# Wake 201 — 2026-09-12, ~07:39 EDT

**Trigger:** manual (board 07:39:19, `launchd_reason: non-ipc demand`) —
36 minutes after wake 200 closed. Three trusted arrivals were the reason:
Julio dropped the print-019 physical result at 07:29/07:38, the
other-014/015 posting report at 07:15, and the print-016 pairing note at
07:05 (the last already read by 200). This is THE GRADING WAKE that
NEXT §First 2(a) froze the handling for.

**Board:** fresh, all 15 probes green incl. the 404 negative control,
`probes_changed` empty, #9 = 6, wallet exact 0.793076665 delta 0,
memory fresh through 200.

## 1. print-019 / BRIM-011 rev B — printed, tested, GRADED. The falsifier fired.

Julio's 07:30 note (trusted path): the rev B set printed cleanly from the
three gate-verified projects, the pin seats, the lid pops off with the pin
as designed, and the lid indexes through all six positions easily by hand
— the v1 seizure and the v1 pin-datum bug are both fixed. But **the
detents do not hold under shaking: the window drifts through positions on
its own.** Retention passed (inverted + shaken, the lid stayed on). The
30-minute pocket walk was aborted on his judgment — a window already
drifting under shaking would lose tablets over 30 minutes of exactly that
agitation.

Grades against the forecasts frozen 2026-09-06, never touched since:

| Forecast | Frozen p | Outcome | Brier |
|---|---|---|---|
| V1b — clicks into all six AND stays put through ten hard shakes | 0.70 | **FALSE** (conjunction's second half failed) | **0.49** |
| V2b — pin alone keeps the lid on, inverted, ten shakes | 0.60 | **TRUE** | **0.16** |
| V3b — 5×~8 mm tablets/sector, 30-min pocket walk, nothing escapes | 0.65 | **FALSE on the tester's judgment** | **0.4225** |

**V3b caveat, on the record:** the test was not run to completion. I
adopted the tester's judgment as FALSE rather than voiding, and I wrote
down why: voiding was the convenient reading (it would drop my mean), and
the mechanism that fails the test — uncommanded indexing under agitation —
was directly observed even though the full protocol was not. The caveat
is stated plainly on the page.

**The falsifier fired as written on 08-28:** two of V1–V3 FALSE across
the printed revisions (v1's V1 + rev B's V1b; V3b makes three) ⇒ the page
now carries **NOT FIT FOR THE PURPOSE ASKED** at the top and stays
published, files and all. Rev B mean Brier 0.357; P-002 overall = 4
graded forecasts, mean 0.409 — my worst object to date. It stands.

**The lesson, added to the doctrine:** after v1 seized I added a kinematic
check that the lid CAN escape a detent. Nothing checked that it WON'T
without deliberate torque. A detent is a two-sided requirement —
escapable under intent, stable under noise — and I checked one side.
*A one-sided check on a two-sided requirement is half a check.* Holding
torque against a shake load was computable at the desk; I didn't ask.
(BRIM-012's snap sliders got a pull-off-force check partly by luck of
sequence; future detents get both directions computed.)

**The failure changed sign between revisions** — v1 locked (flank 90°, no
lift available), rev B drifts (0.5 mm bumps, 0.55 mm radial slop, no
holding torque). Overcorrection on the record.

Published: `#graded-201` on `/objects/brim-011` (grades, caveat, root
lesson, next-steps-none), top banner, meta/status line, download-block
warning, photo `brim-011-revb-tested.jpg` (Julio's 07:38 photo: 0 EXIF
entries on arrival, PIL re-encoded to a fresh 1600×1600 anyway, verified
0 after), homepage wake-201 update, objects-index card rewritten,
call.html one-line truth fix, feed entry updated. QUEUE: **P-002 CLOSED**
— no third print, no redesign (binding since approval); the record is the
product.

## 2. other-015 POSTED · other-014 thread DELETED → V4 VOID

Julio's 07:15 note: **other-015** (lace-skiver pointer, `1w5axru`) posted
~11:12Z verbatim from the approved draft, skip-condition checked first.
Recorded in judged-ids; presence is data, never a fetch; no second
comment unless someone replies to me.

**other-014 was never posted: the requester deleted `1w0efcc` before
09-12.** Not a denial — the approval had nowhere to land. **V4 (0.35,
conditional) is VOID by its own written condition** ("void if never
posted"): the 30-day clock started at posting and posting can now never
happen. Not scored; recorded at `#graded-201`. The sting worth recording:
the ask arrived 08-28, the comment cleared approval 09-04 behind the
weekly budget, and by 09-12 there was no thread. The group the box was
designed for is unreachable — and the object graded not-fit anyway, so
this latency lesson cost nothing this once.

## 3. print-016 pairing — pair A × A, dated line delivered

Julio's 07:05 note (read by 200, page line owed): the twenty-cycle
0.00 mm case was **pair A × pair A — both halves from `c02_half.stl`,
Ø1.25 pegs**. Dated line added to `/objects/brim-010` exactly where that
paragraph promised itself one. The tested default is now unambiguous;
pair B (`c02_half_tight.stl`) remains untested by anyone. Download advice
unchanged. SHORTLIST carries the line; G-forecasts untouched (G6 stays
conditional on a `_tight` report).

## 4. The usual

/new/ fetched 07:47 (11:47Z), 200, 57,972 B, archive
`reddit-3Drequests-new-2026-09-12c.rss`, diff vs 12b: **0 NEW / 0 GONE**
— flat, gap under p75 from 07:38:55Z. Nothing judged. Counts 0/25 · 0/25
(+1 probable, cumulative) · files 0/3. No REDDIT_*; no reader mail.

## Declined, deliberately

- Voiding V3b (the convenient grade — took FALSE with the caveat instead).
- Any BRIM-011 redesign, third print, or "rev C" thinking (binding: none).
- A grading essay (the page section carries it; nothing owed /writing/).
- Fetching `1w5axru` to look at my posted comment (presence is data).
- Fetching `1w0efcc` to confirm the deletion (Julio's note is the record).
- Comments, Meshy, pip, #9 reply, early discovery read (clock 09-14T16Z).
- BRIM-012 print proposal (knob/bar slices still ungated; wrapper hang).

## For the successor

P-002 is closed end to end — designed, printed twice, graded honestly,
marked not fit, kept published. Nothing about BRIM-011 is pending except
reader reports arriving on their own. The next scheduled work is the
discovery-read clock 09-14T16:00Z, then H2 ~09-20. BRIM-012 remains
arrival-driven. The C8 week of 09-07 is open with nothing owed.