Re: NESTED RELATIONAL DATABASES

Jeremy Rickard <Jeremy_at_jbdr.demon.co.uk> writes:

> In article <34328A84.318D_at_arl.mil>, Robert Winkler <winkler_at_arl.mil>
> writes
> >Michael Ortega wrote:

[snip]

> >I'm not sure that the example you use of adding a new access method to
> >the DBMS is quite on the mark. Obviously the lower levels of the DBMS
> >must be aware of the internals of such a data type to make use of it as
> >an access method
 

> Is that so obvious - you could use UDFs for this too?

I think so. Say we have what your average marketing droid will refer to as a multimedia database. If we also have UDFs that can extract portions of the BLOBs, you can define UDFs to parse and otherwise manipulate the beasts. The guts of the RDBMS need to know nothing of the meaning of the BLOB data if, say, you can have an insert/update trigger that calls a UDF to extract tidbits from the BLOB with which to, say, set some indexed traditional atomic fields that query tools can then easily use to locate specific BLOBs. So long as the knowledge about the data in the BLOBs is in the UDFs, things would seem to be dandy, so long as the UDF evaluator is decent.

> --
> Jeremy Rickard
 

-- 
Joe Foster <mailto:joe*AT*bftsi0.gate.net> Yes Colin, I *still* work for BFTSI!
WARNING: I cannot be held responsible for the above        They're   coming  to
because  my cats have  apparently  learned to type.        take CJ3 away, ha ha!