Those with a Windows box, try this experiment: Open the calculator and switch to scientific. Switch the mode to binary. Now try to subtract, lets say, 111111 from 11101011. You should get -01010100, or in place of the negative sign a series of 1's indicating the sign. Yet, for some reason, you don't. What kind of screwed up piece of dried-up dog crap is that? How did Bill's team manage to take a computer, something which natively speaks binary, and get it to forget how? Just to spell it out: if you are running Windows, you are running an operating system(!) designed by programmers who can't get their system to perform mathematics! Depressing, isn't it?
I realize that KCalc displays the number with a series of ones instead of a neg sign. But hit the dec button to switch to decimal and you get a reasonable answer of -172. (63-235=-172, so this is the correct answer. Try it on a windows machine and your answer is definitely not -172 in base ten.
The calculator is not a new computing concept. Microsoft has shipped one with their OS since, what, 3.0 for sure. What a damn embarrassment for the geek community that we are creeping up on 2 decades later and they still cant get it right.