Dan wrote:
> I'm not sure I agree. XOR could be a valid and useful logical
operator in
> conjunction with UNION. Even though the attribute in both relation
> "fragments" have the same name, and we
> presume to they have the same semantics, the fact that they are
qualified in
> two different relations makes the attributes distinguishable,
assuming of
> course that they are of the same type.
"XUNION" perhaps? Where (R1 XUNION R2) is equivalent to (R1 MINUS R2)
UNION (R2 MINUS R1).
Received on Mon Nov 01 2004 - 15:46:57 CST
