Part two of the Long Strange Trip, a thirteen-year memoir with honest dates.

The Fargo shop’s work that shaped me hardest wasn’t websites — it was native mobile apps with GPS in their bones. Years of them. I’ve written the technical retrospective already; this is the memoir version, which is mostly about humility.

Browser work spoils you in ways you don’t notice until a device in a moving vehicle takes the spoiling away. A website fails politely, on a desk, with a console open. A GPS app fails in the field — signal dropping behind a grain elevator, a phone baking on a dashboard in July or freezing in a cab in January, the battery dying at the exact worst moment. North Dakota is a fine place to learn this: the state is practically a laboratory for “what if there’s no signal and also it’s minus twenty.”

What that hammered into me, build after build: the failure case is the product. Anyone can make software work when everything’s working. The craft is what happens when the network vanishes mid-operation — does the app lose the data, or hold it and heal when the signal returns? Getting that right isn’t a feature; it’s the whole difference between a tool people trust in the field and a demo that impresses in a conference room.

I didn’t know it then, but those years installed the conviction that runs this whole practice now: assume the failure, design for the recovery. Backups nobody restored, forms that fail silently, checkouts that error only under load — every rescue I take today is some system that was built for the sunny day. The GPS years taught me there’s no such thing as a sunny day you can count on. Up here, especially.

They also taught me to love constraints. Mobile in that era meant tight memory, tight batteries, tight screens — you couldn’t solve problems by adding more; you solved them by choosing. That’s a muscle, and it still lifts.

Next: Standing up the cloud — in which the servers leave the closet.