Richard Gibson

Since this issue is getting hot, I think it's worth asking if all of those parameters _are_ to be treated as public interface... in short, do we feel comfortable with...

That works, but also fails the convenience criterion in many cases because the outside-in traversal order is backwards from defaults (and again, is not necessary for other similar events).

The prototype isn't null, it's `Object.prototype`.

What's special about arguments objects? %Object.prototype% is used as the prototype for many other built-in objects as well: * Property descriptors from Reflect.getOwnPropertyDescriptor and Object.getOwnPropertyDescriptor{,s} (and also the object from...

@jmdyck Thanks, fixed all three issues.

I agree, but that may be significantly more difficult (at least with its current organization).

Can this also resolve #634 via https://github.com/tc39/ecmarkup/pull/481 ?

I just submitted gh-3487, which should correct `.offset()` and `.position()` behavior in browsers implementing https://github.com/w3c/csswg-drafts/commit/180b348a1ac0931043cd195651fc5164463d2bce (although we should still add new tests for _this_ issue), but I don't want to...

> Looks like your solution is reliant on getBoundingClientRect -- have you accounted for the way Chrome's return value for this API changes when pinch-to-zoomed? (See also https://bugs.chromium.org/p/chromium/issues/detail?id=489206) > Also,...

Hmm, after exploring live examples, I see what you mean. I guess I'd like the other PR to land, then we can update `.offsetParent()` to match the spec and add...