Re: SUO: Category Theory
> For folks thinking of implementing IFF let me say a few words. I suppose
one
> could start at any module. For example, folks interested in
implementations
instead of *implementations* read *applications*
> The central notions in the IFF-TCO are *collection*, *relation*,
*function*
The *relation* here is *binary relation*. N-ary relations are represented in
the IFF (in the IFF-MT and IFF-ONT), but are not used in the IFF. That is,
for the IFF the notion of n-ary relations only appears at the object-level.
There also needs to be some discussion of constraints on the implementation
of the IFF namespaces.
Robert E. Kent
rekent@ontologos.org