[masarray]

Field-driven code fix, one atomic commit, no manual CI reruns

Parent: exact ARSAS #505 head 5e55fa2022bf3f79f1d747a176e75365306bdcd9. The user tested exact merged source 95dcb825a806f72f565198100f7b28d74c7b66dc (merge of #505 into #504) with immutable engine 8b0c0f7e695b016834e40a6f068994dcc945cb1a. No engine changes here.

Real BCUGE evidence: same-PC KM_Test capture 180/180 GOOSE frames, 3 publishers, typed CSWI6 value change, known quality bits; 67/67 static report values with 6/6 exact RCB ingress and zero cyclic MMS process polling. This is the new acceptance base and disproves an ongoing adapter/capture failure.

Changes

• GooseTimelineEventPolicy: event-display diagnostic categories coalesce repeated numeric GOOSE Gap/TAL variations; keep raw per-frame diagnostics, actual capture counts, payload/stNum changes, and explicit Show retransmissions. Initial publisher detection remains one per actual stream. No protocol decoder changes.
• GooseCanonicalLeafProjection: only exact same full LD+LN+DO FCDA owner permits folding adjacent or non-adjacent .q into .stVal engineering row; normalize optional MMS $ST$ notation and preserve independent raw ordered DataSet values. Orphan/ambiguous q is kept visible rather than guessed; explicit 14→7 existing regression remains, with additional F650-shaped fixture.
• GooseSubscriberLiteView: widen inspector column 360→420, Value and Quality grids, compact stable event rows, legible Quality tones and tooltips; no large cards or rearchitecture.
• SntpUserSettingsStore: user-local atomic sntp-server.json, persist exact interface ID+IP and last successful enabled toggle. Restore only when same live local NIC/IP exists; never silently bind another adapter, never persist per-packet telemetry or overwrite closing state. Retain previous explicit startup ON only with exact match; otherwise off. Existing standalone SNTP and heartbeat unchanged.
• Synthetic regression tests for normal/retransmission semantic coalescing, exact FCDA pairing across MMS notation, safe SNTP persistence/corrupt input.

CI/performance: one commit fda92c6b1869c29b0382e25f1dbbe2bab7c68140 triggers natural PR CI once; no workflow dispatch/retries. Do not merge to main or release until Windows build, visual UX at 100/125/150% DPI and same-PC BCUGE event/quality/NTP restart acceptance. R10 physical mainline gate unchanged.

Known scope: this does not claim all diagnostic warnings are false, nor suppress evidence of genuine repeated state transitions. SCL metadata for an unrelated C264 publisher may remain unbound; require the correct publisher CID/SCL. Quality is only joined when exact source identity exists. P2 typed FCD report issue remains ARIEC61850 #159.

[masarray]

BCUGE field acceptance basis: 180 captured / 180 GOOSE / 0 other / 3 streams; one CSWI action also appeared on multiple publisher streams. Raw wire retransmissions/test/supervision diagnostics must remain, but numeric Gap/TAL changes must not flood the operator event grid. Quality must be assigned only by exact same full FCDA owner; raw ordered members stay intact. BCUGE MMS static path showed 67/67 values, 6/6 exact RCB ingress, zero cyclic polling. SNTP restore must require same NIC ID + IPv4. Validate 100/125/150% DPI and application portable EXE before merge/release. Source: user's local screenshots and diagnostic; no customer raw captures committed.

[masarray]

Exact-head CI qualification — 2026-10-10

• Commit fda92c6b1869c29b0382e25f1dbbe2bab7c68140, 1 commit ahead / 0 behind P8 C264 P3A: context-aware exact NIC selection for same-PC GOOSE capture #505; nine scoped files.
• Build ARSAS #38033124690: SUCCESS. Windows Release build, ARSAS regression tests, pinned-engine validation, portable single EXE publish, Windows executable smoke test and artifact upload all succeeded.
• Windows x64 portable single EXE artifact #11663182469 — correct candidate for repeat BCUGE GUI/UX test.
• Smart Discovery Field Capture #38033124675: SUCCESS, separate packaging route; use the main portable artifact above for this feature test.
• Golden Provenance, Golden Budget, Merge Execution, Repeat-Run Stability, Production Promotion, SV evidence: SUCCESS. Mainline Readiness guard stays closed/red as expected, pending unrelated R10 physical authority — do not disable it.
• No manual CI rerun, no engine changes, no main/release merge.

Field acceptance still outstanding

1. Exact same-PC KM-TEST capture: 3 stream identities, actual frame count continues to increment and raw retransmissions remain decoded.
2. With Show retransmissions OFF, numeric Gap/TAL changes alone do not append new event rows; real stNum/payload changes are retained, stream-specific test/supervision evidence remains in Inspector.
3. Check GOOSE1/GOOSE2: exact companion CSWI.Pos.stVal + CSWI.Pos.q collapses to one engineering row with readable Quality while raw 4 wire entries remain accessible; unknown C264 publisher does not acquire false labels.
4. Inspector value/quality columns readable at 100/125/150% DPI.
5. Select PC IP, toggle NTP ON/OFF, restart ARSAS; the same interface/IP and last valid enabled state restore correctly; if IP is removed, no arbitrary NIC is chosen. Must not claim relay time synchronization from SNTP serving.
6. Preserve 67/67 report display, six exact static RCB ingress streams and zero cyclic MMS polling on BCUGE.

Remain DRAFT until those operator tests are accepted.

[masarray]

New BCUGE follow-up implementation is isolated in draft PR #507 (#507), exact one-commit base #506 HEAD fda92c6. Field build of #506 showed remaining lexicographic CSWI1/10 sorting, q still separate/Quality blank, duplicate per-stream GOOSE events and stale historical first-frame inspector. #507 implements natural view sorting, strict full+typed adjacent q pairing with LD conflict rejection, per-stream process change with conservative cross-stream correlation, and live-follow/historical pause. No engine pin, static DataSet, NTP transport, command path or mainline gate changed.

[masarray]

Disposition: Superseded — verified stacked milestone consolidation; no mainline merge and no physical acceptance.

Canonical replacement: #543, open staged tip at 6285867780b8ea79c231d72a3819e9e5f99748d2. This PR HEAD fda92c6b1869c29b0382e25f1dbbe2bab7c68140 is a commit ancestor (verified GitHub compare: ahead=231, behind=0). Later changes in the stack can intentionally revise UI behavior; keep the canonical tip OPEN.

Unique delta: GOOSE timeline quality/event coalescing and persistent SNTP settings. Audited PR patch: 9-file patch contains GooseTimelineEventPolicyTests.cs, GooseCanonicalEngineeringTests.cs and SntpUserSettingsStoreTests.cs. Original changes and tests remain in reachable commit history of #543; no commit from this PR is orphaned. Historical ancestry alone does not certify current semantics, which remain subject to final exact-file and field validation at integration.

Safety: Only close the obsolete intermediate PR tracking record. The PR source branch, original commits, test/evidence history, engine lock, main, stable v1.6.42 and CI/physical gates are untouched. No release/engine authority claimed.

Follow-up: Later GOOSE PRs correct presentation choices; preserve raw-wire provenance and test SNTP persistence on consolidated candidate. Canonical integration/qualification remains explicitly unfinished in #543 and successor release-gated workstreams.