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

Re: [xsl] Re: xsl 2.0?

Subject: Re: [xsl] Re: xsl 2.0?
From: Paul Tyson <phtyson@xxxxxxxxxxxxx>
Date: Sun, 10 Nov 2013 18:39:41 -0600

It is clear there are substantial barriers, itemized by Liam, to W3C
fulfilling the promise of XSL-FO by carrying it through to 2.0 and

I believe this leaves a gap in the processing standards for XML and
HTML, one that will become more troublesome before it goes away.

A toolchain comprised solely of XML processing tools (xslt, xproc,
xquery, etc.) is better in many ways than one that has some XML tools
and some CSS tools. It is a shame to let CSS outpace XSL-FO, forcing
users who want to use advanced formatting features to extend their
toolchain and processing paradigm to deal with the CSS syntax and
processing model.

CSS is terrific for what it does, and as another poster on this thread
said, this is not about either/or, better/worse, or turf war. But beyond
the vocabulary of property names and a largely compatible semantics for
that vocabulary, the paradigms and processing models of CSS and XSL-FO
are different. If all you want to do is "apply style to HTML (or XML)",
then, from the outside, it is immaterial what approach you take. And if
your world is exclusively HTML, then CSS is a good choice.

But many industrial users work in a world where HTML is just a subset of
their requirements--the tip of an iceberg that is mostly about selecting
meaningful chunks from multiple large variegated datasets and presenting
them in useful formats, one of which might be HTML. Those who (like me)
want to use the power and simplicity of XML processing for these tasks,
need a set of standards to support the process end-to-end. Not having
advanced formatting objects capabilities leaves a gap in that processing
chain, as I have only recently discovered.

So far I have been able to work around the deficiencies without much
trouble, because I have only implemented simple features. But I have in
mind more sophisticated formatting and behavior that cannot be specified
in XSL 1.1. Ideally it would be possible for XSL-FO to specify all the
formatting and many of the behavior features possible in HTML5, PDF, and
modern office document formats.


On Sat, 2013-11-09 at 20:40 +0800, Liam R E Quin wrote: 
> On Fri, 2013-11-08 at 09:20 -0500, Charlie wrote:
> > I've followed this thread with great interest.
> > 
> > I'm the in-house XML Developer for a large US defense company. XSL-FO is 
> > used extensively in military technical publishing. We have complex 
> > documents with extensive cross referencing and strict pagination 
> > requirements. I cannot see CSS ever supporting those requirements.
> Since I don't think any MIL-STD people turn up at CSS meetings, I agree
> it's unlikely, but you should not be too dismissive. I remember an
> accountant at a former employer telling me that credit card companies
> would never accept payments over the Internet.

Current Thread