Re: new syntax for datatypes

From: Dan Connolly (
Date: 02/15/01

Ian Horrocks wrote:
> Further to Peter's earlier email about the new proposal, I have now
> updated the web site with the new language specification, semantics
> and example files:

yes, this looks about right; by way of making the
benefit of off-the-shelf RDF tools evident, here's
a little diagram I consed up this afternoon...

generated using from

which is a cleaned up excerpt from

Thursday, 15-Feb-01 16:31:02 GMT

some comments:

*** I think you should have a # at the end of
  xmlns:xsd =""

e.g. the full name of decimal is

er... actually, the namespace name should be

cf. Appendix A of Part 2 of the schema spec
and the discussion of it
  URIs for primitive datatypes and facets?
  Dan Connolly (Fri, Dec 15 2000) 
(esp Martin G's reply) cited from my notes on this stuff

*** [dex] is not well-formed;
	you can only have one root element in an XML document.
	(yes, this sucks. yes, I'd like to change it.
	No, I don't expect a revision of XML 1.0 any
	time soon.)

I'm not quite sure what you're trying to do with XML Schema
stuff and RDF stuff in the same document, but this
attempt doesn't work.

*** There are a few xml-level typos. I highly recommend
you use xmlwf to check things. I invoke it ala
make from emacs's meta-x compile thingy to xml-wf-check
all my XML stuff (XHTML, RDF, etc.)

xmlwf is part of expat

** This isn't how RDF 1.0 syntax works:

  <rdf:type rdf:resource="daml:UniqueProperty"/>
In RDF 1.0, you have to spell out daml: longhand when
you use it inside a resource attribute.

*** hmm... should our cardinality propoerties now
take numbers, rather than numerals, as their values?

> After my recent reprimand I remembered to keep the previous version
> online:
> Regards, Ian

Dan Connolly, W3C
office: tel:+1-913-491-0501
  (put return phone number in from/subject)

This archive was generated by hypermail 2.1.4 : 04/02/02 EST