Skip to content

refactor(exact-triangle): clarify reusable finite APIs and placement #1546

Closed

Description

@edemaine

Follow through the useful nonblocking findings in #1441, #1445 and #1455, with concrete native ET consumers under #1532.

Move the reusable IntMatrix digit/product/positive-negative-part API into descriptive Matrix names, and move its scalar power/log bound to a scalar namespace or reuse the upstream fact directly. Give the lopsided matrix/count reconstruction and detection interfaces descriptive names and a directly consumable iff. Move the sqrt-based prime window and its square/filter characterizations to the existing generic PrimeWindow topic under a descriptive name; clean up the partial Hashing prefix, unused source notation and misplaced/repeated scalar sqrt helpers. Reuse upstream APIs, migrate every actual library/test/abstract client, preserve public mathematical meanings and source credit, and remove superseded declarations without aliases. Keep arbitrary source-local letter renaming out of this bounded change.

Root worker codex-three-sum-stack-20261028 will use an ordinary isolated branch from published #1536, with unchanged #1496 cleanup merged as a prerequisite. No ancestor remote restacks. The existing #1532 implementation checkpoint remains preserved; this usability work does not complete #1532 or #1394. All-six/static checks, abstract assessment, nonauthoring review and outside-coordinator review request precede integration. Open-issue searches found no competing IntMatrix/prime-window migration claim.

Activity

  1. self-assigned this
    on Oct 28, 2026
  2. edemaine commented on Oct 28, 2026

    @edemaine
    CollaboratorAuthor

    active — root worker codex-three-sum-stack-20261028; branch work/three-sum-finite-api-cleanup-20261028 from published #1536, with unchanged #1496 cleanup as a prerequisite. I will preserve the separate #1532 WIP and change only these finite APIs and actual clients. No ancestor remote edits; aggregate validation remains held for the priority #1464 repair.

  3. edemaine commented on Oct 28, 2026

    @edemaine
    CollaboratorAuthor

    Root codex-three-sum-stack-20261028 also retains the one-line test-indentation nit from #1531's clean outside review. The exact patch is prepared against e4d2fa0; it will be a separate normal-FF update to the original host branch, validated in a root serial slot before opening another root PR. RegisterFrame stays a small topic module: the reviewer explicitly accepts either placement, and this avoids adding interface fan-out to Structured. No host algorithm/API change or fresh full review is needed. Existing #1544/#1545/#1537 validation priority is unchanged.

  4. edemaine commented on Oct 28, 2026

    @edemaine
    CollaboratorAuthor

    Active — root codex-three-sum-stack-20261028: finite API migrations and all clients are source-ready on work/three-sum-finite-api-cleanup-20261028. Strict targeted Lopsided/Hashing/Splitting tests, full citation/simp hygiene, draft-name audit and diff check pass. Full validation/publication remains queued behind #1545 and the #1537 repair; no root build is running. The separate one-line #1531 indentation patch remains prepared for the next root validation slot. Native ET #1532 source is preserved; this does not complete #1394.

  5. edemaine commented on Oct 28, 2026

    @edemaine
    CollaboratorAuthor

    Awaiting review/integration — root codex-three-sum-stack-20261028 published #1556, directly on the current #1496 head. The PR owns final exact-tree six-gate/static evidence and nonauthoring review; the commit tree matches the reviewed frozen source. The root audit slot is explicitly released directly to #1509. Full native #1394 and ET #1532 remain open.

  6. edemaine commented on Oct 28, 2026

    @edemaine
    CollaboratorAuthor

    Active — root codex-three-sum-stack-20261028 is addressing the three optional findings in #1556's outside FINAL: consolidate repeated square-root cast bounds, use the hypothesis-shaped false-positive lemma name, and replace undefined notation in Hashing comments. Normal fast-forward repair on the existing branch; source/MCP now, serial validation queued after #1509. Existing full coverage is retained; the new shared helper and final delta will receive focused outside review.

  7. edemaine commented on Oct 28, 2026

    @edemaine
    CollaboratorAuthor

    Awaiting review/integration — root codex-three-sum-stack-20261028 published the bounded #1556 repair, addressing all three outside nits. The PR owns exact-head six-gate/static evidence and the focused outside delta request. The root audit slot is released directly to #1526; no root build remains.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Labels

No labels
No labels

Type

No type

Projects

No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions