Good IT Science

摘要

Practical it science tends to evolve most rapidly from projects that have business objectives in mind rather than scientific ones. When I talk about letting out your inner mad scientist or inventor, I'm not talking about strapping jet engines to your Volkswagen or programming in Lisp. The litmus test for good IT science is simple: It's good if the results of what you create are simpler than those produced by existing solutions. I'll set the stage with an example. I remember being blown away by SonicXQ, the Web services orchestration architecture that became Sonic ESB (enterprise service bus). I saw the surface of it — an engine that pushes smart objects around a choo-choo track using Web services — but I had no idea how beautifully easy it was, easier than using Web services without an ESB. It would take me longer to diagram how Sonic did it than to build you a sample solution that uses it. Good science.