1- anything that can go wrong will go wrong
2- anything that can't go wrong will go wrong
3- any attempt to correct for failures of type 1 or type 2 will, by increasing the complexity of the system, increase the sources of failures of type 1 and type 2
As the "failsafe" nuclear power plant security failures in Japan have demonstrated clearly since 11 March, "failsafe" security measures don't work. We live in a universe ruled by Murphy's Laws!
Take the case of poor Mubin Shaikh.
This fellow, well intentioned and / or hopelessly naive, offered Canadian security agencies (CSIS, RCMP) his services as a mole within the radical Islamic community in Canada. His goal, he says, was to be a better-than-average citizen and personally run the risk of ferretting out terrorists plots. Weird. Pollyanish, maybe, but brave, at least (I mean, these dudes are willing to kill innocent people: what might they do to a mole if they found out? Would you do it..) At any rate, Shaikh's services led to the conviction of theToronto 18 plot members.
Normally, the guy should get a medal, right? He volunteered to put himself at risk to serve his country, etc. So how was Mubin Shaikh rewarded for his time, effort, courage?
According to a wiki-leaks document, Shaikh was amply "rewarded" for his services by being place on one or more American terrorist suspect lists. Note: I don't make this stuff up! - I simply don't have enough imagination.. However, I imagine this type of incredible cockup must have led engineer Murphy to formulate his famous Laws: engineers apply scientific theories to practical situations and thus are forced to deal with all the vicissitudes and perversities of the material world.