Why an analytical platform needs this, not just a tidy codebase

How do we define and fill systemic gaps.

§The gap: presentation drift

Every analytical tool eventually gets asked to change how something looks — a table becomes a chart, a summary gets a new layout, a status badge changes colour. In most systems that change touches the same code paths that computed the thing in the first place, because "compute it" and "show it" were never cleanly separated. Over enough of these changes, a reader can no longer be certain that what changed was the display and not, somewhere quietly, the finding. This is presentation drift, and it is a genuine integrity failure for an analytical system, not a cosmetic one — a report whose numbers might have shifted when someone changed the font is not a report you can stand behind.

§The rule that closes it

The machine's own name for the boundary that prevents this is the formation boundary: information is formed — decided, computed, folded from the log — strictly on one side; every surface only ever presents what was already formed, and never depends on which surface is attached (huid 00 §5, cited throughout the codebase's own comments). Concretely: a reading (the machine's word for a computed view of the log — Part 3 and 5.3 give this its full technical treatment) is built once, inside the inner rings, and every outer ring — the wire, the screen — can only ever ask for a reading by name and render what comes back. A screen component cannot fold the log itself even if it wanted to; the machinery that does that folding is not on its side of any door it can reach.

§What this buys, in standard terms

• Reproducibility. Two different screens — the main workbench and the separate debug board described in Part 4 — show the same finding identically, because both ask the same inner ring for the same reading; there is no second code path that could quietly disagree.
• Auditability of the interface itself, not just the data. Because the rings are enforced mechanically (a test, not a convention), a reviewer does not have to re-read every screen component to be sure it stayed on its side of the boundary — the suite already checked.
• Swappability without risk. The whole screen layer could be replaced — a different UI, a different client entirely — and nothing about what the machine has established would need to change, because the screen never held any of that logic to begin with.

§The honest limit

Ring discipline guarantees where logic is allowed to live; it does not by itself guarantee the logic is correct. A reading can still be wrong in what it computes — the rings only guarantee that once it is right, nothing downstream can quietly make it wrong again, and that the same right answer reaches every surface identically. 6.3 gives the actual rule set this enforces, ring by ring, and 6.4 shows it holding for one real gesture end to end.