Alexander Neumann

> This PR changes the selected BLAS/LAPACK implementation for some targets: @BillyONeal I don't think that table is correct. Before it is basically undefined due to https://gitlab.kitware.com/cmake/cmake/-/blob/master/Modules/FindBLAS.cmake?ref_type=heads#L394-L400 The code iterates...

cc @BillyONeal: All good

My PC: Before: ``` Elapsed time to handle libiconv:x64-windows-release: 1 min Elapsed time to handle gettext-libintl:x64-windows-release: 57 s Elapsed time to handle gettext:x64-windows-release: 4.8 min ``` After: ``` Elapsed time...

@dg0yt: any ideas what the expensive configure checks are? At least the one I used here are basically so fast that they doesn't seem to matter.

> So the expensive thing is to do the same test several times. Passing a cache file into the top-level configure makes the later tests re-use previous results. For me...

> make should really be able to build "debug" and "release" in parallel If the sources are always copied before configure I'll agree with that. > Also "debug" and "release"...

related: https://codereview.qt-project.org/c/qt/qtbase/+/305484

You probably want to avoid the whole code duplication and rather have a variable called `OPENSCAD_QT_VERSION` which can be either set to 5 or 6 (or auto). See e.g. https://gitlab.kitware.com/vtk/vtk/-/blob/master/CMake/vtkQt.cmake?ref_type=heads

Do I read this PR correctly that only CMake changes are required to build OpenSCAD with Qt6? If so, I might try building it with vcpkg on windows.

From experience with VCPKG: ITKTargets.cmake might not be the same if libraries are linked directly and the configuration have different prefixes. Then the `INTERFACE_LINK_LIBRARIES` will contain the library name and...