Should we migrate to pybind11 or merge with can_ada?

[anonrig]

@tktech did an amazing job with https://github.com/TkTech/can_ada. I think we should follow the path, or maybe merge the two implementations, since it is 2 times faster and the goal of Ada project is performance?

What do you think @lemire @tktech @bbayles?

[TkTech]

can_ada is under the MIT license, so feel free to use it however you like. I believe it's complete as-is, or you can copy parts of it, fork it, whatever. There's also an experimental variation using nanobind which has a smaller range of supported versions and a more complicated build, but offers another ~30% speed boost.

[Master-Hash]

What about the URL class? Without it, it may be a break change.

[bbayles]

I don't mind, but will bow out as maintainer rather than do the rewrite. Cheers!