Repository navigation
Declare the dashboard MergeReady semantic-admissibility rung drop - #10157
Merged
Merged
Conversation
Rebuilt from #10039, which carried two hunks against a derived surface that no longer exists: DESIGN.md now states it deliberately does not restate the drop names, and docs/design-ledgers.md was split into design-failure-modes.md and design-rung-drops.md. Only the authority edit was ever real. Projection regeneration is deliberately NOT in this commit: it lands with the regen once the docs/design-rung-drops.md window is clear. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01L9g69G7ZkiCXGCeUUJgo9D
gunbai-bot
Bot
changed the title
briansrls
pushed a commit
that referenced
this pull request
Sep 3, 2026
Ruled contender cleared. The conflict was positional, not semantic: this branch
converts rows from `authored:` to `declaration: AuthoredProse { .. }` while
#10157 appended a row in the same tail region, so git conflicted on adjacency
while nothing about the two changes disagreed.
One genuinely new row -- dashboard_merge_ready_semantic_admissibility --
absorbed at main's content byte-for-byte with one new LegacyProseIdentity arm,
which is the whole cost of a prose row arriving during this migration. The other
four rows on main's side of the hunk were already present in this branch,
converted; the spark paragraph was verified identical to the one already carried
rather than assumed.
MAIN HAD ALREADY ROSTERED THE NEW ROW in the auto-merged region, so adding it
again produced a DUPLICATE that the identity join would have refused --
`every_legacy_arm_is_claimed_exactly_once` fails on two claims of one arm as
readily as on zero. Caught by counting occurrences before regenerating rather
than by the witness, but it is the same question, and it is the second time
tonight this file's controls have named a defect I introduced by hand.
The projection was REGENERATED, never resolved: docs/design-rung-drops.md came
back GeneratedArtifactConcurrentDivergence -- unmerged with the ours bytes in
place and no conflict markers -- which is precisely the shape that would let a
picked side pass unnoticed. 28 sections, one per row.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VdJu3Xkr9PX3gdBqen9Cdn
gunbai-bot Bot
pushed a commit
that referenced
this pull request
Sep 3, 2026
#10154 split floor_cut's trigger into siblings because the original named five capabilities against an eight-capability loss -- the same class as this PR's own found-not-fixed item, a trigger asserting less than the capability it restores. Two lanes independently found instances of one class in one file on one night, which is evidence the class is common in this carrier rather than a coincidence about the file, and an argument FOR consolidating it. Absorbed mechanically: three new prose rows wrapped in AuthoredProse at main's content byte-for-byte with three new arms, and floor_cut itself re-wrapped after main rewrote its paragraph. Main had already rostered all three, so no roster inserts -- the duplicate trap from #10157, checked for this time rather than discovered. gunbc.guarantee_stall took main's authority_target_same_expression_equivalence row beside this branch's prose_declared_rung_drop_stall: two lanes adding different rows at the same position, both kept. Main's row was drafted as a DROP on #10154 and moved to the stall carrier on review, for the reason this PR exists -- a row denying the meaning of the carrier it sits in is a meaning fork, not a caveat. THE STALL'S POPULATION WAS RE-SYNCED, NOT LEFT TO DRIFT. It enumerates every prose identity, so four new prose rows made it silently incomplete -- a bounded population that has stopped being the population is the failure this row is about. Population and arm roster now agree exactly, 26 each, checked by set comparison rather than by count; the header sentence carries 26 instead of 21. Three defects of mine caught before pushing: a truncated closing brace that made the corpus unparseable, a roster insert whose anchor only matched once so two arms went unregistered, and a duplicated population member. Each was found by an identity or set check rather than by reading. All four identity assertions re-executed: green. Projection regenerated. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01VdJu3Xkr9PX3gdBqen9Cdn
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
1 participant
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.
Supersedes #10039, which is CONFLICTING and carried two hunks against a derived surface that no longer exists.
What this is
Three authored lines in
dag/gunbc/rung_drop.dag: thedashboard_merge_ready_semantic_admissibilityrow and its roster entry.Why #10039 could not simply be updated
#10039 carried three hunks. Two were dead:
DESIGN.md— that document now states it deliberately does not restate the drop names, because an index derived from the roster is a second projection of the same rows and made DESIGN.md regenerate on every ledger append.docs/design-ledgers.md— split on main intodocs/design-failure-modes.mdanddocs/design-rung-drops.md. The file the hunk edits does not exist.Only the authority edit was ever real. Rebuilt on current main rather than conflict-resolved, per the side-chat ruling that the obsolete generated-surface hunks be deleted from the candidate.
On the row's own evidence
The row cites
git merge-tree --write-tree origin/main cc1f3ff60returningrc=1beside a dashboardready=true. That direction is sound and deliberately so:rc=1proves a textual conflict exists. The converse would not be sound —merge-treedoes not invoke this repository'sgenerated-artifactmerge driver, so anrc=0from it is not evidence that a merge is clean. The row only ever uses the sound direction.Deliberately NOT in this commit
The regeneration of
docs/design-rung-drops.md. It lands with the regen in one commit once that path's window is clear — #10130 and #10154 both hold live regens of the ledger projections, and a third concurrent one would invalidate one of theirs.Draft until then: a head carrying an authority edit without its projection fails the generated-artifact gate, and it should, so there is nothing here for CI to usefully say yet.