The CSS3 draft says the properties apply to "block-level elements". I can't find a formal definition of this term in CSS, but I think we could easily implement it for the blocks that are the interior of table cells.
See http://www.w3.org/TR/CSS21/visuren.html#q5 for the CSS2.1 definition of "block-level". I would expect CSS3 to have a similar definition. I'm not sure that these properties really apply in a sane way to display:table elements, though....
I see no problem with applying this to floated or positioned blocks.
Then we need to solve the intrinsic width issue, since almost all floated/positioned blocks in fact have shrinkwrap width.
[The meta-strategy is to choose a strategy that will 1) get us to something useful working as quickly as possible while 2) being plausibly extensible to full generality over time.]
Makes sense.
The real issue is that if a float is allowed to float to the left of its column, and interfere with the previous column
Oy. Right. We don't want that.... ;)
-Boris _______________________________________________ mozilla-layout mailing list [EMAIL PROTECTED] http://mail.mozilla.org/listinfo/mozilla-layout
