HTML:html@profile (was Re: Extension Mechanism for HTML)

Folks,

The profile at http://www.muzmo.com/profiles/book.html is not complete.
In particular, I have only spec'd REL/REV values, but you can easily infer
the corresponding CLASS definitions. I will flesh this out a bit if anybody
thinks that it is worth considering as an example profile.

However, please consider:

	<html profile="http://www.muzmo.com/profiles/book.html">
	<head > ...
	<link rel="toc" ...>
	<link rel="top" ...>
	<link rel="next" ...>
	<link rel="previous" ...>
	<body > ...
	<div class ="article"> ...
	<p class ="copyright"> ...
	<div class ="colophon"> ...
	<div class = "appendix"> ...

The fact that I have specified a profile means that my definitions of the 
CLASS
and REL values might be in contention with definitions provided by a 
default profile.

If multiple profiles were in play, we might have further contention, but we 
might also
discover some interesting places where contention gives us a richer 
definition of
an element's meaning.

Consider

	<span class="emphasis"> ...

One profile might define emphasis in terms of stress emphasis and another in
terms of typographic emphasis and yet another in terms of vocal emphasis.
The definition that applies is the one that is appropriate in the given 
context,
or possibly the aggregate definition.

The advantage of using CLASS over ROLE is that CLASS is already used by CSS.

Employing a profile to define your classes reduces the probability of 
contention,
and provides a URI-based methodology for resolving contention automatically
or by human intervention. It's not perfect, but it's a big improvement over 
what
we have today, and it plays well with GRDDL and thus RDF and the Semantic Web.

Regards,

Murray

Received on Tuesday, 8 May 2007 15:49:58 UTC