Commit Graph

37 Commits

Author SHA1 Message Date
Joseph Doherty a59dbd4f07 docs(heidenhain): driver accepted, #829/#853 closed
The ignitionoee CR capture (6d2aec40: 7 h read-only sample of 34 production
controls on the Java LSV2 driver) and the 2026-11-11 iTNC 530 run did not
change the driver's code: ProgramState 0 = cutting (11,865 of 11,869 motion
pairs), 7 = idle; TNC 426/430 BadNotSupported; overrides BadUserAccessDenied
under INSPECT; capabilities Unknown until first poll; alarm classes 1/7/8 seen,
class 5 not. Owner accepted the driver on that basis; remaining limits stay
documented in Heidenhain.md §Hardware-gated.
2026-11-13 07:46:56 -05:00
Joseph Doherty f0c6cf9640 fix(security): running drivers refuse out-of-namespace credential refs at resolve time (#834)
v2-ci / unit-tests (push) Has been cancelled
The #815 warn-only compatibility window is over. A running Galaxy driver
(InitializeAsync, every gateway ReopenAsync, the ack fallback password) and a
running OpcUaClient driver (session open) now refuse an out-of-namespace
credential reference with the same CredentialReferenceException that save,
deploy (CredentialReferenceOutOfScope), Test Connect and browse already raise,
before any store / environment / file read and before any endpoint is dialled.
The one existing scope check (CredentialReferences.Parse/ResolveAsync) is now
the only path; nothing new was added.

Removed:
- LegacyCredentialReferenceTracker (Core.Abstractions), its rate-limited WARN
  and the credential-compat:<field> diagnostics entry.
- The per-driver Legacy* resolve methods and every `legacy` parameter
  (GalaxySecretRef, GalaxyAckCredentialResolver, OpcUaUserIdentityFactory,
  OpcUaUserCertificatePaths, OpcUaClientSecretResolution).
- IDriverDiagnostics on GalaxyDriver / OpcUaClientDriver (it only carried the
  tracker's entries).
- The tracker also kept a pre-#836 absolute UserCertificatePath loading on a
  running OpcUaClient driver; that is refused now too.

Preflight / banner decision: KEPT, as a hard error. The Running source
(driver-health credential-compat entries) and CredentialReferenceBannerModel
are gone with the tracker, but the ConfigDb scan still tells an operator
something no other surface does: a Deployed finding is now a driver that is
not connecting, a Draft finding is a deploy the gate will refuse - named
before anyone deploys, e.g. after restoring an old ConfigDb backup or
importing a pre-#815 config. It also carries the #835 cross-instance and
#836/#850 path checks. CredentialReferencePreflightService now logs one Error
per finding (was Warning) with no warn-only wording, and the /raw banner is
alert-danger.

DEV and QA ConfigDbs hold no Galaxy/OpcUaClient driver (checked 2026-11-13);
no production install exists.

Tests: GalaxyDriverLegacyCredentialTests replaced by
GalaxyDriverCredentialScopeTests; new OpcUaClientDriverCredentialScopeTests;
tracker tests in OpcUaClientSecretResolutionTests /
OpcUaClientUserCertificatePathTests deleted; AdminUI preflight tests drop the
Running/banner-model cases and assert Error level. Against the old code the
new refusal tests fail (Galaxy 5/6, OpcUaClient 4/5; the two positive controls
pass on both).

docs/security.md: the migration table now says running drivers refuse; the
warn-only section is replaced by "The compatibility window is over".

Claude-Session: https://claude.ai/code/session_01GT67ACmpMR7b9LGRPLijMi
2026-11-13 02:06:43 -05:00
Joseph Doherty eec87d3f74 fix(ablegacy): libplctag refuses MicroLogix function files locally instead of a comms fault (#592)
v2-ci / unit-tests (push) Has been cancelled
libplctag 2.6.0 cannot parse RTC/HSC/... names; tag creation failed with
PLCTAG_ERR_BAD_PARAM, which the status translation reported as BadConnection:
evicted handle, degraded health, counted toward marking a healthy device down.
The libplctag runtime now refuses function-file addresses before creating a
native tag (Unsupported -> BadNotSupported, health untouched), matching native.
Docs: function files are unsupported on both transports, not "unverified".

Claude-Session: https://claude.ai/code/session_01GT67AD99zQQqpLCVm3p7xk
2026-11-13 01:00:47 -05:00
Joseph Doherty 2e537dc6ca docs(heidenhain): honest provenance for alarm class/group codes (#853)
v2-ci / unit-tests (push) Has been cancelled
No public source exists for DNC_ERROR_CLASS / DNC_ERROR_GROUP or their numbers
(the table was written from memory). The iTNC 530 Technical Manual's PLC error
table documents the same class and group NAMES without codes; three real
iTNC 530 records corroborate 7/4, 7/3 and 1/3. DNC and LSV2 numberings are
known to differ (ProgramStatus vs R_RI 26), so the numbers stay unproven.
Docs/comments only; no behaviour change.

Claude-Session: https://claude.ai/code/session_01GT67AD99zQQqpLCVm3p7xk
2026-11-13 00:36:11 -05:00
Joseph Doherty f8785cfa2c Merge branch 'worktree-agent-a71443b71c4c01c13' into heidenhain-pylsv2-parity
# Conflicts:
#	.claude/rules/drivers.md
#	docs/drivers/Heidenhain.md
#	tests/Drivers/ZB.MOM.WW.OtOpcUa.Driver.Heidenhain.Tests/HeidenhainPyLsv2ParityTests.cs
2026-11-12 23:57:06 -05:00
Joseph Doherty 3055646688 Merge branch 'worktree-agent-ab2193d7e99435def' into heidenhain-pylsv2-parity
# Conflicts:
#	.claude/rules/drivers.md
#	docs/drivers/Heidenhain.md
2026-11-12 23:56:31 -05:00
Joseph Doherty 4cee05c2f8 feat(heidenhain): R_RI runtime-info parity with pyLSV2 — state names, refusal codes, timer/unit docs
Compared every R_RI decoder we use (22-28, 30-34, 51) byte-for-byte with pyLSV2 (MIT,
97a3d691) and ran the captured iTNC 530 replies through pyLSV2 itself: no decode
disagreement, so no published value changes.

- Additive read-only String leaves Status/ExecutionStateName, Status/ProgramStateName and
  Tool/AxisName (pyLSV2 ExecState / PgmState / decode_tool_info names; out-of-table code =
  "Unknown"). They ride the raw leaf's feature, so no extra wire read; raw Int32 leaves unchanged.
- Refusals: 52/72 -> BadOutOfRange, 58 -> BadTimeout, 56 -> BadInvalidState, 57 (outside the
  alarm walk) -> BadNoData. IsStable and the 57-only alarm-walk end are unchanged.
  Lsv2RefusalCodes names every code (pyLSV2 LSV2StatusCode) in exception/log/probe text.
- Timers 30-34 documented as uint32be seconds (lifetime counters) per pyLSV2, hardware-unproven;
  TimersEnabled stays false. Axis/tool units: none on the wire, none assumed.
- Lsv2ReadOnlyPolicy untouched (pinned by a test); R_RI 20/21 left out.

Claude-Session: https://claude.ai/code/session_01GT67AD99zQQqpLCVm3p7xk
2026-11-12 23:55:34 -05:00
Joseph Doherty 50933d57fc feat(heidenhain): PLC/DATA parity with pyLSV2 — FF bits, INPUT/OUTPUT_DWORD, PLCDEBUG write-up
- PLC bit areas (MARKER/INPUT/OUTPUT/COUNTER/TIMER) accept 00/01/FF: real iTNC 530s
  return FF for the same operands over R_DP, and pyLSV2 decodes them as non-zero = true.
  02..FE stay BadDecodingError.
- New PLC:INPUT_DWORD / PLC:OUTPUT_DWORD areas (pyLSV2 MemoryType 13/12): signed
  int32le, element index x 4 over the input/output-word byte pools.
- R_PR offsets agree with pyLSV2. Its count units (BYTE x2, WORD/IW as word counts)
  conflict with the captured controls, so we keep the captures and pin them in a test.
- Docs: pyLSV2 correspondence table, native-operand mapping (no aliases), and why
  raw R_MB is refused (pyLSV2 logs in as the write-capable PLCDEBUG role). That stays
  an owner decision; Lsv2ReadOnlyPolicy is unchanged.

Claude-Session: https://claude.ai/code/session_01GT67AD99zQQqpLCVm3p7xk
2026-11-12 23:53:16 -05:00
Joseph Doherty ba9015fac8 feat(heidenhain): classify the control family from R_VR; DATA gated to iTNC 530 (per pyLSV2)
- HeidenhainControlFamily + HeidenhainControlFamilies.Classify (Contracts): one authority, ported from
  pyLSV2 (MIT) dat_cls.py VersionInfo.control (strip spaces, case-insensitive prefix, ITNC530 first).
  Families: ITnc530, Tnc (640/620/320/128), Tnc7, Lathe (CNC PILOT 640 / MANUALplus 620), MillPlus,
  LegacyTnc (426/430, not a pyLSV2 class), Unknown.
- New additive read-only leaf Controller/Family (String, Identity feature, derived from the same R_VR
  reply: no extra request). Lsv2Versions.Family is computed, so record equality and the #831
  identity-change rediscovery signal are unchanged. Probe message + /hosts diagnostics (<host>.family).
- DATA gate now goes through SupportsDataPath(family) (pyLSV2 client.py read_data_path is iTNC-only):
  non-iTNC and Unknown => Capabilities/DataLogin Unsupported, DATA login never sent. Replaces the
  "model contains itnc" substring check (same result on the real iTNC530 controls).
- No per-family selector pre-gates: pyLSV2's TNC 640 observations (27/28/51/30-34 unknown telegram,
  25/26 need DNC) are already handled by the reply-driven refusal/suppression path; tests pin it.
- Lsv2ReadOnlyPolicy, HeidenhainTagWritable and the no-IWritable rule are untouched.
- Simulator: per-family ControlProfiles (strings from the pyLSV2 README tables / protocol.rst), both
  R_VR field counts, 124-byte R_PR; HeidenhainControlFamilyTests (classification, gates, decoding,
  wire log inside the read-only set, refused selectors asked once).
- docs/drivers/Heidenhain.md: Control families matrix (iTNC 530 hardware-proven; the rest per pyLSV2,
  hardware-unproven) + open questions (C_CC buffer size, encrypted passwords, R_RI 20/21).

Claude-Session: https://claude.ai/code/session_01GT67AD99zQQqpLCVm3p7xk
2026-11-12 23:52:06 -05:00
Joseph Doherty 233e4bab25 Merge #863 TwinCAT Test Connect: drop the AmsNetId-as-IP TCP pre-check; probe via the local AMS router ADS handshake
Claude-Session: https://claude.ai/code/session_01GT67AD99zQQqpLCVm3p7xk
2026-11-11 12:14:58 -05:00
Joseph Doherty fdc9473819 fix(twincat): Test Connect probe no longer dials the AmsNetId as an IP (#863)
TwinCATDriverProbe ran a first-phase TCP check against the AmsNetId's
first four octets on the AMS port (801/851). An AmsNetId is not an IP and
the AMS port is not a TCP port, so Test Connect timed out against every
real runtime (Z35474: dialled 5.118.56.93, an unrelated public host).

The probe now opens no socket of its own: it takes the driver's path,
AdsClient.Connect + ReadStateAsync through the local AMS router, bounded
by the probe budget (Task.Run + WaitAsync), and classifies failures as
no local AMS router / no route or target unreachable / timeout / other.
The degrade-to-green 'TCP reachability only' outcome is gone - with no
TCP check nothing was contacted, so a missing router is reported red.

Tests: the regression test listens on the AmsNetId-derived endpoint and
asserts no connection arrives (fails on the old probe); classification
and budget tests drive the handshake through an internal seam.

Claude-Session: https://claude.ai/code/session_01GT67AD99zQQqpLCVm3p7xk
2026-11-11 12:13:55 -05:00
Joseph Doherty e6ec5a0948 docs(opcuaclient): keep server host-name check opt-in; surface it in the Security panel (#837)
Owner decision 2026-11-11: certificateValidation.validateServerHostName stays
opt-in (unset = off). The AdminUI toggle moves from the "Subscriptions, limits &
discovery" panel to the Security panel beside mode/policy, with help text on the
NAT/IP caveat, IPv6 zone ids and the loose first-label match. Binding unchanged
(tri-state, "Default (off)" leaves the key absent). OpcUaClient.md states the
decision; drivers rule updated; source-scan test pins placement + caveats.

Claude-Session: https://claude.ai/code/session_01GT67AD99zQQqpLCVm3p7xk
2026-11-11 11:57:59 -05:00
Joseph Doherty bdbc47f5fc fix(mtconnect): serve an empty string-typed value as Good "" (#824)
Owner decision 2026-11-11 (Java parity): a genuinely empty observation on a
String/Reference or unauthored item is Good "" instead of BadNoCommunication.
Non-string items (numeric, bool, DateTime, arrays) keep BadNoCommunication -
the Agent supplied no value; BadTypeMismatch would wrongly blame the authored
type. The UNAVAILABLE sentinel is unchanged, and a structured observation with
no text stays BadNotSupported (structured check now precedes the empty check).
Parity matrix, quality-mapping section and drivers rule updated.

Claude-Session: https://claude.ai/code/session_01GT67AD99zQQqpLCVm3p7xk
2026-11-11 11:56:30 -05:00
Joseph Doherty 26349320df Merge #848 #849 #850 credential name claim/pin, env: per-instance rule, merged preflight 2026-11-11 06:36:06 -05:00
Joseph Doherty ea156265a2 docs(security): id-form matching, ASCII-only name form, one-save pin side-step, rename/device-save race (#848 #849 review)
Co-Authored-By: Claude <noreply@anthropic.com>
2026-11-11 06:35:52 -05:00
Joseph Doherty 340d9d26b7 feat(mtconnect): CommsLossConfirmMs AdminUI field, refused-connection fast path reaches native conditions, dedicated live-gate Agent (#852)
- AdminUI MTConnect driver form: "Comms-loss confirm (ms)" in the Reconnect section, always written
  (0 = first refusal, negative = off is never dropped), case-insensitive load, unknown keys preserved,
  advisory at >= 90000 ms (the watchdog fires first). Form output byte-pinned and parsed through
  MTConnectDriver.ParseOptions on the driver side.
- Driver: a confirmed refused-connection outage now annotates every told CONDITION quality-only
  BadNoCommunication (state/ack kept, once per outage, UNAVAILABLE-Bad skipped); the first delivery
  afterwards re-diffs all subscribed conditions so unchanged ones get a quality-only Good.
- docker-dev/docker-compose.mtconnect-gate.yml: throwaway Agent in its adapter's network namespace so a
  stop is a genuine ConnectionRefused (a plain stopped container is a DNS failure = ambiguous).
- Docs: MTConnect.md comms-loss section + live-gate recipe, docker-dev README, drivers.md rule.

Refs #852
2026-11-11 06:23:18 -05:00
Joseph Doherty 9072777422 docs(security): #835/#836 review — unsaved-endpoint rule, credential-use audit, accurate residuals, AuthType-gated path rule, directory resolution
Co-Authored-By: Claude <noreply@anthropic.com>
2026-11-11 04:12:40 -05:00
Joseph Doherty 87416a230f Merge #835 #836 per-instance credential references, confined user-certificate path 2026-11-11 03:53:12 -05:00
Joseph Doherty 0810e808c9 docs(security): per-instance credential references (#835) and the user-certificate directory (#836)
Co-Authored-By: Claude <noreply@anthropic.com>
2026-11-11 03:52:48 -05:00
Joseph Doherty 19ffb3c902 feat(mtconnect): refused Agent connections mark values BadNoCommunication after a short confirmation; clamp the raised sample count to the Agent's bufferSize (#846)
- A /sample or re-baseline failure whose chain holds SocketError.ConnectionRefused, unbroken for
  CommsLossConfirmMs (new, default 2000; 0 = first refusal; negative = off), publishes every subscribed
  value BadNoCommunication once per outage; the first delivery afterwards republishes all of them from
  the index (unchanged values included). Timeouts/resets/DNS/pause keep the #527 watchdog contract.
- SampleCountCeiling clamps the adaptive raise to the /probe bufferSize (never below SampleCount).
- Docs: MTConnect.md comms-loss section + config key; drivers.md rule.

Refs #846
2026-11-11 03:48:05 -05:00
Joseph Doherty ddc1f29e3a feat(heidenhain): Part 9 condition class from alarm class/group, ALARM: selectors in the browse picker, unclaimed alarms counted + logged (#839)
- Core: optional AlarmEventArgs.ConditionClass (AlarmConditionClass Base/Process/Maintenance/System/Safety)
  carried by NativeAlarmProjector on AlarmConditionSnapshot and written to ConditionClassId/Name by
  OtOpcUaNodeManager (kept across class-less events, carried across a #641 rebuild).
- Heidenhain: HeidenhainAlarmClasses — HEIDENHAIN DNC-interface class/group names (corroborated by the
  captured iTNC 530 record); class 5 => Safety, group 1/2 => Process, 4/5 => System, else Base.
- DiscoverAsync offers AlarmConditions/ selector leaves (IsAlarm Boolean); browse-commit seeds the alarm object.
- Alarms no selector claims: otopcua.driver.alarms.unmatched counter, /hosts diagnostics, rate-limited Warning.

Refs #839
2026-11-11 03:41:14 -05:00
Joseph Doherty 61b9adeb9c Merge branch 'master' into fix/832-docker-dev-batch 2026-11-11 02:32:42 -05:00
Joseph Doherty e9064dbdf4 Merge fix/815-credential-ref-scoping
# Conflicts:
#	.claude/rules/drivers.md
#	docs/drivers/Heidenhain.md
2026-11-11 02:08:24 -05:00
Joseph Doherty 4563940792 fix(security): warn-only compatibility for deployed drivers + upgrade preflight (#815)
User decision: save, Test Connect, live browse and the deploy gate keep
refusing out-of-namespace credential references, but an ALREADY-DEPLOYED
driver must not fault at the upgrade restart or (Galaxy) at the next
gateway reconnect, since ReopenAsync re-resolves Gateway.ApiKeySecretRef.

- LegacyCredentialReferenceTracker (Core.Abstractions): the one, deprecated,
  call-site-chosen compatibility path. Running Galaxy (InitializeAsync,
  ReopenAsync, ack fallback) and OpcUaClient (session open) pass it and keep
  the pre-#815 semantics (case-insensitive prefixes, Galaxy file:, OpcUaClient
  literal pass-through) with a rate-limited Warning (driver/field/allowed
  form, never the value) and a credential-compat:<field> diagnostics entry
  (both drivers now implement IDriverDiagnostics). Browsers and probes pass
  none and stay hard-fail; Heidenhain has none.
- Deploy-gate message names the migration.
- Upgrade preflight: CredentialReferencePreflightService (admin nodes, once
  after ApplicationStarted, ConfigDb only, resolves nothing) logs a Warning
  per Draft row / Deployed-artifact finding; the /raw banner shows them plus
  Running drivers reporting warn-only use (driver-health diagnostics).
- docs/security.md: exact replacement table (incl. the old form default
  env:MX_API_KEY, file:, un-namespaced secret:, casing changes), the
  ReopenAsync reconnect behaviour, the warn-only window and its removal.

Co-Authored-By: Claude <noreply@anthropic.com>
2026-11-11 02:07:33 -05:00
Joseph Doherty d3285af6b3 Merge feat/830-heidenhain-followups 2026-11-11 01:55:10 -05:00
Joseph Doherty 9aff90f884 fix(opcuaclient): never pass checkDomain for a None endpoint; document both SDK check sites (#816 review)
SDK 2.0.1 checks the server host name in TWO places: Session.OpenAsync (after
the chain is trusted) and DefaultSessionFactory.CreateChannelAsync, against the
certificate in the GetEndpoints reply and before any security-mode test. opc-plc
sends a certificate on its None endpoint, so `connect -u
opc.tcp://10.100.0.35:50000 --validate-host-name` was refused on an
unauthenticated channel, where a host check proves nothing.

All three call sites now pass checkDomain only when the selected endpoint's
SecurityMode != None: OpcUaEndpointSelection.CheckDomainFor for the driver
(OpenSessionOnEndpointAsync, so the failover sweep too) and the AdminUI browser;
the same expression in Client.Shared ConnectToEndpointAsync (shared by
failover). New live tests: driver None + flag on vs :50000 reads; browser None +
flag on vs the NAT fixture browses; the secure + flag refusals still hold.

Docs: name both check sites; a zone-qualified IPv6 host never matches (FindDomain
compares the zone too), so link-local needs validateServerHostName=false; the
match is loose (first label alone, subject DC= parts count, no wildcards), which
caveats the CA recommendation.
2026-11-11 01:53:01 -05:00
Joseph Doherty 8a6f0e3c84 fix(mtconnect): the Agent refuses count ABOVE bufferSize, not at it (#825)
Live gate against the cppagent 2.7 fixture: count=16384 with bufferSize
16384 streams; 16385 is OUT_OF_RANGE ('count' must be less than 16385).
Warn only when SampleCount > bufferSize, pin both sides of the edge
against the live Agent, and correct the #804 wording that said 'at or
above'.
2026-11-11 01:38:48 -05:00
Joseph Doherty 2b0eed1a8f feat(heidenhain): controller alarms as native Part 9 conditions; rediscovery on identity change (#831)
- ALARM: selectors (HeidenhainAlarmSelector, Contracts): ALARM:*, ALARM:OTHER,
  ALARM:CLASS=<r>[/GROUP=<r>][/NUMBER=<r>] (<r> = n | lo..hi; negative numbers allowed).
  An ALARM: address and the cross-driver alarm object only work together — the factory
  refuses either alone, so editor, save path, CSV and deploy inspector agree. Data type
  fixed Boolean; never a value (ReadAsync BadNotSupported), never in a data subscription.
- HeidenhainDriver implements IAlarmSource: one condition per ALARM: tag at its RawPath,
  driven by the SAME allowlisted R_RI 27/28 walk on the control's single loop, on the new
  Polling.AlarmMs cadence (default 2000 ms, only for controls carrying an ALARM: tag).
  FOCAS #647 transition rules (Raise / Retrigger / Clear), identity = class+group+number+text,
  deterministic display order, told-record on the driver instance, recorded only while a
  handler listens. A failed walk (refusal, overflow, comms loss) is quality-only, never a clear.
- Acknowledge is LOCAL-ONLY: AcknowledgeAsync is a no-op success (FOCAS/MTConnect posture);
  Lsv2ReadOnlyPolicy is unchanged — no ack or write request exists.
- Opt-in per tag, no driver flag (FOCAS #647 / MTConnect #626 precedent); Alarms/* arrays unchanged.
- IRediscoverable: a reconnect whose R_VR identity (model / NC / PLC / options) differs from
  the previous session raises OnRediscoveryNeeded once (advisory re-browse chip). No extra traffic.
- AdminUI: Heidenhain joins NativeAlarmModel's alarm drivers; the typed editor fixes ALARM:
  tags at Boolean and seeds/drops the alarm object with the selector; driver form gains
  "Alarm conditions (ms)". (HeidenhainDriverFormModel also carries the #830 credential-notice change.)

Tests: HeidenhainAlarmTests (grammar, pairing, rules, simulator e2e with the captured iTNC 530
alarm record, local-only ack = zero bytes sent, quality-only outage/refusal, rediscovery),
Runtime.Tests HeidenhainNativeAlarmEndToEndTests (real DriverHostActor condition), a
DraftValidator case, and the NativeAlarmDriverSetGuard now scans the Heidenhain driver.
2026-11-11 01:37:41 -05:00
Joseph Doherty 9742a66a9e docs(security): credential-reference namespaces + migration note (#815)
docs/security.md gains the canonical table of per-driver namespaces, the five
enforcement seams, the out-of-scope cases (appsettings ${secret:}, Sql
connectionStringRef) and the migration steps for existing configs. Galaxy,
OpcUaClient, Heidenhain, Configuration and driver-specs docs, the Galaxy
options XML docs, the drivers rule file and the docker-dev compose comment
point at it. docker-dev's env:GALAXY_MXGW_API_KEY stays valid (allowed legacy
prefix), so no rig change is needed.

Co-Authored-By: Claude <noreply@anthropic.com>
2026-11-11 01:37:25 -05:00
Joseph Doherty ae91e6b7f7 feat(opcuaclient): opt-in server-certificate host-name check; NAT fixture keeps its PKI (#816, #818)
#816 finding: the SDK checks that the server certificate names the host only
when a session is created with checkDomain:true. The driver, the AdminUI
browser and Client.Shared all called the 8-argument
DefaultSessionFactory.CreateAsync, which forwards checkDomain:false — so no
host-name check ever ran (why a trusted NAT cert for *.invalid browsed fine
over an IP in #813). When enabled the SDK (Session.OpenAsync ->
ValidateDomains -> FindDomain) compares ConfiguredEndpoint.EndpointUrl.IdnHost
— the CONFIGURED host, because OpcUaEndpointSelection/EndpointSelector
rewrote the advertised one — against the cert's DNS/IP SANs; secured
sessions only, and the auto-accept handler accepts a mismatch.

Policy: certificateValidation.validateServerHostName, tri-state. Unset =
OpcUaCertificateValidationOptions.DefaultValidateServerHostName = false, the
pre-#816 behaviour, so nothing changes on upgrade; true = validate the
configured host; false = explicit NAT/port-publish opt-out. Threaded into the
driver, the AdminUI browser (same JSON) and the AdminUI form (a select that
keeps "unset" unset), Client.Shared ConnectionSettings.ValidateServerHostName
and the CLI --validate-host-name. A driver refusal names the opt-out.

#818: the NAT fixture's /app/pki is on the opc-plc-nat-pki volume (verified:
same thumbprint across --force-recreate), its PUBLIC cert is committed as
Docker/nat/opc-plc-nat-server.der, and NatPublishedEndpointBrowseTests run
every secure browse on a throwaway PKI (internal browser seam): refusal with
nothing trusted, positive with the cert trusted (unset/false host check), and
a BadCertificateHostNameInvalid refusal with the check on. The trusted cert is
OPCUA_NAT_SERVER_CERT or the committed one (skip on thumbprint mismatch);
OPCUA_NAT_SERVER_CERT_TRUSTED is retired. CI exports its throwaway NAT
container's public cert. ServerHostNameValidationTests pin the driver against
the plain opc-plc. MIN_EXECUTED not raised (needs a hosted run): +6 in
v2-integration (440 -> 446), documented in scripts/ci/README.md.
2026-11-11 01:33:00 -05:00
Joseph Doherty 7af9c40e41 Merge feat/803-heidenhain-lsv2 2026-11-11 00:03:43 -05:00
Joseph Doherty f05e81710c docs(rules): Heidenhain credential namespace note (#803 review) 2026-11-10 23:54:44 -05:00
Joseph Doherty 8fb03d6a72 docs(heidenhain): driver page, README/probe index rows, drivers rule note (#803)
docs/drivers/Heidenhain.md: read-only boundary + allowlist table, transport, config,
credential references, address grammar (fixed catalog, PLC element/byte-count rules,
DATA types incl. FF Boolean), quality table, lifecycle/health/comms loss, probe, tests,
hardware-gated items and deliberate exclusions.

Also: InitializeAsync clears surviving subscription handles (the actor re-subscribes
after every successful init, so a kept handle would double-publish).
2026-11-10 23:23:25 -05:00
Joseph Doherty f17ff1e643 docs(mtconnect): Java reference parity matrix + live validation record (#804) 2026-11-10 23:08:42 -05:00
Joseph Doherty 2331f7ed14 feat(sql): opt-in writes — allowWrites + per-tag writable, single-row parameterised UPDATE (lmxopcua#753)
v2-ci / unit-tests (push) Waiting to run
allowWrites is no longer inert: SqlDriver implements IWritable, default off.

- Gates (outermost first): node Writable + WriteOperate (server), driver
  allowWrites, the tag's own TagConfig "writable": true (strict JSON boolean;
  anything else refuses the tag, and the deploy gate says why), and a model that
  names exactly one row — KeyValue, or WideRow by where-pair. topByTimestamp and
  Query are BadNotWritable.
- SqlWritePlanner: the only statement a write can become is
  UPDATE <table> SET <value col> = @v WHERE <key col> = @k, built from the
  catalog-gate-canonical identifiers (dialect-quoted) with value and key bound;
  value-before-key marker order so positional ODBC binds correctly. No authored
  write SQL, INSERT, UPSERT or procedure.
- Executed in a transaction that commits only when exactly one row was touched:
  0 rows -> BadNoData (never inserts), >1 -> rolled back, BadConfigurationError,
  unknown count -> rolled back, BadNotSupported. Bounded by operationTimeout;
  provider errors -> BadCommunicationError; health untouched.
- Discovery advertises Operate only where a write could succeed.
- AdminUI: driver form's Allow writes switch is live; tag editor gains a
  per-tag write switch; model validates writable strictly (parity corpus).
- Factory no longer warns "allowWrites ignored"; it logs whether writes are on.
- Tests: SqlWriteTests (planner SQL/bind order/injection/coercion, every gate,
  readback, absent key, two-row rollback, gate-rejected tag, discovery);
  MySqlReadTests write legs against InnoDB (keyed row only, two-row rollback,
  server-side grant refusal, default-off). Seed adds a setpoints table and the
  one UPDATE grant.

(lmxopcua#753)

Claude-Session: https://claude.ai/code/session_01GT67AEHd2Acv1YPwGfXEoV
2026-11-09 17:30:31 -05:00
Joseph Doherty 7f29a75edb feat(mqtt): read Sparkplug B array metrics as 1-D OPC UA arrays (lmxopcua#754)
v2-ci / unit-tests (push) Has been cancelled
R06:M-5 made Sparkplug *Array leaves unbindable because the ingestor refused
them. The ingestor now decodes the packed bytes_value of all 13 array types
(new SparkplugArrayCodec: LE fixed widths, int64-ms DateTime, uint32-count +
MSB-first BooleanArray, NUL-terminated UTF-8 StringArray; golden vectors are
the spec's examples, reproduced byte-for-byte by SparkplugNet 1.3.10) and
publishes a typed 1-D array through the scalar coercion, all-or-nothing.

Shape uses the fleet's one array model: the tag's platform isArray/arrayLength
keys (ValueRank=1, ArrayDimensions=[arrayLength] or [0] = variable length).
The driver reads them with TagConfigIntent's exact semantics and never rejects
on them, so composer, driver, authoring model and deploy inspector agree
(MqttArrayShapeParityTests). Shape mismatch (array metric on a scalar tag, or
the reverse) is BadTypeMismatch - a String tag no longer gets the packed
buffer as base64; a malformed body or a length differing from a declared
arrayLength is BadDecodingError (MTConnect #627 verdict). Template/DataSet
members of an array type take the same path. Arrays stay read-only (encoder
refuses them; MQTT is not in DriverArrayWriteSupport).

Browser: array leaves are bindable again and state isArray in AddressFields;
the commit mapper writes "isArray": true (no arrayLength). Discovery reports
IsArray/ArrayDim. Docs: Mqtt.md "Array metrics", design table, drivers rule.

Tests: codec (golden + malformed), ingest through the real codec + Dispatch
(the project's MiniBroker never forwards a PUBLISH, so there is no in-process
broker harness), browse, discovery, commit mapper, parity corpus. Mutating the
fix back out fails 18 MQTT + 9 AdminUI tests.

Claude-Session: https://claude.ai/code/session_01GT67AEHd2Acv1YPwGfXEoV
2026-11-09 13:58:38 -05:00
Joseph Doherty 8c6067b62e docs: move subsystem gotchas out of CLAUDE.md into path-scoped .claude/rules
Keeps the root CLAUDE.md lean by lazy-loading per-area detail (drivers, topology,
data-plane auth, alarms, historian) only when matching files are touched.
2026-10-15 06:14:58 -04:00