Python 3.12

[dan-fritchman]

Not immediately working. E.g.:

https://github.com/dan-fritchman/Hdl21/actions/runs/8443756151/job/23127988377

(Or at least our CI is not immediately working.)

[dan-fritchman]

Noting https://github.com/Vlsir/Vlsir/pull/98 (basically the same topic for VLSIR) worked, without any real changes.
Presumably just updates to our dependencies in the test environment.

[joamatab]

Awesome!

could we make a new release?

[dan-fritchman]

Should probably actually fix this one first.
You wanna do a quick PR for it?

[joamatab]

the tests worked for me

[dan-fritchman]

So I don't see it passing anywhere on the CI runs?
E.g. the most recent run:
https://github.com/dan-fritchman/Hdl21/actions/runs/9786217659

We explicitly enable "continue on error" for py3.12 here: https://github.com/dan-fritchman/Hdl21/blob/5633bc66b19dfecedbda23f69ec3e4818021afe2/.github/workflows/test.yaml#L43

Maybe just PR that out, and enough stuff "behind the scenes" will have changed to make it "just work".

[dan-fritchman]

#232 "fixes" this.
Which is to say, fixes our test setup.
Particularly I split out the big matrix of Python and pydantic test-versions from the containerized simulation tests.

[dan-fritchman]

Just in time for Python 3.13's release on October 1:
https://peps.python.org/pep-0719/

[dan-fritchman]

OK well spoke to soon, #232 doesn't quite work yet.
Appears the combination of Python 3.12 & Pydantic 1.9 don't play nice. E.g.: https://github.com/dan-fritchman/Hdl21/actions/runs/10733148162/job/29766001200

Noting for Python 3.7-3.8 it's "the other way around". Pydantic v1 works, v2 fails.

I don't see what we're gonna be able to do about that, other than maybe writing a more elaborate dependency specification (if that's possible). Probably just like a note or table.