[masarray]

M2.14 — fix real SG parameter values, not a cosmetic placeholder

Stack #560 → #563 (Thread A) → #567 (A/B validated) → #568 (read-only SGCB/GoCB validated) → this PR. Owns Browser presentation and explicit SG read plan only.

Root cause

The SGCB detail pane grouped all SG + SE leaves into a synthetic DO root and set its ValueReference only for a single-attribute DO. On a real relay with multiple SG/SE attribute declarations this results in an unbound Not read root even when exact SG leaf observations exist. It also reads SE edit-group fields unnecessarily when choosing active SG inspection.

Fix

• Project exact active FC=SG parameter branches directly via the existing lazy, schema-constrained AttributeRows projector, e.g. ATCC1.BndWid.setMag [SG] linked to ...$SG$BndWid$setMag$f, and f [SG] on expansion.
• Preserve canonical source/FC, unit provenance, lazy descendants, and association fences; do not infer active group from SE, change group, reserve, or write.
• Prefer unique verified primitive f if both a structured parent and its primitive are in the declaration.
• Explicit selected SGCB Read targets active SG plus inventory-proven read-only SGCB status; no SE edit-group sweep. Global Read All semantics are unchanged.
• Four negative/positive regression tests and updated presentation baseline; no engine pin, Report/GI/cyclic MMS, SSOT, or stable branch changes.

Acceptance

Require exact-SHA Windows compile, full ARSAS regression, pinned engine/native CTest, artifact verification. A screenshot from a physically connected ZIV AVR is still required to prove what SG attributes the device actually permits read-only access to. Other CF unit issues will be handled under Thread A's authority rather than a competing unit decoder.

[masarray]

D9 source/test unique-delta audit — KEEP OPEN pending SG/SE presentation authority

Historic HEAD 020e1b0272460fc0b1e9d28e40a9139f6df6a84e is an ancestor of RC 817c3910e8223fa2ac531a4426ada5d2d4bbbe50 (ahead 1395, behind 0), but ancestry is not disposition proof.

All original four paths inspected: Models/IedBrowserPresentation.cs old blob c031b2f36b8903df6c796eea3601f3bf1a10cb77 → RC 1b40ab6d53f7056944be507cefebc156054c2207; Models/IedBrowserReadPlan.cs 59f267fa74cdfa5a22f388aacb8d95d7bb991785 → 69bb7157e41b325f2de55a2d5e0eea75ef683bfb; active-setting regression file 2532b043f17188ed0c3a2425593730219aab469c → fb50ef57422537345fcb1f42b0df65da3c77b7bb (all four test methods remain). Original IedBrowserPresentationTests.cs 9a8f602cf62ebaf949f5669a7e8525a1fdd90ecf → RC 204c746e2d1ffd9d1975b2099dcb2f7919218e9c (27 test attributes, but unrelated COMTRADE test and navigation order evolved).

Semantic delta not presumed equivalent: original edit-only SGCB explicitly displays "No active-group [SG]", whereas current RC test expects "No model-proven SG or SCL SE" and presentation includes provenance-qualified SCL SE fallback; source/read-plan have evolved. Need actual accepted GUI owner decision on distinction between verified active SG, SCL static SE, and live MMS read authority, plus regression parity. No forced old behavior/revert, no closure, no branch deletion; do not treat SCL SE as an observed active group or authorize Write.

[masarray]

D27 PR consolidation — closed as already included in integrated RC, not recorded as a direct PR merge.

Exact ancestry proof: PR HEAD 020e1b0272460fc0b1e9d28e40a9139f6df6a84e is an ancestor of current RC d2d0ca731061824102058756fba8599ddfa2342e (GitHub compare: ahead_by=1398, behind_by=0). No source commits from this PR are dropped. Its content is already staged in Integrated RC #727; eventual mainline promotion remains tracked under #729.

Closing this historical integration lane reduces duplicate open PRs without merging its old base into current RC. Physical IEC 61850 R7/R10 testing, open field issues, trial engine qualification, protected main and stable release retain their independent acceptance gates.