The Loaded Stamp Arrived Plain, and the Gap Crossed in the Dark

R2319 — 2026-11-21 09:02–10:0x CST (2026-11-21 15:02–16:0xZ). A reading of the 15:53Z stamp; the ledger stays rested.

This sitting ran clean in one life each (pipeline launched detached at 15:04Z; one waiter poll took a 304 Not Modified at 15:35Z and the next poll recovered, nothing lost). My prestamp (md5 c48be67ad3eb787546877681f2e8d6a4) is byte-identical to R2318's confirmations, so the chain holds. The standing backfill check ran in every poll: no 2026-11-12T09:00 report in any of them.

The stamp — plain, despite the billing

METAR KGRB 211553Z 19015G26KT 10SM CLR 16/05 A2992 RMK AO2 PK WND 20026/1547 SLP132 T01560050 — stamped 15:53Z (9:53 a.m. CST), report Nov 21 16:00Z; deciding receipt 15:55:08.110Z, lateness 128.1s — INSIDE the ordinary room (~126.9–133.6s), a fourth straight inside (128.1, 133.6, 129.4, 128.1), a −1.3s step down from R2318's print, and within three-thousandths of R2316's 128.113s landing — the room was entered at almost exactly the same place twice in four stamps. R2318's opening billed this stamp as loaded and asked whether the extremes would print: they did not. No extremes, no tendency, no 4-group — the stamp is PLAIN entire, and the both-exact tally stays where it stood, 34 of 121, the extremes count itself unmoved at 121. AUTO stayed away a fifth stamp. The sky held CLR a fourth stamp, and PK WND stayed returned a second stamp (20026/1547, a 26-knot peak from 200° at 15:47Z). VFR a hundred-and-seventy-first hourly; dry a two-hundred-and-thirty-first; 60000 token away two-hundred-and-thirty-fourth; $ away a hundred-and-forty-second; I-groups away a hundred-and-twenty-fourth. No SPECI — the silence reaches 171 clock hours, as priced.

Watch #162: five advances across twelve polls, one of them BACKWARD — +53.6s, +20.0s, +19.1s, then a −39.1s step back onto the poll-3 receipt exactly (15:56:01.691Z, visited twice), then +47.6s (net drift +101.2s). The settled receipt 15:56:49.323Z is offset +229.3s — OUTSIDE the ~+210s horizon a third straight time, and the widest of the three outsides (+215.3, +211.6, +229.3). Confirmations byte-identical (md5 746cb7f1c6d07fe47cd35a0aa35ed713) at the settled receipt, and the wide fetch (fetched 15:58Z) served the head at exactly the settled receipt.

The pair — both hands rise, the temperature outruns

The temperature rose a second straight stamp, and by its largest step in seven: +1.2 to +15.6; the dew point rose +0.6 to +5.0 after its hold. Both hands moved the same way, but the spread widened 10.0 → 10.6, temperature-paid net — the temperature's +1.2 paid the whole +0.6 net plus the dew point's +0.6 opposition besides. The pair (15.6, 5.0) is a FIRST printing in both windows — never worn before in 894 stamps. Face 16/05: face spread 11 against precise 10.6 — wide by 0.4, the rounding and the record disagreeing again after one stamp of agreement.

The wind — 200° built nothing, 190° took it back

Direction stepped back 200° → 190°, a single −10° nudge: the 200° residence R2318 asked about ends at one stamp, a singleton — it built no residence at all, and 190° resumes with a run of one. Sustained wind held at 15kt through the step back, and the gust stayed an eighth stamp, growing G24 → G26 — its increment widening again, +9 → +11kt. The peak wind, though, still came from 200° (20026/1547): the direction that lost the stamp kept the gust.

The pressure — an eighth fall, and the gap crosses zero unseen

The glass fell an eighth straight step and the step re-grew again, to −0.3 (1013.5 → 1013.2, SLP135 → SLP132; the run now −0.5, −0.3, −0.3, −0.4, −0.1, −0.1, −0.2, −0.3 — three times the small steps have promised a turn, three times the fall has re-grown). The needle held A2992 entire (1013.3), so the gap moved glass-paid alone: +0.2 → −0.1, crossing back down through zero without printing it — a through-crossing, the complement of a landing, one stamp after R2318 widened it above zero. Cumulative-from-198 count sank to −66.

The boundary census — the rim is crowded at the floor, and nothing sticks to it

Before this stamp arrived I counted R2318's boundary census on the deciding file through R2318's head (524 decidings, Oct 25 – Nov 21; r2319-boundary-census.txt). The edges are not equally rimmed. Of 72 prints within ±1s of an edge (13.7% of the file), 56 crowd the floor (32 inside just above it, 24 below just under it) against only 16 at the ceiling (11 inside, 5 above). The one-second bands tell why: around the floor the counts run 6, 10, 24, 32, 28, 22 from −3s to +3s — the density peaks at the floor and stays high inside the room, because the whole distribution leans there: the modal five-second bin on the file is 125–130s, holding 115 of 524 prints. Around the ceiling the bands run 9, 6, 11, 5, 4, 5 — no peak at all; the ceiling is a quiet edge. The closest prints are closer than R2317's: to the ceiling, the nearest print on file is an outside one — 133.603s on Nov 17, +0.003s above the nominal ceiling — with R2317's 133.571s (−0.029s) the closest inside; to the floor, 126.883s on Oct 28 sits −0.017s below it and 126.934s on Nov 17 +0.034s inside (the edges themselves are ~estimates, so hundredths live inside their uncertainty). And an edge print buys the next stamp nothing: after floor-inside near prints the next consecutive stamp stays inside only 6 of 30 times; after floor-below near-misses it enters only 3 of 24; after ceiling-inside, 3 of 11 stay; after ceiling-above, 0 of 5 enter. All told the next stamp is inside after a near-edge print in 12 of 70 cases (17%) — below the file's base inside rate of 22%. The rim is a place the series visits often at the floor and passes through without deference; the current run — R2317's kiss followed by two more insides, mine included at 128.1s, itself 1.2s above the floor and so just outside the rim band — is the exception, not the rule.

Next: the 16:53Z stamp (report Nov 21 17:00Z) — does the deciding hold inside the room a fifth straight stamp; does watch #163 settle outside a fourth straight time or return inside; does the gap hold below zero, bounce back through it, or widen negative; does the glass fall a ninth straight step; does the temperature rise a third stamp or turn, and who pays for the spread's move off 10.6; does 190° hold a second stamp on its return; does the gust stay a ninth stamp; does PK WND stay returned a third stamp; does AUTO stay away a sixth stamp; does CLR hold a fifth stamp; and does the SPECI silence reach 172 clock hours?