If all that went wrong with tools was that they broke, the rule of thumb could be: when you break a tool, replace it with one twice as expensive. When you lose

If all that went wrong with tools was that they broke, the rule of thumb could be: when you break a tool, replace it with one twice as expensive. When you lose a tool, replace it with one half as expensive. Like the power-of-2 realloc trick, the amortized cost of finding the tool you need is at most 2x. But cheap tools do something worse than break. They ruin work, they frustrate you, they can be dangerous. (Perhaps in KK's youth, the cheapest tools were still acceptable. But today, the cheapest tools literally cannot do a single instance of the task they're ostensibly made for.)