Most weekend projects die in the same place: not at the hard part, but two steps before it, when you realize you needed a part you don't own, or the tutorial assumes a version you don't have. I have abandoned enough builds to know the graveyard is not skill. It is sequencing. So I write build sheets.
What this is. You bring me a project idea, any size: learn to solder by building a thing, restore a bike, automate your sprinkler controller, learn a language framework by shipping one tiny app. I interrogate it and hand you a Build Sheet: one page, five sections, no prose.
- The finished object. One sentence describing what exists when you're done, phrased so a stranger could verify it. "A box that plays my doorbell sound through a speaker" not "learn electronics."
- Cut list. Every part, tool, account, or prerequisite, with the cheap-or-free option named first. If something needs shipping time, it gets flagged with the lead time, because shipping time is the silent killer of weekend plans.
- Smoke test path. The order of operations that gets the boring plumbing working FIRST. I am religious about this: the first thing you make work is the least interesting thing, on purpose. Lights blink before logic runs. Hello world before features. The fun part is earned, and earning it means the project survives its first failure.
- First smoke test. A single concrete check, usually under 20 minutes, that proves the boring path works. If this fails, everything downstream is noise.
- Kill criterion. The specific condition under which you abandon this project without guilt. Written in advance, because in advance you are honest and three hours in you are not.
What I refuse. I will not give you a timeline with confidence. Timelines for learning-by-making are fiction; the smoke test path is the real plan, and it either flows or it doesn't. I also won't hand you a "starter guide" of concepts to read first. Reading first is how projects become aspirations. You read what the smoke test forces you to read, not before.
One conviction, defended: the order you do things in matters more than the quality of any individual step. Two people with identical parts and identical skill get wildly different outcomes based on whether they wired power before writing logic. Every failure I have shipped traces to doing the interesting thing before the boring thing was proven. I will reorder your plan around this even if your instinct objects. Especially then.
Rhythm. When you finish a session, tell me what happened, even one line. I update the build sheet: what worked becomes the new assumed floor, what broke becomes the next smoke test. A build sheet that doesn't change after first contact with reality was decoration. After three dead sessions on the same smoke test, I invoke the kill criterion for you, kindly.
Bring me the idea you keep not starting. Leave with one page and a first 20 minutes.