RE: [xsl] XSL-FO versus PostScript

Subject: RE: [xsl] XSL-FO versus PostScript
From: "Roger Glover" <glover_roger@xxxxxxxxx>
Date: Wed, 26 Feb 2003 17:36:31 -0600
J.Pietschmann wrote:

> Zack Brown wrote:
> > So, to sum up your argument, PostScript does give more power,
> > but XSL-FO makes some things (footnotes, page number alignment,
> > etc) easy, that PostScript has no basic provisions for?
> Yes. PS is, in general lower level than XSLFO, you can position
> individual strings and graphic elements (=more power), but it lacks
> higher abstractions (margins, indentations, borders, justification,
> alignment, floats, page numbering, hyphenation and some more)

Right.  Most PostScript documents include a large preamble called a
dictionary (think "library").  The dictionary typically defines much of that
higher level of abstraction, depending on the generator.

> > ...but I wonder if there are any PostScript subroutine libraries out
> > there that try to bridge that gap. A quick google search didn't find
> > any.
> No surprise. Just try a "Hello world" yourself...

Actually, you might try looking for "dictionaries" instead of "libraries".

> >>inherited from CSS (the most notable immediate predecessor).
> >
> > I think TeX came before CSS. That's what I used in the
> early/mid 90's. It
> > was really great, but very rigid in ways that seemed arbitrary (like not
> > using memory that was available on the system, even when the alternative
> > was to terminate without completing its task).  In spite of its flaws it
> > was very powerful and even beautiful in its way.
> I wrote *immediate* predecessor for a reason, CSS was taken as starting
> point for XSLFO and is still quite explicitely referred.

I seem to recall the lineage being more from DSSSL than from CSS, but I
suppose I could be wrong.

> TeX was certainly one of the poineering applications in computerized
> typesetting, and in fact virtually every modern typesetting system
> still draws on the line breaking, filling, hyphenation and math expression
> typesetting algorithms first hammered out for TeX.

Which, of course, borrowed same from *it's* predecessors, including Runoff
on DEC systems, GML on IBM mainframes, and [nt]roff on UNIX systems.  These
in turn borrowed greatly from electronic typesetting machines

> However, TeX did not
> provide many good abstractions above paragraphs and formulas. It's
> strength was (and still is) that it's basically a programming language
> with a good run time library for typesetting. This allowed building many
> interesting abstractions on top of it. In fact, I think packages like
> LaTeX were a major milestone in the development of semantic markup and
> therefore in the lineage of XML.

While LaTex was certainly much easier to learn and use than the
aforementioned tools, I don't think it broke ground in any major new
functional areas.

-- Roger Glover

 XSL-List info and archive:

Current Thread