[webkit-reviews] review requested: [Bug 136433] REGRESSION (r172832): Poor 2-finger scrolling performance at theverge.com articles (all tiles repaint) : [Attachment 237607] Patch

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Wed Sep 3 20:10:22 PDT 2014


Beth Dakin <bdakin at apple.com> has asked  for review:
Bug 136433: REGRESSION (r172832): Poor 2-finger scrolling performance at
theverge.com articles (all tiles repaint)
https://bugs.webkit.org/show_bug.cgi?id=136433

Attachment 237607: Patch
https://bugs.webkit.org/attachment.cgi?id=237607&action=review

------- Additional Comments from Beth Dakin <bdakin at apple.com>
A few notes about this patch. I took Tim's suggestion, and tried to use Brent's
momentum events to create a test, and I kind of succeeded. The test seems to
work on my machine. I am just dubious because I am not 100% convinced that
these momentum events will do the exact same thing on all computers. So I put a
note in the test about how if it turns out to be flakey, we can skip it. But!
It is effective on my machine.

Another note. Sitting with Simon and Sam about this bug, they also want
ScrollableArea::handleWheelEvent() to return early is the ScrollableArea is not
scrollable or rubberbandable. I am going to implement that in a followup patch
because adding a virtual function to ScrollableArea called
isScrollableOrRubberbandable() turns out to be a bit complicated for
RenderListBox since there is already a virtual function of that same name in
its RenderObject inheritance chain. Ah, the joys of multiple inheritance.
Anyway, there are obviously solutions to that problem, but it seems involved
enough that it should be a separate patch.


More information about the webkit-reviews mailing list