I claim that the first undergrad course in Analysis is not actually there primarily to teach you real analysis. Rather, it's there to teach you the difference b

I claim that the first undergrad course in Analysis is not actually there primarily to teach you real analysis. Rather, it's there to teach you the difference between "forall exists" and "exists such that forall", and to hammer into you that intuition is not by itself a proof. Nonstandard analysis, I claim, is not as good a vehicle for those purposes as epsilon-delta analysis is.