They also tend to do it quite badly in some non-trivial cases. I've seen Wolfram spewing a three-line result which could be further simplified to a simple expre

They also tend to do it quite badly in some non-trivial cases. I've seen Wolfram spewing a three-line result which could be further simplified to a simple expression. It also doesn't always help to simply know the result; a lot of insight about numerical problems can be gained from how the result is derived. I also spent several days churning formulae in bars. Got a lot of weird looks.