I suppose that my negative reaction is mostly because, generally, calls for more math in learning a subject results in additional semesters of calculus or differential equations, which would be of doubtful use in my experience developing software.
> then you're using math every time you write any code at all, whether it's ensuring you don't write any cyclic dependencies, making sure you correctly handle all return values, or making sure an algorithm doesn't require quadratic time or worse before you implement it. You need math if you want to know that your custom comparators won't cause a sort routine to blow up, to recognize certain possible sources of bugs in unsafe languages, and most of all to recognize and assess the validity of potential optimizations.
For the most part, I view this as simple bookkeeping. Probably I've been doing this for too long and have just internalized it.
> then you're using math every time you write any code at all, whether it's ensuring you don't write any cyclic dependencies, making sure you correctly handle all return values, or making sure an algorithm doesn't require quadratic time or worse before you implement it. You need math if you want to know that your custom comparators won't cause a sort routine to blow up, to recognize certain possible sources of bugs in unsafe languages, and most of all to recognize and assess the validity of potential optimizations.
For the most part, I view this as simple bookkeeping. Probably I've been doing this for too long and have just internalized it.