Mike Nolan <nolan(at)gw(dot)tssi(dot)com> writes:
> Is the error message below a bug?
> select substr(memid,1,1) as memtp, substr(memid,2,4) as newx
> from memmast group by memtp, newx
> ERROR: column "memmast.memid" must appear in the GROUP BY clause or be used in
> an aggregate function
Works for me in every branch back to 7.1 ... what version are you using?
regression=# create table memmast (memid varchar(8));
CREATE
regression=# select substr(memid,1,1) as memtp, substr(memid,2,4) as newx, count(*) from memmast group by memtp, newx;
memtp | newx | count
-------+------+-------
(0 rows)
regards, tom lane