fdp parity dashboard
Differential survey of @knowvah/dot-engine fdp xdot output vs the native dot -K fdp -Txdot oracle over the dot-track SVG-conformant corpus set (semantic draw-op comparison at ±0.01 — see test/golden/compare-xdot.ts; per docs/conformance.md, not byte equality). Regenerate: npx tsx test/corpus/engine-walk.ts fdp && npx tsx test/corpus/parity-report.ts.
Summary
- Surveyed: 762 (generated 2026-07-25T04:48:55.247Z)
- pass: 609 (79.9%) · diverged (tracked): 0 · accepted (documented, won't-fix): 4 · accepted (A1-drift class): 144
- oracle-error: 3 · port-error: 0 · timeout: 2
Accepted deltas (4) — documented, not chased
Deliberate, root-caused differences we have chosen not to make conformant. Source of truth: test/corpus/accepted-divergences-engines.json; rationale in Known divergences. Excluded from the diverged table below.
| id | #diffs | class | bound | ref |
|---|---|---|---|---|
1879 | 19496 | A1 | up to 64kpt (whole-layout basin) | plans/decision-journal.md 2026-07-24 T4b: fdp chaos class, perturbation-proven on family |
2620 | 10779 | A1 | up to 5.4kpt, 30/30 nodes | plans/decision-journal.md 2026-07-24 T4b: chaos proven — node width +0.001in moves port-own layout max 3390pt |
graphs-badvoro | 33264 | A1 | median 1433pt (whole-layout basin) | plans/decision-journal.md 2026-07-24 T4b: chaos proven — ONE edge len +0.0001 moves port-own layout median 322pt |
241_0 | 446 | A9 | 11 numeric draw-op diffs confined to edge 0->1#0's unfilled_bezier (maxΔ 3.39pt), under injected-identical node positions ⇒ pathplan routing. Same graph and CDT cocircular incircle 1-ULP tie (V8 vs Apple libm) accepted for circo/twopi/sfdp 241_0; exact-rational incircle 185/185. Levers (fmadd triang.ts, hypot route.ts:198) applied; the tie is irreducible. | known-divergences.md#a9-sfdp-fp-ties |
Accepted class: A1-drift — computed, not enumerated
Membership is computed at report time from the injection-attribution harness output (D2) — every diverged id whose native pre-routing position exonerates it (verdict: drift-exonerated) is subtracted from the Diverged table below and counted in Summary; an id that starts passing outright leaves the class silently on the next report regen.
- A1-drift: 144 members — full per-id evidence in
attribution-fdp.json. Rationale: Known divergences.
Diverged (0)
(none)
Errors and timeouts (5)
| id | status | message |
|---|---|---|
2222 | oracle-error | spawnSync dot ETIMEDOUT |
2239 | oracle-error | Command failed: dot -K fdp -Txdot https://gitlab.com/graphviz/graphviz/-/blob/main/tests/2239.dot |
2717 | oracle-error | Command failed: dot -K fdp -Txdot https://gitlab.com/graphviz/graphviz/-/blob/main/tests/2717.dot |
1652 | timeout | |
2108 | timeout |
oracle errors: 13 native-crash (documented, excluded) / 3 timeout-flake (excluded this run, note to retry) Passing ids (609) are omitted for brevity — the full roster is in parity-fdp.json.