I'm reminded of a questionnaire that went round my college
asking "do you think there is a demand for a photographic darkroom?" (you can
tell my age). To which I responded, "I don't know, why not do a survey to find
out?". The point being, no-one has a bird's eye view of the market,
especially in an area where so many tools are free; you can only find out what's
happening across the board by asking people what they are doing individually,
and aggregating the results: you can't find out (with any level of reliability)
by asking people what their opinion of the big picture is.
I see extensive use of XML Schema and DTDs, perhaps with a
bias towards XSDL in data-oriented work and DTDs in document-oriented
applications. I have to say that I have heard an immense amount of talk about
Relax NG and Schematron (and much praise for both of them) but have very rarely
seen either used in anger. But my worm's eye view may be completely
wrong.
As well as data validation, XML Schema is of course widely
used in data binding applications. Schema-aware XSLT and XQuery processing,
however, is still in its infancy - those who use it recognize the value in
speeding up the development cycle, but those who don't tend not to realise what
they are missing.
Michael Kay
|