On Tue, 22 Aug 2006, Patrick H. Lauke wrote:

> Jukka K. Korpela wrote:
>
>>> <var>i</var> => <code role="compsci:variable">i</code>
>> 
>> Some automatic translation programs recognize <samp> and/or <code> markup 
>> and treat the element's content as something that shall not be translated.
>
> Wasn't XHTML 2.0 supposed to be NOT backwards compatible, or do you mean 
> there are automatic translation programs who currently recognise XHTML 2.0 
> already?

XHTML 2.0 is supposed to be incompatible with any previous version of HTML 
(so that it would be more appropriate and less confusing to give it a 
completely new name and start the version numbering of the new language 
from 1.0, but version number confusion is an HTML tradition, starting from 
HTML 2.0). What I was referring to was that <samp> and <code> elements are 
examples of simple and meaningful markup that has some actual software 
support beyond mere rendering issues. The more complex a markup language 
you design, with semantics hidden obscurely into attributes, the more probable 
is that it will remain effectively as a system of text processing macros:
people use markup to achieve the default rendering that they associate 
with markup elements - but software developers won't use the semantic 
information for anything.

It's not just a matter of the difficulty of finding the semantic pieces; 
more importantly, if the markup system is complex, people won't use the 
complex features or they will use them inconsistently and against the 
specifications.

-- 
Jukka "Yucca" Korpela, http://www.cs.tut.fi/~jkorpela/