At 14:14 +0100 28/5/99, Miles Sabin wrote:
In fact, I think we're getting quite close to what I
think is the chief defect of XSL. Despite what's said
about it, it doesn't really separate data-model from
presentation all that well
Agreed
XML/XSL is a model-view type architecture. MV is good at
breaking the dependency of data-models on presentation
issues. Unfortunately it's very poor at separating
presentation issues from the data model.
I don't think this is the issue.
The big problem I see is that one target audience (graphic designers)
and possibly another ('general' public) don't think in the abstract
symbolic terms required by either traditional procedural langauges OR
declarative.
There is a wide range of literature describing how different people
perceive things differently. I'm primarily kinaesthetic and
secondarily visual - I have a lot of trouble with abstract structures
and have to work hard to get the feel of something described in
symbolic terms.
This reminds me of a discussion I got involved in with the UML language.
Rational were defining a standard for visual communication but
refused to get any visual communication experts involved. They based
the standard on an amalgamation of visual languages designed
primarily by computer scientists and engineers over the past decade
or so.
I'm not advocating that we abandon textual languages altogether but
that our textual languages should be designed with wider appeal.
Where languages are concerned with formatted layout then I believe
the focus of the language should be on the layout, not on the source
data. Where we are manipulating data, *then* is the right time to
focus on the data descriptions.
Andy Dent BSc MACS AACM, Software Designer, A.D. Software, Western Australia
OOFILE - Database, Reports, Graphs, GUI for c++ on Mac, Unix & Windows
PP2MFC - PowerPlant->MFC portability
http://www.oofile.com.au/
XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list