Question

The title speaks for itself. The first query –

SELECT    *
FROM      table _t
WHERE     (id1, id2, id3, id4) <> (x1, x2, x3, x4)

The second query –

SELECT    *
FROM      table_t
WHERE     NOT
          (id1 = x1 AND
           id2 = x2 AND
           id3 = x3 AND
           id4 = x4)

What are the differences between these two queries (i.e., is there any performance gain/loss or something) ?

Edit

To all those who think I am joking or something, I just want to say that I have used both queries in postgresql 8.4. Both queries work just fine.

Answer 1

I would not expect there to be any difference between the two queries, but it really does depend on the optimizer for you DBMS. It may be that one of these forms prevents the optimizer from “seeing” the opportunity to use an index that the other form doesn’t.

As @Jinesh says, you’d be better using whatever facilities are available in your DBMS to examine the plans the optimizer produces in these cases.