|Summary:||REGRESSION: WebKit double-draws selections in some cases|
|Product:||WebKit||Reporter:||Eric Seidel <firstname.lastname@example.org>|
|Component:||New Bugs||Assignee:||Nobody <email@example.com>|
|Severity:||Normal||CC:||firstname.lastname@example.org, email@example.com, firstname.lastname@example.org, email@example.com, firstname.lastname@example.org, email@example.com|
|Priority:||P1||Keywords:||NeedsRadar, NeedsReduction, Regression|
|Version:||528+ (Nightly build)|
|OS:||Mac OS X 10.5|
See the attached image. Go to: http://www.mathewgross.com/about and select the top portion of the page. Notice how the selection is drawn twice in some areas. Firefox leaves large blank spaces (which isn't better). But the double-draw makes the text hard to read. WebKit should instead clip around the existing selection (I think) when drawing new selection rects so that they don't overlap.
Created an attachment (id=22555) [details] example of double-drawn selection
It looks like the selection isn't double drawn but that it is drawn in a translucent way. What I see in your screen shot is a big selection for the 2nd line which overlaps the other line's selection. I think this bug should be named "Overlapping selection highlight". The bug might simply be that the 2nd line ("Submitted by Mathew Gross on February 18, 2008 - 4:01pm.") selection's highlight is too high.
Confirmed. One of the lines is getting a crazy tall selection highlight. This is a regression from Safari 3.1.
Padding and border on inline elements is being incorporated into selection top and bottom since we use overflow. We need to change selection away from using painting overflow and to be its own unique values. This is a regression because an empty inline element is now present (because of fixes to make sure those show up). (<span class=taxonomy>)
I encountered a double-drawn selection when I was hiding and unhiding divs. I have a simple reproduction of this: http://ludios.net/webkit_bugs/selection_bugs.html (it is the first bug listed)
Created an attachment (id=133821) [details] Proposed Patch Patch to fix double-draws selections.
This patch will cause some pixel tests to fail on chromium port because of propagation.
(From update of attachment 133821 [details]) Attachment 133821 [details] did not pass chromium-ews (chromium-xvfb): Output: http://queues.webkit.org/results/12132863 New failing tests: editing/pasteboard/subframe-dragndrop-1.html editing/selection/3690703-2.html editing/selection/3690703.html editing/selection/inline-closest-leaf-child.html fast/repaint/4774354.html fast/replaced/selection-rect-transform.html editing/selection/contains-boundaries.html fast/repaint/block-selection-gap-in-composited-layer.html fast/writing-mode/vertical-lr-replaced-selection.html fast/replaced/selection-rect-in-table-cell.html fast/writing-mode/horizontal-bt-replaced-selection.html editing/selection/3690719.html fast/replaced/selection-rect.html fast/writing-mode/vertical-rl-replaced-selection.html
Created an attachment (id=133831) [details] Archive of layout-test-results from ec2-cr-linux-03 The attached test failures were seen while running run-webkit-tests on the chromium-ews. Bot: ec2-cr-linux-03 Port: <class 'webkitpy.common.config.ports.ChromiumXVFBPort'> Platform: Linux-2.6.35-28-virtual-x86_64-with-Ubuntu-10.10-maverick