I agree with Christian, and think that this is a very important point to make. From doing peer reviews, I've found that it's a very common misconception that DISTINCT takes a parameter which in some way restricts its "distinctness".