Sideprojects get abandoned due to analysis paralysis or choice-dilemma forks. The best way to get them done is to build a frankenstein version of them in 2-3 h

Sideprojects get abandoned due to analysis paralysis or choice-dilemma forks. The best way to get them done is to build a frankenstein version of them in 2-3 hours, something that seriously bad, super bad code, bad design, bad architecture, crashes every 2 seconds, hardcoded things everywhere, unextendable, hard to maintain, but even then something that just boots as a vague identity of your idea. Then focus on fixing it into better iterations and don't forget to go wild on it, instead of sqlite try adding cockroachdb, instead of basic hashmap put a redis there, whatever tech you wanted to learn recently, use it. Then let it be a hulk for a bit and invite some friends to use it. If you are still motivated, focus on best practices, maintainability, remove the overengineering and replace with reasonable changes. Goal of side projects is to learn something new and have fun. If the fun stops, don't push it or feel bad, just walk away with the knowledge you gained. :)