Re: Concern about limiting allowed dataranges in DatatypeRestriction (for ACTION-83)

Although the nesting does not add expressive power, I would prefer that
the nested syntax also be allowed, in part because this would mirror XML
Schema datatypes (which has a notion of the underlying datatype) and in
part because it would pave the way for sharing parts of data range
constructs.

peter


From: Michael Smith <msmith@clarkparsia.com>
Subject: Concern about limiting allowed dataranges in DatatypeRestriction (for ACTION-83)
Date: Wed, 06 Feb 2008 15:53:18 -0500

> To meet obligations of ACTION-83:
> 
> In today's telecon, Boris was arguing for restricting the allowable
> dataranges in datatypeRestriction.  In particular, he was citing
> problematic examples like
> 
> DatatypeRestriction ( DataComplementOf ( xsd:integer ) minInclusive 5 )
> 
> I agree that no use case has been provided for this example, but was
> concerned because limiting the construct to only datatypeURI would
> prevent "nesting" datatype restrictions, such as:
> 
> DatatypeRestriction ( DatatypeRestriction ( xsd:decimal ) minInclusive
> 5.0 ) maxInclusive 6.0 )
> 
> Upon re-evaluating the constructs in light of the change to DataRange
> made for [ISSUE-28] on 2008-11-21 [1], I see that this can be expressed
> without nesting as in:
> 
> DatatypeRestriction ( xsd:decimal  minInclusive 5.0 maxInclusive 6.0 )
> 
> So, having clarified the current state of the draft, my concern is
> resolved and I support Boris' proposal to limit DatatypeRestriction to
> datatypeURI.
> 
> Thank you Boris for addressing my concern 3 months ago!
> 
> -- 
> Mike Smith
> 
> Clark & Parsia
> 
> [ISSUE-28] http://www.w3.org/2007/OWL/tracker/issues/28
> [1] http://www.w3.org/2007/OWL/wiki/index.php?title=Syntax&diff=1341&oldid=1327
> 
> 

Received on Wednesday, 6 February 2008 21:23:11 UTC