- From: Simon Sapin <simon.sapin@kozea.fr>
- Date: Wed, 25 Jan 2012 12:00:10 +0100
- 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 UTC