Issue 356: the operator's command ran a stale engine, and a hand-over had no channel; ADR 0272 gains the operator's rule and words.py the daemons' lines #256

Merged
mesh-admin merged 4 commits from issues/356-the-operators-command-ran-a-stale-engine-and-a-hand-over-had-no-channel into main 2026-10-09 17:57:18 +00:00
Contributor

Registers issue 356 (status: located). Do not merge without review. Number 356 checked free across main and every open branch right before the commit (353 to 355 are on open branches).

Symptom (2026-10-09, 18:01 local, the control-node): the operator typed the node-engine's hand-over, as a module's condition told them to, and the engine on PATH answered "unexpected argument": /usr/bin/nox-mesh-host is the engine a person copied at installation, every delivered engine runs from the versions directory, and nothing ever renewed the file on PATH. The hand-over of issue 339 itself was a terminal act at the machine with no channel. The messenger module crash-looped meanwhile, its data directory used as found.

Located in the node-engine (the launcher's fallback never renewed; hand-over local only), the controller (no node hand-over; the condition's words), the catalogue's check (nothing held the texts to ADR 0272). The diagnosis weighs the two fixes for the stale command (a copy on proof, chosen; an exec-shim, rejected: the launcher runs that file as the engine while nothing is delivered, and which version runs is the launcher's policy) and gives the order of delivery (controller first, for the node's grant).

Fixes: mesh-host #74, mesh-controller #187, mesh-catalog #165 — each with the reviewers' checklist up front (every caller value validated, the trust-changing act terminal-only, every test failing with its check removed, fail closed), and the merge check passed locally on each (the controller's on the merge result with main).

ADR 0272, a dated progressive insight (the decision stands): the operator's rule of 2026-10-09 — the operator types one command, nox; the node-engine and the controller are daemons; every line the mesh addresses to the operator names a nox line; how it is checked (hq's words.py and the catalogue's checks/words); the engine's install-time subcommands documented as such; hand-over has its nox line.

words.py: the daemons' lines (DAEMON_LINES, --list lines, the catalogue's copy retired-lines compared with MESH_CATALOG_DIR), looked for in the same documents from 2026-10-09, code blocks and spans included, quotations taken out. It found eight lines on its first run: mesh-controller token issue four times in the lifecycle design (now nox token issue) and issue 343's quotations of the records (now in quotation marks); the operator's-machine design's mesh-controller node is nox node account <node> <login>. The glossary's nox entry says the rule; the checks' README documents it.

merge-check.sh passes locally (records, index, cycle, words with the catalogue's copy from mesh-catalog #165 compared).

Owed before the issue resolves: a mesh-lab replay (a core issue, ADR 0237).

Registers issue 356 (status: located). Do not merge without review. Number 356 checked free across main and every open branch right before the commit (353 to 355 are on open branches). **Symptom (2026-10-09, 18:01 local, the control-node):** the operator typed the node-engine's `hand-over`, as a module's condition told them to, and the engine on PATH answered "unexpected argument": `/usr/bin/nox-mesh-host` is the engine a person copied at installation, every delivered engine runs from the versions directory, and nothing ever renewed the file on PATH. The hand-over of issue 339 itself was a terminal act at the machine with no channel. The messenger module crash-looped meanwhile, its data directory used as found. **Located in** the node-engine (the launcher's fallback never renewed; `hand-over` local only), the controller (no `node hand-over`; the condition's words), the catalogue's check (nothing held the texts to ADR 0272). The diagnosis weighs the two fixes for the stale command (a copy on proof, chosen; an exec-shim, rejected: the launcher runs that file as the engine while nothing is delivered, and which version runs is the launcher's policy) and gives the order of delivery (controller first, for the node's grant). **Fixes:** mesh-host #74, mesh-controller #187, mesh-catalog #165 — each with the reviewers' checklist up front (every caller value validated, the trust-changing act terminal-only, every test failing with its check removed, fail closed), and the merge check passed locally on each (the controller's on the merge result with main). **ADR 0272, a dated progressive insight (the decision stands):** the operator's rule of 2026-10-09 — the operator types one command, `nox`; the node-engine and the controller are daemons; every line the mesh addresses to the operator names a `nox` line; how it is checked (hq's `words.py` and the catalogue's `checks/words`); the engine's install-time subcommands documented as such; `hand-over` has its `nox` line. **`words.py`:** the daemons' lines (`DAEMON_LINES`, `--list lines`, the catalogue's copy `retired-lines` compared with `MESH_CATALOG_DIR`), looked for in the same documents from 2026-10-09, code blocks and spans included, quotations taken out. It found eight lines on its first run: `mesh-controller token issue` four times in the lifecycle design (now `nox token issue`) and issue 343's quotations of the records (now in quotation marks); the operator's-machine design's `mesh-controller node` is `nox node account <node> <login>`. The glossary's `nox` entry says the rule; the checks' README documents it. `merge-check.sh` passes locally (records, index, cycle, words with the catalogue's copy from mesh-catalog #165 compared). **Owed before the issue resolves:** a mesh-lab replay (a core issue, ADR 0237).
mesh-admin added 1 commit 2026-10-09 16:37:16 +00:00
Issue 356: the operator's command ran a stale engine, and a hand-over had no channel
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 superseded: a newer head of the same pull request
858e5413fa
The engine on every node's PATH was the one a person copied at installation, a build behind every delivery; and the way out of issue 339 was a terminal act at the machine, which ADR 0272 and the operator's rule of the same day make impossible to type. The record, located in the engine, the controller and the catalogue's check, with the fixes' pull requests; ADR 0272 gains the rule as a dated insight: the operator types one command, nox, and every text addressed to the operator names a nox line. words.py now refuses a daemon's line, with --list lines for the catalogue's copy; the lifecycle and the operator's-machine designs and issue 343's quotations are reworded.
Author
Contributor

Delivery novox/hq@2cb779d06f85 — delivered since 2026-10-09T17:57:38Z

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

Transitions

  • 2026-10-09 17:47 (new) → proposed (announced): novox/hq#256's head announced
  • 2026-10-09 17:47 proposed → checked (checked): the verdict names this commit
  • 2026-10-09 17:47 checked → ready (accepted): the gate passed or warned, and the repository's own check did not fail
  • 2026-10-09 17:57 ready → published (merged): on the trunk its modules follow: merged there, and its walk opened — or nothing for a walk to move
  • 2026-10-09 17:57 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@2cb779d06f85` — **delivered** since 2026-10-09T17:57:38Z **Delivery plan** — builds nothing: the change touches no module of the mesh's graph **Transitions** - 2026-10-09 17:47 (new) → proposed (announced): novox/hq#256's head announced - 2026-10-09 17:47 proposed → checked (checked): the verdict names this commit - 2026-10-09 17:47 checked → ready (accepted): the gate passed or warned, and the repository's own check did not fail - 2026-10-09 17:57 ready → published (merged): on the trunk its modules follow: merged there, and its walk opened — or nothing for a walk to move - 2026-10-09 17:57 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`.
jschoubben added 1 commit 2026-10-09 17:46:00 +00:00
Issue 356: the subject proves nothing, the signature does; the apply refuses a link at use
mesh/delivery superseded: a newer head of the same pull request
mesh/merge-gate pass: the change touches no module of the mesh's graph
mesh/repo-check pass: its merge-check.sh passed
19776115ef
The review of 2026-10-09 proved a responder can reach the hand-over's subject through a reply, and that the apply followed a link the hand-over had refused. The record no longer says only the controller may publish there: it says the ask is signed with the mesh's key and verified by the engine, and that a handed-over directory is changed through a descriptor that followed no link. To-be 45's writers table gains the row for a hand-over asked of a machine.
jschoubben added 1 commit 2026-10-09 17:46:10 +00:00
Merge remote-tracking branch 'origin/main' into issues/356-the-operators-command-ran-a-stale-engine-and-a-hand-over-had-no-channel
mesh/merge-gate pass: the change touches no module of the mesh's graph
mesh/repo-check fail: its merge-check.sh failed: words: 2 problem(s)
mesh/delivery superseded: a newer head of the same pull request
d1b42b7430
jschoubben added 1 commit 2026-10-09 17:46:37 +00:00
Issue 354: quote the condition's old words, which the daemons' lines check refuses as a line to type
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
2cb779d06f
Author
Contributor

Merge check at d1b42b74

  • mesh/merge-gate: PASS — the change touches no module of the mesh's graph
  • mesh/repo-check: FAIL — its merge-check.sh failed: words: 2 problem(s)

Run by the build seat on g14 as build-1791567996005526046 (builds --log build-1791567996005526046).

--- its merge-check.sh (go toolchain)
records: 259 decision records checked
records: all checks passed
index: 6 topics, 259 records, no stored list
cycle: 721 documents checked, the chain holds
NOT COMPARED: MESH_CATALOG_DIR is not set, so the catalogue's copy of the list was not read
04-ISSUES/354-a-decided-rename-was-never-scheduled-and-the-operator-believed-it-live/00-report.md:32: `mesh-host hand-over` is a daemon's line — the operator types `nox …` (ADR 0272)
04-ISSUES/354-a-decided-rename-was-never-scheduled-and-the-operator-believed-it-live/01-diagnosis.md:32: `mesh-host hand-over` is a daemon's line — the operator types `nox …` (ADR 0272)
words: 2 problem(s)
**Merge check** at `d1b42b74` - `mesh/merge-gate`: **PASS** — the change touches no module of the mesh's graph - `mesh/repo-check`: **FAIL** — its merge-check.sh failed: words: 2 problem(s) Run by the build seat on g14 as `build-1791567996005526046` (`builds --log build-1791567996005526046`). ``` --- its merge-check.sh (go toolchain) records: 259 decision records checked records: all checks passed index: 6 topics, 259 records, no stored list cycle: 721 documents checked, the chain holds NOT COMPARED: MESH_CATALOG_DIR is not set, so the catalogue's copy of the list was not read 04-ISSUES/354-a-decided-rename-was-never-scheduled-and-the-operator-believed-it-live/00-report.md:32: `mesh-host hand-over` is a daemon's line — the operator types `nox …` (ADR 0272) 04-ISSUES/354-a-decided-rename-was-never-scheduled-and-the-operator-believed-it-live/01-diagnosis.md:32: `mesh-host hand-over` is a daemon's line — the operator types `nox …` (ADR 0272) words: 2 problem(s) ```
Author
Contributor

Review addressed — head 2cb779d (merged with main at b962e87).

  • Issue 356's report (§ How it is checked) and diagnosis (§ The fix) no longer say only the controller may publish the subject so the engine hears nobody else: they say the subject bounds who is granted the publish, a responder can reach it through a reply (the review's finding), the ask is signed with the mesh's key and verified by the engine (node, expiry, nonce), and the apply re-checks for a link at the moment of use through a descriptor that followed none.
  • To-be 45 §1's writers table gains the row "a hand-over asked of a machine" (the controller compiles the same row).
  • Merging main brought issue 354, whose two quotations of the condition's old words the daemons' lines check refused; they are now in quotation marks (one line each, so the check's quotation rule takes them).
  • merge-check.sh on the merge result: exit 0 (records, index, cycle, words with the catalogue's copy compared).
Review addressed — head 2cb779d (merged with main at b962e87). - Issue 356's report (§ How it is checked) and diagnosis (§ The fix) no longer say only the controller may publish the subject so the engine hears nobody else: they say the subject bounds who is granted the publish, a responder can reach it through a reply (the review's finding), the ask is signed with the mesh's key and verified by the engine (node, expiry, nonce), and the apply re-checks for a link at the moment of use through a descriptor that followed none. - To-be 45 §1's writers table gains the row "a hand-over asked of a machine" (the controller compiles the same row). - Merging main brought issue 354, whose two quotations of the condition's old words the daemons' lines check refused; they are now in quotation marks (one line each, so the check's quotation rule takes them). - `merge-check.sh` on the merge result: exit 0 (records, index, cycle, words with the catalogue's copy compared).
mesh-admin merged commit 0e7af9988a into main 2026-10-09 17:57:18 +00:00
mesh-admin deleted branch issues/356-the-operators-command-ran-a-stale-engine-and-a-hand-over-had-no-channel 2026-10-09 17:57:19 +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#256