the part on machine precision (z:=|x-y|, Floating Point Numbers (2), slide 8/16) will be redone on the 18th