On Friday 22 November 2002 9:41 am, DPawson(_at_)rnib(_dot_)org(_dot_)uk wrote:
A few people picked up on the selection and transformation part, but it
wasn't seen as sexy in DSSSL.
I'm intrigued as to why this is, and what the difference was then. For me, I
find the difference between XSLT and XSL-FO is immense - one is simple to
understand (although justifiably incredibly detailed and with a
microscopically anal depth to it matched only by the legal profession) and
the other simply makes zero sense to me whatsoever, no matter how hard I try
and 'get it', no matter which way up I hold it - it contains no 'meaning',
but in an amazingly homogenously impenetrable state. They're about as far
apart as doing the washing up and teaching your dog to plan a lunar landing
mission (including finding the budget).
So, what kind of mentality would ever stumble across the 'T' part of what was
to become XSLT and consider that it remotely has a place in a stylesheet
language, which is after all, concerned with making your fonts bold and other
nice harmless endevours. Surely devils invaded and changed all the plans
overnight, one halloween, and nobody noticed the next morning and simply
carried on with the perverted course of style.
I really take issue with the use of 'style' in the name of XSLT. It's not
about style, it would seem. It seems more about structure. It's essentially a
means of restructuring a document. This isn't style. I've never, as a
designer, needed to specify a wholesale alteration of structure, as part of a
style sheet. Style sheets occur at the phase between a designer and a
finished artist, or between a designer and a typesetter, but generally,
structure isn't communicated within a style sheet. Structure is decided at
the beginning, with thumbnail sketches or marker roughs.
However, in the current and future days of dynamic documents, it would be
sensible to see structural alterations apply at a later part in the process -
make the deviation into alternative structures as late as possible. This
makes sense - use a structural specification (but don't call it a style sheet
- it has zero to do with style) - or use many different ones, all from the
same pool of resources and artistic materials.
In a sense, I see XSLT as a kind of alternative phase-space-like specification
of structure. It, on its own - unapplied - represents a completely different
dimensional view of the journey between a document's raw materials on the
pasteboard, and the end product on the shelves. It's that complicated - the
kind of person who can somehow 'see' the altered dimensional space of an XSLT
structural specification is looking in a completely different direction to
the designer of the original document, who in turn is looking in a completely
different direction to the customer or reader.
XSL-FO on the other hand is an almost pedantically precise mapping of what
things should look like (but not at page-level, rather at pre-imposition
level). How can the same people be responsible for both XSLT and XSL-FO?
Proof, as if any is needed, that the W3C does have contact with alien
technology, I reckon.
--
Ian Tindale
XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list