Drop Python 3.8 support

[seeM]

Addresses #6811. This doesn't really fit the feature/bug fix classification but I decided to go with bug fix.

Release Notes

New Features

• N/A

Bug Fixes

• [Drop Python 3.8 support #6811] Python 3.8 has reached end-of-life and is no longer supported.

QA Notes

It shouldn't be possible to start a Python 3.8 session in Positron. Everything else should continue to work as is.

[github-actions]

E2E Tests 🚀
This PR will run tests tagged with: @:critical

^{readme  valid tags}

[seeM]

I'm happy to consider small refinements to this PR based on Austin's feature spec but don't want the scope to creep too much – unless someone else is able to look into the additional work and we're able to sequence it appropriately. This PR is also a blocker to #6855 which is targeted for this milestone.

[isabelizimm]

should we change nightly to only 3.9+ as well?

[seeM]

Good catch, done in 0453138

[austin3dickey]

LGTM, just some nits.

[austin3dickey]

(putting these comments on a diff somewhere so they can be threaded)

I found some other places where we refer to python < 3.9, that you can probably simplify now:

positron/extensions/positron-python/python_files/posit/positron/tests/test_ui.py

Line 239 in 2a33b2f

@pytest.mark.skipif(sys.version_info < (3, 9), reason="requires Python 3.9 or higher")

positron/extensions/positron-python/python_files/posit/test-requirements.txt

Line 14 in 2a33b2f

polars[timezone]; python_version < '3.9' or sys_platform == 'win32'

positron/extensions/positron-python/python_files/posit/positron/pydoc.py

Line 760 in 2a33b2f

# as is from < Python 3.9, since 3.9 introduces a breaking change to pydoc.getdoc

[austin3dickey]

I'm happy to consider small refinements to this PR based on Austin's feature spec but don't want the scope to creep too much

Yeah, the only thing to call out is that people using 3.8 who update Positron might see some confusing behavior where if they switch away from their interpreter, they won't be able to switch back. I think that's fine for now. In my spec I'm proposing that we're a little more transparent about unsupported versions.

[seeM]

I thought about it some more and I think we should delay the jedi-language-server upgrade to the next milestone, and that we should consider implementing some of your suggestions. For the reason you mentioned as well as that I'd want more time for us to have the jedi changes in daily builds before it goes live. Let's discuss with everyone in the next sync about prioritizing some of the low hanging work for unsupported version UX.