Closed
Bug 64865
Opened 25 years ago
Closed 23 years ago
relative positioning on :hover causes frame misordering (and assertions)
Categories
(Core :: CSS Parsing and Computation, defect, P2)
Tracking
()
VERIFIED
WORKSFORME
Future
People
(Reporter: phil, Assigned: attinasi)
References
()
Details
(Keywords: css2, testcase)
Attachments
(1 file)
|
469 bytes,
text/html
|
Details |
Using 'position: relative' on an 'a:hover' selector,
and placing more than one consecutive (no white space)
instance of such an anchor causes Mozilla to get all
confused about which order they were in.
If you move the mouse across them, it will re-order
them. If you have more than two, it will re-order
across the whole group (not just adjacent pairs)
Musical chairs for your web browser.
The page below will demonstrate the bug. (Note that the padding
and backgrounds are only there to make it easier to see what's
happening.)
Use the included URL if you like, but be warned its on geocities,
so you'll get advertising if you have javascript enabled :/
<html>
<head>
<title>Mozilla position: relative + :hover style bug</title>
<style>
a { padding: 5px; background: red; }
a:hover { background: green; position: relative; }
</style>
</head>
<body>
<a href="#">Far Left</a><a href="#">Left</a><a href="#">Middle</a><a
href="#">Right</a><a href="#">Far Right</a>
</body>
</html>
|
||
Comment 1•25 years ago
|
||
Funky..
Platform: PC
OS: Linux 2.2.16
Mozilla Build: 2001010908
Marking NEW.
Status: UNCONFIRMED → NEW
Ever confirmed: true
|
||
Comment 2•25 years ago
|
||
|
|
||
Comment 3•25 years ago
|
||
In debug builds one gets a whole bunch of assertions. It appears that the
frame constructor is having great difficulty with swapping the frames around:
yikes: 'nsnull != placeholderFrame'
nsCSSFrameConstructor.cpp line 8176
assigning to buster, cc'ing marc and pierre for their input, cc'ing hyatt since
it might interest him (similar things happened(ed?) with XBL flipping into and
out of 'float' state), and cc'ing dbaron for kicks.
Assignee: pierre → buster
Keywords: mozilla0.9,
testcase
QA Contact: chrisd → ian
Summary: funky relative positioning when combined with a:hover → relative positioning on :hover causes frame misordering (and assertions)
hmmm...frame misorder bad. me think, p2.
but I'm over budget for moz1.0, so reluctantly setting to next milestone after that.
Status: NEW → ASSIGNED
Priority: -- → P2
Target Milestone: --- → mozilla1.0.1
|
|
||
Comment 6•24 years ago
|
||
Build reassigning Buster's bugs to Marc.
Assignee: buster → attinasi
Status: ASSIGNED → NEW
|
|
Assignee | |
Comment 7•23 years ago
|
||
Moving Mozilla 1.01 bugs to 'future' milestone with priority P1
I will be pulling bugs from 'future' milestones when scheduling later work.
Priority: P2 → P1
Target Milestone: mozilla1.0.1 → Future
I re-checked the test cases, and it appears this bug has been fixed somewhere
along the way?
(Platform: Win32; Build: 2002022408)
There is still something interesting happening with the test cases, though. If
you compare the one in the original URL (which uses anchors) with the "improved
testcase" attachment (which uses spans), I find that the :hover only seems to
take effect in the latter case when the mouse is over the span's padding. When
the mouse is on the text itself, the span reverts to its normal state.
|
|
||
Comment 9•23 years ago
|
||
cc'ing myself
|
|
||
Comment 10•23 years ago
|
||
WFM: Using 2002-7-1 build on WinXP.
Status: NEW → RESOLVED
Closed: 23 years ago
Priority: P1 → P2
Resolution: --- → WORKSFORME
You need to log in
before you can comment on or make changes to this bug.
Description
•