smaug----

It is unclear if [AllowShared] BufferSource is valid

3

The spec says "A type that is not a buffer source type must not be associated with the [AllowShared] extended attribute." and "The buffer source types are ArrayBuffer, DataView, and...

'process a navigate fetch' doesn't load anything ever?

1

https://html.spec.whatwg.org/#process-a-navigate-fetch 14.5.3 waits for some random(?) task to 'process response'. But 'process response' links just to fetch itself being an option algorithm, and it wasn't provided when fetch was called.

[popup] Consider to dispatch show event as the first step of showPopup

2

Currently there is 1. Move the pop-up to the top layer, and remove the UA display:none style. 2. Fire the show event, synchronously. I'd expect show event to be cancelable...

[popup] It is error prone to have both async and sync 'hide' event

3

Mixing sync and async event dispatch is error prone. If one removes a popup from dom (that triggers async hide) and then adds it back and opens the popup, there...

Consider adding localNameFilter to MutationObserver

21

I was looking at a slow (cross) browser addon code, and one thing they need to do is to figure out when new input elements have been added to the...

[Highlight Events] It is very unsual to conditionally use different interfaces but same event type for events

1

HighlightPointerEvent extends PointerEvent and the explainer says: "When a user performs a pointer action on part of an element containing a highlighted Range, a HighlightPointerEvent object must be fired" In...

[Highlight Events] Because of the super weird event propagation, it is unclear how event.eventPhase works with the events

1

If event.target stays as the element, and when currentTarget == target, that is supposed to be AT_TARGET phase. But since propagation goes through Highlight, the examples seem to hint that...

[animation-frame-rate] attribute (FramesPerSecond or AnimationFrameRatePreset) frameRate; is surprising

2

Having an attribute which value is either a string or number (from JS point of view) is rather surprising.

FileSystemDirectoryHandle.resolve is a very vague method name, consider renaming it

3

"resolve" doesn't hint too much what it is actually doing. One needs to read the algorithm to see what it actually returns.

Should browser UI be able to activate fullscreen even if iframe is missing allowfullscreen

6

...and if so, should fullscreenchange event fire? https://bugzilla.mozilla.org/show_bug.cgi?id=1302976#c8 (I think xidorn didn't mean sandboxed document but iframe without allowfullscreen)