W3C home > Mailing lists > Public > www-style@w3.org > January 2012

Re: [CSS21] Issues with Ch.17 Tables

From: Simon Sapin <simon.sapin@kozea.fr>
Date: Wed, 25 Jan 2012 12:00:10 +0100
Message-ID: <4F1FE0BA.2050201@kozea.fr>
To: Anton Prowse <prowse@moonhenge.net>
CC: www-style@w3.org
Le 24/01/2012 22:57, Anton Prowse a écrit :
> Degenerate or not, they need specifying one way or the other!

Yes. What I meant is that as long as they are specified somehow, maybe 
we don’t need to spend too much effort to make them look nice (like a 
too narrow column not overflowing on the next column).

> Actually, my point was that the space visible between the borders of
> different pairs of horizontally-adjacent cells isn't constant throughout
> the table.

I think this is not the case. When traversing a row horizontally, we 
alternate between "column widths" and horizontal border spacing. The 
horizontal border spacing is always the same. Column widths are the 
result of the fixed or automatic table layout. They match the width of 
border area (as in 8.1) of each cell in the column.

> I wouldn't be at all surprised if there were backward-compatibility
> issues ;-)

I wouldn’t be surprised either. Point was: if the existing UAs agreed 
with each other but disagreed with the spec, the spec should be changed 
to match the real world.

In this case however, it appears that UAs do not agree. So I think the 
spec should be changed so that width and height on table cells is 
consistent with other boxes. (ie. they specify a dimension for the 
content area.)

This is according to Boris, just quoted by Gérard:

http://lists.w3.org/Archives/Public/www-style/2011Apr/0743.html

Regards,
-- 
Simon Sapin
Received on Wednesday, 25 January 2012 11:00:50 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:48 GMT