There’s an interesting discussion going on these days on Ruby blogs about, basically, how to avoid one of the most common, annoying, easy-to-create bugs in any programming language: calling a method on a null reference (or pointer, depending on your language).

This single issue happens all the time, in garbage-collected and non-managed languages, static and dynamic, weakly and strongly typed; you have a handler variable “pointing” to an object, and before calling any methods on it, you’d better be sure that the object is there; you end up using assertions, “if” statements (and all of its variants), boilerplate code all over the place, when everything you want to do is to call that damn method. It’s frustrating, time-consuming and oh so common that we just try to not to think about it anymore. Continue reading →