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

Re: [CSS21] properties for table-column (In HTML: COL) & table-column-group (In HTML: COLGROUP) items.

From: David Woolley <david@djwhome.demon.co.uk>
Date: Thu, 30 Jun 2005 00:03:21 +0100 (BST)
Message-Id: <200506292303.j5TN3Lb02788@djwhome.demon.co.uk>
To: www-style@w3.org

> 
> <layout>
>     <vblock>
>         <item src="body div.header" />

What you have here is a relative of XSLT, not so much of CSS, although
it uses CSS selectors instead of XPATH.  It is a new style sheet 
language.

Actually, it isn't really a style sheet language, as it doesn't seem
to allow general rules to be defined, whereas a true style sheet
encapsulates a house style policy, not the rendering of a particular
document.

An alternative, which to me better fits the way that commercial 
web pages are designed, is the tagged PDF model, where the primary
document is explicitly rendering oriented, and the parallel document
that selects bits of the primary document is the structural document
(tagged PDF inlines the structural information where the structure
doesn't conflict with the rendering).  (However, I think that a document
with real content should always be designed first for structure.)

One does have to ask the question as to why XSL is a low priority
for browsers.  I suspect that answer is that the minimum useful XSLT
"style sheet" is much bigger, and more mathematical, than the mininal
CSS styling (especially given that you can (though shouldn't) simulate
(deprecated) font elements with span and inline styles.

Other reasons are that XSL is not intended for incremental rendering
or dynamic re-rendering.

> 6) Content can be presented in a different order than it is in source.

One of the main features of XSLT and why this language is a relative.
Received on Wednesday, 29 June 2005 23:16:59 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:38 GMT