Since we don't actually rely on any features introduced in 3.10, I would heavily prefer we release with 3.9 support the critical fixes in #376 merged.

Python 3.9 goes EoL next week when Python 3.14 is released. Those who refuse to update their codebases from EoL versions can run cpplint via pipx or uvx. This will at least help them to learn the advantages of modern tooling.

I don't see how dropping an EOL version provides any benefit. It doesn't make cpplint more secure when the user hasn't updated their Python anyway, and the namespace indentation problem was a big regression that I think a lot of projects should have fixed.