Closed
Bug 610966
Opened 14 years ago
Closed 12 years ago
Wrong computed style value when block has width and 'margin: auto'
Categories
(Core :: DOM: CSS Object Model, defect, P3)
Core
DOM: CSS Object Model
Tracking
()
RESOLVED
DUPLICATE
of bug 381328
People
(Reporter: skonstantinov, Unassigned)
References
Details
Attachments
(1 file)
|
301 bytes,
text/html
|
Details |
User-Agent: Mozilla/5.0 (Windows NT 5.1; rv:2.0b6) Gecko/20100101 Firefox/4.0b6
Build Identifier: Mozilla/5.0 (Windows NT 5.1; rv:2.0b6) Gecko/20100101 Firefox/4.0b6
getComputedStyle(elem, null).getPropertyValue('margin-left') returns '0px'. Opera has the same problem. Webkit returns actual margin size.
Reproducible: Always
Steps to Reproduce:
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<title>test</title>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<script>
window.onload = function () {
var body = document.getElementsByTagName('body')[0];
alert(window.getComputedStyle(body, null).getPropertyValue('margin-left'));
}
</script>
<style>
html
{
background: #cff;
}
body
{
width: 600px;
height: 600px;
margin: 0 auto;
background: #ffc;
}
</style>
</head>
<body>
</body>
</html>
Actual Results:
0px
Expected Results:
Actual margin size (depends on screen width)
|
||
Comment 1•14 years ago
|
||
Same behavior on the Beta7 Mozilla/5.0 (Windows NT 5.1; rv:2.0b7) Gecko/20100101 Firefox/4.0b7
Works with Safari, although there are ways to run this script so that Firefox would echo the expected result for using getComputedStyle(elem, null).getPropertyValue('margin-left')
You can read more about getStyles behavior in interaction with browsers here:
http://www.quirksmode.org/dom/getstyles.html
Still, settings this problem to NEW, although I am not sure we will have a fix through Firefox 4. Juanb, what do you think?
Status: UNCONFIRMED → NEW
Ever confirmed: true
|
||
Updated•14 years ago
|
Component: General → Layout
Product: Firefox → Core
QA Contact: general → layout
|
|
||
Comment 2•14 years ago
|
||
I am not sure where this bug would fall under, but I've moved it to layout. Maybe someone in that group could comment or move it to the right bucket.
|
||
Comment 3•14 years ago
|
||
David, it looks like the used-margin property on <body> just has the value we computed in nsCSSOffsetState::ComputeMargin, and isn't adjusted for the changes that happen in nsHTMLReflowState::CalculateBlockSideMargins (which are what handle auto margins of all sorts for blocks). Should we be adjusting it for those?
Is there a right answer to that question? I suppose I could attempt to read tea leaves out of the current specs.
|
|
||
Comment 5•14 years ago
|
||
I dunno about "right", but it seems like if we're trying to return used style we should take it into account.
Note that I'm more interested in whether our internal GetUsedMargin() APIs are doing the right thing given how we use them than in how things interact with the brokenness that is getComputedStyle per se.
|
||
Comment 7•13 years ago
|
||
This change is being made in webkit as we speak: https://bugs.webkit.org/show_bug.cgi?id=73334
To summarize: http://www.w3.org/TR/css3-values/#stages-examples says that when auto: is specified, the used value should be the positive pixel value, not 0px. It's also clarified again in the CSSOM draft here: http://dev.w3.org/csswg/cssom/#resolved-value
If we are to be following CSS3, Webkit, and IE9/10, the conclusion is: when margin is "auto", getComputedStyle should return the actual used pixel value.
|
|
||
Comment 8•12 years ago
|
||
The change has now been made and is live in Webkit and Trident.
In the meantime, it's been a continuing thorn in jQuery's side. I've had several bug reports about this, and the most recent is:
http://bugs.jquery.com/ticket/11813
Of course, you guys shouldn't do work just to lower jQuery's bug count, but I'm just trying to provide context as to how this effects real world usage.
Component: Layout → DOM: CSS Object Model
Priority: -- → P3
(In reply to Boris Zbarsky (:bz) from comment #3)
> David, it looks like the used-margin property on <body> just has the value
> we computed in nsCSSOffsetState::ComputeMargin, and isn't adjusted for the
> changes that happen in nsHTMLReflowState::CalculateBlockSideMargins (which
> are what handle auto margins of all sorts for blocks). Should we be
> adjusting it for those?
I think we probably should be.
We need to test this carefully for all values of 'position'.
Status: NEW → RESOLVED
Closed: 12 years ago
Resolution: --- → DUPLICATE
You need to log in
before you can comment on or make changes to this bug.
Description
•