Some modifications seem to prevent building this fork for a Linux target

[Gwani]

Referring to the most recent version of your 'v3.10.10' branch

I understand that this may be totally out of scope for the goals of this project, but i just want to bring it to your attention.

I'm currently trying to build a development environment for Windows and Linux which uses the (mostly) exact same version and features of each tool on both platforms built using the same GCC/Clang Toolchains, so i'm trying to build eveything from the same sources. For building Python, i gave your fork a try. It builds fine for the MingW-w64 platform, but for Linux there seem to be a quite few things which prevent the build. Examples are:

• this use of towupper() in Python/pathconfig.c needs #include <wctype.h>
• this use of _Py_GetDLLPath() Modules/getpath.c appears to need an #ifdef MS_WINDOWS (same as for definition of _Py_GetDLLPath()).
• Makefile.pre.in probably needs to distinguish between MinGW and other builds, as it inserts Windows specific options into the toolchain command lines which leads to : x86_64-linux-gnu-gcc: error: unrecognized command-line option '-municode', x86_64-linux-gnu-gcc: error: unrecognized command-line option '-mwindows' and ld: cannot find -lws2_32: No such file or directory

At this point, i decided to use the upstream version of cpython 3.10.10 for Linux your fork for the MinGW version. Thats fine with me, i just wanted to let you know that this particular version of your fork unfortunately not just "adds support for Mingw-w64 + clang/gcc" - but i understand that the main goal of this project is probably to supply MSYS2 with Python.

Thanks for all your hard work, i'm still making good use of it!

[lazka]

Thanks. While this is not a priority, I agree that it would be nice to make that work.

If we do, we should test it in CI as well.