CI: windows 3.11-13 fail due to lack of `libdiffpy` on conda-forge

[bobleesj]

@sbillinge while Windows CI runs fail while others succeed

Turns out indeed there is no libdiffpy for windows: https://anaconda.org/conda-forge/libdiffpy/files

Should we proceed with sk-packaging without windows?

[bobleesj]

I see Tieqiong's current work: diffpy/libdiffpy#35

[bobleesj]

(I was looking into sk-packaging clustermining -> srfit needed -> srreal needed -> libdiffpy needed..)

[sbillinge]

@bobleesj I am not sure what your question is, but our current top priority is getting srreal working on windows on conda-forge. This we have never managed to do, and it unlocks everything.

If the question is about repackaging clustermining, let's wait, though we could start the boring part of the process (black, moving files to src, tests out of the package, etc.). We won't have the satisfaction of taking everything all the way to conda-forge at p3.13, but that boring work has to be done anyway, and it tests that part of sk-package. I leave that up to you.

[sbillinge]

clustermining has the addtitional interesting wrinkle that we want to move it to a diffpy namespace package, so it also tests again that part of the workflow (we did it before with Fourigui, but that was much earlier in the life of cookiecutter.....)!

[sbillinge]

closed by #59