It's Time to Reinvent All the Wheels

The worst advice for a software developer is to not reinvent the wheel. This advice comes either from experience or people parrot what they've heard. The latter is way to common in the industry.

But if you look around, you'll see many wheels. None of them are like the original invention. How come? Well, the original wheel was not suitable for many tasks and environments. Additionally, our needs evolved and as such the need for different wheels arose. While not one-to-one applicable to software development, there are still many parallels.

Kathleen Booth invented the assembly language. With it developers were able to write software in a more human readable way. But it wasn't good enough. What followed was an evergrowing collection of high level programming languages. The reinvention of the wheel kept happening. Over and over again. We didn't stop at C. We didn't stop at Java. We didn't stop at Rust and we won't stop at Zig or anything that follows.

We are drowning in software that runs worse than software did 10 years ago. The excuse is that modern hardware is powerful and we don't have to worry about it. The results are web pages that take many seconds to load. We have desktop application that take as long to start. While hardware performance kept on improving, software kept on getting worse.

It's time to reinvent the wheel. Solving a solved problem yields knowledge and experience. It teaches the how and the whys. The task at hand may be bigger than thought beforehand. That leads to further areas to explore, resulting in more knowledge and experience.

You might want to examine existing implementations to find answers or draw inspiration. By doing so, you could discover a problem and report or fix it, improving existing software. You could realize that all existing solutions aren't good enough. You might even come up with a different approach that leads to a better solution. Who knows?

By reinventing the wheel it's also easier to contribute to existing projects. You've done something similar. You've experienced all the problems that come with it. You learned what's important and what's not.

The best products come from iterating, learning from mistakes. So go ahead, try to build everything you ever thought of. But build it from scratch. In the worst case you'll learn something. In the best case you'll enrich peoples lifes. You'll contribute to the art that is software development. And please, don't advocate for not reinventing the wheel. We are in dire need of new ones.