8268 2006-04-08 15:46:03 -0700 REGRESSION (r13639): Caret leaves a small footprint in the focus ring as it moves 2007-12-07 11:26:35 -0800 1 1 1 Unclassified WebKit Layout and Rendering 420+ Mac OS X 10.4 RESOLVED FIXED HasReduction, InRadar, Regression P1 Normal --- 1 justin.garcia webkit-unassigned alice.barraclough darin ddkilzer hyatt ian koivisto oldest_to_newest 38801 0 justin.garcia 2006-04-08 15:46:03 -0700 Attaching a testcase 38802 1 7586 justin.garcia 2006-04-08 15:47:01 -0700 Created attachment 7586 testcase 38805 2 darin 2006-04-08 17:04:27 -0700 Dave, this is what I was worried about re-introducing. I guess it happened! 38882 3 hyatt 2006-04-10 11:10:58 -0700 I have a good way of fixing this that will handle overflow too (the old fix for this didn't). 38892 4 hyatt 2006-04-10 13:38:40 -0700 The last outline-related checkin before revision 13639 is the one darin made to deal with textfield focus rings not matching. My code had not yet landed, so this is not related to my coalesced update changes. 38963 5 darin 2006-04-11 09:25:54 -0700 I've verified that the focus-ring drawing functions are called when the caret blinks, but nothing is being drawn. Continuing to investigate. 38964 6 darin 2006-04-11 09:41:34 -0700 There's some kind of rounding problem here. Setting an outline-offset of either 1px or -1px makes the bug go away. 43114 7 alice.barraclough 2006-05-21 22:33:09 -0700 <rdar://problem/4556558> 31024 8 darin 2007-01-29 09:09:08 -0800 Waiting on a Core Graphics fix for focus ring drawing before revisiting this problem. 22605 9 koivisto 2007-02-18 13:47:28 -0800 It seems that if the clip overlaps inside edge of the focus ring exactly one pixel deep the focus ring does not get drawn. This particular bug can be easily worked around in repaintRectForCaret (in SelectionController.cpp) by doing caret.inflate(2) instead of inflate(1) or removing the inflate line. First option increases overlap and makes the focus ring paint correctly, second one removes overlap. That just hides the actual problem though. I think this is most likely the same problem as bug 11965. 63345 10 ddkilzer 2007-12-05 08:44:51 -0800 I can still reproduce this bug with the attached test case (Attachment #7586), but not with this test case: data:text/html,<input type=text> 63346 11 ddkilzer 2007-12-05 08:45:27 -0800 (In reply to comment #10) > I can still reproduce this bug with the attached test case (Attachment #7586 [edit]), > but not with this test case: > > data:text/html,<input type=text> Using a local debug build of WebKit r28434 with Safari 3.0.4 (523.12) on Mac OS X 10.4.11 (8S165). 63605 12 mitz 2007-12-07 11:26:35 -0800 Fixed in <http://trac.webkit.org/projects/webkit/changeset/28523>. 7586 2006-04-08 15:47:01 -0700 2006-04-08 15:47:01 -0700 testcase 8268.html text/html 205 justin.garcia PHA+Q2xpY2sgaW5zaWRlIHRoZSBlZGl0YWJsZSByZWdpb24gYmVsb3cgYW5kIG1vdmUgdGhlIGNh cmV0IGFyb3VuZCB3aXRoIHRoZSBhcnJvdyBrZXlzLiAgVGhlIGNhcmV0IGxlYXZlcyBhIHdoaXRl IGZvb3RwcmludCBpbiB0aGUgZm9jdXMgcmluZyBhcyBpdCBtb3Zlcy48L3A+IAo8ZGl2IGNvbnRl bnRlZGl0YWJsZT0idHJ1ZSI+Q2xpY2sgaGVyZTwvZGl2Pg==