*   more on some of the set-theoretic operations
    for relations

    union
    intersection
    difference

    *   all of these view the relations as a SET of
        rows/tuples

    *   all of these require that the relations involved
        be UNION-COMPATIBLE

        *   the operand relations have the SAME number
            of columns/attributes,

	    and those attributes have
	    (ideally) the same respective domains
	    (or type-compatible according the DBMS in "syntax" terms)

    *   union -

        A u B
 
        result is a new relation containing all of the
	rows from A and all of the rows from B

    *   intersection -

        A n B

	result is a new relation containing just those
	rows that are in A AND ALSO are in B

    *   difference

        A - B

	result is a new relation containing just those
	rows that are in A BUT NOT also in B

*   how do we express these in SQL?

    (subselect)
    union
    (subselect)
    [order by...]

    (subselect)
    intersect
    (subselect)
    [order by...]

    (subselect)
    minus
    (subselect)
    [order by...]