Improve the contributing guidelines.

[tristanlatr]

fixes #752 fixes #753 fixes #756 fixes #757

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Comparison is base (1b66928) 92.69% compared to head (abb5b87) 92.69%.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## master #760 +/- ## ======================================= Coverage 92.69% 92.69% ======================================= Files 47 47 Lines 8285 8285 Branches 1826 1826 ======================================= Hits 7680 7680 Misses 347 347 Partials 258 258 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[tristanlatr]

Rendered changes can be reviewed here: https://pydoctor--760.org.readthedocs.build/en/760/contrib.html

[github-actions[bot]]

According to pydoctor_primer, this change doesn't affect pydoctor warnings on a corpus of open source code. ✅

[tristanlatr]

@buhtz you can review this change, thanks

[github-actions[bot]]

According to pydoctor_primer, this change doesn't affect pydoctor warnings on a corpus of open source code. ✅

[github-actions[bot]]

According to pydoctor_primer, this change doesn't affect pydoctor warnings on a corpus of open source code. ✅

[tristanlatr]

The information from primer and codecov might be useful for maintainers but not for extern contributors. It just blows up the discussion with noise and drive less experienced contributors away.

Even external contributors should be aware of the missing coverage lines or the potential new warnings triggered after a their changes. So I’de say it’s still relevant.

[github-actions[bot]]

Diff from pydoctor_primer, showing the effect of this PR on open source code:

sdk-python (https://github.com/temporalio/sdk-python): typechecking got 1.12x slower (123.8s -> 138.2s)
(Performance measurements are based on a single noisy sample)

docutils (https://github.com/docutils/docutils): typechecking got 1.13x slower (81.6s -> 92.4s)
(Performance measurements are based on a single noisy sample)

twisted (https://github.com/twisted/twisted): typechecking got 1.08x slower (232.5s -> 251.0s)
(Performance measurements are based on a single noisy sample)

[github-actions[bot]]

According to pydoctor_primer, this change doesn't affect pydoctor warnings on a corpus of open source code. ✅