Инженер шоколадных коробок

While mechanical engineers certainly are responsible for huge amounts of computation, their approach to doing math can be somewhat practical: they need some way to get things done (and please, no metaphysical debates about error representations of floating point numbers, because there in fact is a deadline for this car/airplane/rocket/toaster crash test). Especially in RD, one is faced with all kinds of ever-changing tasks involving dirty real-world processes that obstinately refuse to behave according to theory (and sometimes depend more on the day of the week and the weather than on calculation), but still need to be dealt with. Thus the box of chocolates analogy: sometimes you really don’t know what you’re going to get—but you’d better get it done by next week, thank you very much, chop-chop. On the plus side, no one is going to fire you because you used a horribly inefficient root-finding method (as long as it works, that is). Because many of those tasks do involve really non-standard problems