W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > September 2002

Re: RDF Datatyping PART 1 [was: Agenda for RDFCore WG Telecon 2002 -08-30]

From: Patrick Stickler <patrick.stickler@nokia.com>
Date: Fri, 6 Sep 2002 13:37:02 +0300
To: bwm@hplb.hpl.hp.com, "ext Jos De_Roo" <jos.deroo.jd@belgium.agfa.com>
Cc: w3c-rdfcore-wg@w3.org
Message-ID: <FmmWvCSxLssx.gYnnnmMC@mail.nokia.com>



_____________Original message ____________
Subject:	RDF Datatyping PART 1 [was: Agenda for RDFCore WG Telecon 2002-08-30]
Sender:	ext Jos De_Roo <jos.deroo.jd@belgium.agfa.com>
Date:		Fri,  6 Sep 2002 13:17:59 +0300


[trying to catch up w.r.t.
RDF Datatyping -- Current Working Proposal
Last Modified: 29 August 2002]
-- http://www-nrc.nokia.com/sw/rdf-datatyping.html

> PART 1: Core Proposal
> 2.1 rdfs:Datatype

so I guess we have

========================= ========================= =======
rdfs:Datatype rdf:type rdfs:Class .

	Yes.
	And also, the following was omitted but true:

	rdfs:Datatype rdfs:subClassOf rdfs:Class .

rdfs:Datatype rdfs:subClassOf rdf:Property .

	No. Where are you getting that from part 1?
	There is no datatype property idiom any longer.
		
IF                                    | THEN
-----------------------------------------------------------
?d rdf:type rdfs:Datatype .           | ?d rdfs:domain ?d .

	Again, no, as datatypes are not properties, and
	hence have no domains (or ranges).

	Are you deriving this from somewhere in the wording
	of part 1, or are you carrying it along from older
	proposals?

> 2.3 Typed Literal

I still have to try have running code where that first item
of that pair is "a URI Reference denoting a datatype"
I was/am quite believing that an abbreviated token would
be enough to support XML-Schema primitive datatypes
(and an unquoted numeral for integer numbers
like DanC's :jenny :age 10.
in http://www.w3.org/2000/10/swap/test/dt/typedLit.n3)

	With my editors hat off (which may very well remain
	off permanently), I am opposed to any native or 
	default datatypes for RDF. If folks mean xsd:integer,	
	then they must say xsd:integer, and the denotation
	of datatypes is based on generic mechanisms.


> 4. RDF Datatyping Model Theory

I would add here the imposed semantic conditions like

========================= ========================= =======
rdfs:Datatype rdf:type rdfs:Class .

	I believe the RDF schema specifies this.

rdfs:Datatype rdfs:subClassOf rdf:Property .

	Again, this assertion is false.

========================= ========================= =======

and the closure rules like

IF                                    | THEN
-----------------------------------------------------------
?d rdf:type rdfs:Datatype .           | ?d rdfs:domain ?d .

	Again, no, since a datatype is not a property.

[I wonder if there are other closure rules...]

	I don't believe there are, not for part 1.

	Regards,

	Patrick

-- ,
Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/
Received on Friday, 6 September 2002 06:38:26 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:50:56 EDT