Re: [webidl] Add a [SecureContext] operator attribute (#65)

> Does that make sense?

It does, but I would prefer we do it after the "a security check" bit.  The reason for that is that at least in Gecko's bindings the "security check" bit and "is the this value of the right type" bit is actually a single more or less atomic check on some state of the this value, so inserting steps between them is actually quite difficult without causing performance regressions.

> Those ought to be renamed to "[stack of] incumbent settings object", right?

Yeah, looks like the HTML spec mutated out from under IDL here.

> Should I leave the slimmed-down descriptive text in the [SecureContext] section?

That's probably fine as long as it's explicitly marked informative; the thing I really want to avoid is accidental conflicts, real or perceived, between normative descriptive text and the normative bits in es-operation.  Making the descriptive text non-normative avoids that problem, and I agree that it's helpful to a casual reader.

---
Reply to this email directly or view it on GitHub:
https://github.com/heycam/webidl/pull/65#issuecomment-148712632