Quoting David Woolley <david@djwhome.demon.co.uk>:
>> your liking. In HTML the DOM depends on the way your document is 
>> parsed which
>> might give some unexpected results. (Although it will be defined, one day.)
>
> That's not true for valid HTML.  Valid HTML always has a well defined DOM.
> Missing opening and closing tags are still logically present in the document,
> and their omission is just syntactic sugar.

I guess that's mostly true. Besides that SGMLisms are obviously giving 
different
results than you might expect from the validator I guess there are some other
minor tidbits.


-- 
Anne van Kesteren
<http://annevankesteren.nl/ >