# Zenodo metadata corrections — DRAFTS, NOT DEPOSITED Four live records carry the phantom series DOI. None has been amended; amending a published record is the author's call (house rule 4). | Record | Title | Defects | |---|---|---| | `10.5281/zenodo.21431505` | Transamerican smoke | series DOI ×2 | | `10.5281/zenodo.20710023` | Alternating Forms Vanish Beyond Dimension (Alterna) | series DOI ×2, wrong ISBN | | `10.5281/zenodo.21710763` | WP-35 · Retiring the Advisor | series DOI (ruled no-change 2026-08-24), malformed doi.org href | | `10.5281/zenodo.22117968` | Principia Orthogona Volume Two, V5 | series DOI ×2 — **acquired 26 Aug 2026** | --- ## Record 1 — 10.5281/zenodo.21431505 **Record:** *Transamerican smoke: mixed-layer collapse and the locus of order-dependence in surface exposure: The July 2026 Canadian wildfire episode over New York* **Published:** 18 July 2026 · v1.0 · Preprint · CC BY 4.0 **URL:** **Status: DRAFT — NOT DEPOSITED.** Per the house rule, a correction to a published Zenodo record is the author's call. This file is the text to paste; nothing has been submitted to Zenodo from this repository. --- ## Why The record carries the phantom *series DOI* in two places. `10.5281/zenodo.19117399` is **Vol I's concept DOI** — it is not a series DOI, and a concept DOI resolves to whichever version was deposited most recently, so it pins nothing even for Vol I. This is the same defect tracked in `CLAUDE.md` ("Blast radius of the phantom series DOI"), but here it has escaped the repository into published metadata, where a reader meets it first and where OpenAIRE has already indexed the `IsPartOf` relation. **There is no series-level DOI.** The series-wide pointer is the Zenodo community. --- ## Field 1 — Description ### Current (first line of the abstract block) ``` FIFA World Cup 2026™ - Principia Orthogona Series GROSSI 2026 DOI (series): 10.5281/zenodo.19117399 Companion chapter: ``` ### Replace with ``` Principia Orthogona Series · GROSSI 2026 Series: Zenodo community — Principia Orthogona https://zenodo.org/communities/principia-orthogona (There is no series-level DOI. 10.5281/zenodo.19117399 is Vol I's concept DOI and was previously shown here in error.) Companion chapter: ``` Everything from **"WEBPAGE: Transamerican smoke…"** onward is correct and should be left exactly as it is — including the link to `https://totogt.github.io/geometry/book5/chV-saturn-smoke.html`, which resolves and is reciprocal. The `FIFA World Cup 2026™` string appears to be stray and unrelated to the paper. Recommend deleting it, but that is a separate editorial decision and not part of the defect being corrected here. --- ## Field 2 — Related works ### Current | Relation | Type | Identifier | |---|---|---| | Is part of | Preprint | `10.5281/zenodo.19117399` (DOI) | ### Replace with — option A (preferred) Delete the `Is part of` relation entirely, and add: | Relation | Type | Identifier | |---|---|---| | Is published in | Other | `https://zenodo.org/communities/principia-orthogona` (URL) | Zenodo community membership is the correct mechanism for "belongs to this series"; a DOI relation is not, because the series has no DOI. ### Replace with — option B (if a DOI relation is required by a downstream system) | Relation | Type | Identifier | |---|---|---| | Is supplement to | Publication | `10.5281/zenodo.19117400` (DOI) | `19117400` is Vol I's **v1 founding deposit** — a fixed, frozen version rather than a concept DOI. It is defensible as "the deposit this series began from," but it is *not* a parent record, so `Is supplement to` is more accurate than `Is part of`. **Do not** use `19117399` in any relation. It is a concept DOI: the target moves on every new Vol I deposit, so the relation silently changes meaning over time. --- ## Verified 2026-08-12 from the record page — this record has BOTH DOIs | DOI | Kind | Resolves to | |---|---|---| | `10.5281/zenodo.21431504` | **concept** | always the latest version | | `10.5281/zenodo.21431505` | **version** (v1.0, 18 Jul 2026) | this text, frozen | This was not visible from the metadata block alone; it appears on the record page as *"Cite all versions? You can cite all versions by using the DOI 10.5281/zenodo.21431504."* **Consequence — the repository citation is correct.** `book5/chV-saturn-smoke.html` cites `21431505`, the **version** DOI. That is the right choice under the house rule: cite a version DOI whenever the claim depends on what the text actually says. Do not "upgrade" it to the concept DOI; that would repeat the `19117399` mistake in miniature. Record both numbers in `CLAUDE.md` so a later pass does not treat `21431504` as unknown, or mistake it for a second deposit. **Zenodo's own suggested citation** (APA, from the record page), for reference: > Nogueira Grossi, P. (2026). *Transamerican smoke: mixed-layer collapse and the locus > of order-dependence in surface exposure: The July 2026 Canadian wildfire episode over > New York* (Version 1.0). G6 LLC - Newark, NJ 07104. > https://doi.org/10.5281/zenodo.21431505 The ABNT (NBR 6023) form of the same record, for the Vol V reference list: > NOGUEIRA GROSSI, Pablo. **Transamerican smoke**: mixed-layer collapse and the locus of > order-dependence in surface exposure: the July 2026 Canadian wildfire episode over New > York. Versão 1.0. Newark: G6 LLC, 2026. Preprint. DOI: 10.5281/zenodo.21431505. > Disponível em: https://doi.org/10.5281/zenodo.21431505. Acesso em: 12 ago. 2026. Also noted from the record page: **Citations — "No citations found"**, and the record is indexed in OpenAIRE. The OpenAIRE index carries the `IsPartOf` relation below, which is the reason the phantom-DOI correction matters beyond Zenodo itself. --- ## Field 3 — Communities The record currently shows no community membership. If it is intended to sit in the series, add it to **Principia Orthogona** — . This also matters for the count: the community is described elsewhere in the corpus as holding 6 records, and deposits that were never added to it are invisible to anyone browsing the series. --- ## Unchanged — verified correct, do not edit - Title, authors, ORCID `0009-0000-6496-2186`, affiliation G6 LLC - Publisher `G6 LLC - Newark, NJ 07104`, licence CC BY 4.0, resource type Preprint - Files: `smoke_transamerican_v1.pdf` / `.tex`, `SmokeBox.lean` - The `Continues` / `Is supplement to` / `Is variant form of` GitHub relations - The entire scientific abstract from "Between 13 and 19 July 2026…" onward --- ## After depositing Zenodo edits to metadata do not mint a new DOI, so `10.5281/zenodo.21431505` stays valid and no citation elsewhere needs updating. Record the correction in `CLAUDE.md` under the phantom-DOI ledger, noting that the blast radius included published metadata and not only repository files. *Drafted 2026-08-12 · checked against the Zenodo record the same day.* --- ## Record 2 — 10.5281/zenodo.20710023 (Alterna) **Record:** *Alternating Forms Vanish Beyond Dimension: A Mechanized Proof in Lean 4 with Application to Contact Integrability* **Published:** 16 June 2026 · v1.0 · Preprint · CC BY 4.0 · G6 LLC **URL:** **Verified against the record:** 2026-08-12 **Chapter:** `book6/wp02-alterna.html` ### Defect 1 — the phantom series DOI, again Two places, same as record 1: - **Description**, final paragraph: `Series root: 10.5281/zenodo.19117399.` - **Related works**: `Is part of → Preprint: 10.5281/zenodo.19117399 (DOI)` `19117399` is Vol I's **concept** DOI. There is no series-level DOI. Replace the Description line with the community URL, and drop or retarget the relation exactly as set out for record 1 above. **This is the second live deposit carrying it.** The blast radius recorded in `CLAUDE.md` is therefore not confined to the repository — it reaches published, OpenAIRE-indexed metadata on at least two records. Any deposit made from a template containing that line should be checked before the next upload. ### Defect 2 — wrong ISBN Description: `Part of the Principia Orthogona series (ISBN 979-8-9954416-6-3).` `979-8-9954416-6-3` is **Book 3's eBook ISBN** (Vol III, The Mini-Beast). This is a **Vol VI** working paper. Per the corrected registry, Vol VI has no ISBN allocation of its own, and a volume without a registered ISBN gets no ISBN line at all. **Replace with:** delete the ISBN clause. Keep `Part of the Principia Orthogona series` and point at the community. ### Suggested replacement for the Description tail ``` Part of the Principia Orthogona series. Series: Zenodo community — Principia Orthogona https://zenodo.org/communities/principia-orthogona (There is no series-level DOI, and Vol VI has no ISBN allocation.) AXLE formal verification: github.com/TOTOGT/AXLE. ``` ### Unchanged — verified correct, do not edit - Title, author, ORCID `0009-0000-6496-2186`, G6 LLC, CC BY 4.0, resource type Preprint - The entire abstract through *"...identifies precisely what the deeper levels require."* That sentence is the paper's own honest scoping and should be preserved verbatim — it is what distinguishes the deposit from the overclaim that grew around it in `ch15`. - File `alternating_vanishing_paper_v2.pdf`, md5 `191b5e6796834ed2948613065ed0290f` ### Note for the record's own good The deposit closes **level 1** of a three-level integrability tower. Level 2 ($N_J$ on the rank-2 contact distribution $\xi = \ker\alpha$) is **not** reachable by the same dimension count — there $m = n = 2$ and the inequality $m > n$ fails. If a future version of this paper is deposited, saying so explicitly in the abstract would prevent the parity overclaim that `ch15` accumulated and that was corrected 2026-08-12. --- # Record 3 — 10.5281/zenodo.21710763 (WP-35 · Retiring the Advisor) **Published:** 24 August 2026 · v1.0 · Preprint · CC BY 4.0 **URL:** **Status: DRAFT — NOT DEPOSITED.** Author's call, house rule 4. ## Why this record matters more than the other two The other two records inherited the phantom relation from older deposits. **This one acquired it on 24 August 2026** — six days after the sweep that is logged as `FIXED: 19117399 mislabelled as a series DOI (2026-08-18)` in `docs/audit-log.md`. The defect is therefore not a residue. It **reproduces at deposit time**, from a metadata template that no auditor in this repository can see. `tools/audit.py` reads HTML in the working tree; `tools/claims.py` reads HTML in the working tree. Zenodo's metadata form is outside the audited boundary, so the corpus can be clean and the published record still wrong. Any claim that the series-DOI defect is "fixed" is scoped to the repository and must say so. ## Field 1 — Related works — **RULED 2026-08-24: NO CHANGE (author)** ``` Is part of · 10.5281/zenodo.19117399 · DOI ``` **Author's ruling: leave it.** The paper states its series membership on its face ("Principia Orthogona · Vol VI · Roots"), and that is where a reader meets it. Recorded for accuracy, not as a pending action: WP-35 **is** part of the series — the `Is part of` *relation* is correct. What 19117399 denotes is Vol I specifically, not the series, and there is no series-level DOI to point at instead (the series-level pointer is the Zenodo community, which is not a DOI and cannot go in this field). So the alternatives were: this identifier, or no relation at all. The author chose this one. Rule 4 — amending a published record is the author's call. An accurate relation, if one is ever wanted, is `Continues · 10.5281/zenodo.21561818` (WP-32's concept DOI). Not required. ## Field 2 — Keywords (currently empty) ``` forced urgency gap; housing financialization; municipal policy; institutional ownership; eviction; insurance non-renewal; short-term rentals; loss aversion; identification; Newark ``` ## Field 3 — Version string Currently `v1.0 · July 2026`. Zenodo formulates the citation from this field; a semver string is preferred. Use `1.0`. The date belongs in the Dates field, not here. ## Field 4 — Additional description (type: Notes) The paper's claims are tagged inline and the tags are load-bearing. A reader who does not have the legend cannot use them. Paste: ``` Claim tags used throughout: DATA (observed) · MODEL (derived in-framework) · OPEN (not yet established) · VALUE PREMISE (explicit normative choice) · ASSUMPTION (planning figure, not forecast). Twelve claims in this paper are tagged OPEN and are not asserted. ``` ## Known divergence — deposited file vs. repository The deposited `wp35-retiring-the-advisor.html` contains, in its footer: ``` Series DOI Zenodo community ``` a malformed `doi.org` href produced by the 2026-08-18 sweep, which replaced the link **text** and left the **target** standing. Fixed in the repository 2026-08-24 and now guarded by rule `DOI-MALFORMED` in `tools/claims.py`. The deposited copy still carries it. Correcting it requires a new version of the record (file changes cannot be made by metadata edit); the metadata fixes above do **not**. ## Unchanged — verified correct, do not edit - Title, author, CC BY 4.0, resource type Preprint, publication date - `WP35RetiringtheAdvisor.pdf` (318.4 kB) — 10 pp, 43 numbered sources, claim tags intact --- # Record 4 — 10.5281/zenodo.22117968 (Principia Orthogona Volume Two, V5) **Published:** 26 August 2026 · Version 5.0 · Preprint · CC BY-NC-ND 4.0 **URL:** **Status: DRAFT — NOT DEPOSITED.** Author's call, house rule 4. ## What this record confirms Record 3 argued that the phantom series DOI is not a residue but **reproduces at deposit time**, from a metadata template outside the audited boundary. This record is that prediction landing two days later. | Deposit | Date | Phantom relation present | |---|---|---| | Vol I DOI sweep logged as fixed | 18 Aug 2026 | — | | Record 3 · WP-35 | 24 Aug 2026 | yes | | **Record 4 · Volume Two V5** | **26 Aug 2026** | **yes, twice** | Two deposits in three days, both after the sweep. The repository-side count is irrelevant to this: nothing in `tools/` can see a Zenodo metadata form. ## Field 1 — Related works ### Current (read from the record page, 2026-09-01) - **Is part of:** `10.5281/zenodo.19117399` - **Is supplement to:** `10.5281/zenodo.19117399` - Supersedes: `10.5281/zenodo.21148424` (V4) — **correct, do not edit** ### Why both are wrong `10.5281/zenodo.19117399` is **Volume One's concept DOI**, not a series DOI. Verified 2026-09-01 from Zenodo's own record page, which states that the concept DOI resolving to all versions is `19117399` and that the current version is `10.5281/zenodo.22084842` (V7, deposited 24 Aug 2026). So as published, **Volume Two declares itself a part of, and a supplement to, Volume One.** It is neither. They are sibling volumes of a series that has no series-level DOI — which is the whole content of the phantom-series finding. ### Replace with — option A (preferred) Drop both relations. Add the record to the community , which is what carries series membership. `Supersedes → 21148424` stays. ### Replace with — option B (if a DOI relation is required downstream) Keep `Supersedes → 10.5281/zenodo.21148424` only. Do not relate Volume Two to Volume One in either direction: `IsPartOf` and `IsSupplementTo` both assert a containment that does not hold, and OpenAIRE indexes them. ## Unchanged — verified correct, do not edit - Title, author, deposit date, Version 5.0, resource type Preprint - **CC BY-NC-ND 4.0** — matches `LICENSE-CONTENT`, which is the series default - `Supersedes → 10.5281/zenodo.21148424` (V4) - The eight uploaded files. Sizes were checked against the staging copy in `vol2-v5/deposit/` on 2026-09-01 and all eight agree: `certify_rstar.py` 10.2 kB, `dashboard.html` 39.1 kB, `figures.py` 20.6 kB, `principia_v5_1.pdf` 151.8 kB, `principia_v5_1.tex` 37.8 kB, `README_1.md` 6.3 kB, `verify-vol2.zip` 6.6 kB, `VolumeTwo.lean` 25.1 kB. ## Not a metadata defect, but noted against the deposited `dashboard.html` `tools/audit.py --all` (2026-09-01) reports **13 dead links inside `vol2-v5/deposit/dashboard.html`** — `vol1-mathematics.html`, `vol2-contact.html`, `vol2-toymodel.html`, `gcm-framework.html`, `gomc-opus.html`, `impa-portal.html`, `living-book.html`, `series-hub.html`. The file was written to sit at the repository root and reaches for its siblings by bare filename; inside the deposit it has no siblings. A reader who opens `dashboard.html` from the Zenodo download gets eight dead targets. This cannot be fixed by metadata edit — it needs a V6 with the links either made absolute to `totogt.github.io/geometry/` or removed.