Repository navigation
Combine the grammar and lexicon batches onto release/v0.4 (issue #269) - #532
humuhumu33 wants to merge 17 commits into
Conversation
…(BS-50..52, #269) One batch branch carrying three independent slices of the lexicon closure, as BS-25..34 landed in #480: - BS-50 LexiconSchema: the category/frame/channel/associativity/feature tables and surface_safety/is_lean_name/toml_comment_at/entry_path, over the atoms of BS-03. Six roots. - BS-51 LseScopes: Lse::check_scopes, referenced_consts, and outer_explicit_binders/result, over the expressions of BS-25. Three roots. - BS-52 LseTypes: the conservative checker, check_signature and check_value, structural over a host-given nesting bound. Two roots. Register rows (model/ids.toml, SPEC §31, VERIFICATION.md, features scenarios) are inserted as one whole block per slice, in register-ID order, after BS-34. Counts and totals that the regeneration itself writes (README.md, CHANGELOG.md, the §31 total line, repository.rs's and spec_links.rs's row counts) are left at release's current values for the regeneration to fix. .github/workflows/regenerate.yml is release's version overwritten with the fixed one from ci/regenerate-relocks (the deliver job and a relock step that survives combine), so this branch's own regeneration lands itself onto the PR opened for this batch branch. #522, #523 and #524 stay open as the per-slice record; they are not what lands. Branched from release/v0.4 at 6a02d2c. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
SPEC.md's "Total required capability IDs", xtask/src/spec_links.rs's RP-07 check, and repository.rs's RP-07 assertion still said 496: none of them is written by a recipe (cargo xtask counts --write carries the counts a bootstrap case reports, not the §31 row count), so the finish job would not have fixed them and RP-07 would fail after the whole regeneration. All three now read 499, matching model/ids.toml's 499 rows and SPEC §31's 499 rows. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Carries all eight completed grammar/surface slices (GrammarProof, SemanticPrint, SemanticRead, CoreJson, CorePrint, CoreRead, GrammarMath, PropositionTerms) as one batch, as BS-25..34 landed in #480. Hand-authored register content (SPEC.md Sec31 rows + Sec17.23 prose, model/ids.toml entries, features/suites/bootstrap.feature scenarios, VERIFICATION.md, bootstrap/README.md, bootstrap.rs dispatch wiring, bootstrap/lexlean.toml entrypoints, the two xtask/repository.rs register-count constants, and .github/bdd-parts.json, vv-rows.json, intel-host-lanes.json CI-plan entries) is unioned in register-ID order onto release's current content. Regeneration-written files (CONFORMANCE.md, ERRORS.md, roots.json, lexlean.lock) and README.md/CHANGELOG.md are left untouched -- release's side -- for the one regeneration run to fill in. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Merges the grammar batch (cd571f1: BS-35..40, 42, 43) and the lexicon batch (ebe8911: BS-50..52), both authored directly on release 6a02d2c, into one combined batch so neither lands alone and makes the other stale. Conflicts at the shared BS-34 insertion anchor are resolved as "keep both, grammar then lexicon" for the register-order files: SPEC.md §31 rows and §17.23 prose, model/ids.toml, features/suites/bootstrap.feature, bootstrap/README.md, crates/conformance/tests/conformance.rs, and crates/conformance/src/cases/bootstrap.rs. Three of those unions dropped a syntactic separator at the merge seam (ids.toml a missing [[id]] header, conformance.rs a missing closing brace and #[test] attribute, SPEC.md an extra blank line) -- fixed by hand and reverified: 507 unique ids, 507 §31 rows, 42 feature scenarios and 42 conformance fns, no stray conflict markers. Hand-merged rather than unioned: bootstrap/lexlean.toml entrypoints (41+3=44), the register-count constant repeated in both xtask/src/spec_links.rs and crates/conformance/src/cases/repository.rs (bumped 504/499 -> 507 at every site), .github/bdd-parts.json and vv-rows.json (all eleven new parts and rows), and .github/intel-host-lanes.json (grammar keeps its chosen lanes; the three lexicon parts land in "lower", matching where #525 put them). bootstrap/lexlean.lock keeps release's side (the lexicon batch dropped it; the runner relocks regardless). .github/workflows/regenerate.yml stays byte-identical to ci/regenerate-relocks. Supersedes #525 and #527, which stay open as the per-batch record. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
afflom
left a comment
There was a problem hiding this comment.
Combined-source review credits the 507-entry register, 44 entrypoints, and all eleven scenario/test/routing bindings. Reusing the constituent reviews avoids another module audit or producer.
P1 remains: tests/production-export/bootstrap/roots.json is exactly the 73-row lexicon-batch input (a53b31b8). All eleven lexicon roots remain, but all fifteen grammar roots identified in #527 review5899391887 are still absent. Entrypoints do not add production-export request roots. Add those authored rows in this combined branch and disposition the resulting official outputs before acceptance.
The three BS-35/36/37 coverage findings also persist: their case blobs exactly match the reviewed grammar batch, including file/size/range/modulo sampling, while the register/scenarios still promise every source/module/declaration. Resolve that claim-to-coverage mismatch without silently weakening accepted scope.
#532 is now the owner's combined delivery carrier; no competing patch or separate batch regeneration is requested. Preserve existing outputs with their original bindings, qualify the corrected final head, and retain #269's broader scope. The body should distinguish authored batches from landed release work. Receipt: pr532-combined-source-review/receipt.json. No tests, writers or CI controls were repeated.
All eight new bootstrap-{ir-term,ir-proof,ir-document,semantic-types,
semantic-terms,semantic-decls,semantic-models,semantic-reasoning} parts
had landed in print (already the heaviest lane at 20), leaving lower 12,
print 28, validate 11, moves 6. Moves five of the eight lighter/IR-layer
parts (ir-term, ir-proof, ir-document, semantic-terms, semantic-decls) to
moves, two (semantic-reasoning, semantic-types) to validate, and one
(semantic-models) to lower, spreading what is believed the three heaviest
cases (semantic-reasoning, semantic-models, semantic-terms) into three
different lanes rather than stacking them. New totals: lower 13, print 20,
validate 13, moves 11. Every part release or #532 placed is left alone.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Correcting my own earlier instruction: these two files are not regeneration output -- nothing writes them, every slice PR edits both by hand, and the gate's RP-07/RP-11 checks read them. README.md's count sentence and the bootstrap capability row's BS-01..BS-34 claim, and CHANGELOG.md's Unreleased bullets, were all left on release's side in PR #532 and would fail the gate once its regeneration delivers. README.md: the count sentence becomes "All 507 registered conformance IDs ..."; the bootstrap row's range becomes BS-01..BS-52; and the row's prose gains what each of the eleven slice PRs (505, 509-515, 522-524) added to it, in register order, taken verbatim from each PR's own patch against its merge base. CHANGELOG.md: every bullet each of the eleven PRs added under ## Unreleased, in register order, verbatim (including each independent slice's own max_child_output_bytes bullet, since each PR added it as its own change at authoring time). Built as a separate branch on top of batch/bs35-52's current head rather than on it, since #532's regeneration (an hour in) must not be disturbed and neither file is an input to anything it generates. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
README.md and CHANGELOG.md are not regeneration-written, so they needed the BS-35..52 prose and changelog bullets that #532 (batch/bs35-52) was missing. Keeps 515 registered conformance IDs and the bootstrap row's range at BS-01..BS-72, with both sides' clauses in the row's prose (the fix branch's eleven BS-35..52 clauses, then this stack's own, in register order) and both sides' changelog bullets under Unreleased. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
The write recipes of SPEC.md §9.2, run by regenerate.yml (https://github.com/UOR-Foundation/lexlean/actions/runs/43552216104), rewrote 1721 paths. The counts writer reported: BS-05 COMMITTED_CRATES: 306 -> 317 (crates/conformance/src/cases/bootstrap/rust_validate.rs) BS-05 COMMITTED_MANIFESTS: 330 -> 341 (crates/conformance/src/cases/bootstrap/rust_validate.rs) BS-05 COMMITTED_PROGRAMS: 247 -> 258 (crates/conformance/src/cases/bootstrap/rust_validate.rs) BS-05 EVALUATOR_CORRESPOND: 322 -> 328 (crates/conformance/src/cases/bootstrap/rust_validate.rs) BS-05 EVALUATOR_CRATES: 221 -> 227 (crates/conformance/src/cases/bootstrap/rust_validate.rs) BS-05 EVALUATOR_VALIDATE: 457 -> 463 (crates/conformance/src/cases/bootstrap/rust_validate.rs) BS-05 VECTORS_CORRESPOND: 407 -> 418 (crates/conformance/src/cases/bootstrap/rust_validate.rs) BS-05 VECTORS_VALIDATE: 542 -> 553 (crates/conformance/src/cases/bootstrap/rust_validate.rs) BS-06 COMMITTED_CRATES: 306 -> 317 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-06 COMMITTED_DISTINCT: 612 -> 634 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-06 COMMITTED_MANIFESTS: 330 -> 341 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-06 COMMITTED_PACKAGES: 330 -> 341 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-06 COMMITTED_PROGRAMS: 247 -> 258 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-06 EVALUATOR_CRATE: 612 -> 629 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-06 REFUSED_CRATE: 71 -> 73 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-06 VECTORS_CRATE: 785 -> 809 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-06 VECTORS_EXPR: 2265 -> 2326 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-07 VECTORS: 381 -> 407 (crates/conformance/src/cases/bootstrap/rust_lower.rs) BS-07 VECTORS_WITHIN: 296 -> 303 (crates/conformance/src/cases/bootstrap/rust_lower.rs) BS-08 CRATES: 440 -> 451 (crates/conformance/src/cases/bootstrap/rust_moves.rs) BS-08 WITHIN: 355 -> 361 (crates/conformance/src/cases/bootstrap/rust_moves.rs) BS-11 EVALUATOR_MUTANTS: 283 -> 278 (crates/conformance/src/cases/bootstrap/rust_bodies.rs) BS-11 VECTORS: 157 -> 158 (crates/conformance/src/cases/bootstrap/rust_bodies.rs) BS-11 VECTORS_WITHIN: 156 -> 157 (crates/conformance/src/cases/bootstrap/rust_bodies.rs) BS-12 EVALUATOR_MUTANTS: 152 -> 157 (crates/conformance/src/cases/bootstrap/rust_succeed.rs) BS-13 EVALUATOR_MUTANTS: 143 -> 157 (crates/conformance/src/cases/bootstrap/rust_loops.rs) BS-13 VECTORS: 150 -> 175 (crates/conformance/src/cases/bootstrap/rust_loops.rs) BS-13 VECTORS_WITHIN: 67 -> 73 (crates/conformance/src/cases/bootstrap/rust_loops.rs) BS-14 EVALUATOR_MUTANTS: 548 -> 573 (crates/conformance/src/cases/bootstrap/rust_assemble.rs) BS-14 VECTORS: 385 -> 411 (crates/conformance/src/cases/bootstrap/rust_assemble.rs) BS-14 VECTORS_WITHIN: 300 -> 307 (crates/conformance/src/cases/bootstrap/rust_assemble.rs) BS-15 ACCEPTED: 330 -> 341 (crates/conformance/src/cases/bootstrap/rust_package.rs) BS-15 MANIFESTS: 349 -> 360 (crates/conformance/src/cases/bootstrap/rust_package.rs) BS-15 RENDER_MUTANTS: 330 -> 341 (crates/conformance/src/cases/bootstrap/rust_package.rs) BS-05: §31 row: 221 -> 227 (x1), 247 -> 258 (x1), 306 -> 317 (x1), 330 -> 341 (x1), 407 -> 418 (x1), 542 -> 553 (x1) BS-05: the §31 row does not state 322 (-> 328) BS-05: the §31 row does not state 457 (-> 463) BS-05: register: 221 -> 227 (x1), 247 -> 258 (x1), 306 -> 317 (x1), 330 -> 341 (x1), 407 -> 418 (x1), 542 -> 553 (x1) BS-05: scenario: 221 -> 227 (x1), 247 -> 258 (x1), 306 -> 317 (x1), 330 -> 341 (x1), 407 -> 418 (x1), 542 -> 553 (x1) BS-06: 3050 is VECTORS_CRATE + VECTORS_EXPR: -> 3135 BS-06: §31 row: 71 -> 73 (x1), 247 -> 258 (x1), 330 -> 341 (x3), 785 -> 809 (x1), 2265 -> 2326 (x2), 3050 -> 3135 (x1) BS-06: the §31 row does not state 306 (-> 317) BS-06: register: 71 -> 73 (x1), 247 -> 258 (x1), 330 -> 341 (x3), 785 -> 809 (x1), 2265 -> 2326 (x2), 3050 -> 3135 (x1) BS-06: scenario: 71 -> 73 (x1), 247 -> 258 (x1), 330 -> 341 (x3), 785 -> 809 (x1), 2265 -> 2326 (x2), 3050 -> 3135 (x1) BS-07: §31 row: 296 -> 303 (x1), 381 -> 407 (x1) BS-07: register: 296 -> 303 (x1), 381 -> 407 (x1) BS-07: scenario: 296 -> 303 (x1), 381 -> 407 (x1) BS-08: §31 row: 355 -> 361 (x1), 440 -> 451 (x1) BS-08: register: 355 -> 361 (x1), 440 -> 451 (x1) BS-08: scenario: 355 -> 361 (x1), 440 -> 451 (x1) BS-11: §31 row: 156 -> 157 (x1), 157 -> 158 (x1), 283 -> 278 (x1) BS-11: register: 156 -> 157 (x1), 157 -> 158 (x1), 283 -> 278 (x1) BS-11: scenario: 156 -> 157 (x1), 157 -> 158 (x1), 283 -> 278 (x1) BS-12: §31 row: 152 -> 157 (x1) BS-12: register: 152 -> 157 (x1) BS-12: scenario: 152 -> 157 (x1) BS-13: §31 row: 143 -> 157 (x1), 150 -> 175 (x1) BS-13: the §31 row does not state 67 (-> 73) BS-13: register: 143 -> 157 (x1), 150 -> 175 (x1) BS-13: scenario: 143 -> 157 (x1), 150 -> 175 (x1) BS-14: §31 row: 385 -> 411 (x1), 548 -> 573 (x1) BS-14: register: 385 -> 411 (x1), 548 -> 573 (x1) BS-14: scenario: 385 -> 411 (x1), 548 -> 573 (x1) BS-15: §31 row: 330 -> 341 (x2), 349 -> 360 (x1) BS-15: register: 330 -> 341 (x2), 349 -> 360 (x1) BS-15: scenario: 330 -> 341 (x2), 349 -> 360 (x1) SPEC.md:8624: 283 -> 278 (every holder) SPEC.md:8623: 157 -> 158 (every holder) SPEC.md:8551: 355 -> 361 (every holder) SPEC.md:8550: 330 -> 341 (every holder) SPEC.md:8548: 440 -> 451 (BS-08) SPEC.md:8510: 152 -> 157 (BS-12) SPEC.md:8510: 296 -> 303 (every holder) SPEC.md:8507: 381 -> 407 (every holder) SPEC.md:8425: 330 -> 341 (every holder) SPEC.md:8424: 67 -> 73 (every holder) SPEC.md:8408: 2,265 -> 2,326 (BS-06) SPEC.md:8405: 330 -> 341 (every holder) SPEC.md:8403: 2,265 -> 2,326 (BS-06) SPEC.md:8402: 71 -> 73 (every holder) SPEC.md:8401: 330 -> 341 (every holder) SPEC.md:8400: 330 -> 341 (every holder) SPEC.md:8399: 247 -> 258 (BS-05) SPEC.md:8398: 785 -> 809 (every holder) SPEC.md:8398: 3,050 -> 3,135 (every holder) SPEC.md:8397: 330 -> 341 (every holder) SPEC.md:8340: 306 -> 317 (every holder) SPEC.md:8340: 221 -> 227 (every holder) SPEC.md:8336: 407 -> 418 (every holder) SPEC.md:8335: 542 -> 553 (every holder) SPEC.md:8325: 306 -> 317 (every holder) SPEC.md:8323: 330 -> 341 (BS-05) SPEC.md:8323: 247 -> 258 (BS-05) VERIFICATION.md rows for the landing (counts | before | after): | `BS-05` COMMITTED_CRATES, COMMITTED_MANIFESTS, COMMITTED_PROGRAMS, EVALUATOR_CORRESPOND, EVALUATOR_CRATES, EVALUATOR_VALIDATE, VECTORS_CORRESPOND, VECTORS_VALIDATE | 306, 330, 247, 322, 221, 457, 407, 542 | 317, 341, 258, 328, 227, 463, 418, 553 | | `BS-06` COMMITTED_CRATES, COMMITTED_DISTINCT, COMMITTED_MANIFESTS, COMMITTED_PACKAGES, COMMITTED_PROGRAMS, EVALUATOR_CRATE, REFUSED_CRATE, VECTORS_CRATE, VECTORS_EXPR | 306, 612, 330, 330, 247, 612, 71, 785, 2,265 | 317, 634, 341, 341, 258, 629, 73, 809, 2,326 | | `BS-07` VECTORS, VECTORS_WITHIN | 381, 296 | 407, 303 | | `BS-08` CRATES, WITHIN | 440, 355 | 451, 361 | | `BS-11` EVALUATOR_MUTANTS, VECTORS, VECTORS_WITHIN | 283, 157, 156 | 278, 158, 157 | | `BS-12` EVALUATOR_MUTANTS | 152 | 157 | | `BS-13` EVALUATOR_MUTANTS, VECTORS, VECTORS_WITHIN | 143, 150, 67 | 157, 175, 73 | | `BS-14` EVALUATOR_MUTANTS, VECTORS, VECTORS_WITHIN | 548, 385, 300 | 573, 411, 307 | | `BS-15` ACCEPTED, MANIFESTS, RENDER_MUTANTS | 330, 349, 330 | 341, 360, 341 | For a reader: BS-06 COMMITTED_DISTINCT: 612 is also the value of ["EVALUATOR_CRATE"]; rewrite its statements by hand (612 -> 634) BS-06 EVALUATOR_CRATE: 612 is also the value of ["COMMITTED_DISTINCT"]; rewrite its statements by hand (612 -> 629) BS-14 VECTORS_WITHIN: 300 is also the value of ["EVALUATOR_NODES"]; rewrite its statements by hand (300 -> 307) SPEC.md:8516 still states 150 (-> 175 for BS-13?) before "mutants), and 874 seconds for `siteTree" SPEC.md:8552 still states 150 (-> 175 for BS-13?) before "seeded mutants of the crates of at most" SPEC.md:8587 still states 150 (-> 175 for BS-13?) before "seeded runs, the rendered package on 1," written; next: cargo xtask validate-model --write, then the parts without LEXLEAN_COUNTS
README.md states the register size (507) and claims BS-01..BS-52, and carries each slice's clause; CHANGELOG.md carries each slice's entries. Neither file is an input to anything the regeneration wrote. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
afflom
left a comment
There was a problem hiding this comment.
Current-head adoption review:3ef5 now carries generated products plus the README/CHANGELOG amendment already credited in #497. The product comparison reaches the connector's300-path cap, so this review does not assert complete adoption provenance or full product acceptance.
The existing P1 input gap remains after adoption: tests/production-export/bootstrap/roots.json is still exact bloba53b31b8 with73 rows, retaining all11 lexicon roots but omitting all15 required grammar roots. Generated outputs from that unchanged input cannot establish export coverage for the missing roots. The BS35/36/37 case blobs are also unchanged, while CONFORMANCE.md:94–96 retains the previously identified universal source/module/declaration claims. Earlier review5901581394 therefore remains unresolved.
Smallest owner next step: repair the authored root input and coverage correspondence, then explicitly disposition the retained outputs against those corrected inputs and qualify the resulting exact head. Preserve reusable artifacts and successful stages; no blanket writer repeat or parallel implementation is requested. Current43578688227 stops in honesty with downstream jobs skipped; reproducibility43578688116 success does not supply missing full-gate coverage. Receipt pr532-3ef5fe-adoption-delta/receipt.json; no tests, logs, producers or CI controls repeated.
The gate's honesty job failed at fmt-check on #532's regenerated tree (run 43578688227): check-freshness and the reproducibility workflow both passed, so the regenerated tree itself was sound -- the diffs were purely formatting, in exactly three files, `cargo fmt --all --check` confirmed: - crates/conformance/src/cases/bootstrap.rs: the mod list, which the combined merge appended to rather than inserted into sorted order. - crates/conformance/src/cases/bootstrap/lexicon_schema.rs and lse_types.rs: never formatted on their own slice branches. Also ran the cheap honesty recipes (`just fmt-check model spec-links`) on this tree before pushing, each in this worktree's own isolated CARGO_TARGET_DIR (a shared target dir across two worktrees is a correctness hazard here, not just lock contention: repo_root() is compiled into the binary, so a binary built from one worktree and reused from another checks the wrong tree). Both passed clean: validate-model reports 507 ids and clean audits; validate-spec-links reports 507 table rows bijective with the register (RP-07). Formatting test sources is not an input to anything regenerated, so no further regeneration is needed; only the pull_request gate re-runs. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Release gained two files under docs/proposals (#492). With the head containing release, the merge of this pull request lands the tree the whole gate verifies. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
Release gained #520 (canonical JSON refuses duplicate keys that differ only in their escapes) and #533 (the Intel host tests accept the repository's renamed spelling). CHANGELOG.md is the one file both sides changed: release's entry and the eleven slices' entries are both kept, release's first. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
tests/production-export/bootstrap/roots.json is hand-maintained and the grammar slices never added their roots, so the landing regeneration exported no package for PropositionTerms, GrammarMath, GrammarProof, CoreJson, CorePrint, CoreRead, SemanticRead and SemanticPrint, and each of their cases stopped at the stack-bound step (bootstrap/keyword_entry is no package of the committed exports). Adds their 15 production roots. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
afflom
left a comment
There was a problem hiding this comment.
Existing gate evidence now identifies a concrete instance of the previously reported authored-root/export gap. In run43600860221, job132473860716 (bdd-bootstrap-proposition-terms), BS43 stops at stack_bounds.rs:349: bootstrap/keyword_entry is no package of the committed exports. Retained log line663; SHA2564d45194d2c44f7188f49cc50b3ec2309bf7950fe1940d49c6d7a6d2b3f191658. The logged checkout is synthetic8f7d9c9f merging this51628 head with6b1, not actual0a989. No stopped#431 runtime or source was inspected.
Please correct the authored root/export coverage in this existing owner lane and explicitly disposition applicable outputs before final qualification. Preserve the completed differential/evaluator/plant/rustc evidence from this job; this finding requests no blanket repeat of those stages or a completed producer. It does not waive final-head full acceptance and actual merged-tree equality.
The other eight failing jobs have not been diagnosed here and must not be assigned this cause. One selected log was fetched and retained once; no tests, producer, retry or CI controls were run. Receipt pr532-proposition-job132473860716-diagnostic/receipt.json.
afflom
left a comment
There was a problem hiding this comment.
The authored-input portion of the missing-root P1 is resolved at972c7bb: this is a direct one-commit, one-path advance over51628, adding the exact15 missing grammar requests to roots.json. All73 existing rows remain byte-equivalent and the88 names/root names are unique. keyword_entry now maps to PropositionTerms.keywordEntry as BS43 expects.
No cases, assertions, register or generated products changed in this delta. The separate BS35/36/37 coverage findings remain. Existing producer43612402370 already targets this972c head; retain it and explicitly bind/disposition its outputs before resulting-head review and full qualification. No duplicate producer or repeat of completed differential stages is requested. Prior selected-job failure and scoped successes remain preserved.
Draft/source preparation is not final acceptance. Receipt pr532-972c7-root-input-repair/receipt.json. No tests/builds/producers/CI controls or stopped#431/oldTC03 inspection were performed.
The landing regeneration of 972c7bb lost its runner in "Measure the stack bound of each recursive package the record lacks" after 98 minutes, measuring BS-36, BS-38 and BS-39 four at a time. Alone in the gate, BS-36 left 7 GiB of the runner's 16 GiB available and BS-38 and BS-39 about 8.8 GiB, so together they do not fit. Run BS-36 (and the typed IR's BS-69..72, which lost r16's runner twice) one at a time and the rest two at a time, and write free memory to the log once a minute so a lost runner leaves evidence. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
afflom
left a comment
There was a problem hiding this comment.
Bounded scheduler delta review: bf94 directly follows972c and changes only regenerate.yml's stack-measurement scheduling/logging. The same selected cases remain covered; BS36/69–72 run serially, remaining cases at width2, with strict failure propagation retained. No new source blocker demonstrated. Available-memory snapshots are not per-case/global peaks or proof of the lost runner's cause; minute logging adds sampled evidence only.
Concrete repeated work is now visible: old43612402370 completed all three writer parts and combine before counts failed; new43634296837 has completed those same upstream stages again and is measuring counts. This one-file delta changes no upstream writer definitions or product inputs. Please preserve the original successful products and disposition their applicability/custody before another whole-wave repeat. Reuse requires exact input/toolchain/options/product and artifact-custody binding; this review does not authorize unverified cross-run substitution or rewriting manifests to claim a different producer.
Existing owner retains the active run; no retry, producer or CI control was initiated. Prior15-root repair credit and BS35/36/37 coverage, body, output-adoption and full-final-head obligations remain. Draft CI43634429898 is3success/15skipped, not full acceptance. Receipt pr532-bf94-scheduler-delta/receipt.json.
afflom
left a comment
There was a problem hiding this comment.
The new counts failure has a concrete locus: run43634296837/job132593376021 logs BS39 panicking at core_print.rs:673, whose exact-bf94 source asserts largest <= CERTIFICATE_BUDGET (3,145,728 bytes). BS38 passed; the light group exited101 after4,422.83s, so the serial BS36 phase was not reached. The runner continued to upload failed-job output. This receipt does not establish OOM, lost-agent failure or a timeout.
The workflow's grep kept the panic header but dropped the assertion details, including actual certificate length. Existing owner should inspect the retained CorePrint certificate-part sizes/source correspondence and address the growth or explicitly disposition its budget contract; do not silently raise/remove the assertion. Preserve useful failure details in the existing log path.
Three writer parts, combine, counts/model work and BS38 have completed evidence. Bind applicable inputs/definitions/products/custody before any necessary continuation, retaining the failed receipts and avoiding another blanket writer wave. Artifact14135057639 is reported in the log but was not downloaded or adopted here. Only this new110,195-byte job log was fetched once, SHA653efba917a07c07db72cf4271f38b48e87abe21aa06761c234c360aebb91443. Receipt pr532-bf94-counts-diagnostic/receipt.json. No test, writer, retry or CI-control action was performed.
…532) Replaces .github/workflows/regenerate.yml with the version prepared for #532 (blob 704df4f): heavy cases (BS-36, 69, 70, 71, 72) run one at a time with explicit started/ended lines and elapsed minutes, the rest at --test-threads=2, and the whole step is bounded by a shared timeout against a 320-minute budget so a run that cannot finish everything fails cleanly (and keeps every row already measured in tests/stack-bounds.json) instead of being cancelled by the jobs own 360-minute limit and losing all of it. Also carries the once-a-minute free-memory trace from the same change. No other change; not a release merge. This copies ahead of #532 landing so #497s own stack-bound rows can start banking now, since measuring all of BS-65..72 is the longest pole and the two PRs rows stay valid through a later merge as long as the packages do not change. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
- Hand-resolved .github/bdd-parts.json, intel-host-lanes.json, vv-rows.json to combine both sides' lane/part additions; verified no bdd-part is in two lanes and all 508 model/ids.toml parts are covered. - Fixed generated-file conflicts (README.md, SPEC.md, repository.rs, xtask/src/spec_links.rs) to the post-merge count of 508 registered IDs. - Relocated #532's own CHANGELOG bullets (BS-35..BS-52) from inside the new 0.4.3 section back under Unreleased. - Added two stack-bound rows to tests/stack-bounds.json (bootstrap/core_json, bootstrap/print_core_module). - Certificate survey of the 15 production roots new in #532: measured the 7 over-budget roots exactly via the real pipeline (lexlean verify with max_file_bytes lowered so LLS8002 reports the exact byte count before any Lean child spawns); raised CERTIFICATE_BUDGET in core_print.rs, core_read.rs, semantic_print.rs, and semantic_read.rs accordingly, with measurements documented in VERIFICATION.md. All 15 roots stay under max_file_bytes (6,291,456); none required stopping the batch. - Replaced .github/workflows/regenerate.yml wholesale with the IR session's handoff version, adding a 320-minute budget on the measuring step. - Restored bootstrap/lexlean.toml's max_file_bytes to 6291456 after the measurement loop and relocked bootstrap/lexlean.lock. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
afflom
left a comment
There was a problem hiding this comment.
P1: the changed certificate assertions no longer enforce the registered bounds. At e033, model/ids.toml lines2913/2919/2931/2937 still promise no certificate part above3,145,728 bytes for BS-36/37/39/40. Their case constants now allow3,932,160 (semantic_print.rs:45),5,242,880 (semantic_read.rs:49),4,259,840 (core_print.rs:42) and6,029,312 (core_read.rs:37). Passing these relaxed assertions would not establish the unchanged claims.
Existing owner: reconcile the implementation with the promised bounds. Any intended change to that contract needs explicit scope disposition and consistent normative/register/scenario/test/generated-document correspondence; this review does not authorize silently broadening it. Preserve the measured growth and original failures. VERIFICATION's pre-Lean LLS8002 sizing observations are useful, but do not establish elaboration/kernel/axiom-policy acceptance.
Efficiency/custody: the workflow blob704df4fd is identical to the already reviewed #497 workflow, so that review and its deadline/salvage limitations are reused. The commit credits the IR-session workflow handoff. #532 now also carries release fa271, Lean4.32.2 and semantics4b3ed237; this is not evidence that its producer and #497's active producer have identical inputs. Preserve current owners/results, explain input-specific reuse and future sequencing, and do not add a competing writer. Current body still describes the old507-row/6a02 composition and should be updated with the contract disposition.
Bounded selected-file review, not a full product-heavy integration inventory. Receipt pr532-e033-source-identity-review-20261115.json SHA256975af1bf235d6e361171c69957cb452e79dc009538ac5a54180b92d2a70140f7. No tests, sizing experiment, log fetch, writer or CI control. Standing #431/oldTC03 STOP remains in force; release ancestry does not authorize stopped-work adoption.
bootstrap-toml-read, bootstrap-package-ref, and bootstrap-lexicon-entry move from lower to print: in #532's last gate the Windows lanes (240 min budget) took lower 165 min, moves 80, validate 68, print 54, and lower's cases were stopping early at the stack-bound step, so it runs longer once they complete. The same partition now drives the macOS arm64 lanes too. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
The write recipes of SPEC.md §9.2, run by regenerate.yml (https://github.com/UOR-Foundation/lexlean/actions/runs/43755020672), rewrote 2373 paths. The counts writer reported: BS-05 COMMITTED_CRATES: 317 -> 332 (crates/conformance/src/cases/bootstrap/rust_validate.rs) BS-05 COMMITTED_MANIFESTS: 341 -> 356 (crates/conformance/src/cases/bootstrap/rust_validate.rs) BS-05 COMMITTED_PROGRAMS: 258 -> 273 (crates/conformance/src/cases/bootstrap/rust_validate.rs) BS-05 EVALUATOR_CORRESPOND: 328 -> 329 (crates/conformance/src/cases/bootstrap/rust_validate.rs) BS-05 EVALUATOR_CRATES: 227 -> 228 (crates/conformance/src/cases/bootstrap/rust_validate.rs) BS-05 EVALUATOR_VALIDATE: 463 -> 464 (crates/conformance/src/cases/bootstrap/rust_validate.rs) BS-05 VECTORS_CORRESPOND: 418 -> 433 (crates/conformance/src/cases/bootstrap/rust_validate.rs) BS-05 VECTORS_VALIDATE: 553 -> 568 (crates/conformance/src/cases/bootstrap/rust_validate.rs) BS-06 COMMITTED_CRATES: 317 -> 332 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-06 COMMITTED_DISTINCT: 634 -> 664 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-06 COMMITTED_MANIFESTS: 341 -> 356 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-06 COMMITTED_PACKAGES: 341 -> 356 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-06 COMMITTED_PROGRAMS: 258 -> 273 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-06 EVALUATOR_CRATE: 629 -> 631 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-06 REFUSED_CRATE: 73 -> 75 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-06 VECTORS_CRATE: 809 -> 841 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-06 VECTORS_EXPR: 2326 -> 2329 (crates/conformance/src/cases/bootstrap/rust_print.rs) BS-08 CRATES: 451 -> 466 (crates/conformance/src/cases/bootstrap/rust_moves.rs) BS-08 WITHIN: 361 -> 362 (crates/conformance/src/cases/bootstrap/rust_moves.rs) BS-15 ACCEPTED: 341 -> 356 (crates/conformance/src/cases/bootstrap/rust_package.rs) BS-15 MANIFESTS: 360 -> 375 (crates/conformance/src/cases/bootstrap/rust_package.rs) BS-15 RENDER_MUTANTS: 341 -> 356 (crates/conformance/src/cases/bootstrap/rust_package.rs) BS-05: §31 row: 227 -> 228 (x1), 258 -> 273 (x1), 317 -> 332 (x1), 341 -> 356 (x1), 418 -> 433 (x1), 553 -> 568 (x1) BS-05: the §31 row does not state 328 (-> 329) BS-05: the §31 row does not state 463 (-> 464) BS-05: register: 227 -> 228 (x1), 258 -> 273 (x1), 317 -> 332 (x1), 341 -> 356 (x1), 418 -> 433 (x1), 553 -> 568 (x1) BS-05: scenario: 227 -> 228 (x1), 258 -> 273 (x1), 317 -> 332 (x1), 341 -> 356 (x1), 418 -> 433 (x1), 553 -> 568 (x1) BS-06: 3135 is VECTORS_CRATE + VECTORS_EXPR: -> 3170 BS-06: §31 row: 73 -> 75 (x1), 258 -> 273 (x1), 341 -> 356 (x3), 809 -> 841 (x1), 2326 -> 2329 (x2), 3135 -> 3170 (x1) BS-06: the §31 row does not state 317 (-> 332) BS-06: the §31 row does not state 629 (-> 631) BS-06: the §31 row does not state 634 (-> 664) BS-06: register: 73 -> 75 (x1), 258 -> 273 (x1), 341 -> 356 (x3), 809 -> 841 (x1), 2326 -> 2329 (x2), 3135 -> 3170 (x1) BS-06: scenario: 73 -> 75 (x1), 258 -> 273 (x1), 341 -> 356 (x3), 809 -> 841 (x1), 2326 -> 2329 (x2), 3135 -> 3170 (x1) BS-08: §31 row: 361 -> 362 (x1), 451 -> 466 (x1) BS-08: register: 361 -> 362 (x1), 451 -> 466 (x1) BS-08: scenario: 361 -> 362 (x1), 451 -> 466 (x1) BS-15: §31 row: 341 -> 356 (x2), 360 -> 375 (x1) BS-15: register: 341 -> 356 (x2), 360 -> 375 (x1) BS-15: scenario: 341 -> 356 (x2), 360 -> 375 (x1) SPEC.md:8568: 361 -> 362 (every holder) SPEC.md:8567: 341 -> 356 (every holder) SPEC.md:8565: 451 -> 466 (BS-08) SPEC.md:8442: 341 -> 356 (every holder) SPEC.md:8425: 2,326 -> 2,329 (BS-06) SPEC.md:8422: 341 -> 356 (every holder) SPEC.md:8420: 2,326 -> 2,329 (BS-06) SPEC.md:8418: 341 -> 356 (every holder) SPEC.md:8417: 341 -> 356 (every holder) SPEC.md:8416: 258 -> 273 (BS-05) SPEC.md:8415: 809 -> 841 (every holder) SPEC.md:8415: 3,135 -> 3,170 (every holder) SPEC.md:8414: 341 -> 356 (every holder) SPEC.md:8357: 317 -> 332 (every holder) SPEC.md:8357: 227 -> 228 (every holder) SPEC.md:8353: 418 -> 433 (every holder) SPEC.md:8352: 553 -> 568 (every holder) SPEC.md:8342: 317 -> 332 (every holder) SPEC.md:8340: 341 -> 356 (BS-05) SPEC.md:8340: 258 -> 273 (BS-05) VERIFICATION.md rows for the landing (counts | before | after): | `BS-05` COMMITTED_CRATES, COMMITTED_MANIFESTS, COMMITTED_PROGRAMS, EVALUATOR_CORRESPOND, EVALUATOR_CRATES, EVALUATOR_VALIDATE, VECTORS_CORRESPOND, VECTORS_VALIDATE | 317, 341, 258, 328, 227, 463, 418, 553 | 332, 356, 273, 329, 228, 464, 433, 568 | | `BS-06` COMMITTED_CRATES, COMMITTED_DISTINCT, COMMITTED_MANIFESTS, COMMITTED_PACKAGES, COMMITTED_PROGRAMS, EVALUATOR_CRATE, REFUSED_CRATE, VECTORS_CRATE, VECTORS_EXPR | 317, 634, 341, 341, 258, 629, 73, 809, 2,326 | 332, 664, 356, 356, 273, 631, 75, 841, 2,329 | | `BS-08` CRATES, WITHIN | 451, 361 | 466, 362 | | `BS-15` ACCEPTED, MANIFESTS, RENDER_MUTANTS | 341, 360, 341 | 356, 375, 356 | For a reader: SPEC.md:8419 still states 73 (-> 75 for BS-06?) before "runs with a capacity one byte short of" SPEC.md:8441 still states 73 (-> 75 for BS-06?) before "site plants, which are crates of names," written; next: cargo xtask validate-model --write, then the parts without LEXLEAN_COUNTS
afflom
left a comment
There was a problem hiding this comment.
Delivered output credited ate393ed99, sole child ofe033: producer43755020672 completed all six jobs including counts/stack and delivery. Final artifact14186561615 binds exacte033, digest06a69c743e410411bec9000a4cec618be008c0492a917e7b50611bd98146cc39; custody is metadata-based, not ZIP-byte equality.
Complete2373-path inventory confines changes to generated bootstrap build/verify/production outputs, mirrored BS05/06/08/15 counts and stack rows. Authored bootstrap/compiler/workflow/config/lock/identity and the four previously reviewed certificate-bound cases are unchanged. The new counts log measured only BS36:1pass/0fail/507filtered in6232.25s, adding debug stack rows print_at1572864 and print_semantic_module1179648. These are stack measurements, not certificate sizes; the log does not establish fresh certificate sizes or BS37/39/40 runs.
The existing P1 remains: raised case limits exceed unchanged registered3,145,728-byte promises. Passing BS36 under the inherited3,932,160 limit does not prove the3MiB contract. Existing owner should reconcile that implementation/contract scope and remaining count/body/prose disposition while preserving this completed producer and its products. No duplicate measurement, producer or competing patch requested. Draft/full gate43780347297 still incomplete; no final acceptance or merge.
Receipt /workspace/work/pr532-e393-product-review-20261116.json; SHA256 f1b5ab39972743885dd4edd249f87044da0eae7182b29dbbed2738cd551e2946. New log132969972567 retained once; all STOP boundaries preserved.
afflom
left a comment
There was a problem hiding this comment.
Completed run 43780347297, attempt 1, is credited on e393ed9: plan, honesty and rest passed all 29 steps. The complete listing has 18 jobs, of which 15 were skipped, including bootstrap export, Atlas and auxiliary host/Intel work. This establishes the draft development checks only. The new plan log binds the exact head; completed producer and product evidence remain reusable.
The four certificate-limit P1 findings remain: BS-36/37/39/40 promise parts no larger than 3,145,728 bytes while their checks permit larger limits. Existing owner: reconcile those bounds and the implementation, retain delivered products, then provide final full qualification and actual merged-tree equality. No repeated producer or completed gate is requested. This PR remains draft; all standing STOP boundaries remain.
Receipt: /workspace/work/pr532-e393-terminal-gate-review-20261116.json; SHA256 8141ca15c2e43d0a6198f0325319878b54eda2c9349e2e47d1a9bd5d55954e85.
- Resolved all 1072 conflicts: 1067 generated/lock paths took release's side (bootstrap/expected/**, tests/production-export/bootstrap/**, examples/models/expected/**, stdlib/expected/**, bootstrap/lexlean.lock, including 593 modify/delete conflicts where release's own regeneration replaced a content-addressed file under a different hash, resolved as deletions); the 5 hand-maintained files (CHANGELOG.md, README.md, SPEC.md, crates/conformance/src/cases/repository.rs, xtask/src/spec_links.rs) took both sides' additions. - Register total is 509 (the batch's own 508 plus PF-26's one new row); model/ids.toml merged cleanly to that count. Updated README's sentence, SPEC's total and §31 row count, and repository.rs's/spec_links.rs's two hardcoded mentions each from 508/498 to 509. Relocked bootstrap/lexlean.lock fresh (bootstrap/lexlean.toml itself is unchanged by this merge, so the lock came back identical to e393ed9's own). - Combined both sides' CHANGELOG bullets under the shared ## Unreleased heading (BS-35..BS-52's slice bullets, then release's forward.yml/macOS-lane /memory-budget/PF-26 bullets), updating PF-26's own "the register is 498 IDs" to the merged line's true count of 509. - crates/conformance/src/cases/mod.rs: added BS-36..BS-40 to the heavy-case admission bound (HEAVY, now 9 entries), with the doc comment's own evidence from #532's gate: on a 16 GiB runner, job peaks were BS-37 11.0 GiB, BS-40 9.5 GiB, BS-36 8.7 GiB, BS-38 and BS-39 7.0 GiB each, and the ARM leg lost its runner while BS-35..BS-38 overlapped. - .github/intel-host-lanes.json: moved bootstrap-semantic-print from the Windows "lower" Intel lane to "print" (lower took 165 min of the 240-min budget, print 54); confirmed all 51 lane entries (50 bdd parts plus the "rest" remainder) stay unique and complete. cargo fmt --check, cargo xtask validate-model (no --write, 509 ids, meta-gate and audits clean), and all 18 conformance_rp_* cases pass. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
afflom
left a comment
There was a problem hiding this comment.
The added BS36–40 heavy-case admission and semantic-print lane move are credited as scheduling changes. The certificate-bound P1 remains in current source: the model promises 3,145,728 bytes, while BS36/37/39/40 checks permit 3,932,160 / 5,242,880 / 4,259,840 / 6,029,312. Scheduling does not reconcile that contract mismatch.
Existing producer 43844527381 remains active; no duplicate writer or retry is requested. The owner retains the bounded repair/explicit scope disposition, stale 507/6a02 body correction, delivered-output review and final composed-head qualification. This selected-source review does not qualify the capped product-heavy integration diff.
Receipt: /workspace/work/pr532-497-owner-repair-deltas-20261116.json; SHA256 24812521642be3fba43aa5d31a129971d5e2a2e599b95a65fd3aff16f7c4c076.
The write recipes of SPEC.md §9.2, run by regenerate.yml (https://github.com/UOR-Foundation/lexlean/actions/runs/43844527381), rewrote 2033 paths. The counts writer reported: counts: no count moved
afflom
left a comment
There was a problem hiding this comment.
Bounded adoption review at66552b6b credits delivery from existing producer43844527381: all six jobs succeeded and recorded artifacts remain tied to source394ba49c. The delivered commit has that exact sole parent. Uncapped tree-boundary checks show changes only under bootstrap/expected, examples/models/expected, tests/production-export/bootstrap and bootstrap/lexlean.lock; model, conformance source, workflows and tests/stack-bounds.json are unchanged. The counts report says no count moved. This product delivery therefore leaves the existing P1 unchanged: modeled3,145,728 bytes versus BS36/37/39/40 allowances3,932,160/5,242,880/4,259,840/6,029,312. Existing owner retains contract/implementation reconciliation and stale-body update. Current exact-head gate43863888061 remains active; preserve this delivery and qualify the final composed head without repeating completed producers/checks. No full product semantics or final acceptance claim.
Summary
Combines two already-landed batches, per the user's decision to merge them together rather than land either alone (landing one alone would make the other's regenerated tree stale):
cd571f129a, carried individually in Batch-land BS-35..43 onto release/v0.4 (issue #269) #527ebe8911162, carried individually in BS-50..52: the lexicon schema, LSE scopes and LSE type checker as source, batch (#269) #525A real
git mergewas used (not a text splice) to combine the two authored commits, both sitting directly on release6a02d2c7d5. Conflicts at the shared insertion anchor after BS-34 were resolved as "keep both, grammar then lexicon" for the register-order files (SPEC.md §31 rows + §17.23 prose, model/ids.toml, features/suites/bootstrap.feature, bootstrap/README.md, conformance.rs, cases/bootstrap.rs), and hand-merged (not unioned) for: bootstrap/lexlean.toml entrypoints (41+3=44), the register-count constant repeated in both xtask/src/spec_links.rs and cases/repository.rs (504/499 → 507 everywhere), .github/bdd-parts.json and vv-rows.json (all eleven new parts/rows), and .github/intel-host-lanes.json (grammar's lanes kept; lexicon's three new parts land inlower, matching where #525 put them).Three merge-seam bugs were found and fixed before pushing: a missing
[[id]]header in model/ids.toml, a missing closing brace +#[test]attribute in conformance.rs, and a stray extra blank line in SPEC.md. Verified afterward: 507 unique ids, 507 §31 rows, 42 feature scenarios, 42 conformance fns, all JSON valid, no stray conflict markers.Verified with the maintainer's
check_batch.pyagainst all eleven source PRs (505, 509-515, 522-524) on a scratch branch before moving any real branch — every reported line is in the known-benign set (recomputed one-line arrays, each PR's own stale count, the supersededmax_child_output_bytesparagraph), and PR #522's botchedRP-497/§497line is correctly absent from this batch.Supersedes #525 and #527, which stay open as the per-batch authoring record.
Test plan
🤖 Generated with Claude Code