Some of my friends at the University of Washington have been working with computational geometry and running into quite a few headaches because of floating-point error. Of course, they were expecting my work on Herbie to have taught me how to help, so I've been fielding a few floating-point questions. When they got desperate enough to start using MPFR and talking about computable reals, I had an idea. My friends are working on a programming language for describing shapes in two and three dime...
