Re: [XHTML2] How are UAs to interpret <h> and <hx> elements?

On Fri, 10 Jun 2005, Jim Ley wrote:
> 
> Whilst I agree H and Hx is relatively under specified in the XHTML 2 
> draft - do we really want to make document outlines all the same?  I 
> would consider document outlines the sort of value added functionality 
> that can differentiate between user agents, I certainly can't see any 
> requirement for interopability in them, they're not content, they're 
> just a convenient overview for the user, what the user wants here should 
> be paramount, not some defined algorithm.

If one UA considers section A to start at point X and end at point Y, but 
a second UA considers section A to start at point X and end at point Z, 
and they are both conforming implementations, then obviously the user is 
not going to be guarenteed to receive the intent of the author. And that's 
what interoperability is about -- making sure that regardless of the UA, 
the author and the user are on the same metaphoric page.

At the moment, as an implementor, I would have no idea where the end of a 
section lies, and often would have no idea whether a particular header 
brings in a new subsection or closes the previous section(s).

It's not document outlines that I have a problem with, it's finding the 
meaning of the document (the semantics) -- which is, after all, the whole 
point of XHTML2. Outlines are just one concrete example of where this 
information could be used.

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'

Received on Friday, 10 June 2005 00:34:43 UTC