It's probably of little consolation to the addicts among us who spent several terrifying hours in connectivity withdrawal -- and even less consolation, still, to the newly liberated -- but RIM's apparently figured out how its notoriously reliable back end came to a crashing halt this week. The company is now pointing its finger at "the introduction of a new, non-critical system routine" to its caching mechanism as the culprit. "Non-critical," indeed. Anyway, it seems said system routine was not put through enough testing ahead of its deployment to RIM's production systems -- and to make matters worse, its failover setup (hamsters on wheels, perhaps?) didn't pull through, significantly delaying the amount of time to get everything back online. We suspect most users are still a little too shocked at the chain of events to be steamed (yet), so consider this your strike one and two, RIM; just make sure it never happens again.