When will I be able to use Temporal in production?

[marikaner]

Temporal seems to be what the standard was lacking for a long time and I am very interested in replacing our current implementation that uses moment with Temporal. I assume that once the proposal has reached stage 4 it will be safe to use the polyfills in production, but I couldn't find any indication of when (very roughly) this might be the case. I hope it is ok to ask this question here: Is it realistic that stage 4 will be reached next year?

[ptomato]

Hi, thanks for the inquiry! I cannot give any guarantees or say for sure, because the answer depends partly on how the proposal progresses through the TC39 committee meetings, and how browser implementers prioritize it, both of which are largely outside of the Temporal working group's control. That said, we are on the verge of freezing the proposal in order to give it to committee members for review, which we hope is the last step before reaching Stage 3.

My personal opinion (I'm not speaking for TC39 or my employer Igalia) is that it's not unrealistic that Stage 4 would be reached in 2021.

[ljharb]

In general, proposals are safe to use in production via a fully spec-compliant polyfill/shim in stage 3, since that’s also when browsers ship.

[marikaner]

Thanks a lot for this clarification!

[ptomato]

I hope that answered the question!

[marikaner]

ouf, we're a bit off here... any new thoughts?

[ptomato]

Indeed, 2021 was too optimistic.

Currently, you can follow an overview of the remaining work at #2628.

[cromefire]

Now that everything is done there, is there any guess as to when it might become an official API? I'm eagerly waiting to add support for Temporal to some frameworks / libraries...

[ptomato]

At this point it is up to the browsers and other engines to implement it. I'm guessing that may happen in anywhere from 3 to 9 months, assuming that they do not find any further defects in the proposal during that process.

How fast the implementation goes is up to those respective teams to prioritize, and it's outside of TC39's control, but I will continue to publish graphs every so often such as this one: https://ptomato.name/talks/tc39-2024-07/#4
(Note that test conformance is only a rough measure of how much work there is left to do, and should not be interpreted as "% DONE", but it gives a general idea.)

Once we have 2 working implementations, the proposal can advance to stage 4 at which point it becomes part of the published standard. Then it'll be up to you to consult resources like https://caniuse.com/?search=Temporal and decide on what percentage of availability is enough to start adding Temporal support to your codebase(s).