Bug 80175 - Revert borders to integers for subpixel layout
Summary: Revert borders to integers for subpixel layout
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: Layout and Rendering (show other bugs)
Version: 528+ (Nightly build)
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Emil A Eklund
URL:
Keywords:
Depends on:
Blocks: 60318
  Show dependency treegraph
 
Reported: 2012-03-02 11:33 PST by Emil A Eklund
Modified: 2012-03-02 17:48 PST (History)
5 users (show)

See Also:


Attachments
Patch (28.69 KB, patch)
2012-03-02 11:37 PST, Emil A Eklund
no flags Details | Formatted Diff | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Emil A Eklund 2012-03-02 11:33:14 PST
Change border representation and calculation back to integers to ensure that borders given the same width are rendered with the same actual width. This also avoids unnecessary type conversions.
Comment 1 Emil A Eklund 2012-03-02 11:37:18 PST
Created attachment 129931 [details]
Patch
Comment 2 Eric Seidel (no email) 2012-03-02 14:09:25 PST
Comment on attachment 129931 [details]
Patch

hmm... So even table borders are not sub-pixel precise?  I know individual cell-layout for tables is done with integers... but I'm surprised that borders around said table would be?
Comment 3 Emil A Eklund 2012-03-02 14:14:45 PST
(In reply to comment #2)
> (From update of attachment 129931 [details])
> hmm... So even table borders are not sub-pixel precise?  I know individual cell-layout for tables is done with integers... but I'm surprised that borders around said table would be?

Borders are aligned to device pixels to ensure that a box with border a style such as "0.3ex solid black" has all four borders rendered with the same width regardless of the offset. If we did this with subpixel precision the bottom/right border might be pained width a different width.
Comment 4 WebKit Review Bot 2012-03-02 14:31:46 PST
Comment on attachment 129931 [details]
Patch

Attachment 129931 [details] did not pass chromium-ews (chromium-xvfb):
Output: http://queues.webkit.org/results/11799129

New failing tests:
editing/selection/select-line-break-with-opposite-directionality.html
Comment 5 WebKit Review Bot 2012-03-02 17:48:04 PST
Comment on attachment 129931 [details]
Patch

Clearing flags on attachment: 129931

Committed r109642: <http://trac.webkit.org/changeset/109642>
Comment 6 WebKit Review Bot 2012-03-02 17:48:15 PST
All reviewed patches have been landed.  Closing bug.