Closed zbraniecki closed 7 years ago
Unless we get
Significant in-the-field experience with shipping implementations, such as that provided by two independent VMs
by next week, then no chance.
@domenic If I understand correctly:
What else is needed? I asked the SM crew about this feature riding trains and the response from @syg was that we're waiting for it to get to stage 4, so chicken-egg problem?
Neither engine has shipped to stable. It needs to ship to stable in two engines so that we get significant in-the-field experience. If Firefox is not willing to be one of those engines, then we'll have to wait for another.
@arai-a - can we put it on the trains to ride? :)
after https://bugzilla.mozilla.org/show_bug.cgi?id=1379525 gets fixed (and if it passes all existing tests), we could. but not sure about the policy.
Yep, this hasn't landed in V8 yet either, see https://chromium-review.googlesource.com/c/523844/ . I think it'd be best to wait for Stage 4 until we at least have two implementations of this (regardless of your stance on what qualifies as "in-the-field experience").
cc @caitp
That'll be great if "significant in-the-field experience" will be defined in process document because this question pops up multiple times in various repos
I think the rest of the sentence ("shipping implementations", "two independent VMs") is very clear.
Does this mean that object rest / spread should be promoted to stage 4? Both V8 6.0 and SM from Firefox 55 Beta are stable VMs.
Sorry for off topic
I don't know.
@chicoxyzzy that's a question for that repo and that proposal's champion; and the champion would need to bring it up with the committee.
@ljharb what do you mean? why is it a question for that repo and proposal's champion? i asked that question for object rest spread proposal because i wanted to know what is left. but this Significant in-the-field experience
is very vague. How much significant is exactly? it definitely means 2 different VMs, and being shipped into stable distribution channels. But is just shipped enough? or it needs to be be available for a while, like 2 months, or more?
@iamstarkov this repo is for async iteration. Questions that are unrelated to async iteration don't belong in this repo.
As to making the definition more concrete, (if I recall correctly) I believe that there isn't consensus among committee members about what constitutes a more concrete definition, so therefore it's intentionally left vague. "Enough" is going to depend on the feature, and on the risk of breakage as subjectively determined by the committee.
It appears that there's nothing more to be discussed with this issue with regard to async iteration, and it's turned into some kind of weird off-topic support channel, so I'll close and lock this.
With two implementations (SpiderMonkey and V8), spec issues resolved and quite a number of test262 [0] what's blocking stage 4 request now?
@domenic - any chance to request it at the next week's meeting?
[0] https://github.com/tc39/test262/tree/master/test/language/statements/async-generator