The heuristic I use for choosing a statically typed language vs a dynamically typed language for a task is whether or not the code will sanely fit in a single f

The heuristic I use for choosing a statically typed language vs a dynamically typed language for a task is whether or not the code will sanely fit in a single file. If that is the case, then that means I’ll probably be able to keep the structure of the code in my head and therefore I’ll be able to get by with a dynamically typed language. However, once the code starts to span multiple files, typed method signatures in a statically typed language are invaluable. It sucks having to navigate a codebase with tens of thousands of lines of ruby or python code trying to work out exactly what structure of object can be passed to the method you’re working on.