- From: David Woolley <forums@david-woolley.me.uk>
- Date: Tue, 03 Apr 2007 07:19:20 +0100
- To: www-html@w3.org
Kelly wrote: > Yes, I thought the point of XHTML was to transfer AWAY from HTML-as-tag-soup > by forcing the XML conformance rules on authors... > Not really. The reason that XML requires explicit opening and closing tags, and that XML language processors must give up if this is violated, is that it allows the use of mixed name spaces in user agents that have no a priori knowledge of the syntax of one of the namespaces, or, for general purpose tools, for all of them. HTML was designed so that most unambiguous omissions are still valid (which seems to be how someone earlier wanted their tools to work. The tag soup problem, in part, arises from parsers that treat the datastream as tags, rather than elements. If you want to have HTML that is checked for having all optional tags present, you can make a fairly mechanical change to the DTD and run the file through the sgmls validator (as used by the W3C one) locally.
Received on Tuesday, 3 April 2007 06:19:34 UTC