Bugzilla

(In reply to Alex Jakobi from comment #5)
> I re-tested on Android with Chrome and Firefox, this time **without** the initial scroll. The behavior is identical to the one **with** the initial scroll you see in the screen captures.

Thanks for checking!

I think this tells us that, while `ArePointerEventsConsumable` not checking for `overscroll-behavior` may technically be incorrect from the point of view of the internal API, it doesn't actually lead to any user-visible misbehaviour. That is, while it may return `mHasRoom=true` for a subframe with `overscroll-behavior: none` where the conceptually more correct value would be `mHasRoom=false`, the consequence of this is that pull-to-refresh is **not** triggered, and we don't want it to be triggered in this scenario anyways.

It's probably not worth making a change here given that it's not causing a user-visible bug.

> Regarding the dynamic toolbar behavior, I also agree with you: There should be no movement. Should we address it within the scope of this ticket or create a new one?

I think we can re-purpose this bug to fix the dynamic toolbar behaviour. I'll leave it up to you whether to investigate that first or bug 1806218, since it may be unrelated to `ArePointerEventsConsumable`.

(In reply to Alex Jakobi from comment #5)
> I re-tested on Android with Chrome and Firefox, this time **without** the initial scroll. The behavior is identical to the one **with** the initial scroll you see in the screen captures.

Thanks for checking!

I think this tells us that, while `ArePointerEventsConsumable` not checking for `overscroll-behavior` may technically be incorrect from the point of view of the internal API, it doesn't actually lead to any user-visible misbehaviour. That is, while it may return `mHasRoom=true` for a subframe with `overscroll-behavior: none` where the conceptually more correct value would be `mHasRoom=false`, the consequence of this is that pull-to-refresh is **not** triggered, and we don't want it to be triggered in this scenario anyways.

It's probably not worth making a change here given that it's not causing a user-visible bug.

> Regarding the dynamic toolbar behavior, I also agree with you: There should be no movement. Should we address it within the scope of this ticket or create a new one?

I think we can re-purpose this bug to fix the dynamic toolbar behaviour. I'll leave it up to you whether to investigate that first or bug 1806218, since the dynamic toolbar issue may be unrelated to `ArePointerEventsConsumable`.