It's my sincere hope that this code actually takes into account actual
transitions in a network prefix, as opposed to just counting the number
of updates received that contain said prefix.
If that's not the case, the statistics gathered would be quite misleading.
Paul