[Webkit-unassigned] [Bug 197100] [GTK] Slow scrolling (not matching GTK native scroll amount)

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=197100

--- Comment #2 from Chris Lord <clord at igalia.com> ---
(In reply to Alexander Mikhaylenko from comment #1)
> More info:
> 
> - This doesn't affect touchpad/per-pixel scrolling
> - This happens regardless of async scrolling or not
> - Happens with keyboard too
> 
> An easy reproducer - hold pgdown on https://planet.webkitgtk.org/
> 
> My theory is - it's related to smooth scrolling, when you have quickly
> incoming scroll events, the next event cancels animation from the last one
> and triggers a new animation without compensating for the fact the last
> animation stopped early. So if you press pgdown repeatedly but wait for the
> animation to finish each time, it's a lot faster than when you hold it, BUT
> the first (because key repeat delay) and last animations finish and so it's
> visible how the beginning and end of the scrolling is faster.
> 
> That happens in trunk though - 2.33.91 (sorry, 2.34.0 is not in the flatpak
> platform yet) seems to do something different and pgdown is too fast instead
> (while mouse scrolling is same as in trunk) - no idea here.

Testing on master, I don't see this problem - behaviour between smooth and non-smooth in terms of the distance scrolled per mouse-wheel ticks in very quick succession (so they interrupt the animation) is the same.

However, I do see the interrupted smooth scrolling issue with keyboard scrolling (which is a different code-path, so doesn't surprise me). Could you possibly test again using minibrowser from a master checkout and update here what results you get and what you expect?

For reference, here is the code that stops this from happening with smooth scrolling from mouse-wheel events (and it's shared between sync and async paths): https://github.com/WebKit/WebKit/blob/a7eb93bd9b7414f627d8c7038a7163f3cc3945fd/Source/WebCore/platform/ScrollingEffectsController.cpp#L344

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20211011/eb53004f/attachment-0001.htm>