
// the codex
Wisdom
for the build.
Lessons most developers learn the hard way, written down so you can skip a few levels.
// rules of the grid
Ten rules.
Finish small things.
A finished tiny game teaches you more than an unfinished huge one. Your first ten games are training levels. Treat them that way and ship them anyway.
Scope is the final boss.
Every idea is bigger than it looks. Cut it in half, then cut it again. If it can't be fun with one mechanic, more mechanics won't save it.
Feel before features.
Screen shake, a good sound, a tiny pause on impact. How it feels in the first ten seconds matters more than the feature list.
Watch people play. Don't explain.
Hand someone the controls and stay quiet. Every time you want to say "oh, you're supposed to…" that's a design problem, not a player problem.
Remix ideas. Never steal work.
Borrow mechanics and inspiration freely. That's how the medium grows. Never use art, music, or code you don't have the rights to. Credit everyone who helped.
Make the thing you'd play at 2 a.m.
Trends move fast. Your own taste is the one thing nobody else has. Build what you genuinely want to exist.
Compare yourself to last month.
Not to studios with 300 people. Not to someone's highlight reel. Open your oldest project once in a while and notice how far you've come.
Stuck is part of the job.
Every developer is stuck most of the time. Write the problem down in one sentence, walk away, come back. Asking for help is a skill, not a weakness.
Protect your player profile.
Online, share your work, not your life. No address, school, or schedule. Keep a trusted adult in the loop on accounts and money. Block early, block often.
Rest is a build step.
Sleep, movement, and real-world time are where ideas come from. Burnout doesn't ship games. Consistency does.
// spark generator
Blank page? Roll.
One theme, one mechanic, one twist. Most rolls will be weird. That's the point. Weird is where original games come from.
// the training ladder
Ten classics to rebuild.
Rebuilding old games is how most developers learn. Each one teaches a new system. Click one when it's done. Progress saves in this browser.
// the toolkit
Real tools. Mostly free.
Pick one engine and stick with it for a year. Switching engines feels like progress, but it usually isn't.
Engines
Art + Sound
Ship + Grow
// sunday check-in
Five questions. Every week.
Mentors don't hand you answers. They ask better questions. Ask yourself these, then write your answers in the devlog.