3368 Commits

This Branch

This Branch

All Branches

Author	SHA1	Message	Date
jschoubben	e712cd227f	C5: the integration row reads the dropped step as dropped (ADR 0664 §4, issue 1748 review 1) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 23:44:08 +01:00
jschoubben	12059ad3cc	F5: the failurekit check refuses a build module whose manifest does not ask seat-settle-delay (ADR 0664 §5, issue 1748 review 1) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 23:42:46 +01:00
jschoubben	6a29e41175	F3: mesh-build-go and mesh-build-node follow the drops other holders write, and a module with no watch says so (ADR 0664 §4, issue 1748 review 1) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 23:42:13 +01:00
jschoubben	1948da9cb4	F8, dropped: the build module's drop uses the generated contracts' Drop and ends the step dropped, with the named cause (ADR 0664 §4, issue 1748 review 1) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 23:41:10 +01:00
jschoubben	b84a2f4dc1	F6b: merge main (#647) into the transport retry: unknownRevision reads the transport words, the row says passing (issue 1748 review 1) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 23:33:33 +01:00
mesh-tools	9c42423d58	Merge pull request 'The knowledge base's review knows its round: from round 3 only a wrong record asks changes, at round 5 the operator decides (issue 784)' (#648) from fix/784-review-round into main	2026-11-09 22:26:59 +00:00
jschoubben	c12f1a16db	get shows a round only for a proposed version (issue 784, ADR 0665) ... Claude-Session: https://claude.ai/code/session_01JG73KK8HD1gufnqyFxMFdW	2026-11-09 23:22:46 +01:00
jschoubben	09ea4a40bf	A closing keeps the review round, get says the round, a close waits for the operator's round-5 answer (issue 784, ADR 0665) ... Claude-Session: https://claude.ai/code/session_01JG73KK8HD1gufnqyFxMFdW	2026-11-09 23:20:22 +01:00
jschoubben	6b8e7fc621	C1c, C6: any holder serves drop and every node follows it by a watch of the shared pauses; the build modules serve drop and ask the tool runner's seat-settle-delay (ADR 0664 §1, §4, §5, issue 1748) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 23:19:28 +01:00
jschoubben	3c3fdd3446	C5: an integration test of the delayed redelivery and the drop of a running step against the real bus (ADR 0664 §1, §5, issue 1748) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 23:16:09 +01:00
jschoubben	06f28fbadb	C4: the conformance kit gains Passing with the words of each runner when the network did not answer; each build module's tests call it and checks/failurekit refuses one that does not (ADR 0664, issue 1748) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 23:15:35 +01:00
jschoubben	62644aa012	The knowledge base's review knows its round: from round 3 only a wrong record asks changes, at round 5 the operator decides (issue 784, ADR 0665) ... Claude-Session: https://claude.ai/code/session_01JG73KK8HD1gufnqyFxMFdW	2026-11-09 23:13:55 +01:00
jschoubben	bdfd8db265	C3: mesh-build-node and mesh-build-image say a failure of the network as a transport fault (npm ci, buildah bud, the image seat's mirror) (ADR 0664 §1, issue 1748) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 23:13:45 +01:00
jschoubben	89ec9429db	C2b: a seat verb's no responders is a transport fault when the assigner names a holder of the seat; a seat nothing holds fails at once (ADR 0664 §1, issue 1748) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 23:12:35 +01:00
jschoubben	b9cfa4fe16	C2: mesh-build-go says a failure of the network as a transport fault: one download, no retry of its own (ADR 0664 §1, issue 1748) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 23:10:16 +01:00
mesh-tools	8b184d8322	Merge pull request 'mesh-build-go reaches the forge's git when the public proxy is unreachable, and a network failure is no pin (issue 1966)' (#647) from fix/1966-contracts-goproxy into main	2026-11-09 22:08:33 +00:00
jschoubben	02d57c4a53	C1b, C1c: the seat's verb drop refuses a dropped ask at pull, ends one that runs, and the start sweeps old drops (ADR 0664 §5, issue 1748) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 23:07:49 +01:00
jschoubben	5a7561fe1d	C1: a build step that ends on a transport fault below its last run is written again and handed back to the bus with a delay (ADR 0664 §1, issue 1748) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 23:04:44 +01:00
mesh-tools	5e537f851f	Merge pull request 'Distribution and the forge hold their caller list from its config watch, not an ask per call (issue 1966; group with mesh-sdk)' (#646) from fix/1966-callers-config-read into main	2026-11-09 22:00:48 +00:00
mesh-tools	0a6a5aba68	Merge pull request 'mesh-build-image names its copy wait's ownership mark nonce, so the state takes it and nats images are stored again (issue 1966)' (#645) from fix/1966-mirror-waits-token into main	2026-11-09 22:00:32 +00:00
jschoubben	407fc6c9e6	The Go seat goes on to the public proxy only on a 404, the public proxy to direct on any failure (issue 1966, review 1) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 22:55:53 +01:00
jschoubben	6299b47f4f	A mesh package the Go seat lacks is fetched from the forge's git, not stopped by an unreachable public proxy (issue 1966) ... GOPROXY joined the Go seat, the public proxy and direct with commas, so go went on only after a 404 or 410: a contracts package the seat does not hold, and a public proxy a node cannot reach, ended the fetch before direct. It joins them with | now, so any failure goes on to the next. go words a proxy it cannot reach as "invalid version", which the pin path read as an unknown revision and labelled "on no pull request head"; a network failure is never that now (network is unreachable and no route to host are network failures). Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 22:42:09 +01:00
jschoubben	24bf59219e	Distribution and the forge hold their caller list from its config watch, not an ask per call (issue 1966) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 22:42:00 +01:00
jschoubben	e3b4ecde8c	Name the copy wait's ownership mark nonce, not token, so the state accepts it; the copy test's state refuses credential names as the bus does (issue 1966) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 22:42:00 +01:00
mesh-tools	fd9147087e	Merge pull request 'The bus module reads a consumer's counts and a stream's events, so a lost event can be traced' (#644) from fix/1966-nats-consumer-read into main	2026-11-09 21:27:30 +00:00
jschoubben	a3d141e793	nats_stream_messages shows a body only for the stream EVENTS (issue 1966, review 1) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 21:21:03 +01:00
jschoubben	4fbe6e347f	nats_stream_messages shows a body only on an event stream (issue 1966, review 1) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 21:20:19 +01:00
jschoubben	3d70bd3660	nats_stream_messages refuses a module's kept state (KV_*, OBJ_*), which can hold secrets (issue 1966, review 1) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 21:18:43 +01:00
jschoubben	a4e5a8ffef	Format the bus tools' new test (issue 1966) ... Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 21:14:00 +01:00
jschoubben	13dea6b1aa	The bus module reads one consumer in full and a stream's messages by subject (issue 1966) ... A lost event between a shared stream and a module's consumer could not be located: no tool read a consumer's delivered and acknowledged sequence or a stream's messages for one subject. nats_consumer reads from the monitoring document; nats_stream_messages asks STREAM.MSG.GET on the module's own credential, newest first and bounded. Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 21:13:02 +01:00
mesh-tools	6d2ce96630	Merge pull request 'The resolver module's tests hold its configuration's guards' (#643) from fix/1966-resolver-test into main	2026-11-09 20:09:34 +00:00
jschoubben	a7dc4534d7	Hold the resolver's configuration guards in the dnsmasq module's tests (issue 1966) ... Moved from mesh-sdk's catalogue tests, which read a literal the file no longer has. Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ	2026-11-09 20:59:46 +01:00
mesh-tools	4387e77f28	Merge pull request 'mesh-hand-acts answers bounded, newest first; mesh-healers seeds from a partial answer (issue 1966 S8, recovery)' (#642) from fix/1966-s8-hand-acts-window into main	2026-11-09 17:55:10 +00:00
jschoubben	443a340f06	Listing.Complete comment says what the field means now (issue 1966 S8 review 2) ... Claude-Session: https://claude.ai/code/session_01NcbgfFfQ4Z9UE5jK32PmE5	2026-11-09 18:23:33 +01:00
jschoubben	15f81780b9	hand-acts: no entry skipped by an early-ended page, the answer counted as sent, completeness for the window asked (issue 1966 S8 review 1) ... Claude-Session: https://claude.ai/code/session_01NcbgfFfQ4Z9UE5jK32PmE5	2026-11-09 18:23:33 +01:00
jschoubben	ea4dfaa5a2	mesh-hand-acts answers bounded, newest first; mesh-healers seeds from a partial answer (issue 1966 S8) ... hand-acts read the whole log (84k entries, 47 MB) on every call and could never answer inside the call bound or the bus payload. It now reads from the newest entry backwards, page by page, and stops at the window's start, at 900 KiB of acts or at 20 s, saying complete: false, cut and why. warrants is bounded the same way. mesh-healers' start-up seed takes a partial answer and logs it, instead of retrying forever. Quick fix: this (mesh-catalog only, about 1.5 h). Thorough fix: also the tool runner never freezing on an oversized bundle line and logging unanswered calls (mesh-tools, a separate fix by another developer, lead's ruling Q3), and an SDK-side bound. Chosen here: quick for this module, the thorough part running beside it; the byte cap is the module's own (lead's ruling Q1, a `limit` argument is a later note). Claude-Session: https://claude.ai/code/session_01NcbgfFfQ4Z9UE5jK32PmE5	2026-11-09 18:23:33 +01:00
mesh-tools	7630e3cc9d	Merge pull request 'Sessions merge without a human approval: the hard rule keeps a branch and a pull request for every change' (#641) from docs/no-merge-approval into main	2026-11-09 16:58:54 +00:00
jschoubben	994b25f540	Sessions merge without a human approval: the hard rule keeps a branch and a pull request for every change ... The operator, 2026-11-09: "Remove the instruction, i don't care about merges, and little about deliveries". Claude-Session: https://claude.ai/code/session_01NcbgfFSnn24M3GcqttoU7i	2026-11-09 17:57:51 +01:00
mesh-tools	503411e628	Merge pull request 'mesh-build-go reads a compose ask's gzip-packed request (issue 1966, recovery; group with mesh-delivery)' (#640) from fix/1966-checker-compositions-objects into main	2026-11-09 16:31:22 +00:00
jschoubben	4e2a1a5a1f	mesh-build-go reads a compose ask's gzip-packed request (issue 1966) ... The checker sends a large node's request packed, as it passes the bus's message limit; compose reads either form through ComposeJob.RequestBytes. Only mesh-build-go serves compose. Delivered as a group with the mesh-delivery change of the same branch name: an older mesh-build-go refuses a packed ask naming that it holds no request object. Claude-Session: https://claude.ai/code/session_01NcbgfFfQ4Z9UE5jK32PmE5	2026-11-09 13:36:38 +01:00
mesh-tools	6a24e4d0c7	Merge pull request 'Gate record of mesh-catalog's forge-made merges (issue 1966, recovery)' (#639) from fix/1966-forge-made-merge into main	2026-11-09 08:45:50 +00:00
jschoubben	80186d0798	Record this repository's forge-made merges in the delivery gate, so the indexer takes them in (issue 1966, recovery) ... Merges made on the forge directly since 2026-11-09 07:04 UTC have no gate merge record, and the indexer holds them; an empty merge through the gate covers this repository's head. Second of its kind: the group's mesh-delivery member removes the need (the forge bridge reads a forge-made merge's pull request from the forge). Claude-Session: https://claude.ai/code/session_01NcbgfFSnn24M3GcqttoU7i	2026-11-09 09:17:03 +01:00
mesh-tools	89cd8a2e8e	Merge pull request 'The provisioner keeps a provision while a module's newest composition is unread, and withdraws an ask once a key is needed again (issue 1966, recovery)' (#638) from fix/1966-provisioner-newest-held into main	2026-11-09 08:07:53 +00:00
jschoubben	4658e1cb1c	A newest hold's key is <node>.<module>=newest, which the bus holds, and the test store refuses what the bus refuses (issue 1966) ... Review round 2: '@' is outside the bus's key rule, so every newest hold failed to write on the live bus while the in-memory store took it. Claude-Session: https://claude.ai/code/session_01NcbgfFSnn24M3GcqttoU7i	2026-11-09 09:06:39 +01:00
jschoubben	6e6b325cf1	Newest holds are kept on the bus in refused, and every key needed again withdraws its retire ask by one helper (issue 1966) ... Claude-Session: https://claude.ai/code/session_01NcbgfFSnn24M3GcqttoU7i	2026-11-09 09:02:40 +01:00
jschoubben	bbe107fa2a	A provision only the newest build requires is kept while that build's composition is unread, and a key needed again withdraws its retire ask (issue 1966) ... Not read is not unneeded on the newest scope too (issue 940's rule): a module whose newest composition is unpublished or refused is held at its newest build, said in provisioner_plans newest-held and bounded on the health. A retiring key needed again cancels its ask on operator-channel (accept.cancel), retried until the router takes it. Claude-Session: https://claude.ai/code/session_01NcbgfFSnn24M3GcqttoU7i	2026-11-09 08:52:16 +01:00
mesh-tools	ac94f8338f	Merge pull request 'distribution's program listens on 5100 and the registry container on 5001, so no mapping names the artifact store's port (issue 1966, recovery)' (#637) from fix/1966-store-listen into main	2026-11-09 00:58:10 +00:00
jschoubben	a0459a8553	distribution's manifest test reads the listens with its own struct and checks every mapping's sides for 5100 ... Claude-Session: https://claude.ai/code/session_01NcbgfFSnn24M3GcqttoU7i	2026-11-09 01:42:40 +01:00
jschoubben	6cf7f67a1d	distribution's program listens on 5100 and the registry container on 5001, so no mapping names the store's port ... The composer finds a listen's machine port in a container mapping that names it on either side (MachineSide). The program's listen 5000 matched the registry container's inner 5000 in 127.0.0.1:5001:5000, so the store's address became the container's loopback side and the program's own port was never opened: the move to 83cd1a3d on 2026-11-08 23:25 UTC cut the artifact store off (issue 1966 note 109). The program now binds 5100, where every machine fetches (the artifact-store alias and the seats' reach agree), and the container serves 5001 inside and out. A row holds that no container names 5100. Undo: revert this commit. Claude-Session: https://claude.ai/code/session_01NcbgfFSnn24M3GcqttoU7i	2026-11-09 01:40:49 +01:00
mesh-tools	08fbc51af2	Merge pull request 'The keeper's health reads the consumers it looks for (issue 1966, recovery)' (#636) from fix/keeper-health-consumers into main	2026-11-08 22:55:44 +00:00