Bug 10898

Summary: REGRESSION: Some bit garbage left behind on cnn story page after banner ad appears
Product: WebKit Reporter: John Sullivan <sullivan>
Component: Layout and RenderingAssignee: Nobody <webkit-unassigned>
Status: RESOLVED FIXED    
Severity: Normal CC: mitz
Priority: P1 Keywords: HasReduction, InRadar, Regression
Version: 420+   
Hardware: Mac   
OS: OS X 10.4   
URL: http://www.cnn.com
Bug Depends on: 12079, 12123    
Bug Blocks:    
Attachments:
Description Flags
Screenshot showing problem
none
Reduction - strict mode
none
Reduction - quirks mode none

Description John Sullivan 2006-09-17 07:02:09 PDT
To reproduce:
1. visit http://www.cnn.com
2. click one of the story titles in the "LATEST NEWS" section

Watch the page as it loads. First the entire contents of the page load, and look correct. A moment later, a banner ad appears at the top of the page, and the rest of the page shifts down. When this happens, a few pixels from the left side of the first character in the headline of the story are left behind in their previous position, about an inch above their new position. The first char in the headline is now slightly clipped on the left.

This did not occur on shipping Safari.
Comment 1 John Sullivan 2006-09-17 07:03:38 PDT
Created attachment 10604 [details]
Screenshot showing problem

In the attached screenshot, the "S" from "Shoppers" is slightly clipped on the left, and the missing bits are to the left of the blue background behind "Home World U.S.".
Comment 2 mitz 2006-09-18 12:34:21 PDT
Created attachment 10631 [details]
Reduction - strict mode
Comment 3 mitz 2006-09-18 12:35:39 PDT
Created attachment 10632 [details]
Reduction - quirks mode

This test case is broken in shipping Safari as well as in TOT
Comment 4 John Sullivan 2006-09-18 12:40:16 PDT
Maybe the test case is reflecting a different issue then? The bit garbage problem I see does not occur on shipping Safari.
Comment 5 mitz 2006-09-18 12:56:03 PDT
Comment on attachment 10631 [details]
Reduction - strict mode

(In reply to comment #4)
> Maybe the test case is reflecting a different issue then? The bit garbage
> problem I see does not occur on shipping Safari.
> 

The "strict mode" flavor of the bug *is* a regression. I think that explains the regression on cnn.com.
Comment 6 mitz 2006-09-28 10:09:26 PDT
*** Bug 11045 has been marked as a duplicate of this bug. ***
Comment 7 mitz 2006-09-28 10:12:41 PDT
Also failing manual-tests/left-overflow-repaint.html because of this. As noted in bug 11045, the problem is that tables' repaint rects don't account for cells' overflow.
Comment 8 Stephanie Lewis 2006-11-06 21:41:55 PST
radar 4823134
Comment 9 mitz 2007-02-12 16:24:48 PST
Fixed by fixing bug 12079 and bug 12123.
Comment 10 mitz 2007-02-17 22:52:05 PST
Reopening because r19588 was rolled out.