I think there is something fundamentally wrong. The only relationship between poam and milestone I can see is this:Which, if it were permitted and meant something, would:Code:p.poam_id = w.poam_id (+) AND w.weakness_id (+) = m.weakness_id
EITHER produce a Cartesian product - all rows of poam would match all rows of milestone via the NULL rows of weakness in the outer join (allowing NULL=NULL to be true)
OR lose all the rows where there are none corresponding in weakness (allowing NULL=NULL to be false), in which case you don't have an outer join.
What I have just written boarders on the incomprehensible - you need a picture. I say again: LOOK TO THE DATA MODEL.




Reply With Quote