Bug 45456
| Summary: | REGRESSION (r67001): Scrolling by holding down a scroll arrow is much slower than in IE (or previous versions of Safari) | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Adam Roben (:aroben) <aroben> |
| Component: | Platform | Assignee: | Nobody <webkit-unassigned> |
| Status: | RESOLVED WORKSFORME | ||
| Severity: | Normal | CC: | pkasting |
| Priority: | P2 | Keywords: | InRadar, PlatformOnly, Regression |
| Version: | 528+ (Nightly build) | ||
| Hardware: | PC | ||
| OS: | Windows XP | ||
| URL: | https://bugs.webkit.org/attachment.cgi?id=66975&action=prettypatch | ||
Adam Roben (:aroben)
To reproduce:
1. Go to any long webpage (this one is good: <https://bugs.webkit.org/attachment.cgi?id=66975&action=prettypatch>)
2. Click and hold on the downward scroll arrow
The page scrolls far more slowly than it does in IE, or in Safari 5. This seems to be due to the new smooth scrolling behavior.
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Peter Kasting
I can't repro on Chromium/Win (which shares the scrolling implementation); for me the smooth scrolling build scrolls at exactly the same speed as the non-smooth-scrolling build. (Which is how it's designed; the scroll system still scrolls the same amount per key repeat and processes repeats at the same rate.)
Adam Roben (:aroben)
(In reply to comment #1)
> I can't repro on Chromium/Win (which shares the scrolling implementation); for me the smooth scrolling build scrolls at exactly the same speed as the non-smooth-scrolling build. (Which is how it's designed; the scroll system still scrolls the same amount per key repeat and processes repeats at the same rate.)
Can you reproduce using Safari 5 and a nightly build of WebKit?
Adam Roben (:aroben)
<rdar://problem/8411846>
Peter Kasting
(In reply to comment #2)
> (In reply to comment #1)
> > I can't repro on Chromium/Win (which shares the scrolling implementation); for me the smooth scrolling build scrolls at exactly the same speed as the non-smooth-scrolling build. (Which is how it's designed; the scroll system still scrolls the same amount per key repeat and processes repeats at the same rate.)
>
> Can you reproduce using Safari 5 and a nightly build of WebKit?
No.
I tested that page with Chromium w/ and w/o smooth scrolling as well as Safari 5 vs. last night's nightly (r67080, well past my smooth scrolling landing). With both the down arrow button on the bottom of the scrollbar as well as the down arrow key held down on my keyboard, each browser scrolled in the same time (within 1 sec. error due to my watch) whether smooth scrolling was present or absent.
Adam Roben (:aroben)
My bad. Comparing Safari 5 to ToT I see the same thing as Peter: scrolling from the top of the page to the bottom takes the same amount of time regardless of smooth scrolling.
I don't know what I was thinking before. Sorry for the false alarm! (And the smooth scrolling is pretty!)