But let's not forget hindsight is always 20/20. Problems like the shuttle and ISS result from a common engineering dilemma, the struggle between robustness and versatility: Do you specialize and optimize to maximize success, or do you keep the system as flexible as possible so that you please the most users and get the biggest bang for the buck? The truth is always somewhere in between, but where?

The answer is not always obvious during the design phase. Look at some of the great engineering designs of history: the 1911A1 .45 pistol, the C-47 aircraft, the Model T, the Singer sewing machine, and many others. Only time demonstrated how truly remarkable these concepts were, many other seemingly adequate competitors simply didn't hold up over time, in spite of better performance up front.

These designs do seem to have certain things in common: such as simplicity of construction and use and easy modifiability and adaptability to changing conditions.

But it isn't always apparent ahead of time if the machine you are designing has those properties, only long use will reveal them.

This is why we pay engineers big bucks. Any jerk can design a machine to do a job efficiently. The trick is to make it so there aren't any gotchas, or at least not any easily unfixable ones, in the concept.