RE: Publication of RDF Data Access Use Cases and Requirements

> -----Original Message-----
> From: ext Janne Saarela [mailto:janne.saarela@profium.com]
> Sent: 09 June, 2004 13:49
> To: Stickler Patrick (Nokia-TP-MSW/Tampere)
> Cc: kendall@monkeyfist.com; www-rdf-interest@w3.org; eric@w3.org
> Subject: Re: Publication of RDF Data Access Use Cases and Requirements
> 
> 
> > Even if not all query engines will be required to support arbitrary
> > datatypes, queries should be expressable in terms of any arbitrary
> > datatype, and if the query engine understands that datatype, fine;
> > otherwise, all that can be done is exact comparison of 
> lexical forms.
> 
> What's your view to interoperability if query expression
> uses datatype such as xsd:dateTime and operator
> for the query is not equality nor not-equality?
> 
> How would an otherwise DAWG compliant server - which
> didn't support xsd:dateTime (assuming dateTime wasn't part of
> DAWG rec, which I hope it will be) - evaluate the expression?
> 
> An error message quite simply?

Clearly, it should be specified in the DAWG protocol
how the server should respond if it is unnable to execute
the query (either because the datatype is not recognized
or because the specified comparison is not supported).

An error message would seem appropriate.

Nevertheless, it should still be possible to express
queries including typed literals with arbitrary datatypes.

Patrick

> 
> Janne
> -- 
> Janne Saarela <janne.saarela at profium.com>
> Profium, Lars Sonckin kaari 12, 02600 Espoo, Finland
>

Received on Wednesday, 9 June 2004 07:11:59 UTC