From the schools I've been to, and the places I've worked, I'd say many CS geeks barely struggled through Math. I was borderline, was planning a split honours, until I discovered how hard math became in third year. Mathematicians SHOULD be able to program, first, because they have to to do their work, second because thinking analytically, handling all circumstances, etc, should come naturally to trained mathematicians. However, they may not do so well at the aspects that involve daily experience.

For some aspects of programming, understanding math is usefull, even essential. Obviously things that are mathematical need math; on the other hand, a flight simulator, interplanetary rocket simulator, nuclear controller or simulator, could effectively isolate the math to certain sections. You would need a specialist in mathematical programming to handle those segments, but anyone could do the rest.

--
TTTATCGGTCGTTATATAGATGTTTGCA

Mathematicians SHOULD be able to program, first, because they have to to do their work, second because thinking analytically, handling all circumstances, etc, should come naturally to trained mathematicians

I'd have to say that while I agree with the second half, very few of the mathematicians that I dealt with while I was in graduate school had any need to program to do their work (and at that time in my life, nor did I).

Of course there are some (depending on field of interest) that do.

Update: This is just a comparison. I really do not say that lambda calculus or Turing Machines are useless. To the contrary - as I've explained somwhere else in this thread I believe it is quite importand to know some theory - because only theory would give you sound reasoning why something is impossible.

Please define help, because some people consider proving theories to be quite helpful. My definition was something along the lines of "it doesn't hurt" - so i will stick to what i said: "if you want to get into functional programming, then a good knowledge of lambda calculus will surely help". That is, of course, generally speaking - YMMV.

Please don't discourage anyone from learning something new.

jeffa

L-LL-L--L-LL-L--L-LL-L--
-R--R-RR-R--R-RR-R--R-RR
B--B--B--B--B--B--B--B--
H---H---H---H---H---H---
(the triplet paradiddle with high-hat)

I just compared the helpfullness of lambda calculus to that of Turing Machine. I believe both have some value. Perhaps you are right that mine comment could discourage people to learn lambda calculus - and I should formulate it a bit more carefully. But I just tried to be honest.

Lambda calculus was quite nice, but it is too low level to be any practical and the interesting parts start when you learn the extensions to the lambda calculus like the F-system (where you get the ultimate polimorfism).