To-be 56 names ADR 0326 and ADR 0329 in decisions, and:
the lead and the acceptance say the module owns the whole configuration and shows the operator's own values, never a credential;
§1 lists the settings files, the composed user.js, the lists written only while Betterbird is closed, and each declared account's filter file;
§6 (what the tools say) is rewritten for ADR 0329 §1–§3 and §12, replacing "never an address, a host or a name";
a new §8 describes prefs as one entry per preference (the operator's setup out of the catalogue), the five settings and their matching, trust (tags alone untrusted), adoption, filters owned whole, closed-only writes, and derive-and-propose one layer at a time;
How it is checked adds the controller's, the configuration's and the live checks.
Sections 1 to 7 keep their numbers.
Why
Issue 586 and ADR 0329 (accepted 2026-10-12) change what the betterbird module owns and shows; the design said the opposite.
How it was tested
sh merge-check.sh: records, index, cycle and words pass.
Part of issue 586 (with novox/mesh-catalog and novox/mesh-controller branches of the same issue).
## What
To-be 56 names ADR 0326 and ADR 0329 in `decisions`, and:
- the lead and the acceptance say the module owns the whole configuration and shows the operator's own values, never a credential;
- §1 lists the settings files, the composed user.js, the lists written only while Betterbird is closed, and each declared account's filter file;
- §6 (what the tools say) is rewritten for ADR 0329 §1–§3 and §12, replacing "never an address, a host or a name";
- a new §8 describes `prefs` as one entry per preference (the operator's setup out of the catalogue), the five settings and their matching, trust (tags alone untrusted), adoption, filters owned whole, closed-only writes, and derive-and-propose one layer at a time;
- How it is checked adds the controller's, the configuration's and the live checks.
Sections 1 to 7 keep their numbers.
## Why
Issue 586 and ADR 0329 (accepted 2026-10-12) change what the betterbird module owns and shows; the design said the opposite.
## How it was tested
`sh merge-check.sh`: records, index, cycle and words pass.
Part of issue 586 (with novox/mesh-catalog and novox/mesh-controller branches of the same issue).
The design said the tools answer no address and declared only the accounts. It
now names ADR 0329 and ADR 0326, the five new settings and prefs as one entry per
preference (§8), what the tools show (§6), and how each part is checked.
Deliverynovox/hq@c192576f6b2f — delivered since 2026-10-13T00:22:54Z
Delivery plan — builds nothing: the change touches no module of the mesh's graph
Transitions
2026-10-13 00:03 (new) → proposed (announced): novox/hq#336's head announced
2026-10-13 00:04 proposed → checked (checked): a verdict names this commit and the trunk's head as its trunk commit
2026-10-13 00:04 checked → ready (accepted): the gate passed or warned, and the repository's own check did not fail
2026-10-13 00:22 ready → published (merged): on the trunk its modules follow: merged there, and its walk opened — or nothing for a walk to move
2026-10-13 00:22 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@c192576f6b2f` — **delivered** since 2026-10-13T00:22:54Z
**Delivery plan** — builds nothing: the change touches no module of the mesh's graph
**Transitions**
- 2026-10-13 00:03 (new) → proposed (announced): novox/hq#336's head announced
- 2026-10-13 00:04 proposed → checked (checked): a verdict names this commit and the trunk's head as its trunk commit
- 2026-10-13 00:04 checked → ready (accepted): the gate passed or warned, and the repository's own check did not fail
- 2026-10-13 00:22 ready → published (merged): on the trunk its modules follow: merged there, and its walk opened — or nothing for a walk to move
- 2026-10-13 00:22 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`.
Design 56 matches mesh-catalog #262 at 1835f0efad0f9d9b079ce54af35c3fba47e60221. The files table matches the manifest: the defaults file holds no value of the operator's, there is one mergeable file per group naming its keys, the user.js is composed by the tools, and the prefs.js lists and filter files are written only while Betterbird is closed. Section 6's tool table matches the 17 tools. Section 8 matches the code on these points: prefs as one entry per preference with the four refused prefixes, the five settings with key only in a node's layer and never on tags, tags alone untrusted, adoption, filters owned whole with the kept copy, writing only while closed, and deriving one layer at a time that carries the layer whole. It cites ADR 0326 and ADR 0329, which the frontmatter now lists.
The document names no node, host, address or absolute path. A search for the node names, the domain and private ranges found nothing. Its words are the glossary's: desktop node, layer, seat-free, the controller, the operator.
./merge-check.sh passes with exit 0, run with MESH_CATALOG_DIR set to the pull request's catalogue. records: all checks passed. index: 6 topics, no stored list. cycle: 436 documents, the chain holds. words: none in running prose.
NOT BLOCKING
Section 8 says the waiting writes happen at 'the next start from the launcher'. betterbird_restart writes them too, as ADR 0329 §9 says. Adding it would match §9's own words.
'tags ... is set through the settings verb as well', and the controller's test line, hold only once mesh-controller #286 merges. Until then, the controller keeps the whole betterbird layer for the operator.
The design does not say how the operator's 21 preferences reach the mesh-wide layer without a window of Betterbird's own defaults. My review of mesh-catalog #262 describes an order that avoids the window: propose the layer through the settings verb after the merge and before the push. If that order is adopted, one sentence here would record it.
STATUSES AT THIS HEAD (gitea_commit_statuses): mesh/repo-check success, 'head c192576f6b merged into trunk b236b306c4 — pass: its merge-check.sh passed'; mesh/merge-gate success, 'pass: the change touches no module of the mesh's graph'; mesh/delivery success, 'ready: it delivers once merged'. The combined state is success, and the forge says it is mergeable.
Verdict: approve at c192576f6b2fb0f2170447e26a1067452c3e3c9d
I found no blocking finding.
Design 56 matches mesh-catalog #262 at 1835f0efad0f9d9b079ce54af35c3fba47e60221. The files table matches the manifest: the defaults file holds no value of the operator's, there is one mergeable file per group naming its keys, the user.js is composed by the tools, and the prefs.js lists and filter files are written only while Betterbird is closed. Section 6's tool table matches the 17 tools. Section 8 matches the code on these points: `prefs` as one entry per preference with the four refused prefixes, the five settings with `key` only in a node's layer and never on tags, tags alone untrusted, adoption, filters owned whole with the kept copy, writing only while closed, and deriving one layer at a time that carries the layer whole. It cites ADR 0326 and ADR 0329, which the frontmatter now lists.
The document names no node, host, address or absolute path. A search for the node names, the domain and private ranges found nothing. Its words are the glossary's: desktop node, layer, seat-free, the controller, the operator.
./merge-check.sh passes with exit 0, run with MESH_CATALOG_DIR set to the pull request's catalogue. records: all checks passed. index: 6 topics, no stored list. cycle: 436 documents, the chain holds. words: none in running prose.
NOT BLOCKING
1. Section 8 says the waiting writes happen at 'the next start from the launcher'. betterbird_restart writes them too, as ADR 0329 §9 says. Adding it would match §9's own words.
2. 'tags ... is set through the settings verb as well', and the controller's test line, hold only once mesh-controller #286 merges. Until then, the controller keeps the whole betterbird layer for the operator.
3. The design does not say how the operator's 21 preferences reach the mesh-wide layer without a window of Betterbird's own defaults. My review of mesh-catalog #262 describes an order that avoids the window: propose the layer through the settings verb after the merge and before the push. If that order is adopted, one sentence here would record it.
STATUSES AT THIS HEAD (gitea_commit_statuses): mesh/repo-check success, 'head c192576f6b2f merged into trunk b236b306c42a — pass: its merge-check.sh passed'; mesh/merge-gate success, 'pass: the change touches no module of the mesh's graph'; mesh/delivery success, 'ready: it delivers once merged'. The combined state is success, and the forge says it is mergeable.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
What
To-be 56 names ADR 0326 and ADR 0329 in
decisions, and:prefsas one entry per preference (the operator's setup out of the catalogue), the five settings and their matching, trust (tags alone untrusted), adoption, filters owned whole, closed-only writes, and derive-and-propose one layer at a time;Sections 1 to 7 keep their numbers.
Why
Issue 586 and ADR 0329 (accepted 2026-10-12) change what the betterbird module owns and shows; the design said the opposite.
How it was tested
sh merge-check.sh: records, index, cycle and words pass.Part of issue 586 (with novox/mesh-catalog and novox/mesh-controller branches of the same issue).
Delivery
novox/hq@c192576f6b2f— delivered since 2026-10-13T00:22:54ZDelivery plan — builds nothing: the change touches no module of the mesh's graph
Transitions
The commit's note under
refs/notes/mesh-plankeeps every transition:git log --notes=mesh-plan.Verdict: approve at
c192576f6bI found no blocking finding.
Design 56 matches mesh-catalog #262 at 1835f0efad0f9d9b079ce54af35c3fba47e60221. The files table matches the manifest: the defaults file holds no value of the operator's, there is one mergeable file per group naming its keys, the user.js is composed by the tools, and the prefs.js lists and filter files are written only while Betterbird is closed. Section 6's tool table matches the 17 tools. Section 8 matches the code on these points:
prefsas one entry per preference with the four refused prefixes, the five settings withkeyonly in a node's layer and never on tags, tags alone untrusted, adoption, filters owned whole with the kept copy, writing only while closed, and deriving one layer at a time that carries the layer whole. It cites ADR 0326 and ADR 0329, which the frontmatter now lists.The document names no node, host, address or absolute path. A search for the node names, the domain and private ranges found nothing. Its words are the glossary's: desktop node, layer, seat-free, the controller, the operator.
./merge-check.sh passes with exit 0, run with MESH_CATALOG_DIR set to the pull request's catalogue. records: all checks passed. index: 6 topics, no stored list. cycle: 436 documents, the chain holds. words: none in running prose.
NOT BLOCKING
Section 8 says the waiting writes happen at 'the next start from the launcher'. betterbird_restart writes them too, as ADR 0329 §9 says. Adding it would match §9's own words.
'tags ... is set through the settings verb as well', and the controller's test line, hold only once mesh-controller #286 merges. Until then, the controller keeps the whole betterbird layer for the operator.
The design does not say how the operator's 21 preferences reach the mesh-wide layer without a window of Betterbird's own defaults. My review of mesh-catalog #262 describes an order that avoids the window: propose the layer through the settings verb after the merge and before the push. If that order is adopted, one sentence here would record it.
STATUSES AT THIS HEAD (gitea_commit_statuses): mesh/repo-check success, 'head
c192576f6bmerged into trunkb236b306c4— pass: its merge-check.sh passed'; mesh/merge-gate success, 'pass: the change touches no module of the mesh's graph'; mesh/delivery success, 'ready: it delivers once merged'. The combined state is success, and the forge says it is mergeable.