KDABFloating-point math is hard; compilers will exploit every language loophole to make our FP calculations go faster, sometimes with surprising results. https://www.kdab.com/fma-woes/ #cpp #qtdev #floatingpointmath
nfrgrtNow they're using price values as examples for floating point types. And I keep wondering why I'm frequently sitting in projects where my task is "can you please figure out why our tax calculations are not adding up?"