This post is about a very small Rails design decision. A Rails design decision that I’ve made over and over without thinking about it. You probably have, too. And then a weird test failure made me think about it. And then it made me overthink it. Let me explain. I like the idea of writing Rails code such that there is minimal logic in ActiveRecord classes, and then writing a lot of tests that don’t need to touch the database, and therefore run fast. I frequently find that even if I start ...
