Skip to content

73 gown vertices are inside the body at y 1.153-1.348 m after #714's derived clamp #746

Description

@gidich

#714's derived fold clamp cut the corrected interpenetration count from 129 to 73 and stopped there.
The 73 sit in a band the fold clamp does not govern, and nothing has established what they are.

Measured 2026-08-28, by the orchestrator, on both trees

I ran gown-shard-mechanism-measure.ts against main's mpfb-gown-adult-patient.glb
(sha e6e6f9d4…, 9,450,684 bytes) with OPENCLINXR_GOWN_PIN_SHA/_BYTES overridden, and against
#714's rebake (sha dd2340a9…).

measure main, pre-clamp after the clamp
upper half, two-tests-agree 129 73
lower half, two-tests-agree 0 0
lower half, primary instrument 32 31

"Two-tests-agree" means a parity ray AND the nearest-surface signed distance both call the vertex
inside. It is #714's correction to the single-axis metric, which counted 294 vertices whose
nearest-surface distance put them 12-61 mm OUTSIDE the body.

Where they are, and where they are not

All 73 fall in two adjacent deciles of the gown's own y-range. Every other decile reads zero.

decile y range (m) two-tests-agree
6 1.153 – 1.250 5
7 1.250 – 1.348 68
all others 0

On a 1.776 m figure that band is armpit and sleeve-root height.

The claim this card exists to test

#714's worker reported the 73 as armpit-seam and sleeve-root overlap — a different mechanism from the
fold valleys the clamp governs, and the card's own rejected "push the shell out" candidate.

I could not confirm that. The measurement artifact resolves height only, and the fold band spans
y 0.98–1.51, which overlaps both deciles. A height band is consistent with the armpit reading and
equally consistent with a fold valley that survives at chest height.

The first measurement

Dump the per-vertex rows for the 73 — position, nearest body triangle, and which garment region each
belongs to — and say whether they are armpit-seam/sleeve-root or fold-valley. That answer decides
whether #714's clamp is complete for its own mechanism or whether the fold path still leaks.

THE CAUSE IS NOT KNOWN TO ME BEYOND THOSE NUMBERS. Do not take the armpit reading as fact; it is
the hypothesis under test.

Do not

Report these

If any proof cannot pass as written, OR passes trivially against the ambient range, OR asserts the
opposite direction from the defect, say so in your first report, at the moment you find it.

If the premise above turns out to be false, say so immediately. Never narrow or weaken a clause so
that it passes.

Report any out-of-scope wrongness you saw and are not fixing, naming the object and what is wrong.
The bodice sawtooth is already known and is not this card — name it if you see it, do not fix it.

Write your report with the section headers IN-SCOPE:, OUT-OF-SCOPE:, CLAIM: and NOT TESTED:
in that colon form.

factory_step: body_param

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions