Drop support for Python 3.7

[hugovk]

Python 3.7 is EOL this month:

• https://devguide.python.org/versions/
• https://peps.python.org/pep-0537/

Let's drop it, it'll save us 14 wheels (and we're adding more for 3.12 in #594 :)

I don't mind if wait until 2023-06-27, but we're probably close enough.

[codecov-commenter]

Codecov Report

Merging #595 (92af18b) into main (e1dfadc) will not change coverage.
The diff coverage is n/a.

❗ Current head 92af18b differs from pull request most recent head b042a39. Consider uploading reports for the commit b042a39 to get more accurate results

❗ Your organization is not using the GitHub App Integration. As a result you may experience degraded service beginning May 15th. Please install the Github App Integration for your organization. Read more.

@@           Coverage Diff           @@
##             main     #595   +/-   ##
=======================================
  Coverage   91.66%   91.66%           
=======================================
  Files           6        6           
  Lines        1944     1944           
=======================================
  Hits         1782     1782           
  Misses        162      162           

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[hugovk]

Does dropping a Python warrant a major or minor bump?

I used to do major, but python_requires = >=3.8 ensures the new release only installs for the right group, and I'm now leaning towards a minor, but either is fine with me.

(If we had deprecations to drop, I'd do those at the same time and do major, but we don't.)

A recent poll on the topic favours minor:

https://hachyderm.io/@nedbat/110412912217669532

[bwoodsend]

I agree with the poll. Semantic versioning exists pretty much exlusivley so that people can write package >= 2.x.y; < 3.0.0 and receive automatic improvements without automatic breakages. It's not needed to avoid new versions of packages being installed with too old Python environments because there's a separate mechanism for it.

[hugovk]

I've flipped the labels so the automation will take care of the next version (minor instead of major).

[JustAnotherArchivist]

I'm fine with both merging now and waiting until EOL. As you say, close enough, and even if something urgent comes up that we will fix in the next couple weeks, people should still upgrade to 3.8+ anyway.

And agreed that no major bump is necessary.