Re: XHTML as XML

Jesper Tverskov wrote:
> I am talking about XHTML 1.1 served as xml. If you validate such a page,
> the id attribute in the html element is ok but not the class attribute.
> Since I would like to be able to validate my pages, I can't use the
> class attribute in the html element but it is true that the browsers
> don't care. They only test for well-formedness.

You can still validate your document and ignore that error.


> Nice to know that Opera considers it a bug, what about Amaya? It is much
> more important to convince the web community why such a radical new
> approach to the body and the html elements is necessary when XHTML is
> served as xml.

Most people are convinced. This difference in rendering for HTML and 
XHTML has been advocated on weblogs for several years now.


> Web developers are going to hate W3C for this, and it's going to make
> the transition to an xml based internet more difficult. The arguments
> for such a radical new approach better be good or the new approach will
> help undermine the authority of the W3C.

Untill Microsoft Internet Explorer market share is below 10% there is no 
such thing as XML based internet. (At least not in terms of webpages and 
websites.) Even then, browsers will still support text/html.

Most people using application/xhtml+xml know that there is a difference 
in CSS rendering. It also makes sense. In XML documents like:

  <foo>
   <bar/>
  </foo>

... there is no BODY element. Why should XHTML documents be an exception 
to this?

Personally, I think the CSS WG should have introduced something to 
select only the canvas or the viewport so that when 'background' is 
applied to the root element in XML documents nothing fancy happens.

(I believe there are @canvas or similar proposals out there, they just 
haven't been published yet in official drafts.)


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

Received on Thursday, 10 February 2005 11:18:37 UTC