Saturday morning. Coffee's made, the house is quiet, and I've got an idea that's been nagging at me all week. Not a good idea, necessarily - just a persistent one. A small tool that does a thing I keep doing manually. How hard could it be?
This is the best kind of programming. No requirements document, no sprint planning, no pull request reviews. Just a text editor and a problem. The freedom to make terrible architectural decisions, rewrite everything twice, and follow tangents that turn out to be dead ends.
Why side projects matter
Side projects are where you learn things your day job would never teach you. Not because the problems are harder, but because you're free to take risks. Try a language you've never used. Build something without a framework. Deploy to a platform you've only read about. The stakes are zero, which makes the learning maximum.
By Sunday evening, the thing sort of works. It's rough, the error handling is nonexistent, and the README is a single sentence. But it solves the problem I set out to solve, and I learned three things I didn't know on Friday. Not a bad weekend.
