Short-term prioritization of PRs/issues

[MilesCranmer]

I think some PRs/issues are quite important and should be merged/fixed soon after #508 is done. Here is a ranking of what I think the most important PRs are:

1. High importance:

• Fix failure setting optimize option (#480) #481
• support for "threads" and "min-optlevel" julia options #493
• Turn on compiled_modules=False automatically? #487

2. Simple fixes/typos (could be merged quickly)

• Fix broken link to platform-specific instructions #489
• Fix dead link in #479 #486
• Have sind in docs #412

3. Mid importance

• allow building sysimage with custom PyCall #422
• Better error messages from python-jl #300
• Call jl_sigatomic_begin at the end of init_julia block #204
• Python 3.11 needs find_spec implemented in JuliaImporter #509

4. Other/niche issues

• add unexported variables to module's __dir__ #213
• Warning #383
• allow nested variable interpolation #381
• Added user site bin for julia-py #312
• Fix julia library jl_init_with_image bug #440
• try PYTHONPATH in julia_py_executable to find julia-py #445

cc @mkitti @tkf @marius311

Feel free to suggest any I missed. Once #508 is merged I think the high importance ones should be merged relatively soonish.

[mkitti]

Would you be interested in joining the JuliaPy Github organization?

[MilesCranmer]

Yeah sure why not!

[mkitti]

@ararslan , could you invite Miles to JuliaPy?

[marius311]

Thanks for injecting some life here. My 2c (which can be readily ignored, since I have no free time to devote atm 😢) is that the top priority should be removing roadblocks for a non-Julia user that wants to quickly use some Julia code, or to use a Python package which is using some Julia code under the hood. To that end, happy to see #487 at the top of the list, as, from the experience of setting up several users, the experience is sadly really really bad. You might also consider #496 instead, but I think that's harder, so probably best to focus on the low-hanging fruit first.

[MilesCranmer]

No problem. I have zero time to work on new features unfortunately, but I think helping review/merge maintenance PRs should be ok.