Is this going to succeed?

[letalvoj]

What are the chances this will end up in the upstream? If it is, any idea which major version might contain this change set?

[stonebig]

this should be a safe bet hint it shall land in Python-3.12 https://bugs.python.org/issue46657

[0xaead]

The point is "10 years ago", then another 10 years, and so on...

10 years is a long time. 20 years is a long long time.

And then, the world changed a lot. Look Ruby, Go, etc. (The performance of Ruby has basically caught up with Python.) (A lot of Python has been replaced by Go because of performance and deployment issues.)

Hey! If you really are concerned about the success of this project, why not simply join and provide some contribution? Any thing from test on your local env, filling an issue with logs, to giving a/some pull request is warmly welcomed. :)

[vovavili]

The point is "10 years ago", then another 10 years, and so on...
10 years is a long time. 20 years is a long long time.
And then, the world changed a lot. Look Ruby, Go, etc. (The performance of Ruby has basically caught up with Python.) (A lot of Python has been replaced by Go because of performance and deployment issues.)

Hey! If you really are concerned about the success of this project, why not simply join and provide some contribution? Any thing from test on your local env, filling an issue with logs, to giving a/some pull request is warmly welcomed. :)

I feel like an absolute majority of Python users who really want this project to succeed possess nowhere near enough of technical expertise to help this project in any meaningful way. I consider myself to be above average in Python, and just trying to understand how no-gil works under the hood requires me to re-read the description really slowly, I wouldn't even try to pretend to understand the source code.

[0xaead]

The point is "10 years ago", then another 10 years, and so on...
10 years is a long time. 20 years is a long long time.
And then, the world changed a lot. Look Ruby, Go, etc. (The performance of Ruby has basically caught up with Python.) (A lot of Python has been replaced by Go because of performance and deployment issues.)

Hey! If you really are concerned about the success of this project, why not simply join and provide some contribution? Any thing from test on your local env, filling an issue with logs, to giving a/some pull request is warmly welcomed. :)

I feel like an absolute majority of Python users who really want this project to succeed possess nowhere near enough of technical expertise to help this project in any meaningful way. I consider myself to be above average in Python, and just trying to understand how no-gil works under the hood requires me to re-read the description really slowly, I wouldn't even try to pretend to understand the source code.

imho, complaining, blaming, whining or any other negative emotional feedback to this project won't help much. If you really wish this project heading to success, why not do something? E.g. If you think the documentation of the project is hard to understand, maybe you can give some inline comments, or ask questions and so on to make it better?

Disclosure:::: I am not contributor nor in any way being related to people in this project. I am just an evangelist believing multi-threading makes Python more powerful, which may or may not be fueled by this project. Bottom line, this project is at least an attempt to make Python better. I wonder if criticizing/downvoting the worker of the project, or even the passer-by like me would help...

[guilt]

First of all, I'm mostly a CPython user as-is (but keep looking at these diffs time to time)

Okay, so I know that they posted the 3.13 faster CPython roadmap recently.

How does the rest of the PEPs look like? Any way some of these patch sets be shaped for say 3.14, 3.15?

• Immortal Objects
• Per-Thread Method Cache
• Mutexes and Notifications
• PyThread State Swap
• Attached vs Detached State

Seen in #17

[colesbury]

@guilt, I'm not sure I understand your questions. The PEP is here: https://peps.python.org/pep-0703/ under consideration for Python 3.13. It's being discussed on discuss.python.org, with specific links in the PEP. It's about 15,000 LOC + mimalloc (which is also ~15,000 LOC). There's some information about what the PRs would look like here: https://github.com/colesbury/nogil-3.12/wiki/PR-plan-for-PEP-703.