Aging: Evolving Software, Tools, and Ourselves

If it ain't broke, should you fix it? If it is broke, shouldn't you just rewrite it instead? Evolving software over years (and decades) isn't glamorous or exciting, but it's the necessary work that enables our field to create real economic value. How should we think about – and plan for – the ongoing evolution of our software? How can we know when to evolve and when to push back? When should we or shouldn't we throw it away and go green-field?