Also, I expect this needs a backport to the 3.8 branch. Please confirm.

Also, thanks again for working on this! 😄

Also, I expect this needs a backport to the 3.8 branch. Please confirm.

Yep, I added the needs backport to 3.8 label when I first opened the PR. If there are any merge conflicts (and @miss-islington is unable to resolve them), I'll sort it out manually with cherry pick.

I did leave a couple more formatting-related comments. Sorry for dragging this out. My intention is to be instructive rather than proscriptive. (If you don't find this helpful then please tell me.)

They were definitely quite helpful! Although I have some experience contributing to standard library code changes, it's rather uncommon to encounter situations like this where multi-line arguments are necessary. But, it's great to know for future reference how to format it in a way that's PEP8 compliant and highly readable.

Also, thanks again for working on this!

No problem, thanks for the review! (:

After this is merged, I'll very likely do more investigation into the root cause of the running failure. It can a rather time consuming process to dive into the internals and experiment with different potential solutions. But I've recently found a strong interest in the different implementations of concurrency in the standard library, including asyncio, concurrent.futures, multiprocessing, and most certainly with your upcoming subinterpreters module.

Not to go off on too far off-topic of the PR, but I think CPython has a lot of room for improvement and increased relevance in the world of concurrent programming and building scalable, distributed systems. Many people still unfortunately view Python as "best suited for simplistic, single-threaded applications", despite it's success with concurrent network IO in asyncio. As a result, I'd like to invest as much time as I reasonably can in improving our core models of concurrency, so that CPython can be more competitive with (or even supersede) other ecosystems in that domain.

Thanks for all of your hard work with the subinterpreters module! I'm very excited about the possibilities it will pave the way for with it's drastically increased isolation and potential performance implications for CPU-bound operations.

Thanks @aeros for the PR 🌮🎉.. I'm working now to backport this PR to: 3.8.
🐍🍒⛏🤖 I'm not a witch! I'm not a witch!

GH-18318 is a backport of this pull request to the 3.8 branch.