6.4 KiB
Independent actual-output acceptance runtime review
Final source edf096cd33350865ecfc3c65ec266a6c0d3c4000 satisfies the selected finite regression readback after one explicitly approved target-path spelling exception. All16 inherited failure bodies remain byte-identical after replacing numeric thread identifiers only; no failure-body exception exists. Full core tests still have16 failures and workspace format still exits1. No compiler, test executable, debugger or archived launcher was executed by this reviewer.
Gates and bindings
Prepared read-only checker was run using bundled Python -I -B, args edf096cd33350865ecfc3c65ec266a6c0d3c4000 edf096c 11. Original strict helper refused exact environment comparison; CHECKER-NOTE.md preserves the refusal and original helper. Amended helper SHA256 8fdf8c0ba848d8f61cb09739961239c316fdb0eb17e4b64b88ae34274cce0f53 then completed all checks and exclusively wrote ACCEPTANCE-RUNTIME-READBACK.json.
Readback JSON SHA256 88cb21bc89a1337b32f74ec2fe8adc6dc2835ccea17bb313b5a7e2433fce4fd2.
| Gate | Exit | Actual result |
|---|---|---|
| Focused | 0 | 11 new acceptance tests passed;281 filtered |
| Full | 101 | 292 names:276 passed,16 failed, zero ignored |
| List | 0 | 292 unique names, exactly full runtime inventory |
| Strict core all-target Clippy | 0 | Locked/offline -D warnings completed |
| Workspace format | 1 | Exact prior raw output retained |
All281 predecessor outcomes are unchanged. Focused is exactly the11 added names and every new result is passed. Independently compared raw failure blocks as well as the checker result:16 match after replacement of numeric thread identifiers only, including whitespace and framework lines. Format SHA256 remains c1c6b12861c8117ce6ffc6d6742ebf92618aafb10160ec6f99d370ca72ebab2d.
Baseline2941713 full/list/fmt raw logs and receipts match the locally committed entry evidence ZIP SHA256 19a87583a8dc4c22f35999044de5cac2c2c97b04fb457d2c3df3f6d2c054c865. No remote publication claim is made here.
All five receipts bind current clean source,559 declared inputs and actual Git blob/disk byte equality, unchanged inputs through each run, exact commands and one captured launcher. Input manifest SHA256 6be8ae8911bb3d0aee346b137930d2b27a7d0ed4e77c4c9502b60d07982bd028. Current focused/full/list executable is43,066,880 bytes, SHA256 7bc17993f265bdf7dc79b7ea5191401dd853f1dd47517235663113e4571031bb; checker reread actual bytes. Full raw log SHA256 5df6602c3d62c776720aa97916f0bf740de5488773d25601600aea8de27515e2; Clippy raw SHA256 83248e15864952f5d00a212d9a8ac3e6fc136cec5ff0b927328c20b022c7fff5. Exact per-gate raw/receipt pins are in the JSON.
Actual captured launcher SHA256 a17935d88ccb953f441b8939ccf2dc6aabcd6ea406d9a3754e3e8306a5bd26ce differs from previously reviewed entry launcher only by source pin/description and output directory. It preserves exact source-before/after checks, unfiltered merged process logs, actual exit codes, helper capture and Cargo-reported artifact hashing. Reviewer inspected this delta without running the launcher. Jobs1, incremental0 and all other recorded environment values are exact baseline matches; stack/lifecycle/compiler overrides remain absent. Only CARGO_TARGET_DIR forward slash versus backslash changes, with both exact strings resolving to the same required target. See CHECKER-NOTE; this is not generalized path normalization.
These checks establish declared-input/recorded-artifact concordance, not independently reproducible compiler provenance or coverage of undeclared/untracked dependencies.
Initial failed attempt and correction
Preserved initial source 4872141c036c16825f1c098bd7cf44709897e814 focused run exited101:9 passed/2 failed. Both failures were explicit empty-warning assumptions in rich OA/SA and four-file tests. Raw log SHA256 b26d046e8b6284b5080ecbd6486c04f5556e1866292d5ac52f27893e69906371, receipt ff17f718a831784334443438ff444f6cc441a96ddd896fb1afd8e5602d9ed21a; independently confirmed receipt/log match and counts. Initial artifact43,060,224 bytes, SHA256 cba18de5758cac531dd166c95081fdcb390c2ceb295c06e09a26f3aa998e073c is its own recorded identity, not the final artifact.
Final diff changes only new test expectations/helper and implementation note. Production unchanged. WARNING-FIX-ORACLE derives both legitimate ordered diagnostics and W227/W221 from validator/mapper source and literal expected bytes, not runtime diagnostic output. Final tests SHA256 e42fd54b157a47a69c8de2ea379b375b797a971b5bb604be572ea1f45b55b73d require complete warnings and final rich/four-file inventories. Earlier review's missed empty-warning assumption remains explicitly recorded in TEST-REVIEW.
Scope of runtime and frame evidence
Final full log has eleven stack-named passing tests:ten inherited controls and the new scoped acceptance test. The latter contains five workers:deep128-node success and mismatch, wide129-node success and mismatch, plus real inherited151-byte late refusal; setup outside workers, actual verify/assert/destruction inside128KiB workers. Deep name mismatch may short-circuit before expression equality; deep success traverses actual expressions. These finite results do not establish universal stack headroom or whole predecessor setup on128KiB.
Read initial/final captured acceptance prologues: production verify8648, parse_entry_producers568, equal_target312, warning_inventory776, check_destination456 selected reservation bytes are unchanged by the test-only repair. Final capture has135 records versus initial125, including test/generic machinery, not135 live frames. Existing stack/active phase observations and separate source review retain their own scope. Explicit subtractions exclude pushes/caller overhead/uncaptured callees; captured64-byte ranges may cross symbol boundaries, so no whole-stack bound is inferred. Parsing, resolution and equality/destination transfer are sequential verification phases; summing all phase reservations as simultaneously live is invalid.
Finite actual-output admission, owning model/warnings, exact152/151 and161/160 behavior, nonvacuous actual-E fault control and corrected rich whole-byte/model/warning assertions passed on this artifact. Internal fault controls are not authentic emitted histories or real late-cap proofs. This remains private acceptance, not public source8, installation/history, GUI integration, repeated carried-source editing, all-heap accounting or closure of declared scratch exclusions.