Re: HTML/XHTML

On Apr 14, 2007, at 19:28, Schalk Neethling wrote:

> 2. Will the next version of XHTML be (X)HTML5 or XHTML 2.

If this WG chooses to use WHATWG HTML5 as the starting point and  
preserves this aspect of the draft, the updated version of XHTML 1.0  
will be XHTML5.

XHTML 2.0 is incompatible with XHTML 1.0 but uses the name "XHTML"  
nonetheless.

> 3. A lot of developers were under the impression that XHTML is the  
> future of the web and that HTML will eventually disappear as/when  
> the majority of web browsers fully implement the xhtml+xml mime  
> type and one can therefore fully utilize XHTML. Of course, as there  
> is now a working group officially working on a new functional  
> implementation of a next version of HTML this is obviously not the  
> case.

HTML is not going to disappear.

> So, going forward will the HTML spec and the XHTML spec mirror each  
> other with the only difference being the serialized form and  
> therefore the parser used to parse the code?

In the current draft of WHATWG HTML5, HTML5 and XHTML5 are  
alternative serializations for the same trees for the most part.  
However, there are subtle differences. For example, the HTML5  
serialization cannot represent structured inline content in <p>, so  
an XHTML5 document isn't guaranteed to have an HTML5 serialization.  
In general, the meaning of every conforming HTML5 document can be  
mapped to conforming XHTML5, but in some cases there are slight  
differences in how things are represented:
  * xml:lang vs. lang
  * <html xml:base='...'> vs. <base href='...'>
  * <?xml version="1.0" encoding="..."?> vs. <meta charset='...'>
(I am not considering comments to be part of the meaning of the  
document.)

The corner case where a conforming HTML5 document cannot be mapped to  
an XHTML5 document without dataloss is the case where the HTML5  
document contains (in attribute values or in element content) ASCII  
control characters that XML bans (e.g. form feed).

In addition, some APIs and CSS features behave differently (for  
legacy reasons) if the DOM was flagged to be an HTML document.

-- 
Henri Sivonen
hsivonen@iki.fi
http://hsivonen.iki.fi/

Received on Saturday, 14 April 2007 21:23:36 UTC