Drop Python 3.8 support

[huxuan]

While Python 3.8 is reaching its end-of-life, maybe we should also consider dropping the support (at least in the generated project)? Currently the DevContainer workflow is failed due to Python 3.8 compatibility issue, and it needs extra hack to fix.

More evidence:

1. Sphinx 7.2.0 already drop Python 3.8 support on Aug 17, 2023.
2. pre-commit v3.6.0 alreay drop support for python<3.9 on Dec 10, 2023.

[huxuan]

@msclock Any comments?

[msclock]

@msclock Any comments?

I consider Python 3.8 is not up to the end of life according to security and has not escaped developer selection. Moreover, the scaffold has not yet released a maintainable version. Is the pdm support to resolve the conflicts?

Importantly, why is it not a breaking change when dropping python version support.Ideally, when a breaking change happends, the major version increment can ensure the following fixes and patches having more maintainable release. So I really think it needs to consider carefully.

[huxuan]

the scaffold has not yet released a maintainable version.

Actually, this is also one point for me, since the project is still in alpha stage, and there is only half a year remaining for Python 3.8, maybe it is more ease to propose this breaking change.

Is the pdm support to resolve the conflicts?

Hack are needed in the DevContainer Dockerfile and with that hack, we will lose the renovate update, I will propose a back-compatible hack later.

Importantly, why is it not a breaking change when dropping python version support.Ideally, when a breaking change happends, the major version increment can ensure the following fixes and patches having more maintainable release. So I really think it needs to consider carefully.

Gotcha, thanks for your reminder.

[msclock]

the scaffold has not yet released a maintainable version.

Actually, this is also one point for me, since the project is still in alpha stage, and there is only half a year remaining for Python 3.8, maybe it is more ease to propose this breaking change.

Is the pdm support to resolve the conflicts?

Hack are needed in the DevContainer Dockerfile and with that hack, we will lose the renovate update, I will propose a back-compatible hack later.

Importantly, why is it not a breaking change when dropping python version support.Ideally, when a breaking change happends, the major version increment can ensure the following fixes and patches having more maintainable release. So I really think it needs to consider carefully.

Gotcha, thanks for your reminder.

If it is urgent, the pr also is fine. But it might require a backfoward compatibility later. And it should ensure a minor change at least.🤣

Thanks for accepting the suggestion.

[huxuan]

If it is urgent, the pr also is fine. But it might require a backfoward compatibility later. And it should ensure a minor change at least.🤣

It is absolutely not urgent, but a compatibility strategy. Another question could be, why does sphinx drop Python 3.8 support so early, even before Python 3.12 is released.