Team technology radar — working trial 2.1
A native Bracken Vale workshop tool with eight invented proposals, four quadrants and four editable adoption-ring labels. Projects, roles, trial notes and dates are fictional. This local state is independent of the other applications.
Implemented
- Nominate a tool/practice with project, quadrant, ring, owner, review date, evidence and next experiment.
- Review an existing nomination; every ring move and every recorded review requires a rationale and stores the author role, fictional reporting date and evidence snapshot.
- Record separate sponsor and practitioner suggested rings and reasons. Switch the plotted viewpoint without modifying the agreed placement or durable revision.
- Inspect the whole radar through its numbered table; all changes use keyboard-operable forms. Graph positions are derived from rings, with angle used only for legibility.
- Rename the rings while retaining stable ring IDs and history. Inspect overdue reviews against an adjustable scenario date.
- Shared shell supplies local persistence, import/export and reset.
Reference checks
Run node --test apps/technology-radar/app.test.mjs. Nine tests check JSON roundtrip/reset fixture integrity, mandatory rationale and history, nomination completion, viewpoint isolation, malformed import/invalid dates, escaped values, unique labelled form controls and chronological history/review constraints.
Moving BVR-TECH2 from R2 to R1 adds a second history entry containing R2 → R1, the owner and evidence. A blank rationale rejects the entire change and returns the unchanged original state. Practitioner view changes the plotted coordinates but leaves recorded rings and revision unchanged.
Limits and simpler practice
Positions are explicit workshop judgements, never approval of fitness, security or value. Evidence is authored text, not independently verified. A copied record or role name does not prove that a real sponsor/practitioner participated. There are no shared nominations, votes, live reviews, AI assessments or connections to equipment. Eight proposals provide a readable cross-quadrant sample; no fixed proposal count was required. An ordinary trial register may be enough; the radar supplies a visual conversation aid using the same table data. Human adoption benefit has not been measured.
Reporting dates cannot move before a recorded nomination or review. Imported histories must be chronological and no later than the reporting date. The next review date must be on or after the latest recorded review; new nominations and reviews require a next date on or after their reporting date. Existing overdue records remain visible when time advances.