Repository navigation
Six failure-mode rows, moved verbatim out of a parked branch - #10299
Merged
Merged
Conversation
These were authored on #9725 and are gated on it only by the accident of which branch the author was on — none has anything to do with the wet route, and that PR cannot go green tonight because its receipt's semantic subject has moved. A row nobody can read is not filed. Moved, not rewritten: each row's authored bytes are carried verbatim into its own file under the split shape #10206 established, as a single `receipts` entry. Six roster entries appended at the END, unsorted — roster order is source order and the projection renders in it. per_argument_exhaustive_matrix_blind_to_a_cross_argument_relation closure_subject_with_no_enumerable_membership ceiling_never_exercised_for_a_population_the_census_cannot_plan content_digest_makes_annotations_semantically_load_bearing subject_and_its_digest_as_independent_parameters a_written_row_is_not_a_firing_mechanism Verified in the direction that catches loss, not only duplication: roster 90 lines / 90 distinct, `main-not-mine` EMPTY, `mine-not-main` exactly these six, every identity present in the projection, and row files joined against the roster both ways. The same six are deleted from #9725 in the same motion, so two branches never declare one identity — which is the duplicate-declaration state that took main down for two hours tonight. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01GvVoivi7L449wbh6rjeJY4
gunbai-bot Bot
pushed a commit
that referenced
this pull request
Sep 3, 2026
Deleted here in the same motion that files them off clean main, so two branches never declare one identity — that is exactly the duplicate-declaration state that took main down for two hours tonight, and a deliberate deletion is cheaper than the same deletion discovered later as a conflict. Nothing is lost: the authored bytes move verbatim into six files under the split shape, and #10299 carries the verification in the direction that catches loss. What stays here is the wet-route work these rows were only ever adjacent to. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01GvVoivi7L449wbh6rjeJY4
added 3 commits
September 3, 2026 23:23
…9-six-rows # Conflicts: # dag/gunbc/recurring_failure_mode/roster.dag # docs/design-failure-modes.md
…9-six-rows # Conflicts: # docs/design-failure-modes.md
…9-six-rows # Conflicts: # dag/gunbc/recurring_failure_mode/roster.dag # docs/design-failure-modes.md
briansrls
pushed a commit
that referenced
this pull request
Sep 4, 2026
#10299 moved six failure-mode rows verbatim out of a parked branch. All six are distinct from this branch's row, so all seven are included -- theirs first as the established tail, ours appended last. Projection takes main's side mechanically and is regenerated below from the merged authority. Assertions against both parents: 93 row files, 93 imports, 93 list entries, no duplicates, no cross-surface disagreement, loss arm empty both ways, subsequence holds both ways -- against main we add one row, against the previous head we gain their six. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012imgm3QzXAT6GBn3ifTCDd
gunbai-bot Bot
pushed a commit
that referenced
this pull request
Sep 4, 2026
Two lines on top of main's rendering. All six of #10299's incoming paragraphs are present and this branch's row renders byte-exact against the fold of its receipts. Binary built after the merge. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012imgm3QzXAT6GBn3ifTCDd
gunbai-bot Bot
added a commit
that referenced
this pull request
Sep 4, 2026
…othing (#10308) * The emitted entry point refuses instead of exiting zero having done nothing v1.compiler.emit_rust emit_main_rs rendered an empty-body `main` for every corpus with no workflow function and no `v1.compiler.compile` module. That crate builds, runs, exits 0 and performs none of the program's work -- DESIGN §5's fabricated plausible output, sitting at exactly the point where §7's equivalence-by-execution would run. The same function's has_pipeline arm already refuses per verb through emit_candidate_cli_host, so the fail-closed shape was authorable twenty lines away and the fabricating arm was an inconsistency, not a limitation. emit_entry_point_absent_main_rs now emits a main that names WHICH of the two absences held and exits 2. The refusal marker is one authority (entry_point_absent_refusal_marker) with two readers, so a receipt cannot go green against a message the emitter stopped producing. Same class, second half: the emitted Cargo.toml declared clap unconditionally while the only two things that render `use clap` -- the derived CLI over workflow functions and the generated pipeline dispatch -- are selected by the emission itself. A fixed dependency list is a second authority for what the crate depends on, free to disagree with the sources the same run just wrote. emitted_crate_dependency_lines now derives the set from an EmittedCrateDependencyDemand the emission fills. `ureq` is left in the base list and named as the residue with its next-rung trigger rather than conditioned on a scan of emitted text. Evidence, both directions, on the required `cargo test --release -p v1-compiler --lib` step (gunbc.repo_self_build repo_self_test_command), which runs compiler_tests because it is emitted into the seed lib: a_corpus_with_no_entry_point_emits_a_refusing_main_and_declares_no_clap the_clap_dependency_follows_the_emitted_cli_demand_in_both_directions Both green on a clean checkout of 8bfae78 with only this patch applied. Both RED under the mutation that restores the two prior behaviours: the first fails on the marker's absence, the second on clap appearing under renders_clap_cli: false. Executed half, measured by hand and deliberately not enrolled: emitting dag/std/logic.dag with the built gunbc produces a crate whose Cargo.toml declares clap 0 times, which cargo builds green, whose binary exits 2 and prints the refusal on stderr. It is not enrolled because no per-PR route executes an emitted crate -- every comparable emit-build-run receipt here is classified onto a cadence std.witness_admission answers has-no-scheduled-route, and adding another would be an inert lens. What this does NOT do: it does not make the emitted crate executable, and it is not §7 progress. The verb that would carry that receipt is Compile, whose realization is CliRetainedHostKernel, a declared boundary owned by gunbc.cli_wire_host_admission whose trigger names two substrate capabilities. This converts a silent zero into a refusal a census can count. gunbc.recurring_failure_mode gains one row, emitted_entry_point_succeeds_doing_nothing, stating rung 2 after repair against a ceiling of 4 -- the decidable form emits no entry file at all, and the trigger is an emitted-population model that admits a crate with no bin target. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012imgm3QzXAT6GBn3ifTCDd * chore: regenerate drifted generated artifacts (ci auto-heal) * Regenerate docs/design-failure-modes.md from the merged authority Two lines on top of main's rendering: this branch's identity in the population list and its paragraph. #10286's ANALYST content on stale_base_branch_reads_as_pending_work survives, so the merge lost nothing -- which is the point of regenerating rather than resolving. The binary that produced these bytes was built after the merge. Re-read at the receipt boundaries: 17 receipts, fold equals the rendered line byte for byte at 3470 characters, zero double spaces. The single trailing space that makes this paragraph the one outlier of 69 regenerated identically, as deterministic output rather than drift; it stays deliberately, as the discriminating RED for a receipt-spacing wall that would otherwise be green on arrival and carry no information. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012imgm3QzXAT6GBn3ifTCDd * Regenerate docs/design-failure-modes.md from the merged authority Two lines on top of main's rendering; main's new check_subject_shape_cannot_represent_the_state_the_check_detects paragraph is intact and this branch's row renders byte-exact against the fold of its receipts. Binary built after the merge. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012imgm3QzXAT6GBn3ifTCDd * Regenerate docs/design-failure-modes.md from the merged authority Two lines on top of main's rendering. All six of #10299's incoming paragraphs are present and this branch's row renders byte-exact against the fold of its receipts. Binary built after the merge. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012imgm3QzXAT6GBn3ifTCDd * Split the failure-mode row out: leave only the uncontended repair here Measured cause of three consecutive lost merge races: not one conflict involved the work. The emitter change, the tests and the stage0 mirrors have never collided with anything -- nobody else is touching them. Every conflict was dag/gunbc/recurring_failure_mode/roster.dag and its projection, which eleven open PRs are appending to concurrently and which this PR touched only to carry the §4b ledger row. DESIGN requires the class be FILED, not that it be filed in the same commit as its repair, and the limit of that is worth stating so it is not over-applied: it holds because this row DOCUMENTS a class. It would not hold for a row that is the acceptance evidence for the change, where separating them lands a repair whose wall is absent and leaves exactly the gap a regression walks through. The wall here is the pair of tests, and they stay. So the row, its two registration lines and its two projection lines are removed from this PR and follow in their own, where a lost race costs one cheap cycle and nothing queues behind it. The three contended paths are now byte-identical to main at ca54b2e; what remains is five files no other open PR touches. The evidence stands unchanged: run 33823344142 on d3b9dda was green on every job, and the emitter, tests and mirrors it exercised are exactly what is left here. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012imgm3QzXAT6GBn3ifTCDd --------- Co-authored-by: gunbc-ci-auto-heal <gunbc-ci-auto-heal@users.noreply.github.com> Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
1 task
briansrls
pushed a commit
that referenced
this pull request
Sep 8, 2026
…erent (#10704) * Restore wet_route_model_lags_seed_stall so two 4b triggers have a referent. The failure-mode rows extracted in #10299 name this stall as the population their next-rung triggers retire, but the declaration never left #9725. A trigger whose cited home is missing is retired by nothing. Co-authored-by: Cursor <cursoragent@cursor.com> * Point wet_route_model_lags_seed_stall at a member that resolves. The restored row still named floor_wet_route symbols from #9725, which never landed, so the two citing 4b triggers would have retired against another missing population. The remaining live member is closure_content_digest. Co-authored-by: Cursor <cursoragent@cursor.com> * Split the two citing 4b triggers onto their own populations. wet_route_model_lags_seed_stall now covers only closure_content_digest. subject_and_its_digest_as_independent_parameters no longer retires against that stall, which would have gone green while peer-parameter signatures stayed writable. Co-authored-by: Cursor <cursoragent@cursor.com> * Recover the stall from #9725 and keep warm-moth-142's fabric-M0 receipt. The row text is 3281582, not a re-derivation; climbs_when only wraps the recovered capability. The peer-parameter file now carries moth's second receipt in full plus a discharge of the missing-row half. Co-authored-by: Cursor <cursoragent@cursor.com> * Leave warm-moth-142's fabric-M0 receipt off this PR. They land second: rebase onto this root fix, drop the stale missing-row paragraph, and rewrite against the corrected trigger. Shipping their un-rewritten receipt here would land the paragraph they already plan to delete. Co-authored-by: Cursor <cursoragent@cursor.com> * Keep moth's fabric-M0 receipt and discharge its three false live clauses. The instance, path-vs-hex finding, and not-repointable conclusion stay. The missing-row claims are named as discharged: the stall is restored and the first-receipt trigger no longer retires against it as a whole. Co-authored-by: Cursor <cursoragent@cursor.com> * Regenerate docs/design-failure-modes.md from the merged authority. main_wet_one ran locally; identity join against origin/main lost=0. The projection greps for the split trigger and the fabric-m0 instance. Co-authored-by: Cursor <cursoragent@cursor.com> * Drop the retired param_names argument from callees_from_node call sites. The function no longer declares that parameter, and the floor parse phase refused the whole lane on the three leftover named arguments. Co-authored-by: Cursor <cursoragent@cursor.com> * Enroll OutsideModeledGuarantee stamp citations as trigger sites, not wall tests. Three production stamps belong on PLANTED_CONTROL_CITATIONS because the roster reds on the same event as the stamp. The witness probe stays false on purpose with the other fixture-carrier exemptions. Co-authored-by: Cursor <cursoragent@cursor.com> * Give next-rung trigger citations their own roster and diagnostic kind. PLANTED-CONTROL-RESOLVES still means a lost control. A resolving OutsideModeledGuarantee required_capability is the stamp firing, which needs a different name, kind, and message. Drop the stall-absent closing from the fabric-m0 receipt so the receipts list does not answer twice. Co-authored-by: Cursor <cursoragent@cursor.com> * Regenerate design-failure-modes.md from the updated receipts. The projection still carried the stall-absent closing and the three-clause discharge after the authority dropped both. Co-authored-by: Cursor <cursoragent@cursor.com> * Share the spent-roster join; keep kind and message as parameters. Debt, planted-control, and next-rung trigger citations all ask which roster rows now resolve. Copying that fold minted a third authority for the same traversal. Co-authored-by: Cursor <cursoragent@cursor.com> * Take main's PLANTED_CONTROL_CITATIONS occupancy; drop the trigger roster. #10718 already decided the four-row enrollment. This branch no longer reverts that shape. Remaining work is the stall restore only. Co-authored-by: Cursor <cursoragent@cursor.com> * chore: regenerate drifted generated artifacts (ci auto-heal) Ledger-Repair-Judged: docs/design-failure-modes.md Ledger-Rows-Repaired: docs/design-failure-modes.md subject_and_its_digest_as_independent_parameters Ledger-Repair-Judged: docs/design-rung-drops.md * chore: regenerate drifted generated artifacts (ci auto-heal) Ledger-Repair-Judged: docs/design-failure-modes.md Ledger-Rows-Repaired: docs/design-failure-modes.md subject_and_its_digest_as_independent_parameters Ledger-Repair-Judged: docs/design-rung-drops.md --------- Co-authored-by: gunbc-ci-auto-heal <gunbc-ci-auto-heal@users.noreply.github.com> Co-authored-by: Cursor <cursoragent@cursor.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Six
RecurringFailureModerows authored on #9725, moved here so they are readable. They are gated on that PR only by the accident of which branch the author was on — none has anything to do with the wet route, and #9725 cannot go green tonight because its receipt's semantic subject has moved out from under it. A row nobody can read is not filed.Moved, not rewritten. Each row's authored bytes are carried verbatim into its own file in the split shape #10206 established, as a single
receiptsentry. Nothing was improved in transit — the justification for landing this quickly is that nothing in it is a judgement.per_argument_exhaustive_matrix_blind_to_a_cross_argument_relationclosure_subject_with_no_enumerable_membershipceiling_never_exercised_for_a_population_the_census_cannot_plancontent_digest_makes_annotations_semantically_load_bearingsubject_and_its_digest_as_independent_parametersa_written_row_is_not_a_firing_mechanismRoster entries appended at the end, unsorted: roster order is source order and the projection renders in it, so sorting would reorder the document and destroy the empty-diff oracle the split was checked against.
Verified in the direction that catches loss, not only duplication —
LINES == DISTINCTcannot see a row a migration silently dropped:main-not-mineempty;mine-not-mainexactly these sixgit merge-tree --write-tree origin/main HEADproduces a tree, not a driver refusalThe same six are deleted from #9725 in the same motion, so two branches never declare one identity — that is the duplicate-declaration state that took
maindown for two hours tonight, and a deliberate deletion is cheaper than the same deletion discovered later as a conflict.🤖 Generated with Claude Code
https://claude.ai/code/session_01GvVoivi7L449wbh6rjeJY4