[XSL-LIST Mailing List Archive Home] [By Thread] [By Date]

Re: Formatting Objects considered harmful

Subject: Re: Formatting Objects considered harmful
From: "Simon St.Laurent" <simonstl@xxxxxxxxxxxx>
Date: Wed, 28 Apr 1999 08:50:30 -0400

At 10:49 AM 4/28/99 +0700, James Clark wrote:
>But that's nothing new: there is the same potential for harm already
>with HTML and CSS.
>You have still totally failed to demonstrate that there is any greater
>potential for harm with XSL than there is already with CSS.  
>Both these dangers arise equally with XSL and CSS:
>(i) arises equally with CSS and XSL because the CSS rendering objects
>(or whatever you call them) and XSL formatting objects are at precisely
>the same semantic level;
>(ii) arises equally with CSS and XSL because there because HTML span
>element with a style attribute provides a syntax for CSS rendering
>objects just as XML provides a syntax for XSL formatting objects.

I think some of the XSL folks on this list are forgetting that while abuse
in XML+CSS is _possible_, the architecture of CSS discourages it, unlike XSL.

I have no interest in converting my XML documents to HTML for transmission
over the Web.  While I'll do it for as long as is necessary to support old
browsers, transformation for the sake of transformation isn't my idea of
smart processing.  Annotation is quite adequate for the relatively
straightforward documents I create, and doesn't make the same processing
demands.  If I have XML documents and a CSS style sheet for presenting
them, I'd really rather not go to the trouble of ramming them through an
extra level of transformation to turn them into junk (aka HTML or FOs.)

The very lack of transformation in CSS makes it less prone to such abuse.
Once you've mastered XSLT, you can, of course, do the same evil tricks
converting XML to HTML+CSS or FOs, but that behavior isn't encouraged in
CSS.  Indeed, the focus of CSS on external style sheets discourages such
foolishness - it's much easier to style documents with semantic markup
(even the CLASS attribute of SPAN and DIV elements will do) than with badly
obfuscated HTML.

On the XSL end, we have Guy Murphy's 'semantic firewall' as a real business
case for such 'abuse'.  On the CSS end, we have a practice that encourages
the preservation of semantic information through its foundation mechanisms.

I'd say there's quite a case here, one that really touches on those
fundamental questions of what the Web's all about.

Simon St.Laurent
XML: A Primer
Sharing Bandwidth / Cookies

 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list

Current Thread