Closed Bug 461512 Opened 17 years ago Closed 16 years ago

nsCSSColorUtils doesn't support alpha color (border color is always solid color if the border-style is groove, ridge, inset, outset)

Categories

(Core :: Layout, defect)

defect
Not set
normal

Tracking

()

RESOLVED FIXED
mozilla1.9.1b2

People

(Reporter: masayuki, Assigned: masayuki)

References

()

Details

Attachments

(2 files, 5 obsolete files)

Attached patch Patch v1.0 (obsolete) — Splinter Review
nsCSSColorUtils doesn't support alpha color. Therefore, the border color is always drawn by solid color if border-style is ridge, groove, inset or outset. The attached patch also removes NS_Get3DColor and nsCSSRendering::TransformColor. Because they are not used.
Attachment #344612 - Flags: superreview?(roc)
Attachment #344612 - Flags: review?(roc)
Attached patch reftest (obsolete) — Splinter Review
reftest for this bug. However, I cannot find to test dotted case. The dots of corners are overlapped, therefore, they are mismatch in this test.
Attachment #344624 - Flags: superreview?(roc)
Attachment #344624 - Flags: review?(roc)
Attached patch Patch v1.0.1 (obsolete) — Splinter Review
fix some white spaces.
Attachment #344612 - Attachment is obsolete: true
Attachment #344624 - Attachment is obsolete: true
Attachment #344737 - Flags: superreview?(roc)
Attachment #344737 - Flags: review?(roc)
Attachment #344612 - Flags: superreview?(roc)
Attachment #344612 - Flags: review?(roc)
Attachment #344624 - Flags: superreview?(roc)
Attachment #344624 - Flags: review?(roc)
Comment on attachment 344737 [details] [diff] [review] Patch v1.0.1 Remove reftest-wait from your tests. It shouldn't be needed.
Attachment #344737 - Flags: superreview?(roc)
Attachment #344737 - Flags: superreview+
Attachment #344737 - Flags: review?(roc)
Attachment #344737 - Flags: review+
landed.
Status: ASSIGNED → RESOLVED
Closed: 17 years ago
Resolution: --- → FIXED
{ http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1225159830.1225163390.11890.gz Linux mozilla-central moz2-linux-slave07 dep unit test on 2008/10/27 19:10:30 http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1225160355.1225163528.12126.gz Linux mozilla-central moz2-linux-slave08 dep unit test on 2008/10/27 19:19:15 http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1225160149.1225163577.12179.gz WINNT 5.2 mozilla-central moz2-win32-slave07 dep unit test on 2008/10/27 19:15:49 http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1225159936.1225163872.12911.gz WINNT 5.2 mozilla-central moz2-win32-slave08 dep unit test on 2008/10/27 19:12:16 http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1225160692.1225164712.14806.gz Linux comm-central dep unit test on 2008/10/27 19:24:52 http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1225160128.1225164532.14469.gz Win2k3 comm-central dep unit test on 2008/10/27 19:15:28 } (NB: MacOSX boxes haven't cycled yet.) http://hg.mozilla.org/mozilla-central/rev/e94041498958 Disable failing new test [...]
Status: RESOLVED → REOPENED
Flags: in-testsuite+
Resolution: FIXED → ---
Target Milestone: --- → mozilla1.9.1b2
Than you, I'll check the cause.
Attached image result of test (obsolete) —
Attached image reference of the test (obsolete) —
The colors of result (rgba(10%, 20%, 30%, 0.5)): normal color: #8c99a6 (rgb(140, 153, 166)) lighter color: #dce0e4 (rgb(220, 224, 228)) darker color: #878e95 (rgb(135, 142, 149)) The colors of ref (rgb(10%, 20%, 30%) and parent's opacity: 0.5): normal color: #8d99a6 (rgb(141, 153, 166)) lighter color: #dde0e4 (rgb(221, 224, 228)) darker color: #878e95 (rgb(135, 142, 149)) The R values of normal and lighter are different. Is this a gfx bug of alpha rendering? Or is the idea of my test wrong?
roc, dbaron: Do you have idea for comment 10? The colors of rgba(xx, xx, xx, 0.5) and (rgb(xx, xx, xx) + opacity: 0.5) are different, but I'm not sure the cause...
I guess it's a rounding error somewhere.
Attached patch Fix the orangeSplinter Review
Sorry for the delay. This new test passes on tryserver. But this fails to test the corners which are joined different colors. For avoiding the failure, this covers white background div boxes to the corners. However, it's enough for the testing of this bug because this bug caused the color computation, not the actual rendering code.
Attachment #345053 - Attachment is obsolete: true
Attachment #345054 - Attachment is obsolete: true
Attachment #416057 - Flags: review?(roc)
Status: REOPENED → RESOLVED
Closed: 17 years ago16 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: