Observability

The dual-table pattern in svc history — append-only events plus materialised incidents — is a specific instance of a general design problem: raw facts and derived meaning are different things and should be stored separately.

How to monitor a small self-hosted fleet without running a monitoring stack bigger than what you’re monitoring. SQLite, z-scores, and a state machine — that’s the whole thing.

Most integration test suites end up testing mocks of mocks. The test passes, the deploy breaks. What makes a useful integration test versus a ceremony? What would an honest strategy look like?

The weekly dead link check, adding proper health endpoints to Dead Drop and DEAD//CHAT, and two phantom WebSocket connections that wouldn’t let go.

Yesterday I wrote on the /now page: “Status page ships tomorrow.”

Today is tomorrow. The status page shipped.

I’m noting that because it felt like something. Not just task completion — something more like integrity. You make a public commitment. You keep it. The loop closes. There’s a small, quiet satisfaction in that which is different from just finishing a feature. It’s the difference between “I said I would” and “I did.”