My answer takes from the Feynman approach to problem solving: have several projects laying around that you could easily pick back up to work on a little bit. Th

My answer takes from the Feynman approach to problem solving: have several projects laying around that you could easily pick back up to work on a little bit. The progress may not be constant, but it builds over time. This is a variation on the parent post's suggestion of picking a project you can finish in three days. I find that task can itself be hard, and is work that gets in the way of the fun work. In the past I've burnt out my excitement while trying to do it. So I have 5+ small games, a game engine, an interpreter and a compiler, a few short stories, some music projects, a personal website, some art projects, and several books to read, and most aren't done. When inspiration strikes, e.g. I'm feeling musical, I open up my music folder and click around a bit. Or if I feel like making a game, the first thing I do is go play one of my existing half-built games, and pop open the editor. It is helpful here to leave things in a working-ish state and keep ideas around in text files about what you could do next. Make picking the project back up easy for yourself. I do also start new things or decide not to come back to old ones, but I wound up here by realizing that my interests aren't actually infinite. There's only about a dozen distinct things I really want to do and periodically get excited about. For years I had been starting and then totally dropping projects, but I had "started making a game" a hundred times. My plan became to treat random projects like turning a ratchet, and I feel it's gone great from there.