Demix NavigatorConcurrentHardware

[foolip]

Part of #8929

[foolip]

This is simply splitting the entry without changes. I checked the data against mdn-bcd-collector results and didn't find any differences for Worker. For WorkerNavigator there was a proposed change, but it doesn't look legit, I filed foolip/mdn-bcd-collector#960.

[foolip]

@Elchi3 changes like this will require a lot of fixes on MDN. I'd like to not volunteer for that, but is it useful for me to send PRs like this one?

[Elchi3]

@Elchi3 changes like this will require a lot of fixes on MDN. I'd like to not volunteer for that, but is it useful for me to send PRs like this one?

I'm trying to burn through all mixins on BCD and on MDN. Any help is useful :)

[foolip]

Before this is merged I want to note that WorkerNavigator and WorkerLocation are odd cases and it's not obvious how MDN/BCD should represent them. These interfaces predate the use of [Exposed=(Window,Worker)] and are subset copies of Navigator and Location.

The thing is that even though these interfaces are real, web developers just use navigator.language or location.href in both window and worker scripts. I'm skeptical that it makes sense to document them separately, and am unsure if spitting them on BCD is the right thing.

[foolip]

Or rather I should say I could see us changing this later to use the same mechanism for representing global scope exposure as for interfaces and global methods, assuming a single method for that emerges after WindowOrWorkerGlobalScope demixing.

[Elchi3]

hm, I see. Can we file a separate issue that discusses options for the whole Window/Worker problem with mixins? We can wait to merge this PR until we have had that discussion if you like.

[foolip]

I've filed #9125 for this specific issue, and #9127 for WindowOrWorkerGlobalScope.

[foolip]

I've rebased this now but I don't like it. It's a similar issue to demixing WindowOrWorkerGlobalScope, where web developers can use the same code in windows and workers, and splitting it into two entries seems unhelpful. Finding a way to represent this as one entry meaning navigator.hardwareConcurrency and using notes or partial_implementation if it doesn't work in workers seems better.

[Elchi3]

Thanks Philip! I propose we sort out all other mixins first and then look at everything that touches the window and worker problematic.

[Elchi3]

This was done in #11413