Current status of true async RFC?

[dmitryuk]

Can somebody tell the current status of accepting RFCs? I see this, but poll is closed without any voters

[EdmondDantes]

Can somebody tell the current status of accepting RFCs? I see this, but poll is closed without any voters

Hello.
At the moment, the RFC process has been paused.
It is difficult to say what will happen with the RFC next.

The process has reached a deadlock situation. Essentially, the RFC flow itself does not suit either the PHP community or this project. Therefore, the next efforts will be focused on changing this.

[lucaswitch]

This is the most important feature in PHP, i would say... in the last 10~15y. Not only we have several use cases for it but it's important for the language to keep up with other languages just like you mention on the rfc.
It's almost a weekness of php not having such feature at this point. As users we're really anxious at this point for this feature.

Projects that been growing outside the web ecossystem like https://github.com/nativephp(which i didn't try yet) would hugely benefit from these features as the nature of native applications are to have async features a lot.

As part of this immense community i'm really proud where it already got, wanna see that thing ready and launched.

[EdmondDantes]

Thank you for the feedback. The project is continuing to evolve. I’m personally curious how far it can go.

[macropay-solutions]

@EdmondDantes As mentioned in this comment https://github.com/orgs/true-async/discussions/9#discussioncomment-16038676, php can't be true async without huge breaking changes to avoid structural issues.

Maybe an async-php fork would be a solution (different from TS/NTS), with its own rules, without finally, strtok and other things that share global state.

But...there are already other languages that are async out there, why strugle to learn the "new" PSP's rules?

[macropay-solutions]

Backward compatibility.

[EdmondDantes]

Backward compatibility.

Yes, I know that BC stands for Backward Compatibility.
I don’t understand how BC is related to this discussion, as I don’t see where a BC issue arises.

[macropay-solutions]

Your RFC stated that old code will continue to work as before and those examples with finally and strtok are impacting it, while being linked to the status of this RFC, no?

[EdmondDantes]

Your RFC stated that old code will continue to work as before and those examples with finally and strtok are impacting it, while being linked to the status of this RFC, no?

For TrueAsync, the general rule applies: if the code is synchronous, it will behave the same as before.
If a class is used only within a single coroutine, then the old code will continue to work as expected.

[macropay-solutions]

Thank you for clarifying.

[EdmondDantes]

This is the current position on all related matters.
https://github.com/orgs/true-async/discussions/28