Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
120 changes: 120 additions & 0 deletions docs/DOWNSTREAM_RN_CROSSWALK_20260925.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,120 @@
# Downstream closure crosswalk — historical RN route versus 2026-09-24 successors

**Date:** 2026-09-25
**Disposition:** dependency classification only. This document changes no scientific
Boolean, frozen object, register row, or historical theorem statement.

**Placement:** `docs/DOWNSTREAM_RN_CROSSWALK_20260925.md`. It does **not** live under
`docs/math_status/`. That packet directory is closed (`EXPECTED_NAMES`); an extra file
there fails `tools/math_status_check.py` without discharging or promoting anything.
This note is navigation and classification only. It also does **not** edit
`docs/OPEN_PROBLEMS.md`, which is byte-pinned by `tools/twelve_project_check.py`.

## Purpose

The hardening branch still correctly records the historical D3-LEMMA-RN-UNIF route as
OPEN: ENV-RESCOV → ALLCELL-FDZ-Q4 → LOGQ-TAIL → SYM-Fw-jet → CH-LIFT, with Piece-2,
OBL-H5-JETMOD and FREEZE downstream. Historical objects `rnu_env.py`,
`CL_ANTHROPIC_BUNDLE_2026-09-17_v5.zip`, and `allcell_fdz_enclosures.json` are recorded
ABSENT. This crosswalk prevents newer author-side proofs from being mistaken for those
absent carriers or, conversely, old route-specific walls from being silently treated as
dependencies of every newer draft. The SIDE24 prep triad on this branch stays
ABSENT as a source of truth. This crosswalk does not invent a carrier for it.

This page is the human D0 classification companion to the in-development Math-
[downstream hard gate](https://github.com/d6g8k5htny-coder/Math-/pull/8)
(`DOWNSTREAM-HARD-GATE-20260925-v1`, main [#86](https://github.com/d6g8k5htny-coder/main/issues/86) /
[#90](https://github.com/d6g8k5htny-coder/main/issues/90)). Linking that draft has
scientific effect NONE. As of Math- head `5887a2f8`, the gate still has a load-bearing
own-node eligibility hole: dependency terminality alone can allow
`AUTHOR_SIDE_CANDIDATE` → `controlling`. Do **not** read this crosswalk as evidence
that Math- #8 already enforces every [#90](https://github.com/d6g8k5htny-coder/main/issues/90)
semantic.

## Classification rule

A newer draft can avoid importing a historical-route requirement only when its own
statement does not use that predicate. Avoiding the import does not set the old
predicate TRUE and does not discharge it.
Distinguish historical-route requirements, mathematical interfaces rederived by a
successor, and carrier/replay identities required only for the historical certificate.

Terminal integrity labels used by the Math- gate (not theorem acceptance):
`PROVED_REVIEWED`, `SUPERSEDED_NONBLOCKING`, `REFUTED`, `BLOCKED_ABSENT`. Green CI,
hashes, same-author review, and navigation success are never mathematical discharge.

## Crosswalk

| Historical object / predicate | Math- gate node | Gate classification | Active reading |
|---|---|---|---|
| `rnu_env.py` | `hist.rnu_env.py` | `BLOCKED_ABSENT` | ABSENT; historical-route blocker only |
| `CL_ANTHROPIC_BUNDLE_2026-09-17_v5.zip` | `hist.CL_ANTHROPIC_BUNDLE_2026-09-17_v5.zip` | `BLOCKED_ABSENT` | ABSENT; historical-route blocker only |
| `allcell_fdz_enclosures.json` | `hist.allcell_fdz_enclosures.json` | `BLOCKED_ABSENT` | ABSENT; historical-route blocker only |
| ENV-RESCOV | `hist.ENV-RESCOV` | `OPEN_HISTORICAL` | OPEN for historical D3. A newer author-side draft may restate a qualitative interface; that restatement does not set this predicate TRUE and does not discharge D3-LEMMA-RN-UNIF |
| ALLCELL-FDZ-Q4 | `hist.ALLCELL-FDZ-Q4` | `OPEN_HISTORICAL` | OPEN for historical D3. The fixed-remote author-side candidate is a different statement and does not discharge this predicate |
| LOGQ-TAIL | `hist.LOGQ-TAIL` | `OPEN_HISTORICAL` | OPEN historically. A named successor draft may use a different tail statement; that does not replace or discharge this predicate |
| SYM-Fw-jet | `hist.SYM-Fw-jet` | `OPEN_HISTORICAL` | OPEN historically. An analogous interface written in a successor draft is not a discharge of this predicate |
| CH-LIFT | `hist.CH-LIFT` | `OPEN_ACTIVE` | OPEN and still mathematically relevant for shrinking charts |
| Piece-2 annulus | `hist.Piece-2-annulus` | `OPEN_ACTIVE` | OPEN; active mathematical blocker for shrinking/pin-collision regions |
| OBL-H5-JETMOD | `hist.OBL-H5-JETMOD` | `OPEN_HISTORICAL` | OPEN for historical/certificate route |
| H5-ZBAND | _(not a separate gate node)_ | — | do not conflate with newer normalizer scopes; historical predicate unchanged |
| H5-REMOTE-THRESHOLD | see D4/D5 region nodes | region inventory | Stays OPEN. A fixed-remote height/spatial candidate is a different statement and does not discharge this predicate. The shrinking/global complement stays OPEN |
| `lemma_closed` | `hist.lemma_closed` | `FALSE` | no successor is an operator fold of D3 |
| `certified_C_H` | _(register companion)_ | false | newer qualitative constants are not this certificate |

## D4/D5 region complement (from the Math- gate)

`COVERED_BY_CANDIDATE` below is the Math- draft gate's own label. It is not
acceptance, not a discharge of `OBL-H5-JETMOD` or `D3-LEMMA-RN-UNIF`, and not a
SIDE24 source-of-truth carrier. SIDE24 stays ABSENT. Engineering classification
is not discharge.

| Region | Gate node | Classification |
|---|---|---|
| Fixed-remote covered set | `math.rn-region.fixed-remote` | `COVERED_BY_CANDIDATE` (draft-gate label only; not discharge) |
| Mesoscopic scaled annulus | `math.rn-region.mesoscopic-scaled-annulus` | `OPEN_ACTIVE` |
| Pin-collision | `math.rn-region.pin-collision` | `OPEN_ACTIVE` |
| Intermediate r≪\|x\|≪ρ | `math.rn-region.intermediate-r-to-rho` | `OPEN_ACTIVE` |
| Witness collision | `math.rn-region.witness-collision` | `OPEN_ACTIVE` |

## Author-side candidates (classification only)

These rows name drafts. They are not accepted theorems. No flag is flipped.
`OBL-H5-JETMOD` and `D3-LEMMA-RN-UNIF` stay OPEN. Engineering records are not
discharge. SIDE24 stays ABSENT; this page invents no source-of-truth carrier.

1. UNIFORM-MATRIX-CAP-LIFETIME-20260924-v1 / main [#63](https://github.com/d6g8k5htny-coder/main/issues/63) / gate `math.uniform-matrix-cap-lifetime` (author-side candidate, not accepted here): claims a compact-mark global elder defect O(r³), unrestricted leading finite-bar density, full normalizer and marked Kac–Rice interfaces.
2. LIFETIME-BOUNDED-REMAINDER-20260924-v1 / main [#67](https://github.com/d6g8k5htny-coder/main/issues/67) / gate `math.lifetime-remainder` (author-side candidate, not accepted here): claims an unrestricted cell^(-1/3)+O(1) bound and bounded nonselected density, importing named #63 interfaces.
3. RN-FIXED-REMOTE-WINDOW-20260924-v1 / main [#76](https://github.com/d6g8k5htny-coder/main/issues/76) / gate `math.rn-fixed-remote-window` (author-side candidate, not accepted here): claims a three-determinant O(r⁵) numerator and O(r³) expected count for the between-pin height window at fixed positive spatial distance.
4. RN-MESOSCOPIC-ANNULUS-REDUCTION-20260925-v1 / [Math- PR #7](https://github.com/d6g8k5htny-coder/Math-/pull/7) (open draft) / gate `math.rn-mesoscopic-reduction`: candidate reduction for x = r y on fixed scaled annuli; exact divided-difference determinant ledger still open.

None is a replacement carrier for the historical ABSENT files. None flips
`lemma_closed` away from false. Reading this page does not discharge
`OBL-H5-JETMOD` or `D3-LEMMA-RN-UNIF`. The Math- hard-gate draft is intended to
refuse CONTROLLING promotion while required dependencies stay non-terminal, but
until its own-node eligibility repair lands it must not be cited as a complete
[#90](https://github.com/d6g8k5htny-coder/main/issues/90) enforcer.

## Downstream consequence

If the goal is to reproduce/promote the historical D1/D3 certificate exactly, the
ABSENT carriers and ordered walls remain blockers. If the goal is a new author-side
statement by the 2026-09-24 analytic route, those carrier identities are not
dependencies unless explicitly imported; the named analytic-review interfaces plus
the shrinking-annulus/collision complement stay open. This page does not promote
either route. A future operator may formally retire the
historical route as superseded only through the applicable register/fold mechanism.
This crosswalk does not perform that governance action.

## Related reading

- [RN status (historical route walls)](math_status/STATUS_RN_UNIF.md)
- [Math status packet](math_status/README.md) — OPEN/HOLD; closed directory
- [Open problems §A5](OPEN_PROBLEMS.md) (pinned; do not rewrite for navigation)
- [Research index — RN counting](RESEARCH_INDEX.md)
- Fixed-remote candidate (external file named PROOF.md; not accepted here): [Math- remote_window PROOF](https://github.com/d6g8k5htny-coder/Math-/blob/main/frontiers/remote_window_20260924/PROOF.md)
- Mesoscopic draft: [Math- PR #7](https://github.com/d6g8k5htny-coder/Math-/pull/7)
- In-development D0–D7 hard gate (eligibility repair still open): [Math- PR #8](https://github.com/d6g8k5htny-coder/Math-/pull/8)
- Math- #8 eligibility repair handoff (main cannot push Math-): [MATH_PR8_ELIGIBILITY_HANDOFF_20260925.md](MATH_PR8_ELIGIBILITY_HANDOFF_20260925.md)
- Mesoscopic reduction challenge (not R17): [RN_MESOSCOPIC_REDUCTION_CHALLENGE_20260925.md](RN_MESOSCOPIC_REDUCTION_CHALLENGE_20260925.md)
- Owner surfaces: [#86](https://github.com/d6g8k5htny-coder/main/issues/86), [#90](https://github.com/d6g8k5htny-coder/main/issues/90)
71 changes: 71 additions & 0 deletions docs/MATH_PR8_ELIGIBILITY_HANDOFF_20260925.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,71 @@
# Math- PR #8 own-node eligibility handoff

**Object:** MATH-PR8-OWN-NODE-ELIGIBILITY-20260925-v1
**Against:** Math- head `5887a2f8aaf121760362e6e90f9974f1e0078bc5`
**Author of the patch recipe:** Cursor, from a session that could write `main` and could not push Math-. **This copy** is on `chatgpt/drive-github-hardening-20260919`. It is not a `main` landing. **Scientific effect: NONE.** `OBL-H5-JETMOD` and `D3-LEMMA-RN-UNIF` stay OPEN. No flags are flipped. SIDE24 stays ABSENT. Engineering repair is not discharge.

## Diagnosis

[Math- PR #8](https://github.com/d6g8k5htny-coder/Math-/pull/8) implements the
[#90](https://github.com/d6g8k5htny-coder/main/issues/90) / [#86](https://github.com/d6g8k5htny-coder/main/issues/86)
downstream hard gate, but at head `5887a2f8` dependency terminality alone can
still allow an `AUTHOR_SIDE_CANDIDATE` node to become `controlling`. That fails
the own-node eligibility reading of #90: only a positive reviewed disposition
(`PROVED_REVIEWED`) may become controlling, and terminal dependencies remain
separately necessary.

## Exact fix (ready to apply on Math-)

Portable patch: [`patches/math_pr8_own_node_eligibility.patch`](patches/math_pr8_own_node_eligibility.patch)

SHA-256 `1648237df8525276b17704d12a934cc97c12b1f1184af671ee894b65d60816f1` (8156 bytes).

Applies on top of Math- `5887a2f8`:

1. Add `CONTROLLING_ELIGIBLE = {PROVED_REVIEWED}` and enforce it in
`promotion_allowed()`.
2. Replace fail-open `test_promote_after_all_terminal_deps` with:
- `test_author_side_with_terminal_deps_cannot_become_controlling` (negative)
- `test_proved_reviewed_with_terminal_deps_may_become_controlling` (positive)
3. Add mutant `bypass_own_node_eligibility`; raise `EXPECTED_TESTS` to 28.
4. Refresh `RESULTS.json` / `SOURCE_FILES.json` hashes after the gate change.

Local verification (recorded in
[`patches/math_pr8_eligibility_fix_REPORT.json`](patches/math_pr8_eligibility_fix_REPORT.json)):

```text
28 tests, 8 distinct semantic mutations, modes normal+optimized → passed
```

## Why this is not a Math- push

The session that prepared the patch could write `main` and received **403
pushing to `d6g8k5htny-coder/Math-`** (`Permission denied to cursor[bot]`). The
repair was prepared and locally verified in a Math- worktree; it was not
published on that forge tip. This note and patch are the source-bound
eligibility recipe for an agent or owner with Math- write access. This copy
lives on the hardening branch. It does not land on `main`, does not push to
Math-, and does not merge Math- #8.

## Apply recipe

```sh
git clone --filter=blob:none https://github.com/d6g8k5htny-coder/Math-.git
cd Math-
git checkout 5887a2f8
git apply /path/to/hardening-checkout/docs/patches/math_pr8_own_node_eligibility.patch
python -B -S frontiers/downstream_gate_20260925/hard_gate.py
python -B -S -m unittest discover -s frontiers/downstream_gate_20260925 -p 'test_*.py' -v
python -B -S frontiers/downstream_gate_20260925/run_validation.py --output /tmp/downstream-gate-eligibility
```

Then push to the Math- #8 branch (or open a successor PR) from a credential that
can write Math-.

## Non-claims

- Does not merge Math- #8 or main [#92](https://github.com/d6g8k5htny-coder/main/pull/92).
- Does not flip any scientific Boolean; `lemma_closed` stays false.
- Does not edit Math- PR #7 or award R17 independence credit.
- Does not touch the #91 Drive vault.
- Applying the patch would be an integrity-gate repair on Math- only. It does not discharge `OBL-H5-JETMOD` or `D3-LEMMA-RN-UNIF`, does not flip `lemma_closed`, `prizes_solved`, `discharges_OBL_H5_JETMOD`, `certified_C_H`, or `inventable_attempt_accepted`, and does not invent a source-of-truth carrier. SIDE24 stays ABSENT. Engineering repair is not discharge. The hole on the published Math- tip stays open until a Math- writer applies a repair there.
5 changes: 4 additions & 1 deletion docs/NAVIGATION.json
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,10 @@
"docs/RESEARCH_EXECUTION.md",
"docs/RN_SIDE24.md",
"docs/RN_SIDE24_DENSITY.md",
"docs/RN_SIDE24_CELL.md"
"docs/RN_SIDE24_CELL.md",
"docs/DOWNSTREAM_RN_CROSSWALK_20260925.md",
"docs/RN_MESOSCOPIC_REDUCTION_CHALLENGE_20260925.md",
"docs/MATH_PR8_ELIGIBILITY_HANDOFF_20260925.md"
],
"public_targets": [
{
Expand Down
Loading
Loading