Re: typed containers in RDFS: suggestions about the "long range" problem

> There is no need to use "rdfs:subClassOf ".

the need for "subClassOf" arises, I think, from the fact that I want to
constrain only the pocabulary schema and the processor, but not who writes
RDF models.
That is, I want the long range for a property be declared in the schema as
follows:

<rdf:Property rdf:ID="listApples">
  <rdfs:domain rdf:resource="#theDomain" />

  <rdfs:range rdf:resource="#BagOfApple" />
</rdf:Property>

but then I want to manain the possibility for this property to be asserted
not only through an explicit instance of BagOfApple, but through a generic
rdf:Bag too, e.g.:

<rdf:Description rdf:about="#my Subject">
    <rdf:type rdf:resource="TheSchemaNamespace#theDomain" />
    <voc:listApples>
        <rdf:Bag>
            <rdf:_1>
                    <voc:Apple rdf:ID="item_1">...</voc:Apple>
            </rdf:_1>
             ......
        </rdf:Bag>
    </voc:listApples>
</rdf:Description>

> Your topic (typed container) is very important, and the solution is
enlightening. Thanks!

Tnanks a lot :-)

Regards,
Francesco