If you come across code you don’t understand, just call it “Tech Debt”. If you find even more code you don’t understand, keep calling it Tech Debt. Eventually, you will have enough Tech Debt to be able to declare bankruptcy. Then you can rewrite all the code in a new language with new tools that you do understand.

I’m wary of any time dedicated to resolving tech debt. Instead of building something new, time is spent on refactoring old code. It’s not the rewriting of old code that bothers me, but the time not spent building something new. Spending too much time resolving tech debt suggests that all the new and amazing ideas have stopped flowing and the pipeline is drying up.

When I started Robot Ops, I was using a codebase that was 7 years old. I didn't refactor the old code to get rid of the tech debt. Instead, I reused a lot of concepts that worked, using my improved programming skills to write better code for only the new parts of the game. Refactoring the old code might have been cool, but writing beautiful new code while building the awesome new features of the game was even cooler.

Also, I simply don't have the time to refactor all that code. I'm trying to finish the game as quickly as possible. Once it's done, I'm moving on to even bigger projects. I learn my lessons as I go, and I would rather apply those lessons to new projects than to old ones.