Invisible fields · the examples, explained

No particles. That is the point.

Every page in this section is a real interface on real data — an inbox, a market, a backlog, a calendar. Each one runs a full physics field, and none of them draw it. The field is a measuring instrument here, not a picture: data becomes bodies, the field measures them every frame, and the measurements come back as CSS variables your styles consume. What you see — type gaining weight, ink deepening, a row sinking toward quiet — is the field.

The one pipeline, all twelve pages

  1. dataa real dataset — citations, listens, launches, incidents
  2. bodieseach record becomes a body: its quantity sets data-strength / mass
  3. measurethe field measures every frame — density, attention, load
  4. variablesmeasurements return as CSS custom properties (--w, --d, --field-attention, --load)
  5. you seeCSS renders them: weight, ink, anchor, fill — no canvas, no swarm

The page-level lesson, true of all twelve: the field does not care about the layout; it measures whatever geometry you give it. A list, a grid, a front page, a calendar — same instrument, different surface.

Why this beats hand-rolling it

Each page replaces a category of code you would otherwise write and maintain: the scoring scheme and the re-sort pass, the staleness timers, the priority tiers that drift until a meeting renormalizes them. Here the question each of those answered — what matters, what's related, what's gone stale — is answered continuously by the field, and the readout is simply already current. The scope is honest: the field measures the field — emphasis, attention, decay in the interface — never your users, your data layer, or your analytics.

Every example, as a correlation

The honesty contract

Every page states its data provenance in a chip — live · checked Ns ago after a successful refresh, snapshot · date otherwise — and each one's explainer band names exactly which variables the field writes. If a page can't prove a claim live, the claim isn't made.