Closed
Bug 114611
Opened 23 years ago
Closed 23 years ago
[INVALID]collapsed margins give unexpected results
Categories
(Core :: Layout, defect)
Tracking
()
RESOLVED
INVALID
People
(Reporter: sproctor, Assigned: dbaron)
References
()
Details
From Bugzilla Helper: User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:0.9.6+) Gecko/20011209 BuildID: 2001120908 I've set up a few examples http://www.psa.neu.edu/list_test/ http://www.psa.neu.edu/list_test/example2.html http://www.psa.neu.edu/list_test/example3.html the last one is how I would expect them to perform. basically, if you have a list, without setting the margin, then there is a one line margin at the top. the background color of this is not of the block element behind the list, but whatever element has content. This doesn't make sense to me. the examples I made are the clearest way to explain it. Reproducible: Always Steps to Reproduce: visit links
Comment 1•23 years ago
|
||
All looks fine to me... the <ul> has a 1-em top margin. The margin is transparent, so it shows the background color of the parent element of the <ul> What do you mean by "whatever element has content"?
Reporter | ||
Comment 2•23 years ago
|
||
It isn't the parent element though. the parent element is a div with a gray background, and mozilla shows a blue margin (which is from the div which is the parent of the other div). what I meant by "whatever element has content" is that if you put content in the parent element, it will show that background, otherwise it will show the background of the parent of that element. (check out http://www.psa.neu.edu/list_test/example2.html for an example of what I mean)
Comment 3•23 years ago
|
||
Ah. I see what you mean. This is correct; all that's going on is that vertical margins are collapsing per the CSS spec. So in http://www.psa.neu.edu/list_test/ the margin of the <div> (0em) and the margin of the <ul> (1em) collapse to give a 1em top margin on the <div>. In Example 2, margin collapsing cannot occur because there is content between the margins. So you see what you see.
Reporter | ||
Comment 4•23 years ago
|
||
you're absolutely right. I'd missed the section on collapsable margins. Regardless of that, this behavior doesn't make sense, as one element can be drawn in by the margin of an element inside of it, which doesn't seem to be the purpose of collapsing the margins. The spec is rather vague and gives its reasoning only as collapsing margins will make the site look more like what the designer intended. I think for nested elements it would make more sense to never increase the value of a margin. for example: if you have block A with block B inside of it, and A has a margin of 3 and b has a margin of 5, it would make more sense to me for the margin of A to remain 3, and the margin from B to A to collapse to 2, rather than A to get a margin of 5 and B's margin to collapse to 0 as they do currently. It seems that the current behavior is conformant and any change would break that, but I don't think the standard makes sense.
Comment 5•23 years ago
|
||
Sending this over to Style System; perhaps Hixie or dbaron can mention this to the WG.
Assignee: attinasi → dbaron
Status: UNCONFIRMED → NEW
Component: Layout → Style System
Ever confirmed: true
QA Contact: petersen → ian
Whiteboard: spec poorly written?
Assignee | ||
Comment 6•23 years ago
|
||
Collapsing margins make sense to me, because I expect them, and I think they make it easier to get the correct results by specifying margins without having to use lots of adjacent sibling combinators. Do you really think we should change the spec?
Component: Style System → Layout
Summary: CSS layout with list tags is inconsistent → [INVALID]collapsed margins give unexpected results
Whiteboard: spec poorly written?
Assignee | ||
Comment 7•23 years ago
|
||
Oh, I see, you think the assignment of space within them should be changed. That assignment of space was actually only described in an errata item to the spec (I think), although it may have found its way into CSS2 or the revised CSS1 (I don't know when it was writen).
Reporter | ||
Comment 8•23 years ago
|
||
right. collapsing the margins makes perfect sense to me, though the way it's done doesn't. I've describe what I see as the most logical method for collapsing the margins. I've found nothing in the errata pertaining to this topic. from the CSS1 spec: "Two or more adjoining vertical margins (i.e., with no border, padding or content between them) are collapsed to use the maximum of the margin values. (section 4.1)" and from the CSS2 spec: "the expression collapsing margins means that adjoining margins (no padding or border areas separate them) of two or more boxes (which may be next to one another or nested) combine to form a single margin. (section 8.3.1)" It seems pretty clear their intentions. to me the current implimentation doesn't make much sense. when I say margin 0, I expect the border of the block to touch adjacent blocks regardless of the content in my block. unless there's some more important reason for the current design, I would the spec should change.
Assignee | ||
Comment 9•23 years ago
|
||
INVALID. If you want to change the spec, please raise the issue on www-style.
Status: NEW → RESOLVED
Closed: 23 years ago
Resolution: --- → INVALID
You need to log in
before you can comment on or make changes to this bug.
Description
•