Peter Eisentraut <peter_e@xxxxxxx> writes:
> Am Freitag, 30. September 2005 09:03 schrieb Oliver Siegmar:
>> - Is this requirement (all fields in GROUP BY) based on the SQL standard?
> What you describe is called "functional dependencies" in the SQL standard.
> (The query processor would recognize that the non-primary key columns
> "functionally depend" on the primary key and therefore do not have to be
> grouped separately.) This is definitely a valid feature but PostgreSQL does
> not implement it yet.
Note also that the "functional dependencies" stuff was added in SQL99.
Postgres' current behavior corresponds to the SQL92 standard.
regards, tom lane
---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
choose an index scan if your joining column's datatypes do not
match
