Re: [CSS2.1] Absolute positioning versus inlines and width

On 17 Mar 2004 at 13:39, Boris Zbarsky wrote:

> Worse yet, it requires fully laying out the inline before even attempting to
> lay out the positioned element.  The inline generates multiple boxes, and you
> have to hold of dealing with the absolutely positioned element until all the
> boxes are done... what happens when the inline spans a page break?
> 
Well, there seem to be rather severe bugs with this feature when i 
tested. Sometimes the absolute box just seemed to dissapear. I'm not 
sure if i fully understood the page break reference, could you 
elaborate?

> Frankly, if none of the current UAs implement what the spec says, I propose
> changing the wording to say that the containing block is formed by the first
> box the inline generates (that is, from the start of the inline to the first
> line break or the end of the inline, whichever comes first).

That does seem like a better definition to me. But doesn't that mean 
that you still need to lay out the inline before the absolute box in 
the case i provided? (in case the page break does not occur or 
happens after the absolute span). I've only done one test, but it 
seems the containing-block-width is grabbed from the nearest ancestor 
that is positioned which is also non-inline in all three browsers i 
tested.

 /Staffan