Bug 134249

Summary: The color of the box-shadow CSS property does not inherit correctly from the color CSS property
Product: WebKit Reporter: Xavier Zalawa <xavier>
Component: CSSAssignee: Nobody <webkit-unassigned>
Status: RESOLVED CONFIGURATION CHANGED    
Severity: Normal CC: bfulgham, simon.fraser
Priority: P2    
Version: 528+ (Nightly build)   
Hardware: Mac   
OS: OS X 10.9   
Attachments:
Description Flags
Screenshot of the result of HTML tests in Safari 7.0.4
none
HTML tests none

Description Xavier Zalawa 2014-06-24 07:07:00 PDT
Other browsers tested:
  Chrome 35.0.1916.153 m: FAIL
  Firefox 30.0: OK
  IE 11.0.9: OK
  Opera 12.17: OK

What steps will reproduce the problem?
1. Nest two depths of `div` HTML element twice (one after the other).
2. Apply the same `box-shadow` CSS property without defining the color on the last depth in each one.
3. Apply the `color` CSS property on the first depth with a different value for each one.

What is the expected result?
If it is absent, the color of the shadow should be taken from the `color` CSS property as it is defined in the CSS3 specification of W3C.
http://www.w3.org/TR/css3-background/#shadow-color

What happens instead?
The color of the shadow does not change and is identical to the color of the first drop-shadow painted in the page.

It seems that the problem does not exist/happen when the `color` and `box-shadow` CSS properties are applied at the same time on the same HTML element.
I have opened the same issue for Blink here: https://code.google.com/p/chromium/issues/detail?id=388209
Comment 1 Xavier Zalawa 2014-06-24 07:08:38 PDT
Created attachment 233699 [details]
Screenshot of the result of HTML tests in Safari 7.0.4
Comment 2 Xavier Zalawa 2014-06-24 07:10:23 PDT
Created attachment 233700 [details]
HTML tests
Comment 3 Simon Fraser (smfr) 2014-06-24 23:41:23 PDT
Pretty sure we have older bugs on this.
Comment 4 Brent Fulgham 2022-07-13 17:20:05 PDT
We don't see an issue with this test case at present, though it may have changed in the many years since you filed this. If you believe there is still an issue, could you please REOPEN this bug with a small test case illustrating the issue?