Skip to content

Prove a two-delivery price floor for the selected BIT word - #7739

Open
petalas wants to merge 2 commits into
CrocSwap:mainfrom
petalas:research/w3-1118c-bit-multi-delivery
Open

petalas wants to merge 2 commits into
CrocSwap:mainfrom
petalas:research/w3-1118c-bit-multi-delivery

Conversation

@petalas

@petalas petalas commented Nov 18, 2026 •

Copy link
Copy Markdown
Contributor

Summary

Every original paid arrival in the selected PR7001 BIT word retains its old minimum price when replaced by at most two direct nested deliveries from strictly earlier chart libraries. The scope grants free retention and destination XOR, keeps all independent inputs and old deadlines, and covers both orientations for the tariff s^t, 0 < t <= 1. This is a scoped proof backed by exhaustive finite replay. No new κ follows. Advance claim

A strictly improving bundle must have total rank below the original rank. The sum of all individually lower-rank donor libraries and the free destination library excludes 5,733 of 5,742 arrivals, regardless of bundle size. The remaining nine are forward rank-seven arrivals into chart 0. Quotienting by the free destination library excludes all 4,626 eligible singletons and 478,764 rank-affordable pairs in those nine cases.

Evidence

The replay authenticates the PR7001 native word, derives rational nesting and reconstructs the original causal rows before testing each arrival. A joint-span control detects a row attainable by two donors together when neither can supply it individually. A strict-budget control excludes equal-rank-sum pairs from the cheaper domain. Both direct modes passed locally in 1.32 seconds each. The registered replay and committed-byte/source-history preflight also passed.

From the repository root, with CPython 3.12.15 and its standard library, run the following commands.

python3.12 -B -W error -X dev research/bit-causal-multi-delivery/verify.py
python3.12 -B -W error -X dev -O research/bit-causal-multi-delivery/verify.py
python3.12 -B scripts/research_package.py --package research/bit-causal-multi-delivery
python3.12 -B scripts/check_research_commit.py --package research/bit-causal-multi-delivery --check-source-history

The direct checks report PASS_BIT_CAUSAL_MULTI_DELIVERY arrivals=5742 union_excluded=5733 exceptional_singles=4626 exceptional_pairs=478764.

Scope and limitations

The proof fixes the original libraries, required rows, deadlines, charts and productive targets. The higher-arity search at the nine exceptions remains open. Intermediate relays, adaptive library changes, helper restrictions, and physical compiler delivery, retention and restoration are outside the model. The selected cold replay, predecessor scientific suites, Lean, Docker, and independent human or formal proof-kernel review were not run.

The contribution was produced by Nick Petalas. Credit aliromman for the fixed-library model and the multiple-delivery question in PR #7723, CubeJerry for PR #7001, zxv and the retained authors for the native word, DreamingOfClouds for the chain-root caller, and ToikY for the related causal stock bounds. Pinned predecessor proof

Open next questions

  • Test three or more donors at the nine exceptions. Start at forward event 26861 with a donor subset of total rank at most six.
  • Test a changed causal library from one intermediate relay. Specify its entry, delivery and restoration costs before pricing a complete caller.

Related open work

CI notes

This PR carries the shared CI-doc and replay-gap sync used by other open PRs.

The base fails test_current_ci_section_matches_selected_record because the guide names make frontier6030-verify while the selected record requires make frontier7000-verify. The shared sync updates the guide and runs the triggered historical replay against its pinned snapshot.

The six navigation tests and the registered historical replay passed locally on CPython 3.12.15. Hosted CI passed all three checks on 7e44f44b8078b3788256772533890fbf85503a0c. The plan, BIT multi-delivery replay and historical replay-gap replay all passed. Run 44394177520

Merge danger

Two-way. The patch adds a standalone proof and replay with one pinned input.

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.

1 participant