Brecht Sanders

Found it! Apparently the make sources store the value of CXX set at build time of make in the `.exe` file. I modified my [build script](https://github.com/brechtsanders/winlibs_recipes/blob/main/recipes/make.winlib) to strip the full...

New builds with fixed `mingw32-make` at: - https://github.com/brechtsanders/winlibs_mingw/releases/download/14.1.0posix-18.1.8-12.0.0-msvcrt-r3 - https://github.com/brechtsanders/winlibs_mingw/releases/download/14.1.0posix-18.1.8-12.0.0-ucrt-r3

Thanks for finding and figuring out this issue. If this is indeed an issue in the GCC code it would be best to report your findings there (https://gcc.gnu.org/bugzilla/).

The winlibs build uses official releases and for ncurses the release version is still 6.4. The assumption is that when significant bugs are fixed a new release is issued. Maybe...

For another issue (#168) I released a new build of GCC 10.5.0, but using the ncurses as you proposed. Could you please test if this release works for you?

I'm open to suggestions on how downloads on github.com can be organized in a better way. But for now the different downloads are explained at https://winlibs.com/

I notice you use LLVM/CLang v17 while the latest winlibs release with LLVM is v18.1.1 The focus of https://winlibs.com/ is mainly on GCC. The LLVM supplied uses the GCC standard...

Can you please try if this issue is resolved with the most recent release at: https://github.com/brechtsanders/winlibs_mingw/releases/tag/13.2.0posix-18.1.3-11.0.1-msvcrt-r7 ?

Do you have objective facts about speed? What exactly is slow? Compiling (C/C++)? Linking? I don't write the software (gcc, binutils, ...) myself, I just compile it for Windows, but...

There have been precompiled header issues with GCC on Windows in the past. Do you have issues with the latest version (GCC 12.2.0) ?