- From: Etan Wexler <ewexler@stickdog.com>
- Date: Sun, 23 Mar 2003 02:36:13 -0800
- To: www-html@w3.org
- Cc: Ernest Cline <ernestcline@mindspring.com>
Ernest Cline wrote to <mailto:www-style@w3.org> on 1 March 2003 in "[XHTML2] di element?" (<mid:3E60A05C.19381.AFF55DC@localhost>): > Might it be thought worthwhile to add a <di> element to the List > module? It would absolutely be worthwhile. > This would give a way to > structuraly group items without having to use the class attribute. Your point is understood, but I must note that the 'class' attribute cannot structurally group elements. The 'class' attribute groups elements logically only. The need for a 'di' element type is thus increased, if anything. > Currently, grouping must be inferred from the convention that > (dt+,dd+) is one definition list item. Back when these elements didn't > have any attributes this wasn't really a problem This was always a problem in that a definition list item, having no element of its own, could not posess an ID. For me and for some others, the ability to assign an ID and to refer to it is the main benefit of the new element type. Consider the following document fragments. <a href="#term-XHTML">XHTML definition</a> <dl> <dt id="term-XHTML">XHTML</dt> <dd>Extensible Hypertext Markup Language</dd> </dl> <a href="#def-XHTML">XHTML definition</a> <dl> <di id="def-XHTML"> <dt>XHTML</dt> <dd>Extensible Hypertext Markup Language</dd> </di> </dl> In typical Web browsers, the apparent difference between the two is slight. Regardless, the semantic difference is significant. The effects are felt anywhere that addressing elements is important: style sheets, metadata, and so on. -- Etan Wexler, son of son of scion of scion. <mailto:ewexler@stickdog.com>
Received on Sunday, 23 March 2003 05:36:23 UTC