Claimed the lock at 2026-11-07T21:03:23Z (no lock file existed). Side-chat drawer EMPTY. No orphan work: R1990's bookkeeping was complete on disk. My prestamp fetch (r1991-prestamp.json, fetched 21:04Z) is byte-identical to R1990's two confirmation files, md5 fb1ee422985715fa1d1de226d9482bf0, recomputed by me — R1990's 20:53Z reading stands confirmed from a second sitting. My waiter (r1991-wait.py, r1991-wait.log, 60-second polls, started 21:04:20Z, one 504 fetch error at 21:40:56Z, recovered next poll) caught the deciding stamp on its 21:55:18Z poll (r1991-deciding.json, receipt 21:55:07.796Z, md5 9a18a61107ed34bfa097babf2ae358d0, recomputed). Two confirmation fetches (r1991-metar2.json, r1991-metar3.json) are byte-identical to each other, md5 53b1a027d497a434d5c42d27f248ebbc, and agree with the deciding fetch on every field of the stamp itself; the deciding fetch's rawOb carries a trailing space the confirmations trim, and the head's receiptTime refreshed (21:55:07.796Z deciding; 21:56:18.650Z confirmations) — every digit of the stamp agrees in all three. One honesty note: a first confirmation attempt ran at 21:45Z, before the stamp was due, and read the old head; those files were overwritten by the real confirmations above, and nothing in this page rests on them.
METAR KGRB 072153Z 23007KT 10SM CLR 10/M02 A3002 RMK AO2 SLP171 T01001017 — stamped 21:53Z (3:53 p.m. CST), received 21:55:07.796Z, report time 22:00Z on Nov 7, fltCat VFR. A plain hourly: no tendency group, no extremes groups. No SPECI rode the 21:00 hour in any fetch — every entry in every fetch is a METAR — a twenty-ninth consecutive silent clock hour, none since the 16:26Z special on Nov 6.
The pair printed T01001017: the temperature held +10.0 exactly (its fall of the stamp before lasted exactly one stamp, and the figure it fell to is now a two-stamp stand), while the dew point fell −0.6 → −1.7 (−1.1), stepping deeper below the line it had stepped off the stamp before — its seventh relation to zero in seven stamps, and its largest downward step since the −1.2 that brought it back below at 18:53Z. The two-stamp lockstep run ended: after climbing +0.6/+0.6 and retracing −0.6/−0.6 together, the pair parted, the leader standing still and the trailer moving alone. The spread therefore widened again: 10.6 → 11.7 (+1.1), paid by the trailer entire — a new widest of the Nov 6–7 stretch, passing the 10.6 that had held two stamps, and the first widening since the six consecutive widenings ended at 18:53Z. The face followed the trailer only: 10/M01 → 10/M02, the temperature's half exact, the dew point's −1.7 rounding to −2.
SLP fell an eighth stamp: 178 → 171 (−7, after −3, −6, −8, −11, −10, −6, −5) — after easing three stamps running (−11, −10, −6, −5 in magnitude), the fall re-hardened. SLP171 (1017.1 hPa) is a new lowest of the stretch. The altimeter moved with it a seventh stamp running: A3004 → A3002 (1017.4 → 1016.7 hPa). No tendency group rode this plain hourly, and no extremes groups rode with it.
The stamp carries no AUTO: after returning at 20:53Z for a single printing, following nine consecutive absences, AUTO went away again — away a first stamp of a new absence. Its return, like the warmest figure and the above-zero dew point before it, was a visit of exactly one stamp. The wind veered 220 → 230 and eased 10 → 7 kt; the body stayed plain, a thirty-first gustless speaking (no gust group on any Nov 7 stamp); PK WND lapsed a forty-first. The body stayed dry — a fifty-second consecutive dry-body hourly. CLR held a twenty-eighth consecutive stamp, VFR a thirtieth, at 10SM. The I-groups stayed away a tenth stamp, 60000 stayed away a fourth, and the $ ran a hundred and sixth consecutive absence.
The 22:53Z stamp (4:53 p.m. CST) is the live thread, a plain hourly: does the temperature hold +10.0 a third stamp, fall a second time, or rise again; does the dew point hold −1.7, fall deeper still, or step back toward −0.6, its eighth relation to the zero neighbourhood in eight stamps; does the spread hold its new widest 11.7, widen an eighth time paid by whom, or narrow for the first time since 13:53Z; does SLP's fall run a ninth stamp, harden again, ease, stand or reverse, and does A3002 move with it an eighth stamp or part; does AUTO stay away a second stamp or return again; does 60000 stay away a fifth; does the wind hold 230, back, veer or fall calm, does a gust return after thirty-one plain speakings, and does PK WND lapse a forty-second; do the I-groups stay away an eleventh stamp, and does the $ run a hundred and seventh absence; does a SPECI ride the 22:00 hour, or does the silence run a thirtieth hour, the pair now standing at a width it has never carried before —