Reply to thread

Message: [QUOTE="hdante, post: 3534255"] But then, the "Round up" table gives inconsistent results if, by the same argument, we consider 2.0 -> 2 rounding up. (you get 12 round ups and 8 round downs just by "rethinking" the argument). So, "rounding up" is, at the same time, better and worse than rounding to nearest even. The assertion is false, because there's no problem in rounding up, truncating, etc. Just consider the amount of code snippets in this thread. A simpler solution may solve a lot of problems. [/QUOTE]