bindings: python: v2.5.0 checklist

[vfazio]

• Resolve bindings: python: when do we EOL python 3.9 (and onwards)? #151
• Resolve bindings: python: format and lint examples and other files #171
• Resolve bindings: python: investigate free threaded builds #117 (note that this may be absolutely experimental and is not a blocker)
• Resolve bindings: python: Would type aliases make things easier to read? #150 (if we feel like it, should be trivial)
• If I have time, make some CI pipelines in my fork to build and run tests against 3.10-3.14(t)
• Resolve bindings: python: discuss whether we can close or release multiple times #180
• Resolve bindings: python: make minor modifications to C extension #181
• Consider bindings: python: consider a changelog #182

[brgl]

FYI: I plan to release python3-gpiod v2.4.1 later today with the recent fixes. If you have anything you think should go in, please let me know.

[vfazio]

FYI: I plan to release python3-gpiod v2.4.1 later today with the recent fixes. If you have anything you think should go in, please let me know.

Nothing that can't wait I don't think.

You can see my current list of staged changes on https://github.com/vfazio/libgpiod/commits/vfazio-upstream-proposed/

I still need to find time to test them and massage the commits so they can get past review on the mailing list.

[vfazio]

I'm going to start feeding patches to the mailing list soon. I'll start with the low hanging fruit first (3.10, lint settings, etc).

The free-threading stuff will likely spark some conversation.

[vfazio]

If I have time, make some CI pipelines in my fork to build and run tests against 3.10-3.14(t)

I've got something that seems to work for now on current master. Once I get things massaged, I will make sure everything passes before submitting

https://github.com/vfazio/libgpiod/actions/runs/23690880801

[vfazio]

Just as a status update, I'm making progress on the C extension changes.

Something I just became aware of is https://peps.python.org/pep-0793/

The PyInit_* stuff will be deprecated in favor of multi-phase module initialization that was introduced with 3.5 in https://peps.python.org/pep-0489/

I see no reason to not go ahead and move to the multi phase stuff. The biggest "challenge" right now is the gpiosim._ext module since it uses module state. I need to do a bit more digging into that.

[vfazio]

@brgl the ship has probably sailed a bit on this, but when should I be adding SPDX FileCopyrightText lines to the python bindings?

I don't know if it's appropriate to back date these or if it's a "update them going forward" kind of thing.

[brgl]

@brgl the ship has probably sailed a bit on this, but when should I be adding SPDX FileCopyrightText lines to the python bindings?

I don't know if it's appropriate to back date these or if it's a "update them going forward" kind of thing.

Feel free to backdate them to whenever git history points you.

[vfazio]

We're almost there... i just need more sleep and more caffeine.

[vfazio]

Series sent: https://lore.kernel.org/linux-gpio/20260522200419.105496-1-vfazio@gmail.com/

This should close out the last two checklist items.