Closed Bug 75878 Opened 25 years ago Closed 23 years ago

sibling elements move elements with absolute position

Categories

(Core :: CSS Parsing and Computation, defect)

x86
All
defect
Not set
normal

Tracking

()

RESOLVED DUPLICATE of bug 33784
mozilla1.1alpha

People

(Reporter: wo, Assigned: attinasi)

Details

(Keywords: compat, testcase, Whiteboard: WONTFIX?)

Attachments

(2 files)

If an element has position:absolute, its position should not depend upon the position of its siblings. So the H1 element in this document <body> <p></p> <h1 style="position:absolute;left:0;top:0;">This should be at 0/0</h1> </body> should appear at position 0/0. But in M0.8, M.081 and NB2001041004 (win98) it is at about 0/20. If you remove the paragraph, it is at 0/0.
Attached file testcase
looks like a problem with margin collapse... seeing this on a 2001-04-12 linux build too.
Status: UNCONFIRMED → NEW
Ever confirmed: true
OS: Windows 98 → All
I checked, and this also occurs in strict mode. (So it's not just a quirks thing from people using <p> for line spacing or whatever.)
The bug doesn't really work as described. 1) There is no bug in strict mode. With or without the empty P element, the H1 is always about 20px from the top because html.css defines: H1 {margin: .67em 0} 2) In quirks mode the problem is that without the empty P between the BODY and the H1, the H1 matches the collapsing margin rule in quirks.css that says: body > h1:first-node {margin-top: 0} 3) The difference of behavior happens with any empty element: an empty SPAN does it as well an an empty P. So... the bug is not that in most cases the H1 is at 20px from the top. The bug is that in quirks mode and without empty element, the margin-top of the H1 is set to 0 even though it is an absolutely positioned element. I'm afraid the collapsing margins issues cannot be solved with style rules in quirks.css as we have been doing in bug 97361, bug 83839 and others because they should not apply to fixed or absolutely positioned elements. It is annoying but I think it can be marked WontFix. For a correct absolute positioning of elements that have a margin defined in the UA stylesheet, the author should set (or in that case reset) the margin anyhow. Reassigned to Marc and CCd Dbaron who worked on it too.
Assignee: pierre → attinasi
Attached file Testcase v2
Doing the collapsing margins stuff any other way without causing author style rules that specify margins to be ignored would be very difficult.
Target Milestone: --- → mozilla1.1
Keywords: css2, testcase
WONTFIX?
Keywords: css2compat
Whiteboard: WONTFIX?
This is quirks-only, and we've been wontfixing other bugs related to this issue.. If we ever decide to change the way we handle margins in quirks mode, this bug will automatically be "fixed". Note that the testcase in this bug is invalid to start with (the only way that text could end up at 0,0 with those styles and markup is due to a weird quirk).
Status: NEW → RESOLVED
Closed: 23 years ago
Resolution: --- → WONTFIX
I'd rather mark this is a duplicate...
Status: RESOLVED → REOPENED
Resolution: WONTFIX → ---
*** This bug has been marked as a duplicate of 33784 ***
Status: REOPENED → RESOLVED
Closed: 23 years ago23 years ago
Resolution: --- → DUPLICATE
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: