Kevin S. Clarke wrote:

> On 7/11/05, Leigh Dodds <[log in to unmask]> wrote:
>>I've rarely encountered
>>arguments against RDF that go beyond the syntax.
> I have questions as to whether it will scale to very complicated data
> models (like one that would accomodate all the data in MARC -- and not
> just as "controlfield" chunks).

The experience we've had so far, is that scaling in terms of size
of dataset is variable. We've hit 200 million+ triples in our store.
Querying is slow, but direct access (e.g. by known identifier) is
very fast. The only "chunks" in our data are abstracts, which may
contain some markup.

I expect performance improvements to come quite rapidly
now that RDF is seeing increased adoption. I see Oracle's 10g database
now supports RDF via its "Networked Data Model".

In terms of graph complexity, I'm not aware of any limits or "scaling"