Andreas Rossberg

I suggest that we don't conflate two separate considerations: (1) the API specification, and (2) API implementations. For implementations, I agree it makes perfect sense to provide a best-effort attempt...

Let's take a step back. Of course we can muse about simple checks like mutable globals on a case-by-case basis. But I think we should strive for something more principled....

@alexcrichton, I didn't mean to discount what you said. Rather, I am trying to get clarity on what we are talking about. In particular, I'd like to clarify that most...

@sunfishcode, a specification makes a normative requirement. If we agree that some things are impractical on some engines, then the "to the extend that they can" qualifier means that it...

@alexcrichton, @sunfishcode, well, I think we all have our part in not properly listening. ;) Of course, I'm happy to discuss further, as I don't think it's so much the...

@lukewagner, that's an interesting way to phrase it. FWIW, almost none of the suggested checks are expressible in terms of the API itself, because they either are concerned with meta-level...

@alexcrichton: > This honestly is a baffling statement to me in the sense that it sounds like you're saying that my arguments here follow no principles Sorry if it came...

Yeah, this is one of the issues I mentioned when we discussed the changes to references. To be honest, I still don't know how to address that properly. Conflating the...

Note that this adapter implementation is not part of the proposal but only of the prototype. The only reason I did it in that direction was that it was somewhat...

The global not being mutable qualifies as a type error as well. Both variants of error should never occur in a correct program. Do you have any specific use case...