Skip to content

Parity ledger: CE0, OP0.1, the registry, U2's evidence record and F0/F1 status - #587

Merged
MaxGhenis merged 4 commits into
masterfrom
ledger-delta-20261025
Oct 25, 2026
Merged

MaxGhenis merged 4 commits into
masterfrom
ledger-delta-20261025

Conversation

@MaxGhenis

@MaxGhenis MaxGhenis commented Oct 25, 2026 •

Copy link
Copy Markdown
Contributor

This updates docs/parity-ledger.md for what merged after its last check at f8ffc7d9 (#569). It re-verifies five rows and step 1 of the critical path at 9967dfd6 (master after #577), and step 2's successor entry at 9a344a1a (master after #561). It changes no code, gate, run or progress data, so docs/progress.md and the scorecard are untouched.

Verified at 9967dfd6

  • Covered earnings. CE v2's design note, CE0, merged in #568 (docs/design/covered_earnings_v2.md, which says nothing in it has run). That replaces the clause calling the retired correction design a draft. CE1's coverage-rule registry is in review (#583).
  • Births. OP0.1's source, exposure, access and power audit merged (#567). OP1's harness is still in review (#566).
  • Alignment. The open-population design's revision 14 (#576), §§3 and 6.4: OP6 will align births and emigration and leave death draws to the mortality adapter the mortality lane chooses, the baseline's or SM once admitted. SM is designed to preserve the baseline's expected deaths within each age, sex and year cell, up to a logged clipping residual (mortality design §6).
  • Poverty (U2). #577 committed the pre-registration evidence record the specification pins (docs/design/u2_preregistration_evidence.json). Step 7's fresh forecast was frozen and independently checked in the lane's evidence folder, outside the repository, so it can't be checked here yet. The next step is the Candidate 2 design: latent-permanent conditioned chained QRF (research memo) #42 registration, which is Max's, and then the one-shot.
  • National costs. F0, Registration 18's build, is in review as #585. F1, the income tax on benefits, is in review as #586, after #578 corrected F1's signed-split invariants in the mortality design (§6).
  • Critical path, step 1. The held-out data registry merged in #560. B0.2's plan is now #584, which replaced Track B B0.2: source, exposure, access and power records for B3 through J #554.
  • P2 (at 9a344a1a). P2, the successor's reserve, merged in #561. The forward-earnings row and critical-path step 2 say so; neither S1 nor P2 has registered.
  • Header. It records the third verification.

Seven link definitions are added (#576, #577, #578, #583, #584, #585, #586). Every reference resolves, and no definition is unused.

Not covered here

Merges after 9a344a1a (for example D1a, SSI0 and SM1, once they land) get their own ledger update. The branch merges master 9a344a1a.

Review

Review r1 (20261025-023327-ledger-587-review-r1) asked for two medium fixes, both applied in a7564e4a: the poverty row now says where step 7's record lives, and the alignment row names the death adapter and the cell-level conservation. Review r2 (20261025-041855-ledger-587-review-r2) found round 1's alignment fix good and asked for two more medium fixes, applied in the next commit: the poverty row no longer describes the unposted registration's content, and the forward-earnings row no longer calls P2 a draft. A delta review at the new head follows.

🤖 Generated with Claude Code

…F1 status

Re-verifies five rows and step 1 of the critical path at 9967dfd, after
#560, #568, #576, #577 and #578 merged and F0 and F1 opened as #585 and
#586:

- covered earnings: CE v2's design note merged in #568; CE1 is in review
  (#583);
- births: OP0.1's audit merged (#567); OP1 is in review (#566);
- alignment: OP6 aligns births and emigration, and deaths come from the
  mortality lane's SM (open-population design §6.4, revision 14, #576);
- poverty: #577 committed U2's pre-registration evidence record, and a
  fresh forecast was frozen and independently checked (step 7); posting
  the #42 registration is next;
- national costs: F0 is in review as #585 and F1 as #586, after #578's
  design erratum;
- critical path step 1: the holdout registry merged in #560, and B0.2's
  plan moved to #584.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@vercel

vercel Bot commented Oct 25, 2026 •

Copy link
Copy Markdown

The latest updates on your projects. Learn more about Vercel for GitHub.

Project Deployment Actions Updated
social-security-model Ready Ready Preview Oct 25, 2026 8:38am UTC

Request Review

MaxGhenis and others added 2 commits October 25, 2026 04:17
…dapter; P2 merged

Review r1 (20261025-023327-ledger-587-review-r1) asked for two medium fixes:

- Poverty: say that step 7's forecast was frozen and checked in the lane's
  evidence folder outside the repository, which the registration cites by
  SHA-256 with its headline and interval, rather than imply a repository
  record.
- Alignment: OP6 leaves death draws to the adapter the mortality lane
  chooses (the baseline's, or SM once admitted); SM is designed to preserve
  expected deaths within each age, sex and year cell up to a logged
  clipping residual (open-population design 6.4; mortality design 6).

Also merges master 9a344a1 and records that P2 merged in #561
(critical path, step 2), re-verified at 9a344a1.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…erged in forward earnings

Review r2 (20261025-041855-ledger-587-review-r2) asked for two medium fixes:

- Poverty: drop the clause on what the registration will cite; the
  repository can't support a claim about a comment not yet posted.
- Forward earnings: P2 (#561) is merged, not a draft in review; registering
  either S1 or P2 as attempt 2 comes to the maintainer separately.

The header now names both P2 entries re-verified at 9a344a1.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@MaxGhenis
MaxGhenis marked this pull request as ready for review October 25, 2026 08:53
@MaxGhenis
MaxGhenis merged commit 7e050df into master Oct 25, 2026
12 checks passed
@MaxGhenis

Copy link
Copy Markdown
Contributor Author

Merged at 6f8946630d0f4685dfbe9e532f5ada19be3c8cd3 with --merge --match-head-commit.

  • Checks: gh pr checks exited 0 on the latest run for this head (Tests run 40177370894: four pytest shards, pytest (3.14), graph-integration 3.13 and 3.14; lint and build; Vercel; deploy skipped).
  • State: MERGEABLE and CLEAN, not a draft, no CHANGES_REQUESTED review.
  • Independent reviews (GPT-6.1 Sol, read-only):
    • round 1 at 4dcc0992: two medium wording fixes, applied in a7564e4a;
    • round 2 at a7564e4a: approved the master merge cd0dbe46 (tree equal to git merge-tree --write-tree) and the round-1 fixes, and asked for two medium fixes, applied in 6f894663;
    • round 3 at 6f894663: APPROVE. Both round-2 findings fixed, nothing else changed, all 232 reference uses resolve to 82 definitions. Its one minor finding was a stale sentence in this PR's body, now corrected.
  • Scope: docs/parity-ledger.md only, plus the merge of master 9a344a1a. No tier change.

Merges after 9a344a1a get their own ledger update.

@MaxGhenis
MaxGhenis deleted the ledger-delta-20261025 branch October 25, 2026 09:42
MaxGhenis added a commit that referenced this pull request Oct 25, 2026
Master brings #587 (ledger delta), #575 (D1a) and #570 (SM1). None of
them changes the registry module, its tests or its mutation specs, and
the registry module imports only the standard library and yaml.

The tier conflict resolves to master's counts plus this branch's nine
unit tests: unit 18,549 + 9 = 18,558, total 25,128. The README tier
table now matches the tier file.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

This branch was successfully deployed

1 active deployment
Preview — 6f894663 Deployed Oct 25, 2026 by vercel[bot]
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