Comments on RDF Schema specification

1.	Now that XML Schemas are in Last Call, I think it is appropriate
that the atomic datatypes used in examples as values for rdfs:range be based
on the actual XML Schema Datatypes.  Even if an RDF schema for these
datatypes has not been written yet, this change would make the examples more
realistic.

For instance, Example 3.2 might read:

<rdf:Description ID="rearSeatLegRoom">
  <rdf:type resource="http://www.w3.org/1999/02/22-rdf-syntax-ns#Property"/>
  <rdfs:domain rdf:resource="#PassengerVehicle"/> 
  <rdfs:domain rdf:resource="#Minivan"/>
  <rdfs:range
rdf:resource="http://www.w3.org/1999/XMLSchema-datatypes/decimal"/>
</rdf:Description> 
Ditto for the example in section 7.1.

Even better would be explicit text defining this usage formally, going into
more detail than the last paragraph in Section 1.  Notice that this is also
Recommendation 9 of the Cambridge Communique.

2.	It would be useful to include a discussion of the relationship
between the range constraint as applied to container vs. literal objects.
That is, is it possible to constrain a property to refer to a resource with
rdf:type of rdf:Bag using rdfs:range?  If not, and these concepts are
orthogonal, an explanation of the reasoning.

In other words, some official guidance regarding the "Constraining
Containers" thread last month on www-rdf-interest
(http://lists.w3.org/Archives/Public/www-rdf-interest/2000Apr/0068.html).

Regards,

-Daniel

Daniel Lipkin
Chief Architect
Saba
650-581-2577
dlipkin@saba.com  

Received on Monday, 1 May 2000 18:00:40 UTC