You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
[docs](adr): interim-ratify D4/D5 qaFit and preset calibration (closes OI-2, OI-4)
Establish the model-decision log (docs/adr/, per charter Section 14.4/14.5) and
record the interim ratification of the two open model-value items. No Domain
Advisor is appointed yet (charter D12), so the Owner ratifies using the
documented literature anchors — honestly framed as defensible expert defaults,
not empirically validated facts (empirical validation stays the v3.0 item; an
independent Domain Advisor may revise any value via a superseding ADR).
- docs/adr/README.md — the ADR log + process; distinguishes these model ADRs
(0001…) from the design spec's inline app ADRs (ADR-001…007).
- ADR-0001 — ratify the D4/D5 qaFit baseline (Cockburn/Martin/Bogard/Jackson/
Richards & Ford grounding); flags the weakest cells (SSR security, Hexagonal
security/interop) for future scrutiny. Closes OI-4 / DI-1.
- ADR-0002 — ratify the preset calibration; accepts the four sub-2% margins as
honest close calls with the standing "re-run verify-model on any change"
rule. Closes OI-2.
- Status updates: Model Data Sheet v0.8 (§4 D4/D5 and §6 now ✅ Ratified
(interim); new ✅ legend tier), SRS v0.7 (OI-2/OI-4 closed), design spec v0.5
(DoR items + DI-1 closed), charter v1.6 (ADR-log path clarified to docs/adr/).
- Synced cross-document version pointers; fixed the ADR relative links; both
guard scripts still pass (no numeric change).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
| 1.3 | 2026-06-11 | Editorial clarification (no scope change, no re-approval): designated the Section 22 solo-stage KPI table as the operative v1.0 target set (Section 4 = aspirational); aligned the sign-off date with the v1.2 baseline |
23
23
| 1.4 | 2026-06-11 | Scope clarification: a basic custom-configuration import/export (JSON) is in the MVP (matching Build Spec v3); the richer organization-level config import/export remains deferred to v2.0. Resolves a charter↔Build Spec↔SRS mismatch |
24
24
| 1.5 | 2026-06-12 | Reference hardening (editorial, no scope change): added peer-reviewed and SEI sources [16]–[24] (ADD technical report, SUS and its acceptability threshold, architectural design decisions, technical debt, GQM, Strangler Fig, MADR) and cited them inline where claims rely on them |
25
+
| 1.6 | 2026-06-13 | Editorial: clarified the ADR decision-log path to `docs/adr/` (Section 14.5), where the model-decision records now live |
25
26
26
27
---
27
28
@@ -383,7 +384,7 @@ decisions documents its own.)
383
384
384
385
**14.5 Onboarding new contributors.** Required files in the repo: `README.md`, `CONTRIBUTING.md`
385
386
(how to set up and contribute), `EXTENDING.md` (how to add factors/QAs/dimensions/anti-patterns/
386
-
translations), `CODE_OF_CONDUCT.md` (adopting the Contributor Covenant [13]), `ADR/` (decision
387
+
translations), `CODE_OF_CONDUCT.md` (adopting the Contributor Covenant [13]), [`docs/adr/`](../adr/) (decision
387
388
history), Issue & PR templates. Provide several `good first issue`s for newcomers.
388
389
389
390
**14.6 Recognition.** Record contributors (e.g. an `all-contributors` file), mention them in
| 0.4 | 2026-06-12 | Computation precision: linked the [Scoring Algorithm Specification](../03-blueprint/scoring-algorithm.md); pinned defaults exactly (budget = 2, the no-signal level of the inverted factor) in FR-FACT-5/AC-2; AC-2/AC-3 now state exact expected composites and the expected close-call flag; Section 5.3 presets machine-verified, IoT D5 target widened to SPA / SSR |
24
24
| 0.5 | 2026-06-12 | Reference hardening: added peer-reviewed and standard sources (SUS and its acceptability threshold, MADR, Strangler Fig, additive multi-attribute value model, ADD technical report) and cited them inline at NFR-USE-1, FR-OUT-1, FR-REC-12/13, and Section 5.1 |
25
25
| 0.6 | 2026-06-13 | Calibration-stability review: widened the e-commerce and IoT D4 targets to Hexagonal / Clean (the pair differs only on interoperability and ties exactly whenever that weight is 0); margins for every preset target are now machine-measured (Scoring Algorithm Section 9.4) |
26
+
| 0.7 | 2026-06-13 | Closed OI-2 and OI-4: the preset calibration and the D4/D5 `qaFit` vectors are interim-ratified ([ADR-0002](../adr/0002-ratify-preset-calibration.md), [ADR-0001](../adr/0001-ratify-d4-d5-qafit.md)) |
26
27
27
28
---
28
29
@@ -91,7 +92,7 @@ decision records. The authoritative scope (in/deferred/non-goals) is the charter
@@ -481,9 +482,9 @@ all UX-quality criteria met; KPIs K3 and K5 met at beta; no critical defects.
481
482
| # | Open issue | Owner | Target | Notes |
482
483
|---|---|---|---|---|
483
484
| OI-1 |~~Final factor count~~ — **Resolved: fixed at 14** (Build Spec Section 4 / [Model Data Sheet](../03-blueprint/model-data-sheet.md) Section 2) | Owner | Closed (2026-06-12) | Grouping also fixed in the Model Data Sheet |
484
-
| OI-2 | Preset factor-level values that hit the Section 5.3 outcome targets | Owner + Domain Advisor|M1–M2 |**Baseline levels now recorded**in [Model Data Sheet](../03-blueprint/model-data-sheet.md) Section 6; calibrate against [Section 5.3](#5-data--decision-model-requirements) & ADR-log|
485
+
| OI-2 |~~Preset factor-level values~~ — **Resolved: interim-ratified** ([ADR-0002](../adr/0002-ratify-preset-calibration.md)) | Owner | Closed (2026-06-13)|Levels in [Model Data Sheet](../03-blueprint/model-data-sheet.md) Section 6; all 25 targets machine-verified; independent Domain Advisor may revise|
485
486
| OI-3 | Whether the C4 Mermaid stub (FR-OUT-5) is in v1.0 or deferred | Owner | M1 | Currently **Could**-priority |
486
-
| OI-4 | D4/D5 `qaFit` values (documented as defensible defaults)|Domain Advisor | M1–M3 |**Baseline values now recorded**in [Model Data Sheet](../03-blueprint/model-data-sheet.md) Section 4; ratify & ADR-log per Charter Section 14.4|
487
+
| OI-4 |~~D4/D5 `qaFit` values~~ — **Resolved: interim-ratified** ([ADR-0001](../adr/0001-ratify-d4-d5-qafit.md))|Owner | Closed (2026-06-13) | Values in [Model Data Sheet](../03-blueprint/model-data-sheet.md) Section 4; independent Domain Advisor / empirical study (v3.0) may revise|
487
488
| OI-5 | Confirm/ratify the quantitative performance budgets in NFR-PERF-3 (bundle size, FCP, p95 interaction) | Engineer | M2 (Phase 3 design) | Interim targets already set in NFR-PERF-3; ratify against the real bundle |
488
489
| OI-6 | Empirical-validation study design (deferred per Charter Section 5) | Owner | v3.0 | Out of MVP scope; tracked for v3.0 |
| 0.2 | 2026-06-12 | Realigned to SRS v0.2 / Charter v1.4: added a resilience & edge-case design (Section 5.1) covering FR-EDGE-*, browser-baseline and performance-budget coverage, corrected the design-token source reference, and extended traceability |
22
22
| 0.3 | 2026-06-12 | Build-readiness: linked the new [Model Data Sheet](model-data-sheet.md) (frozen numeric model values), added a Definition-of-Ready gate (Section 11), and updated DI-1 to the recorded baseline |
23
23
| 0.4 | 2026-06-12 | Computation precision: linked the [Scoring Algorithm Specification](scoring-algorithm.md) as the exact contract for `lib/scoring.ts`/`lib/sensitivity.ts`; Definition of Ready updated — scoring computation pinned and preset calibration machine-verified |
24
+
| 0.5 | 2026-06-13 | Definition of Ready: D4/D5 `qaFit` and preset calibration interim-ratified ([ADR-0001](../adr/0001-ratify-d4-d5-qafit.md), [ADR-0002](../adr/0002-ratify-preset-calibration.md)); DI-1 closed |
24
25
25
26
---
26
27
@@ -307,7 +308,7 @@ These app-level decisions are recorded here; model-value changes follow the ADR
307
308
308
309
| # | Open issue | Notes |
309
310
|---|---|---|
310
-
| DI-1 | Ratify D4/D5 `qaFit` values| Baseline now recorded in the [Model Data Sheet](model-data-sheet.md) Section 4; Domain Advisor ratifies & ADR-logs (SRS OI-4)|
| DI-2 | Whether URL-hash state needs a compression library | Decide against a size budget (DI-4) |
312
313
| DI-3 | Final design-token values vs the prototype | Promote the prototype `:root` blocks to the token source of truth |
313
314
| DI-4 | Ratify the performance budgets against the real bundle | Interim targets already set in SRS NFR-PERF-3 (≤ 300 KB gzip, FCP ≤ 2 s, p95 ≤ 100 ms); inherits SRS OI-5 |
@@ -325,8 +326,8 @@ free of guesswork — each line is either fixed or has a usable baseline.
325
326
-[x]**Design tokens** fixed against the prototype (Section 6).
326
327
-[x]**Acceptance criteria** enumerated and ready to wire as tests (SRS Section 7, AC-1…12).
327
328
-[x]**Scoring computation pinned**: formulas, tie-breaking, close-call, sensitivity, rounding, and float-precision rules in the [Scoring Algorithm Specification](scoring-algorithm.md), with machine-verified fixtures (`node scripts/verify-model.mjs`).
328
-
-[x]**Preset levels pass the calibration test** against SRS 5.3 — all 25 targets machine-verified (SRS OI-2; Domain-Advisor ratification still pending).
329
-
-[]**D4/D5 `qaFit` ratified** by a Domain Advisor (SRS OI-4) — *baseline usable now*.
329
+
-[x]**Preset levels pass the calibration test** against SRS 5.3 — all 25 targets machine-verified, **interim-ratified** ([ADR-0002](../adr/0002-ratify-preset-calibration.md), closes SRS OI-2).
330
+
-[x]**D4/D5 `qaFit`interim-ratified** by the Owner ([ADR-0001](../adr/0001-ratify-d4-d5-qafit.md), closes SRS OI-4); an independent Domain Advisor / the v3.0 study may still revise the values.
330
331
-[x]**Factor content authored (EN/ID)**: labels, level labels, and help for all 14 factors — [Model Data Sheet Section 2.1](model-data-sheet.md) (Translator review pending).
331
332
-[x]**Option & message content authored (EN/ID)**: educational metadata for all 21 options, the 7 anti-pattern messages, and the 12 fitness-function templates — [Option Content Sheet](option-content-sheet.md) (Translator & Domain-Advisor review pending).
332
333
-[ ]**C4 stub** scoped in or out of v1.0 (SRS OI-3).
| 0.5 | 2026-06-13 | Linked the new [Option Content Sheet](option-content-sheet.md): the bilingual educational metadata for all 21 options, anti-pattern messages, and fitness-function templates are now authored |
25
25
| 0.6 | 2026-06-13 | Calibration-stability review: widened the e-commerce and IoT D4 targets to Hexagonal / Clean (exact tie whenever the interoperability weight is 0); target margins are now measured by the verification script, with the four sensitive targets documented in the [Scoring Algorithm Specification](scoring-algorithm.md) Section 9.4 |
26
26
| 0.7 | 2026-06-13 | Factor-content consistency (Section 2.1): made the Indonesian QA names uniform across all help texts (no more mixed EN/ID tokens) and established the baseline ID QA vocabulary; fixed two English level labels so they are truly verbatim from Build Spec Section 4 (lifespan "prototype", dataVolume "big data") |
27
+
| 0.8 | 2026-06-13 | Interim ratification: D4/D5 `qaFit` ([ADR-0001](../adr/0001-ratify-d4-d5-qafit.md)) and preset calibration ([ADR-0002](../adr/0002-ratify-preset-calibration.md)) accepted as v1.0 defaults by the Owner (interim Domain-Advisor role, charter D12); values remain editable, pending independent review and the v3.0 empirical study |
27
28
28
29
---
29
30
@@ -38,10 +39,12 @@ every value the scoring engine needs, in one place,** so Phase 4 development has
38
39
39
40
- 🔒 **Fixed** — taken verbatim from Build Spec v3; change only via the model-change ADR process
0 commit comments