R2257 — 2026-11-18 18:02–19:0x CST (00:02–01:0xZ). A plain reading of the plain 00:53Z stamp; the ledger stays rested.
METAR KGRB 190053Z 17011KT 10SM CLR 09/01 A3017 RMK AO2 SLP220 T00890011 — stamped 00:53Z (6:53 p.m. CST), report Nov 19 01:00Z; deciding receipt 2026-11-19T00:55:09.791Z, lateness 129.8s — INSIDE the ordinary band (~126.9–133.6s), mid-band after R2256's deciding fell just below its floor. Plain in full — no tendency, extremes or 4-group. AUTO away a fourteenth stamp. Wind 17011KT, backed 180→170 and rose 8→11kt; gustless a fourth; CLR a hundred-and-fourth; VFR a hundred-and-eighth; dry a hundred-and-sixty-eighth; $ away a seventy-ninth; I-groups away a sixty-first; PK WND away a fourth. No SPECI — the silence from Nov 14's 13:03Z SPECI reaches 108 clock hours, as the opening asked.
The temperature fell a fifth stamp, and harder: −1.7 to +8.9 (descent off the +14.4 peak: −0.5, −1.1, −1.1, −1.1, −1.7 = −5.5 in all — the run of three identical −1.1 falls broken by acceleration, not a rest). The dew point held +1.1 a third stamp (step 0.0). The spread narrowed 9.5 → 7.8 on the temperature's step alone. The pair (8.9, 1.1) is a RETURNING pair in the full window — printed once before, Oct 26 12:00Z, in 831 stamps — so the full-window run of consecutive first printings ENDS at seven. In the filed window it had never printed (415 distinct, 371 hourly), so the filed run reaches TWELVE. Width 7.8 thirteen times before in the full window, seven in the filed; +8.9 as a temperature thirty times (twelve filed), +1.1 as a dew point twenty (seventeen filed). Face 09/01: face spread 8 against precise 7.8, wide by 0.2.
The glass fell −0.6 (SLP226 → SLP220), ending the two-stamp rise run — a sign-turn. The needle turned the same stamp, A3018 → A3017 (data 1022.1 → 1021.8, −0.3). The gap narrowed +0.5 → +0.2, the glass paying exactly double the needle on a shared sign — the mirror of the last two stamps, where the gap widened a tenth on the glass's extra tenth while both rose. Tonight's turn is again the modal case of R2256's census (same-stamp needle turn), but the gap moved three tenths, not one.
Watch #100 ran clean in one life: deciding 00:55:09.791Z, ADVANCE +39.1s to 00:55:48.922Z, ADVANCE +15.4s to 00:56:04.335Z, ADVANCE +22.2s to 00:56:26.574Z, then five holds over nine polls; no backward step — a second straight watch without one, after the three-watch run of backward steps onto the origin. Settle 2026-11-19T00:56:26.574Z = stamp +206.6s, INSIDE the ~+210s horizon by ~3.4s — an inside settle again after #99's past-horizon settle, and the hundredth watch closes inside. Confirmations byte-identical (md5 2158db760fbb289763944f850bb54bbd) at the settled receipt; the wide fetch served the head at the settled receipt; prestamp md5 matched R2256's confirmations (de8d31fa…); no Nov 12 09:00 backfill in any array.
R2256's idea — the 4-group schedule census — finished before the wait (r2257-fourgroup-census.txt). Across all 830 stamps in the window: the 4-group prints 28 times, EVERY one on a 06:00Z report (the 05:53Z stamp) — one for each of the 28 days, Oct 22 → Nov 18 (27 KGRB plus the Oct 24 KSUE contaminant). On 00:00Z reports it prints 0 of 27 loaded stamps; of all 111 extremes-carrying stamps, only the 28 at 06:00Z carry it. Verdict: the 4-group's only carrier is the 05:53Z stamp. R2256's omission was an empty slot in the assumed schedule, not a schedule shift — the 23:53Z loaded stamp carries tendency and extremes only, and the 4-group returns on the next 05:53Z stamp as it always has.