- From: Ernest Cline <ernestcline@mindspring.com>
- Date: Sat, 13 Dec 2003 23:24:00 -0500
- To: "Boris Zbarsky" <bzbarsky@MIT.EDU>
- Cc: www-style@w3.org
> [Original Message] > From: Boris Zbarsky <bzbarsky@MIT.EDU> > To: <ernestcline@mindspring.com> > Cc: <www-style@w3.org> > > > This implies that if it were to be extended so as to be able to serve > > for a generic table, that instead of having anonymous table-cells > > span multiple children, they should only span a single child > > (including possibly anonymous inline boxes > > How would this make the situation any better for a "generic table"? > The column still depends on CSS properties, no? The intended advantage was that for any element, its potential column placement would not be affected by styling. However, a closer examination convinces me that my proposed solution while a step in the right direction is insufficient for a generic case. With total abandonment of the collapsing of adjacent children into anonymous table objects and if all cells had a rowspan and colspan of 1, it would be enough to enable :first-column, :last-column and :only-column to be supported, but not :nth-column(an+b) or the like. This is because CSS would be able to determine if it element would either be in the kth column of n columns or the only column before layout, but not which of the two I'm beginning to think that it will require a complete rewrite of the rules for building anonymous table objects to have any hope of being able to support columns with the current table model. Another possibility would be for the HTML and CSS groups to work together to devise an entirely new set of table syntax and table styling that would support styling tables in XHTML2 with CSS3, but I'm not yet ready to concede that it will require something that drastic to be able to support CSS styling by columns.
Received on Saturday, 13 December 2003 23:24:06 UTC