W3C home > Mailing lists > Public > www-style@w3.org > November 2005

RE: Why Binding Scripting in Style Layer Conflates Semantics

From: Shelby Moore <shelby@coolpage.com>
Date: Fri, 25 Nov 2005 20:21:00 -0500 (EST)
Message-ID: <4767.203.168.9.135.1132968060.squirrel@webmail6.pair.com>
To: www-style@w3.org

I still didn't cover the entire essense of Mark Birbeck's posts.  I want
to argue against my own logic to see if it will hold up.

I believe that Mark Birbeck is also implying that if we subclass markup
semantics at the contained children, then the styling of the parent markup
is not semantically bound to the semantics of the parent, but to the
subclassed semantics of the contained children:

<select>
   <option class='city'>Shanghai</option>
   <option class='country'>China</option>
   <option class='longitude'>119.9843</option>
</select>

So then why not do from CSS?

select { bind SmartDataSelect.xml }

The above style layer binding would _hopefully_ bind to code
(SmartDataSelect.xml), which detects the semantics of the contained
children when doing the presentation.

But for example if the code (SmartDataSelect.xml), implements (subclassed)
semantics other than that implied by the semantics of an unspecialized
<select> coupled with specialized contained children, then we will have
obscured semantics from the markup layer.  Oh sure, the CSS could be
parsed to see what semantics we are applying, e.g. SmartDataSelect.xml,
but this architecture of embedding markup semantics in style would not
interopt well, specifically with the other description languages of
Semantic Web (RDF, etc) which are necessary to give ontology to semantics.
 Not to mention the complexity it adds to both style and markup by
conflating the two.

Whereas, we can accomplish the same effect with XSLT, and avoid the
problems mentioned in previous paragraph.  The output of an XSLT
transformation could be:

<select x:Class="SmartDataSelect">
   <option class='city'>Shanghai</option>
   <option class='country'>China</option>
   <option class='longitude'>119.9843</option>
</select>

Thus via XSLT, we get the rich styling via subclassing, without the
obscuring markup semantics.

Again I thank Mark Birbeck for his very important post.  Mark, you've
helped clear up a few issues in my mind.

So how does CSS style arbitrary subclasses?  Does CSS properties only
apply to normative type and those that inherit from them?  How do we
decide which properties are style and which are semantic?  That is what I
am hoping to see in "Future of CSS in Semantic Web?" thread.


-- 

Kind Regards,
Shelby Moore
http://coolpage.com
Received on Saturday, 26 November 2005 01:21:27 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:41 GMT