[masarray]

M2.14 field acceptance feedback — operator UI + honest IEC evidence

Base: verified #579 07b7b182 (10/10 CI), stacked above #577 (DO hierarchy), #575 (icons/first-run), #573 (A SSOT stress) and B #570 (scalar/vector guard). No parent branch, engine pin, static Report, MMS writes or public stable changes.

Field feedback and targeted correction

• Navigation was actually clipped by the runtime layout math, not the initial XAML width. Widen the responsive eight-button shell to available desktop space (bounded 842–1190 DIP); preserve selected-pill calculations using actual shell width. No duplicate tab control.
• Use the operator-supplied identical wireless SVG contour for IEC 61850 Explorer and GOOSE; use the supplied distinct filled monitor contour for Live Monitor. Existing path licensing and other operator icons remain intact.
• Remove "independent, not paired" from Value cells; retain clear separate real magnitude and angle observations, with mismatch reason in provenance. The phasor symbol ∠ is reserved for actual compatible Report/Read samples; never print false pairs or fake missing values.
• IEC 61850-7-3 Vector.ang normative unit is degrees. Render ° for exact cVal/instCVal.ang.f in MX when no conflicting or overriding explicit unit exists. Never infer angle for Hz, TotW or unrelated ang.f; unit CF for magnitude still requires model/read evidence.
• SGCB "empty" was exacerbated by presentation selecting only FC=SG from the live discovery model, ignoring metadata available in an attached SCL. On SGCB pages only, fall back to exact-domain SCL SG directory if the live SG inventory is absent, with an explicit SCL declaration only, not proof of active value or MMS read authorization. When only SCL SE exists, label SE as editing declarations, not as active SG. Live SG declarations take precedence. Read-plan remains restricted to inventory-proven targets and must not probe SCL-only references.
• Negative and positive tests for supplied nav icons, angle semantics and stale/mismatched samples, exact-domain SG fallback, no implicit SG read, and IEC-specific degree display.

Still open

Real CF.units magnitude metadata and read access are relay-dependent; unsupported reads remain Not read. The screenshot BCUGE still requires physical validation of value/angle provenance and SG entries. This PR does not claim full reference GUI parity. Exact-SHA Windows CI plus physical BCUGE/ZIV acceptance are required before review/promotion.

[masarray]

PR #580 field GUI corrections now have a green descendant build: draft PR #581 at a9e6c1c4a4dbe1c26b661e4a4fe709109f105ebe, one test-only follow-up commit over #580 head 96f1c80301af. Old M2.13 vector regression falsely expected cVal.ang.f to omit the IEC normative degree; changed only its test expectation, not the renderer. Build ARSAS #38699911526 PASS: 2,002/2,002 app tests (includes the four WPF STA cases that had timed out in prior CI), 19 pinned engine, 9 native CTest, 6 packaged managed, portable smoke. Exact-SHA unexpired ARSAS-win-x64-portable-single-exe artifact ID 11956446253.

Field acceptance still needs physical BCUGE/ZIV: the SCL fallback only proves SG declarations (not live active settings); CF magnitude units need exact verified data; vector magnitude/angle cannot display as paired when timestamp/source authority mismatches. No engine pin or main/stable change. This is the proper candidate for next screenshot comparison, not red #580 SHA.

[masarray]

D27 retired redundant historical PR, source is already in integrated RC. GitHub ancestry 96f1c80301af43903f8defff954dc4478578eeb5..d4598619b1b8fc21487cf5beb83a8a8284d0e9cf: ahead_by=1376, behind_by=0; no source commits omitted. Closed as absorbed into RC history, not as a direct merge. Canonical integration remains #727; physical/engine acceptance and mainline #729 stay separate. Do not infer field release authority.