(no longer active) Reporter
	Description • 10 years ago

Boris mentioned that these currently generate code which access the pref service from the worker.  We should explicitly generate a WebIDL compiler error for now before someone starts to need it without realizing the implications.

	Boris Zbarsky [:bzbarsky]
	Comment 1 • 10 years ago

The problem is the codegen doesn't know what's exposed to workers....

	Kyle Huey (Exited; not receiving bugmail, old account, do not use)
	Comment 2 • 10 years ago

Yeah, this is simply not possible today :(

We do need to move the information about which interfaces are visible in workers into codegen eventually to generate RegisterBindings.cpp.  Perhaps we should do that now.

	(no longer active) Reporter
	Comment 3 • 10 years ago

(In reply to comment #1)
> The problem is the codegen doesn't know what's exposed to workers....

Hmm, can't it check [ExposedTo] from the AST?

	Kyle Huey (Exited; not receiving bugmail, old account, do not use)
	Comment 4 • 10 years ago

(In reply to :Ehsan Akhgari (lagging on bugmail, needinfo? me!) from comment #3)
> (In reply to comment #1)
> > The problem is the codegen doesn't know what's exposed to workers....
> 
> Hmm, can't it check [ExposedTo] from the AST?

What is [ExposedTo]?

	Boris Zbarsky [:bzbarsky]
	Comment 5 • 10 years ago

> What is [ExposedTo]?

Ehsan mean [Exposed], from <http://heycam.github.io/webidl/#Exposed>: A pretty new addition to the spec.

But yes, we should implement [Exposed].  We can also use it to do things like skip codegen for some things on workers.

That said, I think the current spec for [Exposed] is wrong in an obvious way.  See https://www.w3.org/Bugs/Public/show_bug.cgi?id=25495

	(no longer active) Reporter
	Comment 6 • 10 years ago

(In reply to Boris Zbarsky [:bz] from comment #5)
> > What is [ExposedTo]?
> 
> Ehsan mean [Exposed], from <http://heycam.github.io/webidl/#Exposed>: A
> pretty new addition to the spec.
> 
> But yes, we should implement [Exposed].  We can also use it to do things
> like skip codegen for some things on workers.

Ah, yeah, sorry I didn't realize that we haven't implemented that yet.  And I got the name wrong too!!

> That said, I think the current spec for [Exposed] is wrong in an obvious
> way.  See https://www.w3.org/Bugs/Public/show_bug.cgi?id=25495

Can we somehow help improve the situation there?  I think supporting [Exposed] would be beneficial to the readability of our WebIDL files as well.

	Boris Zbarsky [:bzbarsky]
	Comment 7 • 10 years ago

I filed bug 1017988 on implementing [Exposed].

> Can we somehow help improve the situation there? 

https://github.com/heycam/webidl/pull/12

	(no longer active) Reporter
	Comment 8 • 10 years ago

(In reply to comment #7)
> I filed bug 1017988 on implementing [Exposed].
> 
> > Can we somehow help improve the situation there? 
> 
> https://github.com/heycam/webidl/pull/12

Thanks!

	(no longer active) Reporter
	Comment 9 • 10 years ago

Is this possible to fix now?

	Boris Zbarsky [:bzbarsky]
	Comment 10 • 10 years ago

Bug 1017988 part 10 fixes it.