Bug 21888 - REGRESSION: [Canvas] Transparent fillStyle stops line rendering
Summary: REGRESSION: [Canvas] Transparent fillStyle stops line rendering
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: DOM (show other bugs)
Version: 528+ (Nightly build)
Hardware: Mac (Intel) OS X 10.5
: P1 Normal
Assignee: Nobody
URL: http://www.marcuscobden.co.uk/stuff/2...
Keywords: Regression
Depends on:
Blocks:
 
Reported: 2008-10-25 15:05 PDT by Marcus Cobden
Modified: 2008-11-12 23:20 PST (History)
0 users

See Also:

Attachments
fixed strokePath() (1.25 KB, patch)
2008-10-25 23:55 PDT, Dirk Schulze 		eric: review-
Details | Formatted Diff | Diff
strokPath() fix with LayoutTest (4.10 KB, patch)
2008-10-29 03:15 PDT, Dirk Schulze 		darin: review+
Details | Formatted Diff | Diff
Show Obsolete (1) View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Marcus Cobden 2008-10-25 15:05:00 PDT 
If the fillStyle of the 2d context is set to a fully transparent colour, lines will not render, if it is set to anything else, the line renders correctly.
Lines should (iirc) be affected by strokeStyle rather than fillStyle.

This is a regression, as the testcase works in Safari 3.1.

See the provided link for a testcase.
(Also, apologies if I've selected the wrong component.)
Comment 1 Alexey Proskuryakov 2008-10-25 15:08:06 PDT 
Confirmed with r37887.
Comment 2 Dirk Schulze 2008-10-25 23:55:50 PDT 
Created attachment 24680 [details]
fixed strokePath()

fixed strokePath() in GraphicsContextCg.
Comment 3 Darin Adler 2008-10-27 14:23:50 PDT 
Comment on attachment 24680 [details]
fixed strokePath()

Great fix, but we also need a regression test!
Comment 4 Eric Seidel (no email) 2008-10-28 23:07:44 PDT 
Comment on attachment 24680 [details]
fixed strokePath()

LGTM.
Comment 5 Eric Seidel (no email) 2008-10-28 23:12:24 PDT 
Comment on attachment 24680 [details]
fixed strokePath()

Ah true... we do need a regression test.  You can make one with getImageData so it doesn't need pixel tests.
Comment 6 Dirk Schulze 2008-10-29 03:15:31 PDT 
Created attachment 24742 [details]
strokPath() fix with LayoutTest

added a LayoutTest to the patch.
Comment 7 Darin Adler 2008-10-29 09:12:25 PDT 
Comment on attachment 24742 [details]
strokPath() fix with LayoutTest

Excellent!

r=me
Comment 8 Dirk Schulze 2008-11-12 23:20:01 PST 
landed in r38364.