Say in to-be 27 how a value several modules share is set once (issue 415, ADR 0309, ADR 0319) #327

Merged
mesh-admin merged 1 commits from design/415-shared-value-trust into main 2026-10-12 11:54:02 +00:00
Contributor

Issue 415. In to-be 27, one paragraph now carries both ADR 0309 (a module's setting left unset reads the mesh-wide value of a shared name before its own default) and ADR 0319 (a shared value reaches no trusted setting; it changes ADR 0164 §4 and §9 and ADR 0309 §2, §3, §5 and §6). It replaces the ADR 0309 paragraph that #324 added, and #324's ADR 0312 paragraph stays as it is. ADR 0319 is added to the one decisions list, beside 0312 and 0309.

Where the records stand: ADR 0319 was accepted by the knowledge base's review at 11:44 UTC on 2026-10-12, at version 10. ADR 0309's home insight already names to-be 27 and was accepted at version 13.

Rebased onto main (d0e23b1). merge-check.sh passes locally.

Issue 415. In to-be 27, one paragraph now carries both ADR 0309 (a module's setting left unset reads the mesh-wide value of a shared name before its own default) and ADR 0319 (a shared value reaches no trusted setting; it changes ADR 0164 §4 and §9 and ADR 0309 §2, §3, §5 and §6). It replaces the ADR 0309 paragraph that #324 added, and #324's ADR 0312 paragraph stays as it is. ADR 0319 is added to the one `decisions` list, beside 0312 and 0309. Where the records stand: ADR 0319 was accepted by the knowledge base's review at 11:44 UTC on 2026-10-12, at version 10. ADR 0309's home insight already names to-be 27 and was accepted at version 13. Rebased onto main (d0e23b1). `merge-check.sh` passes locally.
Author
Contributor

Delivery novox/hq@e8cff1646a9a — delivered since 2026-10-12T11:54:27Z

Delivery plan — builds nothing: the change touches no module of the mesh's graph

Transitions

  • 2026-10-12 11:50 (new) → proposed (announced): novox/hq#327's head announced
  • 2026-10-12 11:52 proposed → checked (checked): the verdict names this commit
  • 2026-10-12 11:52 checked → ready (accepted): the gate passed or warned, and the repository's own check did not fail
  • 2026-10-12 11:54 ready → published (merged): on the trunk its modules follow: merged there, and its walk opened — or nothing for a walk to move
  • 2026-10-12 11:54 published → delivered (done): nothing for a walk to move

The commit's note under refs/notes/mesh-plan keeps every transition: git log --notes=mesh-plan.

<!-- mesh-delivery:view --> **Delivery** `novox/hq@e8cff1646a9a` — **delivered** since 2026-10-12T11:54:27Z **Delivery plan** — builds nothing: the change touches no module of the mesh's graph **Transitions** - 2026-10-12 11:50 (new) → proposed (announced): novox/hq#327's head announced - 2026-10-12 11:52 proposed → checked (checked): the verdict names this commit - 2026-10-12 11:52 checked → ready (accepted): the gate passed or warned, and the repository's own check did not fail - 2026-10-12 11:54 ready → published (merged): on the trunk its modules follow: merged there, and its walk opened — or nothing for a walk to move - 2026-10-12 11:54 published → delivered (done): nothing for a walk to move The commit's note under `refs/notes/mesh-plan` keeps every transition: `git log --notes=mesh-plan`.
Author
Contributor

Verdict: request changes at a5dbc0945751ac047826f8a80ff48058a5d228e4

Content against ADR 0309 and ADR 0319 (both accepted): the new to-be 27 paragraph states all five things the records decide — the shared name (desktop.font-size example), the resolution order (default, then shared value, then the module's mesh-wide layer, then the node's, last winning), the shared value as a fourth place of change beside ADR 0164's three, that no trusted setting may declare or read a shared name (and registration refuses one that does), and that typing is left open to issues 346/410. Frontmatter (decisions: ADR 0319, ADR 0309, updated: 2026-10-12) is correct, the addition is prose only, it stays inside the public-repo rules (no node names or addresses), and it uses established glossary words (preference, trusted setting, warrant).

Two problems, both blocking:

  1. Real conflict with #324, not just a stale description. #324 merged into main first (d0e23b1) and already added its own ADR 0309 paragraph plus decisions: ADR 0309/ADR 0312 to this same file, in the same place. This branch's merge-base with main (ca9d11e) predates that merge, so it is stale. I trial-merged origin/design/415-shared-value-trust onto origin/main in an isolated worktree: git merge reports CONFLICT (content): Merge conflict in 03-DESIGN/01-to-be/27-a-module-requires-the-mesh-resolves.md, with a frontmatter conflict (ADR 0312 vs ADR 0319 both inserted at the same point) and a body conflict where #324's merged ADR-0309-only paragraph sits head-to-head with this PR's fuller ADR-0309-and-0319 paragraph. The forge's own mergeability flag on the head (merge.mergeable: true from commit-statuses) does not reflect this — it looks stale against the current main tip, so it should not be trusted here. The branch needs rebasing onto current main, and the new paragraph should replace #324's now-merged one (which it supersedes) rather than sit beside it.

  2. The PR description is stale, as flagged: it still says "ADR 0319 is proposed in the knowledge base and under its review; merge only once it is accepted," and that ADR 0309's home insight is "re-proposed after that." ADR 0319 is accepted (knowledge-base, approved 2026-10-12T11:44:03Z) and ADR 0309's home insight already names to-be 27. Please update the description before merging so it doesn't tell a future reader to wait on something already done.

Checks: merge-check.sh passes on this PR's own head in isolation (confirmed in an isolated worktree: records/index/cycle/words all pass, matching the PR's own claim), and the forge's mesh/repo-check and mesh/merge-gate both read success — but both were computed against a base this branch no longer matches, so green here doesn't mean mergeable now. Please rebase, resolve the conflict by merging the two paragraphs' content, re-run merge-check.sh, and update the description, then I'll re-review.

Verdict: request changes at a5dbc0945751ac047826f8a80ff48058a5d228e4 Content against ADR 0309 and ADR 0319 (both accepted): the new to-be 27 paragraph states all five things the records decide — the shared name (`desktop.font-size` example), the resolution order (default, then shared value, then the module's mesh-wide layer, then the node's, last winning), the shared value as a fourth place of change beside ADR 0164's three, that no trusted setting may declare or read a shared name (and registration refuses one that does), and that typing is left open to issues 346/410. Frontmatter (`decisions: ADR 0319, ADR 0309`, `updated: 2026-10-12`) is correct, the addition is prose only, it stays inside the public-repo rules (no node names or addresses), and it uses established glossary words (preference, trusted setting, warrant). Two problems, both blocking: 1. **Real conflict with #324, not just a stale description.** #324 merged into main first (d0e23b1) and already added its own ADR 0309 paragraph plus `decisions: ADR 0309`/`ADR 0312` to this same file, in the same place. This branch's merge-base with main (ca9d11e) predates that merge, so it is stale. I trial-merged origin/design/415-shared-value-trust onto origin/main in an isolated worktree: `git merge` reports `CONFLICT (content): Merge conflict in 03-DESIGN/01-to-be/27-a-module-requires-the-mesh-resolves.md`, with a frontmatter conflict (ADR 0312 vs ADR 0319 both inserted at the same point) and a body conflict where #324's merged ADR-0309-only paragraph sits head-to-head with this PR's fuller ADR-0309-and-0319 paragraph. The forge's own mergeability flag on the head (`merge.mergeable: true` from commit-statuses) does not reflect this — it looks stale against the current main tip, so it should not be trusted here. The branch needs rebasing onto current main, and the new paragraph should replace #324's now-merged one (which it supersedes) rather than sit beside it. 2. **The PR description is stale**, as flagged: it still says "ADR 0319 is proposed in the knowledge base and under its review; merge only once it is accepted," and that ADR 0309's home insight is "re-proposed after that." ADR 0319 is accepted (knowledge-base, approved 2026-10-12T11:44:03Z) and ADR 0309's home insight already names to-be 27. Please update the description before merging so it doesn't tell a future reader to wait on something already done. Checks: `merge-check.sh` passes on this PR's own head in isolation (confirmed in an isolated worktree: records/index/cycle/words all pass, matching the PR's own claim), and the forge's `mesh/repo-check` and `mesh/merge-gate` both read success — but both were computed against a base this branch no longer matches, so green here doesn't mean mergeable now. Please rebase, resolve the conflict by merging the two paragraphs' content, re-run merge-check.sh, and update the description, then I'll re-review.
jschoubben added 1 commit 2026-10-12 11:50:36 +00:00
Carry ADR 0309 and ADR 0319 in one paragraph of to-be 27, replacing the 0309 paragraph alone (issue 415)
mesh/merge-gate pass: the change touches no module of the mesh's graph
mesh/repo-check pass: its merge-check.sh passed
mesh/delivery delivered
e8cff1646a
jschoubben force-pushed design/415-shared-value-trust from a5dbc09457 to e8cff1646a 2026-10-12 11:50:36 +00:00 Compare
Author
Contributor

Verdict: approve at e8cff1646a

Re-review against my comment 10200. Both problems there are fixed:

  1. Merges cleanly now. I fetched origin/main (d0e23b1) and origin/design/415-shared-value-trust (e8cff16) and merged in an isolated worktree: fast-forward, no conflict. The branch now carries #324's ADR 0312 paragraph unchanged, and the 0309+0319 paragraph replaces #324's ADR-0309-only paragraph in the same spot, with one decisions list (ADR 0312, ADR 0319, ADR 0309, …) instead of two competing insertions.

  2. The description is accurate now: it says ADR 0319 was accepted at version 10 (11:44 UTC 2026-10-12) and that ADR 0309's home insight already names to-be 27, and it says what replaced what.

Content still matches both records: the paragraph states the shared name, the resolution order (default, shared value, mesh-wide layer, node, last winning), the shared value as ADR 0164's fourth place of change, that no trusted setting may declare or read a shared name, and that typing stays open to issues 346/410. It cites both records' changed clauses correctly (0309 changing ADR 0164 §7/§8 and ADR 0262 §3/§7/§8/§10; 0319 changing ADR 0164 §4/§9 and ADR 0309 §2/§3/§5/§6).

merge-check.sh passes on the merged tree in the isolated worktree (records, index, cycle, words all pass). The forge's mesh/repo-check now reports fresh (created 11:52:12, after this push): pass, its merge-check.sh passed; mesh/merge-gate also passes; merge.mergeable: true with nothing blocking against current main.

Verdict: approve at e8cff1646a9a4adda0a17720b50e968578ae1d05 Re-review against my comment 10200. Both problems there are fixed: 1. Merges cleanly now. I fetched origin/main (d0e23b1) and origin/design/415-shared-value-trust (e8cff16) and merged in an isolated worktree: fast-forward, no conflict. The branch now carries #324's ADR 0312 paragraph unchanged, and the 0309+0319 paragraph replaces #324's ADR-0309-only paragraph in the same spot, with one `decisions` list (ADR 0312, ADR 0319, ADR 0309, …) instead of two competing insertions. 2. The description is accurate now: it says ADR 0319 was accepted at version 10 (11:44 UTC 2026-10-12) and that ADR 0309's home insight already names to-be 27, and it says what replaced what. Content still matches both records: the paragraph states the shared name, the resolution order (default, shared value, mesh-wide layer, node, last winning), the shared value as ADR 0164's fourth place of change, that no trusted setting may declare or read a shared name, and that typing stays open to issues 346/410. It cites both records' changed clauses correctly (0309 changing ADR 0164 §7/§8 and ADR 0262 §3/§7/§8/§10; 0319 changing ADR 0164 §4/§9 and ADR 0309 §2/§3/§5/§6). `merge-check.sh` passes on the merged tree in the isolated worktree (records, index, cycle, words all pass). The forge's `mesh/repo-check` now reports fresh (created 11:52:12, after this push): pass, its merge-check.sh passed; `mesh/merge-gate` also passes; `merge.mergeable: true` with nothing blocking against current main.
mesh-admin merged commit 361d503034 into main 2026-10-12 11:54:02 +00:00
mesh-admin deleted branch design/415-shared-value-trust 2026-10-12 11:54:03 +00:00
Sign in to join this conversation.
No Reviewers
2 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: novox/hq#327