This approach has helped me, above all, to manage my expectations, which were too high. When it's clear that I only have an hour or two, I rarely start projects

This approach has helped me, above all, to manage my expectations, which were too high. When it's clear that I only have an hour or two, I rarely start projects or parts of projects that are too big. Instead, I can focus better on the core problems in this short time. I lay down software architecture ideas, boil down problems to core algorithms, implement core functionality, write outlines of blog posts or summaries etc. Most of the time this means I'm 80% done. If the small project still attracts me then I'll automatically continue it the next day and start working on the remaining 20% that take much more time.