Re: URIs for primitive datatypes and facets?

There seem to me two possible alternatives:

1.  Change it so that #name is the primary address
- This is obviously much more straightforward and easier to use
-  This will mean that content negotiation on the schema would allow
   versions in non-XML languages to be returned with preserved frag id
- Yes you need in principle to use a schema to know that id is an id
   but the reservation of "id=" should perhaps be reconsidered to
    obviate this need for this in future.

2. Consistently use #xpointer(id("date")) as  the identifier.
 - within the document, replace id=foo with ref="#xpointer(id("date"))"
 - outside the document recommend the use of
    #xpointer(ref("#xpointer(id("date")) ")) in order to be sure to
    match with the above even without a schema.

But seriously, there is a certain level of usability and simplicity which
is necessary here.  Also, propose a requirement that an application
be able to use XML schema datatypes without being Xpointer-aware,
on the general principle of data hiding.

Tim
sans chapeau.


----- Original Message -----
From: "Dan Connolly" <connolly@w3.org>
To: <www-xml-schema-comments@w3.org>; <www-rdf-logic@w3.org>;
<joint-committee@daml.org>
Sent: Saturday, February 17, 2001 2:20 PM
Subject: Re: URIs for primitive datatypes and facets?


> > From: "Martin Gudgin" <marting@develop.com>
> > Date: Mon, 18 Dec 2000 16:01:39 -0000
> [...]
> >
> > You're correct that the schema annotation says that the explicit
xpointer
> > version of the URI is the 'correct' one. Personally I think I prefer the
> > shorthand. Perhaps you/we could raise this as an agenda item and try to
get
> > it changed to be the shorthand version?
>
> I'm very interested to know how this turns out; I see that
> the chair proposes to decline our request,
> http://www.w3.org/2000/12/xmlschema-crcomments.html#datatype-uri
> but I don't see the final disposition/wg-decision.
>
> www-rdf-logic and joint-committee folks, if you have an opinion
> about whether #xpointer(id("date")) is acceptable, or whether
> it's an undue implementation burden, compared with #date, please
> speak up now.
>
> The implementation burden is something like what Sandro described:
>
> [[[
>     We can solve this easily for the standard datatypes by putting all
>     the classes in some ontology, publishing the classes you've been
>     using (and using a nicer identifier syntax).
> ]]]
>
> --        www-rdf-logic@w3.org from February 2001: a few issues with
> daml
> http://lists.w3.org/Archives/Public/www-rdf-logic/2001Feb/0091.html
> Fri, 16 Feb 2001 14:42:07 GMT
>
>
> > Gudge
> >
> > ----- Original Message -----
> > From: "Dan Connolly" <connolly@w3.org>
> > To: "Martin Gudgin" <marting@develop.com>
> > Cc: <www-xml-schema-comments@w3.org>; <joint-committee@daml.org>
> > Sent: Monday, December 18, 2000 3:36 PM
> > Subject: Re: URIs for primitive datatypes and facets?
> >
> > >
> > >
> > > Martin Gudgin wrote:
> > > [...]
> > > > >From Appendix A of Part 2[1]
> > > [...]
> > > > [1] http://www.w3.org/TR/2000/CR-xmlschema-2-20001024/#schema
> > >
> > > Thanks for the pointer. But... which is it?
> > > ...XMLSchema#date or ...XMLSchema#xpointer(id(date)) ?
> > >
> > > They're defined to mean the same thing, in the case of
> > > an XML representation of the schema, but I'm looking
> > > for opaque identifiers.
> > >
> > > My reading of the spec says it's the latter
> > > (ugh!).
> > >
> > > > Data type    URI
> > > >
> > > > string       http://www.w3.org/2000/10/XMLSchema#string
> > > > boolean      http://www.w3.org/2000/10/XMLSchema#boolean
> > > > float        http://www.w3.org/2000/10/XMLSchema#float
> > > > double       http://www.w3.org/2000/10/XMLSchema#double
> > > > decimal      http://www.w3.org/2000/10/XMLSchema#decimal
> > > > timeDuration http://www.w3.org/2000/10/XMLSchema#timeDuration
> > >
> > > or
> > >
> > > >       For example, to address the date datatype, the URI is:
> > > >
> > > >         http://www.w3.org/2000/10/XMLSchema#xpointer(id("date"))
> > >
> > > ?
> > >
> > > > Hope this helps,
> > >
> > > Quite a bit; thanks!
> > >
> > > > Gudge
> > > >
> > > > [1] http://www.w3.org/TR/2000/CR-xmlschema-2-20001024/#schema
> > >
> > > --
> > > Dan Connolly, W3C http://www.w3.org/People/Connolly/
>
> --
> Dan Connolly, W3C http://www.w3.org/People/Connolly/
>

Received on Tuesday, 20 February 2001 00:24:56 UTC