[masarray]

Stacked on R1A PR #136 and milestone #129. Do not merge independently into main.

Problem fixed

The Dataset Signals → bound RCB selection could be made from the canonical offline/discovered model, but entering Reports attaches the live report service. That attach rebuilds the live DataSet/RCB inventory and previously reset selection to row 0, so the operator's exact static DataSet choice could silently drift.

R1B behavior

• Persist the chosen canonical static DataSet reference as a report route.
• On live report-service attach/re-attach, restore only an RCB whose live probe succeeded and whose DatSet binding matches that exact canonical reference; when several RCBs bind the same DataSet, use the existing pool selector's best selectable/preferred candidate rather than first-row order.
• If there is no verified live bound RCB, leave reporting unselected/disabled and surface a diagnostic instead of choosing another RCB.
• Selecting another DataSet or manually choosing another RCB clears the armed route, so user intent always wins.
• UI now calls the action Static Reporting… and shows a compact STATIC · <DataSet> indicator before the still-explicit Enable + GI command.
• No automatic GI, no automatic RCB enable, no DatSet write, and no cyclic MMS polling fallback is introduced.

Behavioral QA

ied_report_workspace_qa now selects the Dataset Signals route, disconnects/re-attaches the Reports association, verifies the same canonical DataSet + a live-probed selectable RCB are restored, then proceeds to the existing explicit GI test. Success emits STATIC_REPORT_ROUTE_RESTORE_PASS.

Acceptance: exact-head Qt/C++/Release Hardening green. Real AA1E1F06R4 acceptance remains required for full R1.

• Rendered Browser QA can now save and upload an exact 1024px Browser screenshot from the same functional route test; this replaces relying on the Simulator-tab screenshot for Browser UX review.
• Qt/Release workflow concurrency now deduplicates push + PR validation by head branch so stale duplicate jobs no longer consume runners.

[masarray]

Stage-main stacked dependency audit 2026-10-23

Parent R1A source PR #136 is closed unmerged solely as redundant stacked review tracking. Verified: #136 head 1d3899e is a true ancestor of this OPEN #137 head b47ac16 (behind=0). Original feature branch remains intact; this PR still targets feat/browser-dataset-signals-r1a and is not independently mergeable into main as the whole R1A+R1B stack without deliberate validated integration. You now own review of inherited R1A as well as this R1B route preservation; track actual Open SCL, Dataset Signals, bound RCB, field and Windows/Qt/Release acceptance in OPEN issue #129. This closure does not imply merge, R0 acceptance or hardware FAT. Keep #130 open: it is diverged with distinct commits.

[masarray]

Stage→Main stacked Workbench source audit · 2026-10-23

GitHub ancestry: CLOSED unmerged PR #127 head 3a3ceef9149fc0dd31b27f4c9232418a68fae0e6 is a true ancestor of this still-OPEN cumulative PR #137 head b47ac165f4ae9a3927cb81641cefc88df3e477dc (ahead=45, behind=0); earlier CLOSED unmerged #136 R1A is also ancestor (ahead=27, behind=0). Source branches intact, nothing from these closures automatically merged into public main.

This #137 now carries R0 Browser recovery + R1A canonical static Dataset Signals + R1B exact-bound-RCB route. Old PR base feat/browser-dataset-signals-r1a must not be mistaken for a complete eventual main integration. Deliberate full-stack restack/review is needed, with new exact-head Qt/Windows+1024px and physical AA1E1F06R4 acceptance in OPEN #128/#129. Independent #130 diverges and has not been closed.

[masarray]

Historical PR reconciliation / archival — 2026-10-24

Disposition: CLOSED AS SUPERSEDED HISTORICAL SOURCE — NOT MERGED, NOT ACCEPTED AS COMPLETE. Original review head preserved: b47ac165f4ae9a3927cb81641cefc88df3e477dc. This candidate is significantly diverged from current main, so merging the old branch would risk reverting newer Workbench, IED Browser and Process Bus integration.

Remaining requirement: exact static DataSet/RCB choice preserved through live Reporting attach, fail closed on mismatches. Canonical open feature owner(s): #129. Cross-PR migration index: #308.

All source commits remain accessible through this PR. Port only behavior proven missing in the present mainline, in reviewed and tested small patches with exact-commit CI and external evidence where required. Closed PR is an archival/reconciliation status, not a claim that functionality passed a test. No M3.H Stable binary, tag or checksum is changed.