[ENH][systoleos] First commit

[oeramo]

Progress up to this point tracked in SystoleOS/guix-systole#52

[oeramo]

@RafaelPalomar I have made a mistake with the other branch that is currently set as default, https://github.com/OUH-MESHLab/BrainlabMirror/tree/supercede-systoleos-issue. As the admin, please set main as the default branch and delete the supercede-systoleos-issue branch. Thank you.

[RafaelPalomar]

@oeramo, please add a README.md with some notes on how to build and run the image with qemu

[RafaelPalomar]

This probably needs to be updated to the new path

[RafaelPalomar]

Also. How about the guix vm command

[oeramo]

@RafaelPalomar, I haven't managed to find a way to build the image from within this repository.

I had done my previous tests on the guix-systole repository before moving the PR here, and there building the image was trivial because all dependencies were under the same repository, and I could use the -L flag of guix system image to specify where to look for Scheme packages.

When the operating-system is defined in a different location, to avoid having to manually clone dependencies in place, we should leverage Guix Channels to find the dependencies. I have tried to set this up in a few ways, but the builder cannot find 3D Slicer, despite it being correctly imported, and despite the guix-systole channel pulling successfully before building starts:

guest@gnu ~/systole-playground/BrainlabMirror$ guix time-machine -C systoleos/channels.scm -- system image -t iso9660 systoleos/brainlabmirror.scm
ice-9/eval.scm:223:20: In procedure proc:
error: slicer-5.8: unbound variable
hint: Did you forget `(use-modules (guix-systole packages slicer))'?

Furthermore, when the operating-system definitions are set to depend on each other (one inherits the other, like in 23636a2), the builder cannot find the inherited module:

guest@gnu ~/systole-playground/BrainlabMirror$ time guix system image -L . -t iso9660 systoleos/brainlabmirror.scm
ice-9/eval.scm:223:20: In procedure proc:
error: systoleos-base: unbound variable
hint: Did you forget a `use-modules' form?

[RafaelPalomar]

I guess you could use as many -L flags as needed. Wouldn't this be enough?

[RafaelPalomar]

@oeramo, what's the status of this? Is it ready for review?