P5-01: Add drag/re-layout stress QA harness with saved metrics report #24

New Issue

Rbanh · 2026-02-19T13:41:28-05:00

Rbanh commented

2026-02-19 13:41:28 -05:00

Objective

Track real-world drag behavior and post-tidy quality with reproducible automated QA.

Scope

Extend browser QA flow with component drag stress (large displacement + repeated tidy).
Capture compile status metrics before drag, after drag, and after auto tidy.
Save machine-readable QA report artifact for triage.

Acceptance Criteria

QA run asserts no regressions on crossings/overlaps after drag+tidy.
Report artifact is generated for each run with metric snapshots.
CI can surface report on failures.

## Objective Track real-world drag behavior and post-tidy quality with reproducible automated QA. ## Scope - Extend browser QA flow with component drag stress (large displacement + repeated tidy). - Capture compile status metrics before drag, after drag, and after auto tidy. - Save machine-readable QA report artifact for triage. ## Acceptance Criteria - QA run asserts no regressions on crossings/overlaps after drag+tidy. - Report artifact is generated for each run with metric snapshots. - CI can surface report on failures.

Rbanh added this to the Phase 5 - Layout Robustness and JSON Import UX Hardening milestone 2026-02-19 13:41:53 -05:00

Rbanh commented

2026-02-19 13:42:20 -05:00

Initial QA notes collected for this milestone:

Current expanded sample compile metrics (local QA run): 0E, 1W, 0 crossings, 0 overlaps, 1 bends, 11 tie-nets, 2.86x detour.
Dense analog fixture metrics: 0E, 2W, 0 crossings, 0 overlaps, 4 bends, 6 tie-nets, 1.97x detour.
Main risk pattern: readability remains acceptable for crossing/overlap, but detour/label-tie behavior can still feel non-intuitive after manual drag + tidy.

Proposed hard targets for Phase 5 (initial):

crossings: 0
overlaps: 0
default sample detour <= 2.2x
dense analog detour <= 2.0x
post-drag tidy detour <= 2.4x

Initial QA notes collected for this milestone: - Current expanded sample compile metrics (local QA run): `0E, 1W, 0 crossings, 0 overlaps, 1 bends, 11 tie-nets, 2.86x detour`. - Dense analog fixture metrics: `0E, 2W, 0 crossings, 0 overlaps, 4 bends, 6 tie-nets, 1.97x detour`. - Main risk pattern: readability remains acceptable for crossing/overlap, but detour/label-tie behavior can still feel non-intuitive after manual drag + tidy. Proposed hard targets for Phase 5 (initial): - crossings: `0` - overlaps: `0` - default sample detour <= `2.2x` - dense analog detour <= `2.0x` - post-drag tidy detour <= `2.4x`

Rbanh commented

2026-02-19 14:00:41 -05:00

Delivered in commit 8b6c959.

What shipped:

Extended browser QA flow with drag stress and post-tidy checks in tests/ui-regression-runner.js.
Added machine-readable metrics artifact output/playwright/ui-metrics-report.json generated every UI run.
Captured initial, post-drag, post-drag-tidy, and dense-fixture metric snapshots.

Verification:

npm test passed
npm run test:ui passed
Report now includes thresholds + measured metrics for regression triage.

Delivered in commit `8b6c959`. What shipped: - Extended browser QA flow with drag stress and post-tidy checks in `tests/ui-regression-runner.js`. - Added machine-readable metrics artifact `output/playwright/ui-metrics-report.json` generated every UI run. - Captured initial, post-drag, post-drag-tidy, and dense-fixture metric snapshots. Verification: - `npm test` passed - `npm run test:ui` passed - Report now includes thresholds + measured metrics for regression triage.

Rbanh closed this issue

2026-02-19 14:00:50 -05:00

Sign in to join this conversation.

No Label

No Milestone

No project

No Assignees

1 Participants

Notifications

Due Date

The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: Rbanh/schemeta#24