Euclid and Pythagoras are not, and cannot, be patented. So it seems somewhat illogical to argue against patents with the hypothetical of those algorithms being patented.