Bug 112967 – Touch dragging should not send fake mouse down event as webapps may misinterpret it

• Status: RESOLVED
• Resolution: WONTFIX
• Component: New Bugs

[Varun Jain]

Touch dragging should not send fake mouse down event as webapps may misinterpret it

[Varun Jain]

Created attachment 194345 [details]
Patch

[Antonio Gomes]

Comment on attachment 194345 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=194345&action=review

> Source/WebCore/ChangeLog:8
> +
> +        Reviewed by NOBODY (OOPS!).
> +
> +        Tests: Existing tests in fast/events/touch/gesture/logn-press* cover this functionality.

I would like here a description of your solution. is there any real world webpage affected? why can not it be reduced to a layout test so it does not regress?

[Varun Jain]

Created attachment 195376 [details]
Patch

[Varun Jain]

(In reply to comment #2)
> (From update of attachment 194345 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=194345&action=review
> 
> > Source/WebCore/ChangeLog:8
> > +
> > +        Reviewed by NOBODY (OOPS!).
> > +
> > +        Tests: Existing tests in fast/events/touch/gesture/logn-press* cover this functionality.
> 
> I would like here a description of your solution. is there any real world webpage affected? why can not it be reduced to a layout test so it does not regress?

Done. There is no real world webpage that I know of. But you can image that a page listening to mouse down events may mess up its state on receiving the fake mouse down. This fake mouse done is wrong because we already send a mouse down when we receive a touch down.

[Antonio Gomes]

Comment on attachment 195376 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=195376&action=review

> Source/WebCore/page/EventHandler.cpp:2612
> +        if (subframe && subframe->eventHandler()->handleGestureLongPress(gestureEvent))
> +            return true;
> +        m_mouseDownMayStartDrag = true;
> +        dragState().m_dragSrc = 0;
> +        TemporaryChange<IntPoint> mouseDown(m_mouseDownPos, m_frame->view()->windowToContents(mouseDragEvent.position()));

Please add a test for this change as well.

[Varun Jain]

moving to https://code.google.com/p/chromium/issues/detail?id=168165