- From: Andrew Fedoniouk <news@terrainformatica.com>
- Date: Thu, 05 Feb 2009 09:18:41 -0800
- To: "www-style@w3.org List" <www-style@w3.org>
- CC: David Hyatt <hyatt@apple.com>
Tab Atkins Jr. wrote:
> On Wed, Feb 4, 2009 at 10:46 PM, Andrew Fedoniouk
> <news@terrainformatica.com> wrote:
>> I think we have bug in specification regarding negative margins and
stacking
>> order.
>>
>> Attached is the illustration that demonstrates layout of three
>> DOM elements with the one in the middle having
>>
>> margin-top: -10px;
>> margin-bottom: -10px;
>>
>> I think that its stacking order is slightly larger (as shown on the
>> illustration) than its normal siblings thus it will be drawn on
>> top of *both*, previous and next, its siblings.
>
> Do you mean you *want* it to have a higher stacking order? Right now
> it seems to definitely follow normal stacking order.
I meant "it should" draw that way.
>
> Is there any particular reason that you think margins should affect
> stacking order here? I would find that *highly* unintuitive, not to
> mention probably page-breaking at this point.
Consider this:
<div style="background:green; height:100px">
<div #neg style="margin:-10px 0; background:red; height:100px">
<div style="background:yellow; height:100px">
Here effect of drawing elements with negative
margins is non-symmetrical for top and bottom margins of div#neg element.
>
> If you *do* want your behavior, it can be achieved by simply setting
> the middle element to be position:relative and z-index:2, in addition
> to the margins.
Specification introduced overlapping of elements by using negative
margins. That overlapping *must* be specified. As simply as that.
Current "not defined" spec produces following overlap (lateral view of
tree elements with the one in the middle having negative margins):
<pre>
\
---\
\-----
\
</pre>
But possible options are:
Negative on top:
<pre>
=====
---- ----
</pre>
Negative underneath:
<pre>
---- ----
=====
</pre>
As many people as many ideas of what is more intuitive here.
>
> ~TJ
>
>
--
Andrew Fedoniouk.
http://terrainformatica.com
Received on Thursday, 5 February 2009 17:19:12 UTC