Start with two empty tables:
LHS is This KTable
RHS is Other KTable
1- RHS is updated to Y|bar
2- LHS is updated to A|Y,1
   -> sends Y|A+ subscription message to RHS
3- LHS is updated to A|Y,2
   -> sends Y|A- unsubscribe message to RHS
   -> sends Y|A+ subscription to RHS
4- RHS processes first Y|A+ message
   -> sends A|Y,bar back
5- LHS processes A|Y,bar and produces result record A|Y,2,bar.
6- RHS processes Y|A- unsubscribe message (update store only)
7- RHS processes second Y|A+ subscribe message
   -> sends A|Y,bar back
8- LHS processes A|Y,bar and produces result record A|Y,2,bar
Thus, the first result record, that should have been `A|Y,1,bar`, is now
`A|Y,2,bar`. Furthermore, we update result table from `A|Y,2,bar` to
`A|Y,2,bar`.
By
 using a hash function to compare in step 5, we can instead see that the
 message is stale and should be discarded, lest we produce double output
 (or more, in the case of a rapidly changing set of events).