Re: Expected behavior when releasing a button after pointer capture

On Mon, Apr 27, 2015 at 3:06 PM, Scott González <scott.gonzalez@gmail.com>
wrote:

> On Mon, Apr 27, 2015 at 1:47 PM, Rick Byers <rbyers@chromium.org> wrote:
>
>> FWIW, I'd say it's a bug not to send pointerleave to an element the mouse
>> is no longer on after release of capture (including by button up).  It
>> would be wrong for a framework to consider such elements still hovered.
>>
>
> You get the pointerout/pointerleave events as you cross the boundary since
> those are still fired on the capturing element. There's just no
> accompanying pointerover/pointerenter events, even after releasing the
> button. This is the behavior that I think should change. Once capture is
> lost, if the pointer is over another element, I think there should be
> events to indicate that.
>

Oh right, I was thinking of my proposed design where we wouldn't send
leave/out during capture (to avoid the hit testing) <grin>.  Yes I agree
there should be an enter/over too (it's just less obviously a bug the way I
was thinking a missing leave/out event would be).

The question of what should happen when content appears underneath is very
>> similar to the question of what should happen when the content scrolls.  WebKit
>> has just changed their behavior here
>> <https://bugs.webkit.org/show_bug.cgi?id=99940>, and we're planning on
>> changing blink
>> <https://code.google.com/p/chromium/issues/detail?id=317007> - probably
>> to something similar to Firefox
>> <https://code.google.com/p/chromium/issues/detail?id=317007#c26> (or
>> possibly WebKit).
>>
>
> That sounds good. I've been waiting a few years for browsers to be
> consistent on this.
>