Comment by leelou2

This article hits on something I've been wrestling with in our codebase. The absence of static typing in Clojure makes refactoring feel like walking a tightrope without a safety net sometimes. I've found that building a solid test suite before any major refactoring is absolutely critical - it's the only reliable way to ensure you haven't broken anything when reorganizing code. What's worked well for us is treating functions as the primary refactoring unit rather than trying to impose OO-style refactoring patterns. Breaking down large functions into smaller, well-named pure functions has provided the most bang for our buck. We've also found that leveraging spec for runtime validation gives us some of the safety of types without giving up Clojure's flexibility. The tooling situation has improved significantly too. Cursive's refactoring tools have been solid, and I'd be curious if others have had success with the newer REPL-integrated refactoring approaches mentioned in the article. Has anyone managed to set up effective continuous integration that catches runtime errors in untested paths?