- From: Ian Hickson <ian@hixie.ch>
- Date: Thu, 24 Apr 2008 06:22:06 +0000 (UTC)
- To: Ben 'Cerbera' Millard <cerbera@projectcerbera.com>
- Cc: HTMLWG <public-html@w3.org>
On Wed, 23 Apr 2008, Ben 'Cerbera' Millard wrote: > > Basically, I'm suggesting we allow: > > * <td scope> to work like <th scope>. > * <td id> to work like <th id> when the headers attribute points to it. The current algorithm makes a clear distinction between what's a header and what's a data cell, and I really would like to keep that distinction if at all possible. While I agree entirely that we could make this more generic (and more like HTML4) and just blur the lines between the two to the point where any data cell can be a header and have headers and so forth, and while I agree that there are cases where this would make sense in some literal way, I think this would make things more confusing for authors without much gain. In practice few tables need this. If we want to help accessibility and get authors to really use this features right more often than they get them wrong, I think we should hold back from giving them too much rope. (For the same reason, I think we should not bother with providing axis="" either.) Originally I was against adding headers="" at all for similar reasons, but after studying the data collected by you and others, as well as doing my own research, I concluded that headers="" could indeed be more gain than pain on the long run. I looked at making data cells into headers at the same time, and at the time the conclusion I came to was that it really wasn't worth it. -- Ian Hickson U+1047E )\._.,--....,'``. fL http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,. Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'
Received on Thursday, 24 April 2008 06:22:49 UTC