Re: [heycam/webidl] Modernize invoking user code (#113)

Ah, I had missed the "if the stack is now empty" thing.

I think the difference in behavior between the dispatch from click and dispatch from script is pretty weird.  In particular, it means that if a capturing listener grabs the event and redispatches it you get different behavior than if you just allow the event to propagate.... :(

Relying on "stack is empty" also involves other problems if a UA doesn't want `alert` and company to completely block the UI (and yes, I know the spec model of alert is that it blocks everything; this is user-hostile and broken): suddenly your microtasks stop running correctly while the alert is up.

It seems like we need to sit down and figure out how this stuff should actually work in the real world of browsers, which doesn't match the simplistic world of the spec and will keep diverging from it (e.g. iOS Safari recently moved to more-nonblocking alerts), as I understand.

---
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/heycam/webidl/pull/113#issuecomment-220452981