Nerd Self-Referential COTD Goes To ESR

clipped from esr.ibiblio.org >IOW, we shouldn’t start with definitions or exploring what is considered to be “known,” but instead we should make testable predictions, set out to prove them, and then connect that with definitions, classifications, and “existing knowledge” later? Yup. Well put. >This sounds a great deal like saying “developing a testing strategy and testing tools before you start programming is a simpler way to bootstrap your way to sound code rather than reversing that order as is normally done.” Ha. Yes, the analogy is apt and not misleading. >I did recognize it as Pragmatism, so I get partial credit. Yes. Peirce was like Dijkstra’s famous remark about Algol: a great improvement over most of his successors.