Automatically install latest julia version

[mathbunnyru]

Describe your changes

• Automatically determine the latest stable (but non-LTS) version of Julia
• Julia 1.9.4 is gonna be installed after this PR

Issue ticket if applicable

Checklist (especially for first-time contributors)

• I have performed a self-review of my code
• If it is a core feature, I have added thorough tests
• I will try not to use force-push to make the review process easier for reviewers
• I have updated the documentation for significant changes

[mathbunnyru]

@yuvipanda @benz0li please take a look.
This should get rid of PRs which only update Julia version.

[mathbunnyru]

After this operation, 1,043 kB of additional disk space will be used.
jq is lightweight

[yuvipanda]

This looks good to me, but what do you think about rewriting this script in python instead? That would allow us to not use jq. I love jq too, but I always find complex bash gets out of hand fast.

[mathbunnyru]

This looks good to me, but what do you think about rewriting this script in python instead? That would allow us to not use jq. I love jq too, but I always find complex bash gets out of hand fast.

Thanks, I will rewrite this script in Python. I might need to use plumbum package though, it makes my life much easier when I need to call some utilities from Python.

[mathbunnyru]

@yuvipanda done. If you like the Python implementation, I will get rid of jq and setup-julia.sh.
Kept them for now, so you can compare.

[yuvipanda]

Yay for rewriting shell scripts in python :) I think that's the way to go, especially if we find ourselves parsing JSON.

Can you add even a simple docstring to all the functions? I find that helps tremendously to both clarity when reading as well as in keeping them well scoped during modification.

I understand the desire to use plumbum, but I find it pretty difficult to read and understand, especially because it's overriding array syntax for command execution. One of the big reasons for me for writing things in python over bash is that it makes it much easier for newer contributors to onboard and make contributions, and I think plumbum sort of turns python into a hybrid shell situation that makes it a little difficult. So I do have a strong preference for just using subprocess, especially when there's no piping or complex shell-like things involved. Here, we're only using it to call subprocesses and nothing else (as far as I can tell). So what do you think about dropping plumbum and just using subprocess, with the line being 'are we only calling a process, getting its output (optional)? If so, use subprocess, else can use plumbum'.

[mathbunnyru]

@yuvipanda done, please, take a look.

[yuvipanda]

yay this is awesome! There's one missing docstring, but otherwise this looks good to go.

Pleasure working with you, @mathbunnyru