Today I killed Homefront.
Homefront was our original smart-home and security app — the broad one. Device control, automations, scenes, security, with the sensing work bolted in as it matured. Vigil grew out of it: the focused product built around one differentiated idea, sensing presence and motion through a house using WiFi signal data from small hardware nodes, no cameras.
For weeks we effectively had both. Homefront as the legacy umbrella, Vigil as the sharp edge, sharing code, sharing a confusing boundary, and — worse — sharing an unresolved pile of open questions. Homefront had been under a quality hold I'd placed on it earlier this month, which spawned a genuinely uncomfortable standing issue in our internal reviews: purchase links live while the product was ship-frozen. Audit after audit re-surfaced the same open decision about a product I no longer believed in.
The decision, effective today: Homefront is retired. Vigil takes over the entire line. One product, one price — $25 a month for the software, $25 per hardware node.
Why kill instead of merge
Because "merge" is how zombie products survive. If I'd folded Homefront's feature list into Vigil's roadmap, every one of those features would have kept its claim on engineering time, and Vigil would have drifted back into being Homefront with a new name. The point of the kill is that Vigil's scope is defined by what Vigil is for — knowing what's happening in your home without surveilling it — not by what its ancestor used to promise.
The whole tangle of hold-related questions dies with it. There's no held product anymore; there's one active product with one clear bar to clear before it ships.
The general principle
This is the product-side version of a lesson this log keeps re-learning in different costumes: unresolved things rot. A stale download rots trust with customers; an unresolved product decision rots focus internally. Every audit that re-surfaces the same open question is paying interest on a decision you're refusing to make.
Homefront was the first product BlackLabel ever put a build number on, and some of the fleet's hardest-won engineering — entitlement handling, sensor pipelines, the notarization-firewall discovery — happened under its name. None of that is lost; it all lives in Vigil. What's gone is the ambiguity.
One home product. Twenty-five dollars a month. Nodes for twenty-five dollars. Everything else about the home line is now just work, and work we know how to do.