I recently gave a talk at the Haverhill Hackers Meetup about “real refactoring” where I went through a few refactorings I had made in our codebase at QueueDr. The examples were dumbed down a bit for simplicity sake, but the motivations and changes made were clearly displayed. The talk went well and I think those who attended(including myself!) learned a few tricks to bring into their own codebases.
When you start a new project and you are the only developer working on it the possibilities seem endless. No worrying about other hands getting into the codebase and messing everything up. No dealing with shitty legacy code. Then a funny thing happens. Your hands get into the codebase and mess everything up. Your code becomes the shitty legacy code you feared. You quickly realize that you are capable of all the things you complain about, and that completely fine.
Business happens. We rush things, skip testing, and write horrible hacks(that we intend to fix sooner and not later). Outside pressures have a direct effect on code quality. At the end of the day though, having something working and shipped today is far better than something perfect weeks/months/years from now. All code can end up being legacy code, even if you wrote it.
This is why refactoring has to be part of your process even if you are the only one touching the code. It’s unrealistic to think you’ll write everything the way it should be the first time around. So, put on your refactoring hat and get going. Start small. There is no need to be afraid and become paralyzed at the thought of embarking on this refactoring journey, no matter how horrible the codebase has become. It’ll be worth it.