Re: Input sought on datatyping tradeoff

Brian McBride wrote:


>
> Because the A tests have no range constraint.  We either have to decide
> that literals are self denoting - they always denote themselves in which
> case the answer to D must be NO, or there denotation depends on a range
> constraint in which case the answer to A must be NO.

I don't follow. Assuming literals always denote themselves (which is the
"duh!" test I was refering to), why MUST D be NO?

That is to say, a _literal_ is defined as something _without_ an rdfs:range,
and a _value_ is something _with_ an rdfs:range, so why can't each have
their own "eq" rules?

>
>
> >one could have two different types of equality -- string eq and value
equal
>
> Yes.  We are talking about value-eq here.
>
> [...]
>
>
> >note that "value-equal" might be non-monotonic if the <rdfs:range>
propery
> >got detatched from the other triples
>
> that would not be non-monotonic - if you remove a triple then of course
you
> are free to remove some inferences that depend on it.  My understanding of
> non monotonicity is that you must never withdraw an inference because of
> adding new triples.
>

Perhaps I am being dense so help me, suppose I have a set of triples
_without_ an rdfs:range and I say "not equal" and then I add a triple with
an rdfs:range, and then I say "equal" how isn't that non-monotonic? Aren't
_both_ "equal" and "not equal" inferences in this specific case?

Jonathan

Received on Friday, 12 July 2002 09:50:18 UTC