatever completion defect we have in python.el for Python 2 on macOS is affecting a lot of people. It seems usable enough to me, given the circumstances.
> 
> > As I mentioned before, I am not opposed to prioritize "python3".  I
> > don't think anyone is arguing that "python" should precede "python3"
> > anymore.
> 
> Thank you. I don't want to steamroller a change, but there is a real risk of paralysis from too much deference to code details that might just as well have been a simple oversight, which I believe is what happened here.

I'm not sure what you are proposing.  What I want to do now is not to
fix the "bug", but to accept the current situation and express it as
the ERTs.  Otherwise, even if we change the order of the interpreters,
it will fail when some Mac user sets EMACS_PYTHON_INTERPRETER to
"python" and runs ERTs for example.

Even if we could fix this "bug" right away, or even if we made the
decision to stop Python 2 support on Mac, I believe we would still
need a mechanism to skip an ERT based on the interpreter version
because similar stories would still come up.

Unless you are opposed to adding such a mechanism to ERT, I don't see
much difference of opinion between us.  If you want to change the
order of the interpreters as soon as possible, I am not opposed to
that either.