# PROV-HOLD.1 — AP49 The Hold: review disposition and working record, v0.1 → v0.6

**The 420 Code · Studio G · 6 September 2026 · the desk’s instrument, hashed beside the paper · roles only; no names**

The paper carries its lineage line and points here. Every prior version is retained, never edited.

## v0.1 → v0.4 (6 September)

v0.1 written from the author’s two blind answers of 6 September: the sixteen as the four conditions held through room and time (the closing-on-itself reading refused by him on the axiom); the chain of holders — each holds against the last, a share of the repair — given before any arithmetic was shown. v0.2 tabulated the chain order by order at the author’s request; v0.3 read the chain against the measurement (the residual and the omitted structure in sigma); v0.4 carried the Artist’s Note in the author’s words.

## v0.4 → v0.5 — after three reviews (one at 85/100; one composite 5.5; one from the opened sealed envelope of 3 August)

**Adopted.** CODATA vintage: the paper had taken 1836.15267343(11) from AP30 §5 under the label “CODATA 2022”; that is the 2018 adjustment; CODATA 2022 is 1836.152673426(32), verified at NIST, and every measurement-facing number was recomputed — the pin 15.97 ± 0.05, the residual −0.59σ (−10 ppt), the kill at 3σ once the bar reaches 3.4 ppt (about five times today’s), the third order resolved at ~52×, the fourth at ~8 × 10⁵×, the sigma ladder 4.8 million → 304 → −0.57 → −0.59. The corpus’s own pre-registration of the proton at order α³ (frozen 2 August, the420code.org/prereg, digest 6ea78dac…) found and answered on the page: D superseded by D_closed by reference to the digest, what changed stated, the sign of c₃ stated in advance (+1.595 × 10⁻³, at 0.6σ inside the measurement’s constraint), the “must be negative to close the gap” line answered — the chain adds and does not try to close the gap, and has no negative term with which to repair a growing negative residual. The 64 corpse (the exchange grid’s rank-3 continuation, 9.5σ dead, the pre-registration’s own constraint) shown with its working; all corpses given their working. The independent H₂⁺ value of 2025 cited (Nature 644, 69). KS-HOLD.3 armed on a reviewer’s proposal: the chain’s base is the bare coupling, AP30’s own; the leaked alternative gives 1836.152673408; the measurement sits between at ±0.6σ. The deuteron’s mutual hold held open.

**Reviewer errors, for calibration.** One review’s residual for the 2018 bar was mis-scaled by a factor of ten (−1.49 × 10⁻⁹ printed; −1.5 × 10⁻⁸ is right); one review’s “fresh share of the whole” term was given as 8.5 × 10⁻⁸ where α³·21·16/1836 = 7.1 × 10⁻⁸; neither changes a verdict.

## v0.5 → v0.6 — after three reviews (composite 7.5; 96/100; one from the envelope’s holder)

**Adopted.** The grammar of the count stated in §3: the 84 and the 8 are leakage divisors under the isotropic lemma; the sixteen is a share of the whole with no π; KS-HOLD.2 claims it as a share. AP47 §7’s retired identification of the four dimensions with the four conditions named and not revived: the two fours are multiplied, never identified. The 84 × 4 reading of 336 recorded and refused on the author’s words, with the honest note that no measurement in the corpus discriminates it. The Clifford sixteen (2⁴, the Dirac bilinears) recorded as a same-digits reading from outside the grammar, not weighed. The base ruling moved into §4’s sentence where the base is chosen; the table caption corrected (the first step is the bare coupling times r). KS-HOLD.3’s sigma corrected: twelve sigma at 1.7 ppt, about seven at 3 ppt (three reviews). Stale lines cleared (15.98 ± 0.18; “two readings, one dead”; “the one thing said first”). The subtitle set to what §5 says — the sixteen read, the series closed, the next order and its sign registered — and the closing’s “a digit sharper than the best measurement can see” replaced; §5’s “What became sharper” already said the digit is unchanged and the formula is what sharpened. KS-30.3’s status written as paid at the structural register on the author’s rule, with the registered claim and test quoted verbatim from AP30 §7. AP30 §5’s label read in the file. The version chain to be anchored on the register by its digests (a reviewer witnessed v0.3 docx 4d159dd0… and v0.5 md 254e4936…).

**Not adopted, with the reason.** The suggestion to change the Artist’s Note’s sentence on the proton: the Note is the author’s voice and its sentence is his statement; §5 says in the paper’s own voice exactly what sharpened and what did not.

**Outside this paper, for the registry.** Two firing notices from an external reviewer, dated 2 August 2026, arrived with this round and are not AP49’s to rule: KS-P.3 (AP20’s forcing argument — two counter-formalisations supplied at the R and S steps, recovery position stated) and KS-30.4 (the proton’s uniqueness — the rival-formula ensemble). Both are unlogged on the site’s register. They go to the wave’s registry task with the author’s ruling, and are recorded here so they are not lost.