Question

Struggling with a SQL query of mine here.
I have a table:

APPS(id, game)
Where id is the primary key.

What I was trying to do is show all the id’s who use at least all the games that the person with the id “Tim” does.
I have a query that works, and returns what i need it to, but i cobbled it together from forums.
So what I’d really like is a quick and basic rundown of how the query works. Heavy detail isn’t necessary, as i understand most of the terminology, but just what the query does, and how it determines the correct records.

My query is:

SELECT id
FROM APPS X
WHERE NOT EXISTS
     (SELECT *
      FROM APPS Y
      WHERE id='Tim' AND NOT EXISTS
           (SELECT *
            FROM APPS Z
            WHERE Z.id = A.id AND Z.id = Y.id));

Thank you so much for any help!

Edit: The query works fine, but what i was after is how it works and how it returns the records it does.

Answer 1

The query works fine, but what i was after is how it works and how it
returns the records it does.

Right, so you’ve hit the problem called Relational Division. Joe Celko has written a good article about that.

The query that you posted is really strange:

WHERE Z.id = A.id AND Z.id = Y.id

In relational division, the innermost not exists query has two equality statements for different columns. For example, from Joe Celko’s post:

WHERE (PS1.pilot_name = PS2.pilot_name)
      AND (PS2.plane_name = Hangar.plane_name)

Comparing pilot_name and plane_name makes sense, where comparing id twice makes no sense at all.