A Plea for Schemas

[Matthew Gertner]

Rick Jelliffe wrote:
> I certainly agree that a discovery mechanism for schemas is needed, but
> I don't agree on the importance of having a single standard schema
> language.
> 
> Every schema language is built on a zillion trade-offs. The best we can
> hope for is that a schema language will be wonderful on some major
> application domain and not entirely useless generally.  Would EXPRESS be
> wonderful for describing TEI?  Would XML Schemas be wonderful for WAI
> guidelines? Would Schematron be wonderful for VML?  Probably not.
> 
> We need the equivalent of the XML Stylesheet declaration (PI) so that
> alternative schemas can be available  (perhaps with one being the
> canonical schema) and so that use of a document is not dependent on
> buying into a particular schema language.

I buy somewhat that completely different metalanguages might be needed
for very different uses of XML. This could be plausibly said of XML used
for documents and XML used for data. However, this is a trend that
should certainly be avoided if at all possible. It is true that it is
quite easy to convert from one schema language to another, but this
doesn't exactly increase the convenience of using the stuff. Imagine you
have a graphical schema editor, an instance editor, a Java class library
and a browser and they all use different schema languages, for example.
Also, schema languages are not going to be entirely equivalent. The
specificities of a particular approach will be an important guide as to
how schemas should be designed and modularized, and these won't
necessarily map well from one variant to another.

An XML-syntax-based schema language based on element type descriptions
(rather than assertions or whaterver) that includes features for
combining and extending schemas is certainly not impossible to design
(there are a bunch of variants out there already), and this would be
perfectly fine for the examples you mention. Certainly specialized
languages may emerge, but when you consider how entrenched DTDs have
been, despite their significant weaknesses, it is plausible that the
current glut of approaches is more a response to a currently unfilled
market need than an indication that we really need a so many
alternatives.

Matthew


xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To unsubscribe, mailto:majordomo at ic.ac.uk the following message;
unsubscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)