1952402 - `pointermove` for the scond touch won't be fired smoothly once the preceding `pointerdown`s are consumed

Status: NEW

(Core :: Panning and Zooming, defect, P3)

Description

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(sickish, maybe slower respose)]

STR

1. Load https://codepen.io/mariusgundersen/pen/xxrRoxp on a device which supports touch input
2. Check the checkbox labeled "e.preventDefault()"
3. Touch with 2 fingers and move the fingers to draw 2 lines

AR

2nd touch does not draw the line smoothly, but sometimes when I stop 2nd touch temporarily, suddenly the line is drawn.

ER

both lines are drawn smoothly

Notes

I checked the code (not debugged actually).

First, when pointerdown for the first touch is canceled, PointerEventHandler::PostHandlePointerEventsPreventDefault prevents default of the soucre event, probably eTouchStart.

Next, APZEventState::ProcessTouchEvent stores that.

Then, CancelableBlockState::ShouldDropEvents() starts returning true.

Finally, it changes the behavior of eTouchMove at multi-touch here.

Ikezoe-san, could you take a look?

Comment 1

[Hiroyuki Ikezoe (:hiro)]

Is this a blocker for interop-2025?

Comment 2

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(sickish, maybe slower respose)]

(In reply to Hiroyuki Ikezoe (:hiro) from comment #1)

Is this a blocker for interop-2025?

I think currently "No", but looks like really annoying bug, and potentially this can cause serious web-compat issue.

Comment 3

[Hiroyuki Ikezoe (:hiro)]

I added this bug into the agenda of the next team meeting.

Comment 4

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(sickish, maybe slower respose)]

I guess here should check aEvent.DefaultPreventedByContent() instead of aContentResponse. I think the name of aContentResponse is not correct.

Comment 5

[Botond Ballo [:botond]]

Tracking in FFXP-2522 (APZ Correctness Sprint).