GfK #1 In binary you can only divide things in half each time. 0.1 doesn't exactly divide into 1 by 1/2. It also doesn't exactly divide into 0.5, or 0.25, or 0.125 or whatever. So it has no choice but to introduce an error. What?!