On Tuesday, 20 April 1999 1:35, Herouth Maoz [SMTP:herouth@oumail.openu.ac.il]
wrote:
> At 15:57 +0300 on 14/04/1999, Jon Barnett wrote:
>
>
> > I had expected that an empty result set would be
> > returned if the history table is empty (0 rows returned for the select),
> > and
> > queryResult.next() would be false. Is this an incorrect interpretation
> >on my
> > part?
>
> It is. In PostgreSQL, when an aggregate function is used, it always returns
> one row containing one field. If it had no values that matched the query,
> that one field is NULL.
The question is less one of "what is required to correctly interpret the result
from the postgresql jdbc driver" and more a case of determining whether this
result conforms to the jdbc api definitions. To re-phrase this; if I employed
a different JDBC driver, would I obtain the same result? For after all, the
goal of the JDBC layer is to provide a consistent interface, independent of the
database to which you are connecting, particularly when you can late bind the
driver (specify the driver you want to use at run-time).
Best regards,
Jon.
