Looking for the easiest / most scalable way to make a “difference” in SQL Server, see below. 
If you can’t tell from the picture, I’m looking for everything that is not at the intersection.
I saw one way to do this:
select * from ( (select 'test1' as a, 1 as b) union all (select 'test2' as a , 2 as b union all select 'test1' as a , 1 as b ) )un group by a,b having count(1)=1
But I'm afraid what will happen if I use two large sets (I won’t ask from select '' constant operators, my queries will be pulled from real tables.)
EDIT:
Possible Solution...
drop table
RESULTS:
1 String int purple
3 dog int cat
1 String dec purple
3 dog dogs
sql-server sql-server-2008
Nix
source share