Skip to content

Characterize live-row and physical-slot reserve compatibility - #2420

Merged
Mister-Meeseeks merged 1 commit into
CrocSwap:mainfrom
RossRichardson:research/sprint-20261015-0824
Oct 16, 2026
Merged

Mister-Meeseeks merged 1 commit into
CrocSwap:mainfrom
RossRichardson:research/sprint-20261015-0824

Conversation

@RossRichardson

Copy link
Copy Markdown

PR2195 already separates arbitrary live input from private work and calls its lcm-based layout an overreserve. This package makes the resulting interface boundary explicit: under the complete live-packet contract, B = WNH supplies the depth reserve B^D, but an unchanged physical-slot format accepts t B^D exactly when (P/gcd(P,W))^D divides t. For W=28416, P=28425, that gate is 9475^D | t. The historical K_lcm^D remains the minimum positive reserve compatible with both fixed formats.

The contribution is a credited conditional corollary, a complete-row codec, a written all-depth proof, an independent mathematical and symbolic-payload review, and defect-sensitive replay. It preserves all arbitrary input rows, high fibers, private work and parking charges. It does not implement a full selected caller or paid format adapter, change the selected invoice, or establish a new kappa. Minimality is only within the stated complete Cartesian formats.

Sources are pinned to Jose Manuel's PR2195 at 3a1cf78a1ba6322129e42d80c50d21765524d612 and PR2234 at d047ef86fbc5b1614d3d8e2b8d3441413beab73f; six source files are preserved with SHA256/Git-blob checks, original notices and attribution. Prepared for Ross Richardson with substantial OpenAI Codex assistance, including a separate recursive-interface specialist. Independent agent review is not human or formal-kernel review.

Validation at this head:

  • Python 3.12 registered package replay passes: 54,840 primary codec rows, 73,710 exact compatibility equivalences, 12 semantic/domain controls, and nine public-CLI test groups including normal/optimized execution, read-only Git-free replay, source/receipt corruption and actual restoration-guard removal.
  • Independent replay passes normally and optimized: 148,929 complete rows, 24,192 distinct symbolic payloads, 18,000 compatibility equivalences, and 720 direct minimum-common-reserve searches; its finite domains differ from the primary checker.
  • make selected-record-check passes. The selected construction and its records are unchanged; its complete cold replay is outside this package's scope. Hosted CI is not claimed.

Canonical investigation: Discussion2179 claim. The next step is to identify an actual downstream consumer and determine whether it needs the physical-slot format before substituting the smaller divisor. TASK.json records the scoped departure from the stock-envelope card and the overlap limitations.

@Mister-Meeseeks
Mister-Meeseeks merged commit 81ff1f5 into CrocSwap:main Oct 16, 2026
55 of 61 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants