Scenario 1 fails

[twestley]

Scenario 1 tests two nodes with the same time between creating objects (lambda 3) so it should result in 2 keyspaces. This is because long term allocation ratio (% keyspace) for both very quickly becomes 0.5. They each have 50% of the keyspace. But if we set short term allocation to 1 when it is near zero (as described in ADAK Keying Scheme.pdf), it results in 3 keyspaces. This is because the second (of two) nodes receives a message from the first that its long term allocation ratio is 0.5 too late, after it (the 2nd node) gives half of its keyspace to the first node. This is then stable because they both see that the other has <1 long term allocation ratio.