Bug#1026104: longstanding problem with dependencies of python3-numpy in testing
Dear all:
Sandro wrote:
> Please compare the numpy popcon numbers (which we know are not
> detailed, but it's a start) with the handful of people having reported
> this issue. While maintaining numpy, my work caters towards the actual
> numpy users, not people who just want to not have 2 python
> interpreters installed at the same time during a time of a Python
> interpreter transition, in a not-stable release.
Debian testing is not stable by definition. However, in scientific
computing it is not unusual that Debian testing is used routinely
for productive work because Debian stable tends to be too outdated.
The issue for me is not that I do "just want to not have 2 python
interpreters installed". The issue for me is the Depends of the
transitory python3-numpy releases break my dependent software,
as described in https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1025169.
> What Depends are you referring to here? as a reminder, they are
> auto-generated by the Python helper `dh-python` and none are
> explicitly defined in the package Depends field
Would it make sense to submit an issue against dh-python?
> the burden of proof should be on the proponent(s) to come up with a
> sound technical solution
How about: just not release transitory versions to Debian testing
until the dependency on the old Python interpreter is gone?
Helmut wrote:
Given a Recommends, we'd likely have to wrap these scripts in a way that
produces a useful error message. Has anyone seen a patch for making this
happen? Joachim, would you volunteer to write one?
Sounds not too difficult ... for somebody who has full understanding
of these scripts and all context. Which, unfortunately, is not even
remotely the case for me.
Personally, I have just given up on python3-numpy. Instead, I am
now installing numpy and matplotlib using pip3. However, I would
happily return to an all-Debian installation if a solution to the
transition mess can be found. Thank you all for engaging in this
debate.
Kind greetings, Joachim
Reply to: