[masarray]

Problem / evidence

Two actual F650 discovery PCAPNGs show ARSAS stopped GetNameList after 64 pages with moreFollows=true, omitting CSWI1–16 and eveGGIO1 typed LN roots despite 124/124 static DataSet memberships being known. Operator screenshot/log: Open SCL 67 report-backed values, IP Discovery 24; missing feedback and EVEGGIO ST semantics; both 6/6 static RCBs received reports. Engine #155 adds source-neutral DataSet-backed LN schema hints before the one existing type-probe pass.

One atomic consumer integration

• Base exact P7.6F: restore report-backed whole-FCD WYE/DEL current and voltage phases #494 head 3cf821428efd954e415e7455acbd2936ff0019cd, retaining the 13 phase magnitude fix, AuxV, UX recovery, canonical F650 command single-flight and evidence-graded diagnostics.
• Pin ARIEC61850 engine exact 43f8935eb636967359028a0d9df293aac1812d99, engine draft Fix IEC 61850 millisecond timestamp rounding #155 (CI-only engine P0: make FAT real-time and add clean evidence retest boundary [superseded by #159] #156 SUCCESS), without changing historical field/golden references.
• Switch Discovery to engine ProbeSmartAsync(session, discovery, options) overload; this accepts verified DataSet memberships beyond capped name pages and performs ONE bounded hierarchy GVA pass.
• Preserve MaxNameListPages=64, max additional DataSet-proven LN roots=48, max member hints=2048, negotiated max concurrency, association single-flight, zero cyclic MMS process polling, zero dynamic DataSet writes.
• Update beginner-friendly progress wording and add exact-source regression plus simple physical acceptance checklist.

Scope control

No decoder changes, no report RCB write-plan changes, no control operation / feedback code changes, no web UI, no guessed IEC names, no customer PCAP or IID in repository. All public tests synthetic.
No extra CI reruns manually triggered. Draft until exact-HEAD native Build ARSAS/tests and real F650 Open SCL vs IP Discovery captures qualify equality by exact leaf identities and protocol semantics. Naming of IED remains confidence-graded rather than fabricated. More complete engineering-model discovery is separate from static report member semantic closure.

Related ARSAS Issue #495, engine PR #155, consumer PR #494.

[masarray]

D28 archival: old F650/IEC 61850 interoperability prototype is no longer an open PR integration candidate. The operator's more recent successful evidence and accepted ARSAS baseline supersede the prior iteration; the operator will NOT perform a new capture to qualify historic branches. Closing this PR unmerged preserves exact HEAD 0bd1d493ebabb184f2813e2bb3bbc1ae8548a22a, all captured discussion and patch for traceability. Its unique code is not presumed present in the current RC, and closure is not a declaration of new R7/R10 hardware qualification. Working build, main branch, engine lock and stable release are untouched.