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

> the need for "subClassOf" arises, I think, from ....

Yuzhong,

sorry for the incorrect reply to your post.
The fact is that when approached the problem I detected different possible
solution and finally chose that one I posted as the most suitable. In others
of my possible solutions (that however are too much articulated and about
which I had many dubts) the subClassOf seemed to me as necessary (even if,
looking at them now after your comments, perhaps it was not). Recently I
only tried to investigate again whether it was possible to do the same thing
in the RDFS domain without the need of OWL. But I have done it hurriedly and
absent-mindeldy, and then I made some banal errors when I unconsciously
applied the semantic I detected to resolve the problem (and that I
implemented in the OWL domain) into a domain where that semantics does not
apply.

Thanks for yuor comments.

Francesco