Re: [CSS21] Make XHTML <body> magic just like HTML <body>

Anne van Kesteren wrote:

> How is that ideal? You'd always need to know the markup language for the 
> structural elements and the nesting of them, etc. Besides, for all text 

The structural elements are identified in the style sheet.

The reason for requiring explicit closing tags in XML, is
not the popularly believed one that it means that browsers
will reject invalid documents, or that HTML syntax is in some
way ambiguous, but so that the nesting structure of the document
is explicit in the document and the parser can recover it without
knowing anything other than XML syntax rules.

The optional tags in HTML mean that whole elements can only be
inferred by knowing the detailed syntax of HTML.  (Although, those
who want the supposed validation advantage of XML are free to make
all tags in HTML explicit (some older, tag soup, browsers can get
upset by having some of the end tags, though).)

> documents distributed over the web you'd use a non-proprietary language, 
> ideally.

But those may be domain specific, e.g. an XML invoice's detailed syntax
needs to be known to an accounting program, but not when simply being
displayed to a human.  That allows one to have efficient EDI formats, 
with published specifications, which are still usable by people with 
only generic software.  It also means that generic software can edit a 
document and provide visual feedback, removing the need to have editors 
for every domain specific format.

Received on Tuesday, 17 April 2007 07:07:20 UTC