how to make software﹖

recognition-primed decision | > Always start from the problem, not the solution