WebKit Bugzilla
New
Browse
Log In
×
Sign in with GitHub
or
Remember my login
Create Account
·
Forgot Password
Forgotten password account recovery
RESOLVED FIXED
103278
[chromium] Touchscreen fling handling
https://bugs.webkit.org/show_bug.cgi?id=103278
Summary
[chromium] Touchscreen fling handling
Alexandre Elias
Reported
2012-11-26 12:13:23 PST
[chromium] Simplify fling handling
Attachments
Patch
(8.91 KB, patch)
2012-11-26 12:24 PST
,
Alexandre Elias
no flags
Details
Formatted Diff
Diff
Patch
(10.21 KB, patch)
2012-11-26 15:32 PST
,
Alexandre Elias
no flags
Details
Formatted Diff
Diff
Patch
(22.82 KB, patch)
2012-11-26 17:23 PST
,
Alexandre Elias
no flags
Details
Formatted Diff
Diff
Show Obsolete
(2)
View All
Add attachment
proposed patch, testcase, etc.
Alexandre Elias
Comment 1
2012-11-26 12:24:07 PST
Created
attachment 176045
[details]
Patch
James Robinson
Comment 2
2012-11-26 12:39:25 PST
On ChromeOS we want the synthetic wheel behavior. On Android you do not, but that doesn't mean you can just delete the other path.
James Robinson
Comment 3
2012-11-26 12:41:20 PST
Comment on
attachment 176045
[details]
Patch I think you want to pick between wheel-like and touch-like flings based on the input event's SourceDevice. If it's touchpad, do the wheel fling. If it's touchscreen, do a touch fling.
Alexandre Elias
Comment 4
2012-11-26 12:44:20 PST
Robert, could you clarify what parts of this behavior you actually want to keep on ChromeOS and why? I don't know what's intended behavior and what's historical cruft, so I erred on the side of deletion.
James Robinson
Comment 5
2012-11-26 12:46:25 PST
(In reply to
comment #4
)
> Robert, could you clarify what parts of this behavior you actually want to keep on ChromeOS and why? I don't know what's intended behavior and what's historical cruft, so I erred on the side of deletion.
The code currently here implements touchpad-style fling (for lack of a better term). It's what you get on a mac - a sequence of wheel events targeting a specific point on the page (not an element). The sort of fling you want in android (which I'll call touchscreen fling) is different - there you get a sequence of scrolls (not events) targeting a specific scrollable layer (not a point).
Alexandre Elias
Comment 6
2012-11-26 15:32:24 PST
Created
attachment 176093
[details]
Patch
Alexandre Elias
Comment 7
2012-11-26 15:32:59 PST
PTAL, Imade the patch a no-op for touchpad flings.
James Robinson
Comment 8
2012-11-26 15:50:39 PST
Comment on
attachment 176093
[details]
Patch View in context:
https://bugs.webkit.org/attachment.cgi?id=176093&action=review
Could you update WebCompositorInputHandlerImplTest.cpp? It has several tests for the touchpad behavior - it would be good to have some for the touchscreen behavior as well.
> Source/WebKit/chromium/src/WebCompositorInputHandlerImpl.cpp:320 > + TRACE_EVENT2("cc", "WebCompositorInputHandlerImpl::scrollBy", "x", increment.x, "y", increment.y);
i think this should be in the "webkit" trace category, not "cc" (the other ones in this file are wrong too, methinks)
Alexandre Elias
Comment 9
2012-11-26 17:23:08 PST
Created
attachment 176123
[details]
Patch
Alexandre Elias
Comment 10
2012-11-26 17:23:35 PST
Added tests for touchscreen fling.
James Robinson
Comment 11
2012-11-26 17:25:09 PST
Comment on
attachment 176123
[details]
Patch R=me
WebKit Review Bot
Comment 12
2012-11-26 19:24:43 PST
Comment on
attachment 176123
[details]
Patch Rejecting
attachment 176123
[details]
from commit-queue. New failing tests: inspector-protocol/debugger-terminate-dedicated-worker-while-paused.html Full output:
http://queues.webkit.org/results/15003371
WebKit Review Bot
Comment 13
2012-11-26 20:09:38 PST
Comment on
attachment 176123
[details]
Patch Clearing flags on attachment: 176123 Committed
r135808
: <
http://trac.webkit.org/changeset/135808
>
WebKit Review Bot
Comment 14
2012-11-26 20:09:42 PST
All reviewed patches have been landed. Closing bug.
Note
You need to
log in
before you can comment on or make changes to this bug.
Top of Page
Format For Printing
XML
Clone This Bug