Use of the word "should" in OWL Rec

Hi all,

I'm a little puzzled by the use of the word "should" in the OWL 
Reference specification.

Usually this word is used according to RFC 2119, and usually also W3C 
recommendations adhere to this convention, as do the XML rec. 
(http://www.w3.org/TR/2004/REC-xml-20040204/#sec-terminology), XPointer 
rec. (http://www.w3.org/TR/2002/PR-xptr-framework-20021113/#dt-must), 
and also RDF Semantics rec. (http://www.w3.org/TR/rdf-mt/). In the OWL 
Reference, however, I miss such a clarification on the meaning of "should".

Moreover, it seems that the OWL Reference in itself is not clear what 
the word means, or at least it seems to contradict itself. Taking 
Section 3.1.2.1.1 owl:allValuesFrom 
(http://www.w3.org/TR/owl-ref/#allValuesFrom-def)
it says

"In other words, it defines a class of individuals x for which holds 
that if the pair (x,y) is an instance of P (the property concerned), 
then y should be an instance of the class description or a value in the 
data range, respectively."

which seems to contradict the second sentence following the above one

"[..]; just that this is true for individuals that belong to the class 
extension of the anonymous restriction class."

Well, shouldn't that be written as "[..]; just that this SHOULD BE true 
for individuals ..."

Should one have an answer to this?

Greets, Martin

Received on Tuesday, 13 July 2004 04:05:55 UTC