C++ would be a very different language if you couldn't use floats: (NaN + 0.0) != 0.0 + NaN Inf + -Inf != Inf I suspect the algebraists would also be pissed

C++ would be a very different language if you couldn't use floats: (NaN + 0.0) != 0.0 + NaN Inf + -Inf != Inf I suspect the algebraists would also be pissed if you took away their overloads for hypercomplex numbers and other exotic objects.