John W. Parent

> It looks like you know best what the issue with CMake, FindCUDA and kokkos-kernelnis. I've not seen any clear description or analysis what the problem with CMake 3.28.3 is,...

> It's recommended to use `spec.satisfies()` instead of `in spec` these days since the latter can result in unintended consequences see the docs [here](https://spack.readthedocs.io/en/latest/packaging_guide.html#testing-spec-constraints). Thanks for the heads up, didn't...

@alecbcs I refactored so that all platforms now use the `libs` variant (single valued, default to `shared` on Windows) and updated the spec checks to use `spec.satisfies` with the `libs`...

@alalazo > This seems Ctrl+C Ctrl+V code. Can we have proper unit tests to poke paths with spaces on Windows (instead of running the same tests again, but in a...

What is `write_install_manifest` responsible for doing?

> I wasn't able to test on Windows in the case bugs pop up there I think @scheibelp and @johnwparent have access to Windows machines to test. @alecbcs Thanks for...

Thanks for the upkeep Dan! Changes in general still look good to me, so pending CI passing, this is ready for another round of review.

@SMillerDev I have addressed most comments. The remaining point of uncertainty from the previous review is the "magic". I've removed all that I can while still allowing MUMPS to build...

> The errors in the compilation might be due to the fact that the present formula compiles with clang. Given the fact that it is a (mixture) of fortran and...

@fahasch I believe I have made all requested edits and the package seems to be working now.