Semantic versus Structure for XHTML 2.0

Hi,

most of the time, the semantics of elements is misused, or not used 
at all. My concerns is that often people who care about the semantics 
and develop tools for it, have finally a bunch of document which are 
not useful on the Web.

On the Authoring tool side:
	- You have only a few requirement in the XHTML 2.0 spec about 
Authoring tool. I think, it's very important to improve that side and 
to give some user scenarios when it's possible.
	For example, Should an authoring tool propose when it detects 
an address, "do you want to mark it as an address?"

My point is that the semantics element: kbd, code, samp, address etc 
are an inconsistent set of elements and not enough complete.
	We don't have elements like street, city, date, poem, verse, 
abstract, introduction, author, etc.

We could try to identify all the elements we need to put in HTML, but 
I think it will a huge amount of work and necessary useful.

I would encourage a solution where the XHTML spec becomes just a 
structure spec, with Paragraphs, lines, etc and not semantics at all.

We should put the semantics in an attribute with to extend a set of 
normative values outside of the spec.

So it will become an extensible mechanism.

<p sem="address">
<l sem="person">Haruki Murakami</l>
<l sem="street">Omote-Sando</l>
<l sem="city">	Tokyo</l>
</p>

The values of sem attribute and their meaning will be defined in a 
external extensible document.

-- 
Karl Dubost / W3C - Conformance Manager
           http://www.w3.org/QA/

      --- Be Strict To Be Cool! ---

Received on Thursday, 15 May 2003 11:00:28 UTC