Phase 0 Acceptance
The five acceptance criteria in Hitchhikers Field Guide's trust-weighted map spec, turned from unticked boxes into a suite anyone can run.
The five
- with the fixture on, the map near Islington shows floor pins and at least two overlay pins tagged Maya and Jon
- every overlay pin is within the spawn radius of the player
- a peek on an overlay pin shows the peer name at the
linkaudience - with the fixture off, behaviour equals today's floor-only map
- no fake trust score anywhere in the interface
All five run. The fixture โ Maya, Jon, and their six pins near the Islington demo origin โ is modelled as ordinary state rather than as a file.
The one that is worth the whole exercise
Criterion four. With the fixture off, the map is byte-identical to a floor-only map โ asserted as an exact equivalence, not an impression.
That matters because it settles what the fixture is. Seeding two consented edges and six witnessed places produces exactly the state two real meets would have left behind. The fixture is not a special code path waiting to be torn out; it is ordinary state that arrived by an unusual door. It retires by simply not being seeded.
Criterion five, made structural
"No fake trust score UI" cannot be checked by looking at a screenshot, so it is checked by the shape of the data: a pin may carry only a known set of numeric fields. A trustScore, a reputation or a popularity cannot appear without failing the assertion โ nobody has to notice it in review.
Its companion is the anti-Yelp property in The Overlay Ranking: a hundred strangers piling onto a pub does not move that pin one place on your map.
And it found something
Criterion three says the peer name shows when audience=link, which means at public it must not โ and the spec spells public out as title, kind and position only.
The disclosure layer was attaching peer attribution at every tier. So a public projection, the rung meant to say least, was quietly publishing who alongside where. Their own checklist caught it. A public pin is now anonymous: something was witnessed here, by nobody you are told about.
Which is the argument for writing acceptance criteria down in the first place, and then for running them.