[JoyboyBrian]

Change and mathematical scope

Independently verifies Agastya Sondhi / @TMM2009’s PR727 at 71ba94c981477b3527ef21dd0f8140e795a7664d, using a third footprint algorithm: sorted interval endpoints. Status: independent verification and scoped tooling evidence; no new optimization or κ claim.

owner intervals → sort endpoints → reject overlap → sum lengths → check prefix/idle width

The new oracle uses neither bit masks nor per-coordinate occupancy sets. It reconstructs all 16,128 owner bins, 280,592 occupied coordinates and the complete idle histogram for 35,424 helper roles. Original sets, PR727 masks and the endpoint oracle agree. The entire inherited checker receipt and relevant internal states are compared, including exact moment fractions.

This adds a different verification method, invalid interval enumeration, stateful sequence checks and same-input replay evidence. PR727 already supplied its own set/mask tests; this package credits that work and vendors four exact source/provenance files with their original notice. Only research/owner-footprint-crosscheck/ is added, so it can be reviewed independently of PR727’s package.

Exact base: 66849e1475c930fd7732b1a0067ffee7e1e95f60. The selected #626/#543 result remains 51110230636731/31250000000000000. SOURCES.json pins the 17 original inputs and four PR727 files with SHA-256 and Git blob identities. Native artifacts are archived PR622 outputs, not a fresh reconstruction.

Validation

python3 -I -B research/owner-footprint-crosscheck/verify.py

Expected: PASS_OWNER_FOOTPRINT_CROSSCHECK ... sweep_owners=16128.

• Complete-map replay; 21,500 exhaustive cases, 6,000 sequential operations, nine full-map corruptions, five actual-code mutations and four source-pin controls. Both complete engines reject with matching diagnostics; the independent oracle rejects all four footprint corruptions.
• Normal and optimized Python 3.12.12/3.14.7 checks pass. A minimal export replays from a path with spaces and an unrelated working directory. The four vendored files match PR727’s exact remote Git blobs.
• Fifteen alternating fresh-process pairs per interpreter: full-check median wall time 0.303022 → 0.274185 s (9.52% less) on Python 3.12 and 0.297138 → 0.272085 s (8.43% less) on Python 3.14. All 60 timed receipts match byte for byte. Both modes include the same source-admission harness, unlike the author’s direct-CLI reference setup; no claim to reproduce their absolute timings or percentages. Raw samples are retained. No RSS reduction or full-frontier speedup is claimed.
• Selected-record binding, all five existing selected-record tests, seven historical source inventories and staged whitespace checks pass. A final inventory check caught one agent-generated .pyc; only that cache was removed, with no reviewed source changes. Details are in VALIDATION.md.

No full repository suite, cold supplier replay, full controller/47-inequality assembly, Lean or remote CI pass is claimed. The oracle checks footprint accounting, not physical legality of its input windows. All inherited compiler/common-label/mutable-basis/hosting/residual/precision/tape/analytic assumptions remain.

Seed 89, verification engineering. The final overlap refresh found PR727, so the overlapping optimization was retired and this independent audit was completed instead. No pinned coordination issue existed; the local claim contains the result and source link. Refresh through #767 found no endpoint-sweep audit of PR727; its pinned head and base main were unchanged.

Credit Agastya Sondhi/TMM2009 for the mask optimization; Chafik Boukhalfa/chafreaky for the independent checker; Jeremy Kothe/jeremykothe2, DreamingOfClouds, Rohan Arun, icekylinx, eumemic and the retained construction lineage. Prepared by Brian / JoyboyBrian with substantial OpenAI Codex assistance, Apache-2.0. “Independent” describes the endpoint algorithm and replay, not independent human review.

Merge Danger

Door: two-way. Blast Radius: isolated research package. No existing source, default command, selected-frontier pointer or upstream/ file changes.

[Mister-Meeseeks]

100-PR batch review at 9f68dff1e1cccad2bd5b3b51f4ced0bba9765862.

Merged to main, preserving the original contributor commits and package attribution.

Reviewed scope: Independent endpoint sweep reproduces all owner footprint masks and sizes; finite allocation audit, not construction/analytic recertification.

The pinned scoped command passed; the batch ledger contains its command, input hashes, environment and log. This incorporates the research/tooling contribution and does not change the selected κ or discharge the inherited multiplication-theorem assumptions.

Batch review, attribution and evidence. Thank you for the contribution.