Don Isgitt <djisgitt(at)soundenergy(dot)com> writes:
> I am trying to find rows in a table that are duplicates only in the
> sense of certain columns; i.e, not all columns are the same, and then
> for each duplicate entry, I want to retrieve one of the columns that is
> not part of the duplication criteria.
Try SELECT DISTINCT ON. There's a simple example in the SELECT
reference page (something about weather reports I think).
regards, tom lane