Building Moments

A phone call with a mate turned into a wager: could AI actually build a real application from scratch? Five months, 898 commits and 40,000 lines of Rust later, here’s the verdict — and the app. It started with a phone call Like most of my questionable decisions, this one started on the phone with a mate, talking about technology the way we always do. The conversation drifted, as every technical conversation eventually does now, into AI. And we ended up circling the same question everyone circles: is this stuff actually capable of building real software, or is it just very confident autocomplete? ...

There is a moment in every codebase where someone draws a box on a whiteboard, labels it EventBus, and draws arrows fanning out to every other box. The room nods. It looks like architecture. It looks like the right architecture — loosely coupled, extensible, the textbook answer to “how do we let many components react to one thing happening?” That diagram is a trap. I drew it for Moments, my photo manager, in late March 2026. I had the design reviewed by an external UI architect. I shipped it in April across six phased PRs. And on the third of May I deleted the whole thing in a single commit titled, with some satisfaction: ...

Half of the library code in Moments, my photo manager, calls a one-method trait after every database write. That trait has two implementations. One of them returns Ok(()) and does nothing else. From inside the library, there is no way to tell which one is wired up. That is the entire architecture of how Moments syncs to Immich — and the entire reason the local backend, which has no sync at all, is the same program as the Immich one. ...

In the previous post I described an architecture where the Moments library doesn’t know that sync exists. Services record Mutation values into a one-method trait; one implementation writes to an outbox table, another does nothing. The library code is identical regardless of which backend it’s running under, and a third backend could be added without touching the library at all. The architecture is sound. I still believe in it. The first six weeks of running it were a different story. Four bugs in particular are worth describing, because each one tells you something the diagrams don’t. Two are about the abstraction — assumptions the trait quietly makes about how the world works, and what breaks when the world doesn’t cooperate. Two are about the substrate — the boring queue-and-retry machinery underneath the trait, which has its own opinions about how things should be done. Clean architecture doesn’t relieve you of the substrate’s problems. It just gives you one place to deal with them. ...