Anne van Kesteren wrote:
> Quoting David Dorward <david@us-lot.org>:
>> They can't. That is the main difference between HTML and XHTML. The
>> point is that XHTML does not create a better defined parse tree
>> because of this.
> 
> It does. When I use some random element in XHTML I know what the tree 
> will look like. In HTML you don't really know that.

Perhaps it's not so obvious with the way broken HTML parsing has been 
widely implemented in browsers, but SGML does have parsing rules that do 
define what the tree should look like.

-- 
Lachlan Hunt
http://lachy.id.au/