XSchema Question 1: RDF

[Peter Murray-Rust]

At 04:08 30/05/98 -0400, Liam Quin wrote:
>On Fri, 29 May 1998, Simon St.Laurent wrote:
>> How will XSchema use/relate to RDF?
>
>"There's a difficult and complex piece of work going on at W3C that
> has clear and direct impact on what we are doing, but we're a little
> intimidated by it and want something simpler, sooner"

I think this is a useful summary. The 'clear and direct impact' may,
however, depend on how closely we limit our goals. In we are trying to
express non-DTD constraints then we would be foolish to ignore RDF. If we
are simply translating current DTDs into XSchema then we may not need it
*at present*. The question is whether a current simple approach may
seriously undermine future extension. I can't answer that since I don't
have a full enough appreciation of what RDF would require of us.

>I rather think that the best long-term result will come by working
>closely with the RDF people.  It isn't clear to me that an XSchema
>should be an RDF application, but it _does_ seem clear that where
>there is overlap, there should be compatibility.

Agreed. The question is how do we determine overlap and compatibility. For
this we have to have input from people who understand RDF, think they know
where it is going, and think they know what the timescale of the various
bits are.

[...]
>
>Perhaps the answer is, see where the overlap turns out to be, and
>make the most of it.  This in practice means (1) understanding the
>current RDF draft, (2) keeping an eye on both, (3) not being afraid
>to merge.
>
>People _will_ have to deal with RDF, so let's not make the world more
>complex than it already needs to be by doing the same things
>differently...
>
>Where XSchema does different things from RDF, though, where the overlap
>is not complete, clearly it will need to go its own way.

Here's a practical suggestion. Let's:
(a) ask any RDFophiles reading this discussion if they can suggest what the
*simplest possible* XSchema would look like in RDF notation (a small sample
would suffice). They should be able to give a flavour of this fairly shortly.
(b) *simultaneously* define what our simplest possible schema would look
like. A number of us have posted bits of such a schema (Element, Attribute,
ContentSpec and no relations/constraints). I have also been sent privately
an example of a complete proposed xschema.dtd

In this way we shan't be held back. If xschema.xsc and xschema.rdf are
algorithmically transformable, then this could be a starting point.
[Although we must remember we have required XSchemas to be simple for
humans to read and author - I'd hate to lose that.] If xschema.rdf is
considerably more complex than xschema.xsc then we will have to make a
pragmatic decision. If xschema.rdf requires tools that don't yet exist we
may have to decide on the importance of short-term progress. [I am not
averse to creating xschema.xsc as a limited-term project which will help
people manage document creation until full-strength RDF tools come along.]

	P.

Peter Murray-Rust, Director Virtual School of Molecular Sciences, domestic
net connection
VSMS http://www.nottingham.ac.uk/vsms, Virtual Hyperglossary
http://www.venus.co.uk/vhg

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/
To (un)subscribe, mailto:majordomo at ic.ac.uk the following message;
(un)subscribe 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)