Fix python_requires

[IAlibay]

Fixes #3103

Changes made in this Pull Request:

• Fixes python_requires in setup.py by specifying the minor version
• Also sets the max python version to <3.9 (we aren't testing 3.9 for master right now, so if we release it, it's best we avoid allowing users to install 3.9+).

To do

For some reason we don't let 3.5 on master anymore. I was under the impression that we still allowed py3.5 on 1.0.x. I think @orbeckst made the change (from the blame history)?

PR Checklist

• Tests?
• Docs?
• CHANGELOG updated?
• Issue raised/referenced?

[pep8speaks]

Hello @IAlibay! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

There are currently no PEP 8 issues detected in this Pull Request. Cheers! 🍻

Comment last updated at 2021-01-24 00:01:03 UTC

[codecov]

Codecov Report

Merging #3104 (57da1f7) into master (658a73d) will increase coverage by 0.93%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #3104      +/-   ##
==========================================
+ Coverage   90.89%   91.82%   +0.93%     
==========================================
  Files         149      154       +5     
  Lines       20694    21199     +505     
  Branches     3194     3194              
==========================================
+ Hits        18809    19466     +657     
- Misses       1250     1641     +391     
+ Partials      635       92     -543     

Impacted Files	Coverage Δ
package/MDAnalysis/topology/tpr/obj.py	96.82% <0.00%> (-3.18%)	⬇️
...onality_reduction/DimensionalityReductionMethod.py	97.14% <0.00%> (-2.86%)	⬇️
package/MDAnalysis/analysis/legacy/x3dna.py	0.00% <0.00%> (ø)
package/MDAnalysis/due.py	56.09% <0.00%> (ø)
package/MDAnalysis/tests/__init__.py	100.00% <0.00%> (ø)
package/MDAnalysis/analysis/legacy/__init__.py	0.00% <0.00%> (ø)
package/MDAnalysis/tests/datafiles.py	31.25% <0.00%> (ø)
package/MDAnalysis/core/selection.py	99.52% <0.00%> (+<0.01%)	⬆️
...sis/analysis/encore/clustering/ClusteringMethod.py	96.96% <0.00%> (+0.04%)	⬆️
package/MDAnalysis/analysis/base.py	95.60% <0.00%> (+0.04%)	⬆️
... and 85 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 658a73d...57da1f7. Read the comment docs.

[orbeckst]

I don't remember why I removed 3.5 in PR #2738. You are correct in that I shouldn't have done it because for 1.0.1 we need to support whatever 1.0.0 supported, and in 1.0.0 we did not remove support; in 0.20.0 we removed support for >2.7 <3.5.

The metadata also still showed support for 3.5.

Thanks for catching my mistake.

[orbeckst]

👍

[IAlibay]

Interesting something about this change is causing py3.5 to break 😕

[orbeckst]

The pip version is not up to it (from https://travis-ci.com/github/MDAnalysis/mdanalysis/jobs/474702826)

  File "/home/travis/miniconda/envs/test/lib/python3.5/runpy.py", line 193, in _run_module_as_main

    "__main__", mod_spec)

  File "/home/travis/miniconda/envs/test/lib/python3.5/runpy.py", line 85, in _run_code

    exec(code, run_globals)

  File "/home/travis/miniconda/envs/test/lib/python3.5/site-packages/pip/__main__.py", line 21, in <module>

    from pip._internal.cli.main import main as _main

  File "/home/travis/miniconda/envs/test/lib/python3.5/site-packages/pip/_internal/cli/main.py", line 60

    sys.stderr.write(f"ERROR: {exc}")

                                   ^

SyntaxError: invalid syntax

A f-string is used in this pip version.

I wonder why does conda install a pip version that is incompatible with the python version?

(Maybe I ran into the same thing before and couldn't take it anymore and upped the version...)

[IAlibay]

Ooo that's really strange, it was running 20.3 just this morning when I re-ran one of the failed jobs, but it's picking up 21 now...

😫 yeah looks like they released 21 just 9h ago..

Related issue: pypa/pip#9500

[orbeckst]

I sometimes deeply despise new software...

[IAlibay]

@orbeckst New release bugs are just a good way keep us on our toes (and burn off HPC allocations) 🙃

Pinning on CI seems to do the trick here. I'm assuming upstream will yank/amend accordingly, so it hopefully shouldn't impact 1.0.1.

py2.7 CI stalled again 🤯 I've restarted it so hopefully that completes....

[IAlibay]

py2.7 CI is very flaky, had to keep restarting until it finally worked. I don't think there's much we can do here since we're at a point where py2.7 conda installs is a long mess of frozen solves :/ If someone thinks it's relevant please do open an issue though