W3C home > Mailing lists > Public > www-style@w3.org > September 2005

Re: Are there W3C definitions of presentation and content?

From: Gavin Kistner <gavin@refinery.com>
Date: Mon, 19 Sep 2005 07:14:16 -0600
Message-Id: <062BA3E6-FC0D-486F-85D1-388661CDF90C@refinery.com>
Cc: www-style@w3.org
To: White Lynx <whitelynx@operamail.com>

On Sep 19, 2005, at 3:14 AM, White Lynx wrote:
> I would pose question otherwise. If LaTeX typesetting system is  
> capable to generate hyperlinks,
> if XSL formatters can generate hyperlinks, if DSSSL renderers can  
> do this, why CSS formatters
> should not be able to do the same?

My initial reaction is:
   * XSL is all about creating content based on other content
   * LaTeX seems to muddy the waters somewhat. Obviously TOC  
generation is beneficial.
   * I have no idea what DSSSL is
   * But CSS is tryin' really hard to separate content from markup.

> After all CSS is style language that is inteded to present XML/SGML
> document to user, making page user friendly is not just changing  
> colors and font styles, it may require
> generating extra notes, links, embedding external content etc.

But now you're tempting me, because I really would like to be able to  
use CSS to shove ancillary sidebar content as footnotes and create  
links to it.

So I revise my initial reaction to this belief:
   * Author-specified hrefs are absolutely content. They belong in  
XML/HTML/content source.
   * Auto-generated hrefs (by CSS or XSL or script) are non-content  
if such are merely a way to provide UI to the user
     * (But if abused, you could use CSS to include informational,  
content-related URLs.)

Still, IMHO, the 'correct' way to do this XML->XSL->XHTML+CSS,  
leaving the link generation to the XSL stage. But I can't say, from a  
purist standpoint, why it makes more sense to me. Should CSS be able  
to decide whether a list of items should be rendered as radio buttons  
versus a drop-down select? No, in my opinion. This is the realm of XSLT.
Received on Monday, 19 September 2005 13:14:22 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:27:20 UTC