RE: incomplete datatyping (was: Re: datatypes and MT)

> Thanks for the info.
>
> This actually makes the example of the size property
> even more meaningful, as it shows the ambiguity even
> more clearly. Since there is no single derived type
> "responsible" for disambiguation, there is no explicit
> means to decide *which* range should be used to interpret
> the non-locally typed literal.
>

RDF Schema does not indicate how one chooses which schema to use.

I assume its the natural one, at the end of the apppropriate URL.
Alternatively there should be a mechanis, like xsi:schemaLocation for
specifying it. As a third possibility the application just knows.

Any single schema should be clear enough, i.e. it is a bug with the schema
if it is not possible to decude which *range* should be used.

A system for selecting the schema then resolves the "different people say
different things" problem.




Jeremy

Received on Wednesday, 7 November 2001 09:20:05 UTC