Assessment 244: late contraction receipts and current routing

13 October 2026. Scheduled completion 149. Argument and failed extension; independent checker.

Outcome and scope

The attempted unrestricted cross-scale extension did not produce a new native theorem. Construction 85 remains the strongest native construction. The investigation separates two requirements that a semantic owner-to-target map does not supply: enough service after a receipt becomes available, and access to the selected current destination.

A global committed trie plus a bounded pending tuple works under an explicit successive-service gate. The next actual same-source owner must finish the previous batch before replacement. Common owner width or nondecreasing per-source owner lengths, with no target longer than its owner, are sufficient examples. This is a direct adaptation of the established pending-overlay mechanism, not a new numbered construction or a necessity criterion.

Without the gate, a backwards-chosen descending family leaves more pending targets than any fixed overlay capacity. Every target may be shorter than its owner by any prescribed factor. The failure assumes literal insertion first released at the exact owner endpoint, no credited earlier speculation, and at least one job per target bit. It does not bound compressed aggregates, prepared alternate routes or unrestricted native recognition. Short queries may reject long pending words without traversing their whole spelling.

Draining all smaller buckets has a geometric linear service bill. A common-prefix family nevertheless leaves the existing serial directory window a growing distance from its late selected target class. Frozen template handles do not automatically refer to a current per-source version. A new paid selector, alternate current index or shared aggregate is still required.

Independent review and finite checks

Independent adversarial review compared section 15, construction 85 and the proposed service extension. The review supplied the global-overlay control, descending-contraction witness and late class-window test, then checked the written proof and authored the standalone finite checker independently of the exposition.

One material reconciliation concerned the deadline allowance H. It must cover the releasing record’s own remaining punctuation as well as any extra final fixed batch. The argument now chooses a uniform generous H before choosing the witness lengths. Later records contribute their full actual symbol counts. The final boundary is before the pins; no unbounded later query budget is counted.

Fresh checker execution passed 270 symbolic backwards witnesses, 109 symbolic late serial-class-route cases, 900 finite service attempts (41 accepted and 859 rejected), 2,697 retained-snapshot checks, 5,424 membership checks over an image of 158 target words, and a two-version stale-static-handle control. The displayed example uses H=17; abstract algebra cases with H=0 are not claims that ordinary punctuation costs zero. Semantic tests use sets and strings, plus optional rollback controls. They neither implement a native graph machine nor verify the whole input syntax.

Contribution, controls and source custody

Section 15.7 already identifies the late distant-bucket gap and gives an expansion-side insertion obstruction. This investigation does not present that old gap as a new theorem. It tests the tempting all-shorter-class repair, supplies a contraction-only descending failure for fixed late-receipt capacity, and makes the successive-service positive control explicit. These are supporting arguments in a sustained unsuccessful extension. Mathematical and native result counts are unchanged.

Construction 85 and all prior shared-index, replay, affine, causal and unary controls remain intact. Imported entry 129 bounds retain their matching/positive-path hypotheses, growing alphabets and state/port scope. Root CLAIMS.yaml is unchanged. No primary-proof review is counted here; the next scheduled assessment must perform both the two-paper minimum primary review and editorial work.

GitLab maintains the argument, checker and this review. Generated HTML/KaTeX readings and highlighted code with exact downloads project those same bytes to the existing Site. No primary PDFs or scratch are published. The homepage research entry and publication ranking remain unchanged because this is a no-advance investigation. Scoped checks and exact publication receipts are recorded in integration state.

Next decisive test

Construct a paid current-version selector for late contraction receipts, or a shared aggregate that bypasses bucket insertion. Ordinary preparation and activation must make the selected shorter answer accessible by its deadline, without supplied routes, stale handles or final traversal. Test both service-admissible sequences and descending contractions. Preserve arbitrary whole-suffix permutations, arbitrary-name routing, cyclic sharing and heterogeneous service as open.

Scheduled 148 to 149 once; mathematical 148, native 85, window 2 and last reassessment 146 unchanged. Next 150 combines primary review and editorial duties while preserving this mathematical target. Last primary 240 and editorial 239 remain recorded. Installed reverse-writer retirement remains separate and unverified.