1308 Commits

This Branch

This Branch

All Branches

Author	SHA1	Message	Date
jochen	7628b0e5b9	mesh-knowledge: give the bus test's bundle a fixed clock, so it passes on any day (issue 540) ... The bundle dated a reservation from the real day while the proposal kept the fixture's 2026-10-11, so the date rule refused it on every later day.	2026-10-12 12:23:38 +02:00
mesh-admin	b707f27911	Merge pull request 'gitea: an agent's commit on a branch of its own, and the reads around it (issue 375)' (#233) from fix/375-a-gitea-commit-tool into main	2026-10-12 10:14:05 +00:00
jochen	9eb2ea8d69	gitea: let an agent put a commit on a branch of its own, and read history, compares, reviews, directories and code (issue 375) ... An agent could not get a finished change onto the forge without someone pushing by hand, so the branch / pull request / approval rule could not be followed. A Go bundle serves gitea_commit, held to gitea_api's rule (never the default branch, never a protected one), and the summarised reads agents otherwise made through gitea_api as raw JSON.	2026-10-12 12:06:23 +02:00
mesh-admin	da296cacc6	Merge pull request 'mesh-delivery: give a waiting walk its word or say what holds it (issues 529, 362)' (#229) from fix/529-a-waiting-walk-is-answered-or-named into main	2026-10-12 09:55:20 +00:00
jschoubben	1de89cd702	Name a walk of nothing-to-move merges instead of letting it go (issue 529) ... The review showed the give-the-word rule skipped a failed or missing check and group order, and let go moves nothing judged; to-be 47 has no such path. Such a walk is now said, by stalled with walks, as held by nothing, why.	2026-10-12 11:33:51 +02:00
jschoubben	2bee70a478	Give a waiting walk its word or say what holds it (issues 529, 362) ... A merge whose plan sends nothing is delivered at once, yet the controller can still open a walk for it; no live delivery then held the walk and it waited for a person (the nzbgeek case). Give such a walk its word. Name each held delivery in the status, and let stalled say, for every waiting walk, what holds it, so the controller's plans line points at the person's decision instead of a hand start.	2026-10-12 11:24:30 +02:00
mesh-admin	6546c05f6e	Merge pull request 'mesh-issues: the issues are kept in the log changes, and the bucket is their fold (hq ADR 0300, issues 501, 502)' (#228) from adr-0300-issues-on-a-log into main	2026-10-12 08:46:10 +00:00
jochen	d413da203c	mesh-issues: resume an interrupted carry by revision and fault one issue instead of the whole start, so no kept value is lost and the tracker always starts (review of PR 228) ... An open's lost answer is judged by the issue's first entry; an entry that does not fold faults its issue on start; a value written beside the log is never folded over; one issue is kept in at most 252 KiB so its whole operation fits a log entry.	2026-10-12 10:35:57 +02:00
jochen	d33f3e3f71	mesh-issues: pin mesh-sdk at main 9747f3e, where the log and the history are both merged	2026-10-12 10:22:19 +02:00
jochen	dcc7510cd8	mesh-issues: carry each value the issues bucket kept, read through the runtime's history, so the log begins with all the bucket still knew (hq ADR 0300 §6, ADR 0307)	2026-10-12 10:20:51 +02:00
jochen	39da2cd793	mesh-issues: keep the issues bucket at 8 values until the carry can read them, so the bus drops nothing before it is carried (hq ADR 0300 §6)	2026-10-12 02:43:58 +02:00
jochen	983d434623	mesh-issues: keep every issue in the log changes and the bucket as its fold, so the record outlives the bucket's cap and a week of events (hq ADR 0300, issues 501, 502) ... Every write appends one operation at the issue's last sequence before the bucket is written; a refused append writes nothing. The start carries what the bucket holds, refolds what is behind and indexes every issue in memory, so list answers from memory with a ranked search and a before cursor.	2026-10-12 02:41:54 +02:00
mesh-admin	8a92a47bf7	Merge pull request 'Hold a leaf delivery to three minutes in times (hq ADR 0304 decision 4, issue 525)' (#227) from feat/525-leaf-budget-three-minutes into main	2026-10-12 00:02:25 +00:00
jschoubben	2711e812c8	Hold a leaf delivery to three minutes in times (hq ADR 0304 decision 4, issue 525) ... ADR 0304 lowered a leaf module's delivery budget from five minutes to three; times says each class against the controller's same budget.	2026-10-12 01:45:14 +02:00
mesh-admin	a0fb555f79	Merge pull request 'Read the tunnel through sudo where the tool runner is not root (hq issue 522)' (#226) from fix/522-tunnel-escalates into main	2026-10-11 22:07:54 +00:00
jschoubben	b1280f81a2	Read the tunnel through sudo where the tool runner is not root (hq issue 522) ... On novox the tool runner runs as the operator's account, and wg show answered 'Operation not permitted'. Root is the module's concern: the command runs as given when the process is root, else through sudo -n, which the sudo module allows the runtime's account, as nftables does.	2026-10-12 00:05:04 +02:00
mesh-admin	d2dff99bcf	Merge pull request 'Show the tunnel's peers as each machine sees them, and say on the bus when they move (hq issue 522)' (#225) from feat/522-tunnel-peers into main	2026-10-11 21:55:14 +00:00
jschoubben	c31e29f9a6	Keep watching the tunnel after a failed look, and remember peers' names	2026-10-11 23:51:40 +02:00
jschoubben	4fc769ed29	Show the tunnel's peers as each machine sees them, and say on the bus when they move (hq issue 522) ... ace's tunnel leg lost most packets while its own check called the tunnel healthy, and nothing on the mesh could say which address and port the hub saw for each peer, which is what tells a home router's port mix-up from anything else. netcheck_tunnel answers each peer's endpoint, last handshake and traffic, never a key, and netcheck now says when a peer's endpoint changes or its handshake goes stale or fresh again.	2026-10-11 23:50:17 +02:00
mesh-admin	11be8a548a	Merge pull request 'mesh-delivery: a stop ends only its own change on a shared walk (hq issue 459, ADR 0299)' (#221) from feat/459-a-stop-re-walks-the-others into main	2026-10-11 19:49:16 +00:00
mesh-admin	f7fdb5e87b	Merge pull request 'restic: nothing declared has no overdue night; a count's age is judged by awake time (issue 489)' (#224) from fix/489-nothing-declared-is-healthy into main	2026-10-11 19:31:05 +00:00
jschoubben	7f05fad58a	restic: judge no night overdue where nothing is declared, and a count's age by the time the machine was awake ... A machine that backs up nothing read as missing its night after 26 hours up, and a laptop's count, last written before it slept, read as stale on waking until the next look (novox/hq issue 489).	2026-10-11 21:10:11 +02:00
mesh-admin	fca2f92216	Merge pull request 'mesh-issues: reserve's description says it asks the knowledge base (issue 493)' (#223) from fix/493-reserve-says-what-it-reads into main	2026-10-11 19:07:27 +00:00
jschoubben	c0710e2db7	Say that reserve asks the knowledge base for the highest number, not hq (hq issue 493) ... Since #214 reserve reads knowledge_highest; its description still named hq's folders.	2026-10-11 20:53:21 +02:00
mesh-admin	3fa076829f	Merge pull request 'mesh-knowledge, claude-code: a long record is reviewed within a bound sized to it, and a proposal waiting on its author leaves the module healthy (hq issue 508)' (#222) from fix/508-a-long-record-is-reviewed into main	2026-10-11 18:41:53 +00:00
jschoubben	fcab25604e	Review a long record within a bound sized to it, and let a proposal waiting on its author leave the knowledge base healthy (hq issue 508) ... A judgement over one bus call's 30 seconds is collected by its ticket from claude-code, so an eleven-clause record can be reviewed; a proposal out of review tries is told to the operator, not raised as the module's fault, so it no longer holds the module's own rollouts.	2026-10-11 20:32:33 +02:00
jschoubben	3d7f9a3e66	mesh-delivery: say a shared-module leave-out truly, hold back only what was stopped by an act, and say a held walk where it is read (review of #221) ... A merge left out for sharing a module read as containing the stopped merge; the safety net also held back merges stranded by their walk, which nobody stopped; a held walk said why only in the log; and a stop of a merge in a batch did not ask the batch.	2026-10-11 20:27:55 +02:00
jschoubben	8d7f60c9e2	mesh-delivery: never ship a stopped merge walked again, and stop only once the walk is (review of #221, hq issue 459) ... A merge walked again into a batch and then stopped by its group stayed in that batch, which was later let go. Any walk or batch that names a stopped delivery's merge is now asked to leave it out and is never let go; a refused leave-out stops the whole walk. A person's stop marks the delivery stopped only after its walk is, and the answer says what stays on the machines and what will still carry the change.	2026-10-11 20:16:37 +02:00
jschoubben	aeab5b80b4	mesh-delivery: a stop ends only its own change on a shared walk (hq issue 459, ADR 0299) ... A group stop ended the stopped member's walk, and with it every other delivery the batch put on that walk. It now asks delivery-stop to leave the stopped merges out, so the others are walked again; a person's stop on a shared walk first says what ending the walk would take down and acts only on a way.	2026-10-11 19:59:43 +02:00
mesh-admin	977e7a6fd2	Merge pull request 'mesh-knowledge: a review owed says when it is due, survives a restart, and its last try is told to the operator (hq issue 497)' (#219) from fix/497-a-review-owed-outlives-a-restart into main	2026-10-11 17:49:34 +00:00
jschoubben	c394ecd9fd	mesh-knowledge: give a never-reviewed notice its resolved line when raised, and test a refused clear and local time in a set zone (review of #219) ... The operator's channel edits a cleared message with the resolved line it was raised with, so one sent only with the clear was never shown.	2026-10-11 19:41:45 +02:00
jschoubben	d45761678a	mesh-knowledge: clear a never-reviewed notice once the record moves on, key it by version and run of tries, say times locally (review of #219) ... A notice left open kept telling the operator a record was proposed and not in force after it was proposed again or accepted, and a later version's notice merged into it.	2026-10-11 19:36:14 +02:00
mesh-admin	9f63231db9	Merge pull request 'mesh-issues: take every condition key the controller raises and every delivery id mesh-delivery gives (hq issue 475)' (#218) from fix/475-a-delivery-condition-can-be-tracked into main	2026-10-11 17:28:34 +00:00
jschoubben	98d494619f	mesh-issues: take a delivery's id as mesh-delivery forms it, owner/repo@commit (hq issue 475) ... The delivery link's own pattern refused the slash and the @ every delivery id carries, so no delivery could be linked either. Mirror mesh-delivery's IDOf.	2026-10-11 19:24:23 +02:00
jschoubben	9cfb5f9edb	mesh-knowledge: keep when an owed review is due, say it in the history, and tell the operator once after its last try (hq issue 497) ... A failed review's try count was already kept with the version, but not when it was asked again, so a 10-minute pause read as a review lost to a restart; and after the last try only the history and the health said so.	2026-10-11 19:24:07 +02:00
jschoubben	7714f164c3	mesh-issues: take every condition key the controller raises, a delivery's with its slash among them (hq issue 475) ... The tracker's own pattern refused the slash a delivery's id carries, so no issue could track a delivery's condition and the operator's alert kept saying no issue tracks it. Mirror the controller's key rule instead of a pattern of our own; a router test proves a slashed key is matched and named end to end.	2026-10-11 19:22:18 +02:00
mesh-admin	fccad8c0a6	Merge pull request 'mesh-issues: say a delivered fix left in fixing, and flag an abandoned issue (hq issue 483, ADR 0294)' (#212) from fix/483-a-delivered-fix-is-said into main	2026-10-11 16:52:58 +00:00
mesh-admin	58f077e296	Merge pull request 'restic: a backup taken by hand outlives the next thinning (hq issue 453)' (#215) from fix/453-a-hand-backup-is-kept into main	2026-10-11 16:13:45 +00:00
mesh-admin	0a55831992	Merge pull request 'docker: the module's state is in the mesh's directory, never /var/lib/docker (hq issue 452)' (#217) from fix/452-docker-state-is-its-own into main	2026-10-11 16:07:35 +00:00
jschoubben	7e6889a646	docker: keep the module's state in the mesh's directory, never Docker's own ... #205's "place": "." resolved to /var/lib/docker, where the node-engine places nothing (issue 339), so ace refused the apply and put the module back. The state directory now says "place": "mesh", <root>/mesh/docker, and a test holds every directory and path of the module out of the container runtimes' data (novox/hq issue 452).	2026-10-11 18:01:24 +02:00
jschoubben	7211227c76	restic: let a restore point taken by hand go only after a night recorded good, say a missing hand= as a file to push, and say what a failed thinning already let go (novox/hq issue 453)	2026-10-11 17:57:43 +02:00
mesh-admin	b687a98770	Merge pull request 'mesh-knowledge: the review is shown every link kind propose accepts, and a research effort's status apart from its state (hq issue 492)' (#216) from fix/492-the-review-sees-what-propose-accepts into main	2026-10-11 11:12:39 +00:00
jschoubben	a6e2881448	mesh-knowledge: the review is shown every link kind propose accepts and a research effort's status apart from its state (hq issue 492) ... The first live review sent research 046 back for a touches link the rules never named and a status proposed that is the record's state, not the effort's. The rules now list LinkKinds, the one list propose holds a link to, each with its meaning.	2026-10-11 13:07:29 +02:00
mesh-admin	73367192c1	Merge pull request 'Nothing reads decision records from hq: records_decisions retired, reserve asks the knowledge base, an empty import refuses (hq issue 488)' (#214) from fix/488-nothing-reads-records-from-hq into main	2026-10-11 10:55:19 +00:00
mesh-admin	28a9821c2e	Merge pull request 'mesh-knowledge: the import keeps every field, takes clauses only from the Decision section's own list, and can repair what it held (hq issue 487)' (#213) from fix/487-the-import-keeps-every-field into main	2026-10-11 10:55:15 +00:00
mesh-admin	94bc18bc3f	Merge pull request 'mesh-knowledge: the knowledge base reviews a proposal itself, and serves no review verb (hq issue 425)' (#211) from feat/425-the-knowledge-base-reviews-itself into main	2026-10-11 10:55:10 +00:00
jschoubben	1b7f79f761	mesh-issues: a counter not started never starts without the knowledge base's answer, floor or no floor (review of #214)	2026-10-11 12:44:45 +02:00
jschoubben	156fda4b33	mesh-knowledge: refresh repairs records the knowledge base drafted, keeping their clauses while their source is unchanged; ADR 0114 and 0148 as fixtures (review of #213)	2026-10-11 12:44:08 +02:00
jschoubben	6fb827605a	mesh-knowledge: bound every review failure per record, make one out of tries loud, and let only its author revise a pending insight (review of #211)	2026-10-11 12:41:24 +02:00
jschoubben	879bc4bca2	restic: keep a backup taken by hand past the next thinning, until a later night and keep-hand-taken-days, so the restore point taken against a risky change is still there when it is needed (novox/hq issue 453)	2026-10-11 12:28:17 +02:00