John Lewis wrote:
>Kevin wrote on Monday, January 20, 2003 at 9:10:04 PM:
>>Well I'm a bit confused. If an element has a lower stack level than
>>its parent (or ancestor), then my browsers render it underneath the
>>parent. Mozilla seems to allow an element to have a lower stack
>>level than the root element, and it thinks the root element has a
>>stack level of zero.
>I shouldn't be talking about stack levels since I'm no expert.
>However, I think I know where the confusion is arising. If the root
>element is the only element in its stacking *context*, that means no
>descendant of the root element may ever be rendered below the root
>element by changing it's stack *level*.
Elements belong to 2 stacking contexts
* relative it's siblings
* relative it's children
A child with a negative z-index is placed BEHIND it's parent (parent is
always at 0 relative it's children).
That is why IMO Gecko is implementing the spec correctly and the others
However I feel this is actually a bug in the SPEC, becuse there is no
reason why anyone would even want to place something "behind the app".
The possibility to do this adds noting to usablility, but does add a big
risk of accidentally breaking you pages and that is just bad SPEC design
Thus the spec IMO should be changed to clearly specify that the app
background is always at an "infinitly low negative z-index".
Software never has bugs. It just develops random features. =)