Shipping a tiny puzzle game after three years of false starts

What I learned about scope, playtests, and finally publishing something.

I started this project four times. The first three versions died because I kept adding features. The fourth one shipped because I kept removing them.

Playtests with five people in my kitchen taught me more than a year of internal iteration. Watching someone get stuck on a puzzle you thought was easy is a useful kind of pain.

The launch was quiet. A few hundred downloads, a handful of nice emails, one bug report. I am calling it a success and starting the next one smaller.