Add a code "map" section to the developer documentation

[abergeron]

This should be merged after #964 to make sure all the links will work correctly, which is why I publish it as a draft.

Otherwise it is mostly ready if you want to read and comment.

Checklist

Tests

The new pre-commit hook fails for unrelated files.

Documentation

• I have updated the relevant documentation related to my changes

Quality

• I have read the CONTRIBUTING doc
• My commits messages follow this format
• My code follows the style guidelines ($ tox -e lint)

[bouthilx]

Thanks a lot for documentation the flow of execution during a call to orion hunt! I think it is quite complete, despite maybe a short mention of experiment version check/experiment creation race condition handling and a last section on termination criteria and how runners may exit (ex: broken experiment, completed experiment, reached max trials per worker).

[abergeron]

Sorry it took a lot of time, but I think I've addressed the comments now.

[abergeron]

I don't know what is wrong with tests/unittests/client/test_runner.py::test_runner_inside_dask, but I'm reasonably certain I didn't touch anything related to it and it keeps failing, but only on python 3.8.

[bouthilx]

There are 2 levels of max_trials/max_broken. There is at the level of the experiment. If we reach either max_trials or max_broken, all Runners will stop. And there is at the level of the Runner (under the config name worker, that's a bit confusing since the introduction of the Runner which now control multiple workers). If max_trials or max_broken is reached within the execution of this Runner, it will stop, but the other runner working on the same experiment may continue.

See for instance in doc:
https://orion.readthedocs.io/en/stable/user/config.html#max-trials
vs
https://orion.readthedocs.io/en/stable/user/config.html#config-worker-max-trials