Commit Graph
446 Commits
Author SHA1 Message Date
jschoubben e2bc0678c6 Design 47 H1: proof 2, a two-module change along a dependency edge delivered by its pipeline (H §5 step 7) 2026-11-20 20:26:17 +01:00
mesh-tools d76d9e0b1a Merge pull request 'Design 47 H1: the tool runner re-serves a membership without dropping a call; its integration level fits its bound (H-A-158, H-A-160)' (#80) from fix/47-H1-node-tools-integration-bound into main 2026-11-20 16:49:43 +00:00
jschoubben 8ee4fdfb8e H1 integration bound: a reload's stand-in serves no repairs; stop waits its Say
Review round 1. A ledger binding heard while a module was loaded again
made refreshRepairs serve the repairs on the reload's stand-in, whose
subscriptions nobody stopped: the repairs were served twice, and still
answered after the runtime stopped. The stand-in is now marked as
reloading and the refresh leaves it alone; the reload serves the module's
repairs as its binding is when the load ends. The reload test holds a
bundle's start, issues the binding inside the reload, and asks a repair
after the runtime stopped: without the change it still answered (6 of 6
runs, GOMAXPROCS=1 and 24).

No test failed without keep's stop waiting for the heartbeat's goroutine.
The heartbeat test now holds one Say and stops meanwhile: stop must not
return before that Say has; without the wait it failed 50 of 50 runs at
GOMAXPROCS=1 and at 24.

Builds: design 47 H §774, H-A-160; principle 11
2026-11-20 17:45:50 +01:00
jschoubben 05deea5952 H1 integration bound: the repairs' refresh keeps the tools it still serves
When a module's ledger binding came or went, refreshRepairs stopped every
subscription of its tools and served them all again, so a call in that
gap was dropped the same way as on a membership (H-A-158). serve is now
the one routine that changes a module's subscriptions, by tool address:
it keeps those still named, subscribes the new, then stops those no
longer named; refreshRepairs and the first serve both go through it, so
no tool is served twice or not at all.

The ledger test's later binding now comes and goes ten times, and the
module's own tool must answer the first call after each, within the
tests' 2 s bound; it failed without the change at GOMAXPROCS=1 and 24.

Builds: design 47 H §774, H-A-158, H-A-160; principle 11
2026-11-20 17:28:34 +01:00
jschoubben 3850eb8eac H1 integration bound: the heartbeat and the counter tests hold under load
TestTheRuntimeSaysItIsThereEveryInterval slept 70 ms and wanted three
beats at 20 ms: a loaded machine that did not schedule the heartbeat in
time failed it (main.20's integration step on the build node); and its
stop returned before the heartbeat's goroutine ended, so a beat could be
said after it ("said after it was stopped"). The ticks are now the
caller's (keep, beside Keep's own ticker), the test drives them and waits
on each beat, and stop returns once nothing more is said. Pinned to one
processor at idle priority beside six busy loops, the old test failed 13
of 100 runs; the new passes 100 of 100.

TestTwoWritersIncrementingOneCounterOnTheBusLoseNoUpdate bounded a
writer's tries at 1000, a guess: on one processor the scheduler lets one
writer win every round and the other lost its 1000 tries every time. A
write is lost only to the other writer's write, so the bound is the
other's count plus one.

Builds: design 47 H §774; principle 11
2026-11-20 17:23:59 +01:00
jschoubben eb174d853f H1 integration bound: the runner follows a membership without dropping a call
On every membership of a module, HandleCalled stopped each subscription
of a tool and subscribed again on the same subjects; a call the bus
routed in that window was dropped by the client in silence, and its
caller waited the runner's whole bound (30 s). On the build node this
was the 30 s of TestAReplacedBundlesOldToolStaysGoneOnTheNextMembership
and of TestTheLedgerRepairsAreToolsOfEveryLedgerOwner, and most of the
integration level's 72 s. Now a membership keeps every subscription whose
subject and queue it still names, subscribes the new ones, and only then
stops those it no longer names (make before break).

The test asks the served tool once after each of five memberships, with
the tests' own call bound of 2 s (principle 11); it fails without the
change. The runtime tests' call helper and the ledger listing wait that
bound, not the runner's 30 s.

Builds: design 47 H §774, H-A-158; principle 11
2026-11-20 17:18:09 +01:00
mesh-tools b1d10dd67e Merge pull request 'The tool runner binds a module's consumer once and hands each event to the follower and the bundles (design 47 H, H-A-132, H-A-133)' (#79) from fix/47-H1-one-consumer-loop into main 2026-11-20 10:35:33 +00:00
jschoubben fb4fe282c7 H1 one consumer loop: serve's first views wiring in one tested function
Review round 1: no test failed when serve went back to firstview.Start.
startViews starts the follower and answers the options runtime.Run is
given; its test holds that on a module's own account the follower binds
nothing and the runtime is given the hearer, and on the machine's
account the follower binds its own.

Builds: design 47 H §752, H-A-113, H-A-132, H-A-133

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-20 11:33:19 +01:00
jschoubben 13ff8041b8 H1 one consumer loop: the follower hears on the module's one consumer
In a module's own runtime the branch follower (first views) and the
bundles each bound the module's durable <node>_<module>_tools with a pull
loop of their own, so each event reached one of the two: the catalog
revision of main.1 reached the follower and never the indexer's bundle,
whose revision verb answered none-yet. The runtime now binds the consumer
once, for the bundles, and hands each event to the follower after the
bundles took it (runtime.WithHearer); the machine's runtime keeps the
follower's own binding.

Builds: design 47 H §752, H-A-113, H-A-132

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-20 11:26:31 +01:00
mesh-tools f92ec7d24f Merge pull request 'Design 47 H1: the tool runner hears every event and serves every seat's verbs' (#78) from feat/47-H1-tool-runner into main 2026-11-20 07:13:07 +00:00
jschoubben 8a534ba436 T2b first-view tool by name form: the gate env test names no MESH_VIEW_CALLERS word
The runner sets no MESH_VIEW_CALLERS env; the gate knows a first-view tool by its name forms (mesh-sdk 77a7b187).

Builds: design 47 H §752, H-A-69

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-20 02:47:04 +01:00
jschoubben d9fe2568de T3 whole: the tool runner is started by its entry against an embedded bus
main's body is `serve(ctx)`, which the process calls with the service manager's signals and one integration test calls with a
credential file and the node-engine's environment: it serves a bundle's tool and seat verbs over a seat list read from a stand-in of
the seat mesh-catalog, waits a verb its definition's timeout, and rereads the list for a verb the first read does not name.

Builds: design 47 H §774, §760, H-A-16, H-A-61

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-20 02:44:40 +01:00
jschoubben 81663a4d9e T5 checker-name-gone: the state-watch tests use a state of a module that stays
The fixture state mesh-delivery-checker.owed, which the retiring checker keeps, is the assigner's `pins`: the three tests of the
file assert the same of a state watched by a module that is granted it.

Builds: design 47 H §764 (The checker), H-A-61

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-20 02:38:55 +01:00
jschoubben a118559515 T4 no-held-back: the approval gate's table no longer holds a row for an ask held back by a count
The row of an ask the router holds back as the fourth open one goes, and the router's refusal row names another refusal: the
operator channel keeps no limit on open asks, and its record no state `held-back` (mesh-sdk, mesh-contracts).

Builds: design 47 H §764 (The operator channel's limit), H-A-57

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-20 02:38:31 +01:00
jschoubben f2a05fb440 T2 no-callers-hook: the runner admits every caller by serving, with no hook that judges who may call
The `admitted` hook, its parameter of serveSeats and its comments go; the test rows that built definitions with Callers and Class
go with the field mesh-sdk removes. A seat's verb is answered for every caller.

Builds: design 47 H §752, §764 (Every seat contract), H-A-41

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-20 02:38:21 +01:00
jschoubben 5d88eecea9 T1c verb-timeout-console: mesh_call waits the timeout its verb's definition names and says it
The console's surface holds the runner's seat definitions; a seat verb's call waits its `timeout`, and a timeout's text names the
bound waited, not a fixed 30s.

Builds: design 47 H §760, H-A-43

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-20 02:37:41 +01:00
jschoubben 618d8953ed T1b verb-timeout-serve: a verb served from a bundle waits the timeout its definition names
Tool gains RunWithin and Within; a bundle's tool call waits `within` where the seat verb it serves names a timeout, else
CallTimeout. serveSeats takes the bound of a verb's key beside `chunked`.

Builds: design 47 H §760, H-A-43

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-20 02:36:48 +01:00
jschoubben 5314231c0d T1a verb-timeout-ask: a bundle's ask waits the timeout its verb's definition names
The runner reads the `timeout` (whole seconds) of a seat's verb from the seat list's definitions beside `chunked` and hands it to
the bus connection's ask (AskWithin, AskAsWithin, AskAsChunkedWithin); a verb that names none, a seat the definitions do not hold
and a module's tool wait the runner's own bound, RequestTimeout.

Builds: design 47 H §760, H-A-43

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-20 02:35:28 +01:00
jschoubben 76df3cb420 node-tools: the unanswered end of the seat list's first read is said also when it ends after the bound (A-490m)
Builds: design 47 G A-490m

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-19 23:22:40 +01:00
jschoubben d1dc482b05 node-tools: a runner rereads the list only for a verb of a seat it defines, one read shared; asks of undefined seats and tools unchanged (A-490m)
Builds: design 47 G A-490m

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-19 23:07:23 +01:00
jschoubben 2f9690fe80 node-tools: the runner reads the seat list again for a seat verb it does not name, and never asks a chunked verb without its call
A bundle's ask of seat:mesh-wants.input reached the holder with no `call` and was refused ("call is one token of 16 to 64 ..."), so no
composition step passed: the runner chooses the call only for a verb its seat list says sends chunks, and a verb the list did not name
(a list read once, unread, or stale) was taken for one that does not. chunkedAsked reads the list again once and, where the verb is still
not named, refuses the ask saying why. A first read of the list that ended without an answer is now said in the log.

Builds: design 47 G A-249 (the runner receives a chunked seat answer for the bundle), ADR 0636 §1 (the seat list is read, not guessed)

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-19 23:00:04 +01:00
mesh-tools 30b2a72f9b Merge pull request 'node-tools: a probe line, so node-tools gets a pipeline on a revision that is kept' (#76) from probe/node-tools-pipeline-4 into main 2026-11-16 22:49:02 +00:00
jschoubben 470e57124e node-tools: a probe line, so node-tools gets a pipeline on a revision that is kept 2026-11-16 23:48:42 +01:00
mesh-tools e554402e47 Merge pull request 'node-tools: a probe line, so node-tools builds against mesh-sdk go/v0.1.16' (#75) from probe/node-tools-pipeline-3 into main 2026-11-16 22:17:42 +00:00
jschoubben ef0a61d348 node-tools: a probe line, so node-tools builds against mesh-sdk go/v0.1.16 2026-11-16 23:17:31 +01:00
mesh-tools 42a752f052 Merge pull request 'node-tools: a probe line, so node-tools gets a pipeline again' (#74) from probe/node-tools-pipeline-2 into main 2026-11-16 21:55:42 +00:00
jschoubben 4043dd1176 node-tools: a probe line, so node-tools gets a pipeline again (main.970 met node-engines older than G) 2026-11-16 22:55:30 +01:00
mesh-tools b232e39c96 Merge pull request 'node-tools: a probe line, so node-tools gets a pipeline again' (#73) from probe/node-tools-pipeline into main 2026-11-16 20:38:28 +00:00
jschoubben 20b6b02341 node-tools: a probe line, so node-tools gets a pipeline again (its composition step at main.969 was lost with the assigner it ran on) 2026-11-16 21:38:17 +01:00
mesh-tools 72b0d4ca4e Merge pull request 'Design 47 G start: tool runner fixes placed by hand' (#72) from fix/47-G-start-3 into main 2026-11-16 14:15:09 +00:00
jschoubben 4333d5b944 node-tools: the tool runner admits every caller of a seat's verb
The operator, 2026-11-15 23:25 CET: every module, agent and operator may call anything; gates and guards can be added
later. The runner refused every call of the knowledge base and the issue tracker while it had not read their seat
definitions. The approval mark stays, held by the gate.
2026-11-15 23:27:12 +01:00
jschoubben 9e8bf6dc0c Runner: a get session's object reader reads on a context that ends with the session, not with the call that opened it
Builds: design 47 G §490, §518, §620 (a get over one chunk is a session read chunk by chunk)
2026-11-15 01:45:37 +01:00
jschoubben 45fbc813df Runner: a page of a state reads every current value in one stream, not a bus request per key: a page of the assignments took 13 s from a node far from the bus 2026-11-14 21:47:28 +01:00
jschoubben 3d6daad925 Runner: reads each module's events on its own durable <node>_<module>_tools, beside the program's (A-397) 2026-11-14 18:58:20 +01:00
jschoubben 903ac4ce93 Runner: a module's marks are checked once against all its bundles (A-398)
Live on novox every gitea tool was refused: each of gitea's eight bundles overwrote the gate's registered set, so the last one, which does not carry gitea_branch_protection_loosen, made the mark unserved.
2026-11-14 14:36:42 +01:00
jschoubben 9fa29ad801 Tool runner: serve a node-scoped seat's verb also with no machine, in a queue group
A verb asked with no machine (the indexer's seed asks node-build-go's steps) is
answered by any one holder, on the tool and the caller-named subject.

Builds: lead answer A-393 (design 47 G)

Claude-Session: https://claude.ai/code/session_01NcbgfGKaxw984XyAPfLm4h
2026-11-14 11:33:57 +01:00
jschoubben 30027347e9 Tool runner: hear an adopted seat event by its key's prefixes, say an unheard key once
The mesh-catalog seat definition publishes catalog-revision on
mesh.seat.mesh-catalog.event.catalog-revision.<branch>, so the bus key is
mesh-catalog.catalog-revision.<branch>; the exact-key lookup missed and the event
was dropped silently. Look up the key and its shorter prefixes (as mesh-sdk's
views.adoptedHears does) and say an event no view hears once per key.

Builds: design 47 G (catalog half's revision events)

Claude-Session: https://claude.ai/code/session_01NcbgfGKaxw984XyAPfLm4h
2026-11-14 04:26:14 +01:00
jschoubben 7a6174bb3e Reach an old objects entry by the store grants creates for it, not by a membership field (A-380)
Builds: design 47 G A-380 (corrects A-379)

Claude-Session: https://claude.ai/code/session_01NcbgfGKaxw984XyAPfLm4h
2026-11-13 23:19:04 +01:00
jschoubben faa1064214 Fill the gate's Call.Env from the bundle's composed environment (A-377)
Builds: design 47 G A-377

Claude-Session: https://claude.ai/code/session_01NcbgfGKaxw984XyAPfLm4h
2026-11-13 23:17:54 +01:00
jschoubben 27a45d414d Reach a module's old objects section by its membership when no binding names it (A-379)
Builds: design 47 G A-379

Claude-Session: https://claude.ai/code/session_01NcbgfGKaxw984XyAPfLm4h
2026-11-13 23:17:29 +01:00
mesh-tools a96df478d4 Merge pull request 'Milestone 47 G: mesh-tools's part of the one release' (#71) from feat/47-G-join12 into main 2026-11-13 13:09:37 +00:00
jschoubben 6be5679a0d Merge main into feat/47-G-join11: the tool runner's mesh/stream.publish (issue 1979)
mesh/delivery merged: source.main moved to novox/mesh-tools@a96df478d498
mesh/delivery-group group feat/47-G-join12 merged: every member is merged
mesh/check failed (2): mesh-delivery-checker: no published of the tagged version of go-meshconditions of me…; and 1 more | 6be5679a0d58 on a96df478d498
The capability stream-publish is stated under mesh-sdk's word (buscalls.StreamPublishWord) and in the runtime artifact's states, so G's test that the stated words are the manifest's states holds; the publish uses G's exported JetStream.

Builds: design 47 G §545, §622; design 42 zfs-owns-zfs A-42, A-44

Claude-Session: https://claude.ai/code/session_01NcbgfGKaxw984XyAPfLm4h
2026-11-13 09:10:44 +01:00
mesh-tools af83854956 Merge pull request 'The tool runner publishes a bundle's event onto its own bound events stream (issue 1979)' (#70) from feature/1979-zfs-owns-zfs into main 2026-11-13 08:00:53 +00:00
jschoubben df2162e381 tool runner: mesh/stream.publish is the capability stream-publish, taken out of bus-state-calls (A-44)
mesh/delivery merged: source.main moved to novox/mesh-tools@af8385495688
mesh/delivery-group group feature/1979-zfs-owns-zfs merged: every member is merged
mesh/check being carried to df2162e381f5 on af8385495688 | from 447a8d035935, mesh-catalog 117f933dacf0, mesh-delivery 7221a1e5cd7f, mesh-host 0193949…
Builds: design 42 zfs-owns-zfs §25, lead answer A-44
2026-11-13 01:52:15 +01:00
jschoubben a775b42e4d tool runner: mesh/stream.publish says an event on the calling bundle's own bound stream (A-42)
Builds: design 42 zfs-owns-zfs §25, §24, lead answer A-42
2026-11-13 01:49:02 +01:00
jschoubben 481a83e11c Round 2 T2, T3, T4, T5: the tool runner ends G with its merge-base test count; the wait test runs its rows side by side
mesh/delivery-group group feat/47-G-join11 checking: 15 of 17 member(s) ready
mesh/delivery closed: a newer head of the same pull request
mesh/check checking 481a83e11c4f on af8385495688 | novox/mesh-tools#69 in group feat/47-G-join11
The tests G added fold into existing ones as subtests (design 47 G §626, A-306: 171 tests down to 154, integration 9
down to 7). TestMeshWaitRefusesLoudlyAndNeverLoops runs its groups of one-second rows as parallel subtests, each
on a runtime of its own, 15 s down to 5 s (A-311: the one-second floor of a wait's timeout stands, no clock is
injected). The go.mod no longer requires meshconditionsclient, which no package loads (T4). Node names in the
lines G adds are role names (T5).

Builds: design 47 G §626, design 34 every-event §19, A-306, A-311; review round 2 findings T2, T3, T4, T5
Claude-Session: https://claude.ai/code/session_01NcbgfGKaxw984XyAPfLm4h
2026-11-12 22:24:02 +01:00
jschoubben 28bece26b0 Join 11 (try 2): the chunked-answer tests flush the holder's verb before another connection asks it
The tests of A-249 and A-257 served the verb on one connection and asked it from another at once; the
subscription was still in the holder's buffer when the ask reached the bus, so the ask was answered 503
no responders (13 of 300 runs with -cpu 1 before, 0 of 2000 after). The runner flushes after serving
(runtime.go:352, :807); the tests now do as it does.

Builds: design 47 G A-249, A-257
Claude-Session: https://claude.ai/code/session_01NcbgfGKaxw984XyAPfLm4h
2026-11-12 15:14:01 +01:00
jschoubben 816df217cd Join 11: every go.mod requires, at the zero version, each mesh module its packages load
The checker's fetch pins only the zero-version requirements a directory's go.mod names
(mesh-build-go workspace.go goModOf), so a module reached only through another, such as
meshhandacts through meshhandactsclient or mesh-delivery-sdk through meshcatalogclient,
was never downloaded and the shut-off load failed (fetch.go:267). The join's merges had
dropped these indirect requirements that the trunk's go.mod files list.

Builds: design 47 G (join 11, the check's fetch of the group's modules)

Claude-Session: https://claude.ai/code/session_01NcbgfGKaxw984XyAPfLm4h
2026-11-12 14:07:29 +01:00
jschoubben c35c27f664 Area 1 (part): the tool runner's manifest states approval-mark/1
The runner states the approval mark's word beside its call families, so its
artifact's states are exactly what it says (the join of A3's and A5's lines);
TestTheWordsTheRunnerStatesAreItsManifestsStates failed without it.

Builds: design 47 G §545, design 46 7-approval-mark §6; join 11 area 1

Claude-Session: https://claude.ai/code/session_01NcbgfGKaxw984XyAPfLm4h
2026-11-12 10:44:02 +01:00
jschoubben d350c98823 Join 11: the merged heads compile where no area's fix is owed
The join's importers name the delivery repository's moved contracts and sdk
(MV1), node-tools' tests use mesh-sdk's gate, the duplicate node-engine claim
rule of the reader release gives way to G's, and the assigner names its first
watched branch, never a constant. The seat holders' AnswerSeat callers are
area 6's.

Builds: design 47 G join 11 (merge of fix round 11)

Claude-Session: https://claude.ai/code/session_01NcbgfGKaxw984XyAPfLm4h
2026-11-12 10:41:44 +01:00