[SCL] RDF in an SCL world

[Jim Hendler]

At 11:35 -0400 7/16/04, Bill Andersen wrote:
>On Jul 15, 2004, at 11:53 PM, Jim Hendler wrote:
>
>>>  My prediction:  People will extract RDF data, put it
>>>  into SQL databases, use SQL constraints instead of
>>>  OWL, and use SQL as the query language.
>>
>>  OWL is less expressive than SQL???  Where does that follow from John?
>
>In that pretty much all commercial implementations of SQL have 
>iteration, I believe (am not 100% -- Enrico?) that they are 
>computationally complete.  That said, that makes them more 
>expressive than any flavor of OWL.
>

nope, SQL is not Turing equivalent unless you include recursion, 
which is not generally allowed - also, SQL is closed world and OWL 
(that is, OWL Full, which is the only one I care about) is open world 
- unless I misunderstood some things Pat has said.

>Now, why anyone would want to use SQL DML to express constraints I 
>don't know.  Jim is right if he's trying to make the point that OWL 
>is declarative and declarative is the way to go.

I agree w/Bill on this - OWL is more like the "data schema" language 
for RDF DBs than like the query language for them

-- 
Professor James Hendler			  http://www.cs.umd.edu/users/hendler
Director, Semantic Web and Agent Technologies	  301-405-2696
Maryland Information and Network Dynamics Lab.	  301-405-6707 (Fax)
Univ of Maryland, College Park, MD 20742	  240-277-3388 (Cell)