pcbjam/tests/tools/screenshots/changelog.ts

172 lines
7.5 KiB
TypeScript
Raw Normal View History

feat(tests): screenshot regression + Discord review, perf-tracked New tooling in tests/tools/screenshots/ (TypeScript via tsx): - compare.ts: one pixelmatch engine (AA-excluded), connected-component "where to look" boxes, old|new+boxes|heatmap triptych, per-engine floors. - promote.ts: churn-free updater — overwrite a baseline only when decoded pixels differ beyond the floor, copying CI bytes verbatim (no re-encode churn); pulls a CI run via `gh run download` or a local --from dir. - post-discord.ts: always-on CI-on-main report (SHA + e2e status + the track-only runtime-perf table), then screenshot triptychs, batched + size-capped + flood-collapsed + 429-aware. - perf-report.ts: perf table with Δ vs the previous main run (via gh). - changelog.ts: no-build git-history baseline differ (Discord trigger B). - noise.ts / gen-manifest.ts: calibration + manifest generation. CI wiring: - wasm-build.yml: post-test step runs the gate + report on the already- produced test-results (no extra build); report-only (continue-on-error), posts only on push to main, inert without DISCORD_WEBHOOK_URL. - ci-ubicloud.yml: secrets: inherit (pass the webhook through). - screenshot-changelog.yml: ~30s no-build changelog on baseline changes. screenshot-manifest.json: canonical 354-name set + best-effort engine tags (313 chromium-swiftshader / 41 firefox-llvmpipe). Normalize scale:'device'->'css' across 18 spec files (no-op at CI DSF=1) so committed baselines are uniformly css-scaled. Design: CI's Linux render is the single source of truth; no pinned container (accept rare env drift -> re-promote); dev commits via promote. Replaces the byte-cmp compare-screenshots.sh + file-size-proxy update-baseline-screenshots.sh (kept for now until the first re-baseline). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-01 18:01:15 +02:00
/**
* Baseline changelog (Discord trigger B) — no build, no GPU.
*
* On push to main, diff the committed baseline PNGs between two commits and post
* a "Baseline changelog" to Discord: a triptych (old | new+boxes | heatmap) for
* each CHANGED file, the image for each ADDED, and a titled line for each REMOVED.
* "added / removed" only mean anything as a git-history diff, which is why this is
* separate from the re-render drift gate.
*
* CLI (from tests/):
* tsx tools/screenshots/changelog.ts [--base REV] [--head REV] [--dry-run] [--force]
*/
import * as fs from 'fs';
import * as os from 'os';
import * as path from 'path';
import { execFileSync } from 'child_process';
import { PNG } from 'pngjs';
e2e/CI: dual-engine suites, per-engine screenshots, SwiftShader retired, prod web suite, CI-coverage gate Squash of experiment/ff-big-modules vs main. Big-module routing removed: native-EH shrank kicad_editor below SpiderMonkey's x86-64 code budget (runs 29355049705/29356152413 green on stock Firefox), so BIG_MODULE_SPECS routing and the baseline-only-JIT crutch are gone — kicad-firefox and kicad-chromium both run the full suite, with the module compiled the way real users' browsers compile it. Per-engine screenshots end to end: stableShot/shotPath write test-results/<engine>/<name>.png; baselines move to baseline-screenshots/{chromium,firefox}/ and the whole tools/screenshots pipeline (compare/promote/manifest/spec-map/changelog/Discord) keys on <engine>/<name>. Previously Firefox and Chromium renders of one spec overwrote each other and Firefox renders were never actually gated. Seeded from CI run 29421380806 (92 new firefox baselines, +24 chromium web-suite shots); manifest generated from the baseline tree. One merged playwright.config.ts (kicad/asyncify/coroutine/perf as projects); ~25 dead npm scripts dropped. The web suite is gated in CI for the first time ever (4 rotted specs fixed, 5 broken lib-bridge specs triaged as fixme in docs/features/web-e2e-rot/); cheap lint step after npm ci; last 26 blind-sleep violations fixed. SwiftShader retired: CI Chromium renders WebGL on ANGLE → Mesa llvmpipe (--use-gl=angle --use-angle=gl --ignore-gpu-blocklist; the blocklist flag is mandatory — llvmpipe is blocklisted and WebGL is silently unavailable without it) in BOTH configs. Under WORKERS=4 congestion SwiftShader transiently failed the first post-board-load draw and the recovery cascade ended in a silent permanent Cairo fallback — that engine flip was the "~1.2% changedRatio both directions" occ-export baseline flake. Validated 160/160 across two 80-repeat rigs; full analysis in docs/features/wx-parity-bugs/occ-export-context-eviction.md. Chromium baselines shift slightly on llvmpipe — promote once from the first green run. Deflakes the new coverage exposed: presence baselines settle before capture; presence fixtures declare current file formats; perf gets its own outputDir so CI evidence survives; occ-export settles the board paint before the export dialog; menu-item waits (waitForRenderedByLabel before clickMenuItem) in 4 specs + the TESTING.md rule. Web suite runs the PROD build, in parallel: webServer becomes backend `start` + the standalone's e2e:preview (build-preview.mjs: link-wasm → stash the public/wasm symlink aside during vite build, build-demo.mjs's move — then vite preview as the persistent server). The wasm middleware serves /wasm/* in preview and emits COOP/COEP/CORP itself (a pthread worker script's own response must carry COEP or Chrome kills it with ERR_BLOCKED_BY_RESPONSE). VITE_* flags bake at build time; VITE_ALLOW_USER_OVERRIDE joins turbo globalEnv. fullyParallel + default workers: 5.2m → 1.4m. Determinism fixes the parallel run exposed: shared-page specs become serial groups; locks.spec grabs alice's exact item via the new kicadCollabTestSelectByUuid hook (cross-tab "first footprint" order is not a ysync invariant); quit specs poll page.url() (quit supersedes its own navigation — NS_BINDING_ABORTED on Firefox). Suite: 51 passed / 12 skipped / 0 failed in 1.6m. CI-coverage gate (lint:ci-coverage): every tests/**/*.spec.ts must be reachable from the npm scripts the workflows invoke — scraped from .github/workflows/, resolved through package.json, coverage asked from playwright --list itself. Rules: uncovered-spec + orphan-project (with a documented LOCAL_ONLY_PROJECTS allowlist). Gating next to lint:determinism; 138 spec files / 13 projects accounted for. Product fixes kept from the investigations (reachable on real GPUs too): wx 7799fd1be5 — paint flags clear before dispatch + Invalidate always propagates; kicad 3dcfea5e45 — SwiftShader pass-boundary flush + per-instance font texture + first-frame GL-error drain (GAL recovery recovers instead of falling back to Cairo) + the user-facing eeschema switch navigates again under __EMSCRIPTEN__ (project-sync's FaceRegistered gate had rerouted it into the hidden sync player; caught by the newly-gated web suite). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018eUxiPApHgGiu9NFyQfhAq
2026-07-17 12:10:40 +02:00
import { BASELINE_ROOT, DIFF_OUT_DIR, floorFor, labelText, splitKey, LABEL } from './config';
feat(tests): screenshot regression + Discord review, perf-tracked New tooling in tests/tools/screenshots/ (TypeScript via tsx): - compare.ts: one pixelmatch engine (AA-excluded), connected-component "where to look" boxes, old|new+boxes|heatmap triptych, per-engine floors. - promote.ts: churn-free updater — overwrite a baseline only when decoded pixels differ beyond the floor, copying CI bytes verbatim (no re-encode churn); pulls a CI run via `gh run download` or a local --from dir. - post-discord.ts: always-on CI-on-main report (SHA + e2e status + the track-only runtime-perf table), then screenshot triptychs, batched + size-capped + flood-collapsed + 429-aware. - perf-report.ts: perf table with Δ vs the previous main run (via gh). - changelog.ts: no-build git-history baseline differ (Discord trigger B). - noise.ts / gen-manifest.ts: calibration + manifest generation. CI wiring: - wasm-build.yml: post-test step runs the gate + report on the already- produced test-results (no extra build); report-only (continue-on-error), posts only on push to main, inert without DISCORD_WEBHOOK_URL. - ci-ubicloud.yml: secrets: inherit (pass the webhook through). - screenshot-changelog.yml: ~30s no-build changelog on baseline changes. screenshot-manifest.json: canonical 354-name set + best-effort engine tags (313 chromium-swiftshader / 41 firefox-llvmpipe). Normalize scale:'device'->'css' across 18 spec files (no-op at CI DSF=1) so committed baselines are uniformly css-scaled. Design: CI's Linux render is the single source of truth; no pinned container (accept rare env drift -> re-promote); dev commits via promote. Replaces the byte-cmp compare-screenshots.sh + file-size-proxy update-baseline-screenshots.sh (kept for now until the first re-baseline). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-01 18:01:15 +02:00
import { comparePair, type Report } from './compare';
import { loadPng, savePng, withBottomLabel } from './image-ops';
import { buildSpecResolver } from './spec-map';
feat(tests): screenshot regression + Discord review, perf-tracked New tooling in tests/tools/screenshots/ (TypeScript via tsx): - compare.ts: one pixelmatch engine (AA-excluded), connected-component "where to look" boxes, old|new+boxes|heatmap triptych, per-engine floors. - promote.ts: churn-free updater — overwrite a baseline only when decoded pixels differ beyond the floor, copying CI bytes verbatim (no re-encode churn); pulls a CI run via `gh run download` or a local --from dir. - post-discord.ts: always-on CI-on-main report (SHA + e2e status + the track-only runtime-perf table), then screenshot triptychs, batched + size-capped + flood-collapsed + 429-aware. - perf-report.ts: perf table with Δ vs the previous main run (via gh). - changelog.ts: no-build git-history baseline differ (Discord trigger B). - noise.ts / gen-manifest.ts: calibration + manifest generation. CI wiring: - wasm-build.yml: post-test step runs the gate + report on the already- produced test-results (no extra build); report-only (continue-on-error), posts only on push to main, inert without DISCORD_WEBHOOK_URL. - ci-ubicloud.yml: secrets: inherit (pass the webhook through). - screenshot-changelog.yml: ~30s no-build changelog on baseline changes. screenshot-manifest.json: canonical 354-name set + best-effort engine tags (313 chromium-swiftshader / 41 firefox-llvmpipe). Normalize scale:'device'->'css' across 18 spec files (no-op at CI DSF=1) so committed baselines are uniformly css-scaled. Design: CI's Linux render is the single source of truth; no pinned container (accept rare env drift -> re-promote); dev commits via promote. Replaces the byte-cmp compare-screenshots.sh + file-size-proxy update-baseline-screenshots.sh (kept for now until the first re-baseline). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-01 18:01:15 +02:00
import { buildAttachments, paginate, postMessage } from './post-discord';
function git(root: string, args: string[]): string {
return execFileSync('git', ['-C', root, ...args], { encoding: 'utf8' }).trim();
}
function gitBlob(root: string, rev: string, repoPath: string): Buffer | null {
try {
return execFileSync('git', ['-C', root, 'show', `${rev}:${repoPath}`], { maxBuffer: 1 << 28 });
} catch {
return null;
}
}
function parseArgs(argv: string[]): Record<string, string | boolean> {
const out: Record<string, string | boolean> = {};
for (let i = 0; i < argv.length; i++) {
const a = argv[i];
if (a === '--dry-run') out.dryRun = true;
else if (a === '--force') out.force = true;
else if (a === '--base') out.base = argv[++i];
else if (a === '--head') out.head = argv[++i];
}
return out;
}
async function main(): Promise<void> {
const args = parseArgs(process.argv.slice(2));
const cwd = process.cwd();
const repoRoot = git(cwd, ['rev-parse', '--show-toplevel']);
const head = (args.head as string) || 'HEAD';
let base = (args.base as string) || '';
if (!base) {
try {
base = git(cwd, ['rev-parse', `${head}^`]);
} catch {
console.log('[changelog] no parent commit — nothing to diff');
return;
}
}
e2e/CI: dual-engine suites, per-engine screenshots, SwiftShader retired, prod web suite, CI-coverage gate Squash of experiment/ff-big-modules vs main. Big-module routing removed: native-EH shrank kicad_editor below SpiderMonkey's x86-64 code budget (runs 29355049705/29356152413 green on stock Firefox), so BIG_MODULE_SPECS routing and the baseline-only-JIT crutch are gone — kicad-firefox and kicad-chromium both run the full suite, with the module compiled the way real users' browsers compile it. Per-engine screenshots end to end: stableShot/shotPath write test-results/<engine>/<name>.png; baselines move to baseline-screenshots/{chromium,firefox}/ and the whole tools/screenshots pipeline (compare/promote/manifest/spec-map/changelog/Discord) keys on <engine>/<name>. Previously Firefox and Chromium renders of one spec overwrote each other and Firefox renders were never actually gated. Seeded from CI run 29421380806 (92 new firefox baselines, +24 chromium web-suite shots); manifest generated from the baseline tree. One merged playwright.config.ts (kicad/asyncify/coroutine/perf as projects); ~25 dead npm scripts dropped. The web suite is gated in CI for the first time ever (4 rotted specs fixed, 5 broken lib-bridge specs triaged as fixme in docs/features/web-e2e-rot/); cheap lint step after npm ci; last 26 blind-sleep violations fixed. SwiftShader retired: CI Chromium renders WebGL on ANGLE → Mesa llvmpipe (--use-gl=angle --use-angle=gl --ignore-gpu-blocklist; the blocklist flag is mandatory — llvmpipe is blocklisted and WebGL is silently unavailable without it) in BOTH configs. Under WORKERS=4 congestion SwiftShader transiently failed the first post-board-load draw and the recovery cascade ended in a silent permanent Cairo fallback — that engine flip was the "~1.2% changedRatio both directions" occ-export baseline flake. Validated 160/160 across two 80-repeat rigs; full analysis in docs/features/wx-parity-bugs/occ-export-context-eviction.md. Chromium baselines shift slightly on llvmpipe — promote once from the first green run. Deflakes the new coverage exposed: presence baselines settle before capture; presence fixtures declare current file formats; perf gets its own outputDir so CI evidence survives; occ-export settles the board paint before the export dialog; menu-item waits (waitForRenderedByLabel before clickMenuItem) in 4 specs + the TESTING.md rule. Web suite runs the PROD build, in parallel: webServer becomes backend `start` + the standalone's e2e:preview (build-preview.mjs: link-wasm → stash the public/wasm symlink aside during vite build, build-demo.mjs's move — then vite preview as the persistent server). The wasm middleware serves /wasm/* in preview and emits COOP/COEP/CORP itself (a pthread worker script's own response must carry COEP or Chrome kills it with ERR_BLOCKED_BY_RESPONSE). VITE_* flags bake at build time; VITE_ALLOW_USER_OVERRIDE joins turbo globalEnv. fullyParallel + default workers: 5.2m → 1.4m. Determinism fixes the parallel run exposed: shared-page specs become serial groups; locks.spec grabs alice's exact item via the new kicadCollabTestSelectByUuid hook (cross-tab "first footprint" order is not a ysync invariant); quit specs poll page.url() (quit supersedes its own navigation — NS_BINDING_ABORTED on Firefox). Suite: 51 passed / 12 skipped / 0 failed in 1.6m. CI-coverage gate (lint:ci-coverage): every tests/**/*.spec.ts must be reachable from the npm scripts the workflows invoke — scraped from .github/workflows/, resolved through package.json, coverage asked from playwright --list itself. Rules: uncovered-spec + orphan-project (with a documented LOCAL_ONLY_PROJECTS allowlist). Gating next to lint:determinism; 138 spec files / 13 projects accounted for. Product fixes kept from the investigations (reachable on real GPUs too): wx 7799fd1be5 — paint flags clear before dispatch + Invalidate always propagates; kicad 3dcfea5e45 — SwiftShader pass-boundary flush + per-instance font texture + first-frame GL-error drain (GAL recovery recovers instead of falling back to Cairo) + the user-facing eeschema switch navigates again under __EMSCRIPTEN__ (project-sync's FaceRegistered gate had rerouted it into the hidden sync player; caught by the newly-gated web suite). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018eUxiPApHgGiu9NFyQfhAq
2026-07-17 12:10:40 +02:00
// Repo-relative baseline root prefix (e.g. tests/baseline-screenshots).
const prefix = path.relative(repoRoot, path.join(cwd, BASELINE_ROOT));
const diff = git(repoRoot, ['diff', '--name-status', base, head, '--', prefix]);
feat(tests): screenshot regression + Discord review, perf-tracked New tooling in tests/tools/screenshots/ (TypeScript via tsx): - compare.ts: one pixelmatch engine (AA-excluded), connected-component "where to look" boxes, old|new+boxes|heatmap triptych, per-engine floors. - promote.ts: churn-free updater — overwrite a baseline only when decoded pixels differ beyond the floor, copying CI bytes verbatim (no re-encode churn); pulls a CI run via `gh run download` or a local --from dir. - post-discord.ts: always-on CI-on-main report (SHA + e2e status + the track-only runtime-perf table), then screenshot triptychs, batched + size-capped + flood-collapsed + 429-aware. - perf-report.ts: perf table with Δ vs the previous main run (via gh). - changelog.ts: no-build git-history baseline differ (Discord trigger B). - noise.ts / gen-manifest.ts: calibration + manifest generation. CI wiring: - wasm-build.yml: post-test step runs the gate + report on the already- produced test-results (no extra build); report-only (continue-on-error), posts only on push to main, inert without DISCORD_WEBHOOK_URL. - ci-ubicloud.yml: secrets: inherit (pass the webhook through). - screenshot-changelog.yml: ~30s no-build changelog on baseline changes. screenshot-manifest.json: canonical 354-name set + best-effort engine tags (313 chromium-swiftshader / 41 firefox-llvmpipe). Normalize scale:'device'->'css' across 18 spec files (no-op at CI DSF=1) so committed baselines are uniformly css-scaled. Design: CI's Linux render is the single source of truth; no pinned container (accept rare env drift -> re-promote); dev commits via promote. Replaces the byte-cmp compare-screenshots.sh + file-size-proxy update-baseline-screenshots.sh (kept for now until the first re-baseline). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-01 18:01:15 +02:00
if (!diff) {
console.log('[changelog] no baseline changes between', base.slice(0, 7), 'and', head);
return;
}
const outDir = path.join(cwd, DIFF_OUT_DIR);
fs.mkdirSync(outDir, { recursive: true });
const tmp = fs.mkdtempSync(path.join(os.tmpdir(), 'changelog-'));
const report: Report = {
generatedFor: process.env.GITHUB_SHA || head,
changed: [],
added: [],
removed: [],
unchangedCount: 0,
driftLikely: false,
};
const { specFor } = buildSpecResolver(cwd);
// Load a git blob (a committed PNG at `rev`) as a decoded PNG, or null if absent.
const blobPng = (rev: string, p: string): PNG | null => {
const b = gitBlob(repoRoot, rev, p);
if (!b) return null;
const f = path.join(tmp, `${rev.slice(0, 7)}-${path.basename(p)}`);
fs.writeFileSync(f, b);
return loadPng(f);
};
e2e/CI: dual-engine suites, per-engine screenshots, SwiftShader retired, prod web suite, CI-coverage gate Squash of experiment/ff-big-modules vs main. Big-module routing removed: native-EH shrank kicad_editor below SpiderMonkey's x86-64 code budget (runs 29355049705/29356152413 green on stock Firefox), so BIG_MODULE_SPECS routing and the baseline-only-JIT crutch are gone — kicad-firefox and kicad-chromium both run the full suite, with the module compiled the way real users' browsers compile it. Per-engine screenshots end to end: stableShot/shotPath write test-results/<engine>/<name>.png; baselines move to baseline-screenshots/{chromium,firefox}/ and the whole tools/screenshots pipeline (compare/promote/manifest/spec-map/changelog/Discord) keys on <engine>/<name>. Previously Firefox and Chromium renders of one spec overwrote each other and Firefox renders were never actually gated. Seeded from CI run 29421380806 (92 new firefox baselines, +24 chromium web-suite shots); manifest generated from the baseline tree. One merged playwright.config.ts (kicad/asyncify/coroutine/perf as projects); ~25 dead npm scripts dropped. The web suite is gated in CI for the first time ever (4 rotted specs fixed, 5 broken lib-bridge specs triaged as fixme in docs/features/web-e2e-rot/); cheap lint step after npm ci; last 26 blind-sleep violations fixed. SwiftShader retired: CI Chromium renders WebGL on ANGLE → Mesa llvmpipe (--use-gl=angle --use-angle=gl --ignore-gpu-blocklist; the blocklist flag is mandatory — llvmpipe is blocklisted and WebGL is silently unavailable without it) in BOTH configs. Under WORKERS=4 congestion SwiftShader transiently failed the first post-board-load draw and the recovery cascade ended in a silent permanent Cairo fallback — that engine flip was the "~1.2% changedRatio both directions" occ-export baseline flake. Validated 160/160 across two 80-repeat rigs; full analysis in docs/features/wx-parity-bugs/occ-export-context-eviction.md. Chromium baselines shift slightly on llvmpipe — promote once from the first green run. Deflakes the new coverage exposed: presence baselines settle before capture; presence fixtures declare current file formats; perf gets its own outputDir so CI evidence survives; occ-export settles the board paint before the export dialog; menu-item waits (waitForRenderedByLabel before clickMenuItem) in 4 specs + the TESTING.md rule. Web suite runs the PROD build, in parallel: webServer becomes backend `start` + the standalone's e2e:preview (build-preview.mjs: link-wasm → stash the public/wasm symlink aside during vite build, build-demo.mjs's move — then vite preview as the persistent server). The wasm middleware serves /wasm/* in preview and emits COOP/COEP/CORP itself (a pthread worker script's own response must carry COEP or Chrome kills it with ERR_BLOCKED_BY_RESPONSE). VITE_* flags bake at build time; VITE_ALLOW_USER_OVERRIDE joins turbo globalEnv. fullyParallel + default workers: 5.2m → 1.4m. Determinism fixes the parallel run exposed: shared-page specs become serial groups; locks.spec grabs alice's exact item via the new kicadCollabTestSelectByUuid hook (cross-tab "first footprint" order is not a ysync invariant); quit specs poll page.url() (quit supersedes its own navigation — NS_BINDING_ABORTED on Firefox). Suite: 51 passed / 12 skipped / 0 failed in 1.6m. CI-coverage gate (lint:ci-coverage): every tests/**/*.spec.ts must be reachable from the npm scripts the workflows invoke — scraped from .github/workflows/, resolved through package.json, coverage asked from playwright --list itself. Rules: uncovered-spec + orphan-project (with a documented LOCAL_ONLY_PROJECTS allowlist). Gating next to lint:determinism; 138 spec files / 13 projects accounted for. Product fixes kept from the investigations (reachable on real GPUs too): wx 7799fd1be5 — paint flags clear before dispatch + Invalidate always propagates; kicad 3dcfea5e45 — SwiftShader pass-boundary flush + per-instance font texture + first-frame GL-error drain (GAL recovery recovers instead of falling back to Cairo) + the user-facing eeschema switch navigates again under __EMSCRIPTEN__ (project-sync's FaceRegistered gate had rerouted it into the hidden sync player; caught by the newly-gated web suite). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018eUxiPApHgGiu9NFyQfhAq
2026-07-17 12:10:40 +02:00
// Engine-qualified key (`<engine>/<name>.png`) from a repo path under the baseline root.
const keyOf = (repoPath: string): string => path.relative(prefix, repoPath).split(path.sep).join('/');
// Save a single captioned image (added/removed) and record it in the report.
e2e/CI: dual-engine suites, per-engine screenshots, SwiftShader retired, prod web suite, CI-coverage gate Squash of experiment/ff-big-modules vs main. Big-module routing removed: native-EH shrank kicad_editor below SpiderMonkey's x86-64 code budget (runs 29355049705/29356152413 green on stock Firefox), so BIG_MODULE_SPECS routing and the baseline-only-JIT crutch are gone — kicad-firefox and kicad-chromium both run the full suite, with the module compiled the way real users' browsers compile it. Per-engine screenshots end to end: stableShot/shotPath write test-results/<engine>/<name>.png; baselines move to baseline-screenshots/{chromium,firefox}/ and the whole tools/screenshots pipeline (compare/promote/manifest/spec-map/changelog/Discord) keys on <engine>/<name>. Previously Firefox and Chromium renders of one spec overwrote each other and Firefox renders were never actually gated. Seeded from CI run 29421380806 (92 new firefox baselines, +24 chromium web-suite shots); manifest generated from the baseline tree. One merged playwright.config.ts (kicad/asyncify/coroutine/perf as projects); ~25 dead npm scripts dropped. The web suite is gated in CI for the first time ever (4 rotted specs fixed, 5 broken lib-bridge specs triaged as fixme in docs/features/web-e2e-rot/); cheap lint step after npm ci; last 26 blind-sleep violations fixed. SwiftShader retired: CI Chromium renders WebGL on ANGLE → Mesa llvmpipe (--use-gl=angle --use-angle=gl --ignore-gpu-blocklist; the blocklist flag is mandatory — llvmpipe is blocklisted and WebGL is silently unavailable without it) in BOTH configs. Under WORKERS=4 congestion SwiftShader transiently failed the first post-board-load draw and the recovery cascade ended in a silent permanent Cairo fallback — that engine flip was the "~1.2% changedRatio both directions" occ-export baseline flake. Validated 160/160 across two 80-repeat rigs; full analysis in docs/features/wx-parity-bugs/occ-export-context-eviction.md. Chromium baselines shift slightly on llvmpipe — promote once from the first green run. Deflakes the new coverage exposed: presence baselines settle before capture; presence fixtures declare current file formats; perf gets its own outputDir so CI evidence survives; occ-export settles the board paint before the export dialog; menu-item waits (waitForRenderedByLabel before clickMenuItem) in 4 specs + the TESTING.md rule. Web suite runs the PROD build, in parallel: webServer becomes backend `start` + the standalone's e2e:preview (build-preview.mjs: link-wasm → stash the public/wasm symlink aside during vite build, build-demo.mjs's move — then vite preview as the persistent server). The wasm middleware serves /wasm/* in preview and emits COOP/COEP/CORP itself (a pthread worker script's own response must carry COEP or Chrome kills it with ERR_BLOCKED_BY_RESPONSE). VITE_* flags bake at build time; VITE_ALLOW_USER_OVERRIDE joins turbo globalEnv. fullyParallel + default workers: 5.2m → 1.4m. Determinism fixes the parallel run exposed: shared-page specs become serial groups; locks.spec grabs alice's exact item via the new kicadCollabTestSelectByUuid hook (cross-tab "first footprint" order is not a ysync invariant); quit specs poll page.url() (quit supersedes its own navigation — NS_BINDING_ABORTED on Firefox). Suite: 51 passed / 12 skipped / 0 failed in 1.6m. CI-coverage gate (lint:ci-coverage): every tests/**/*.spec.ts must be reachable from the npm scripts the workflows invoke — scraped from .github/workflows/, resolved through package.json, coverage asked from playwright --list itself. Rules: uncovered-spec + orphan-project (with a documented LOCAL_ONLY_PROJECTS allowlist). Gating next to lint:determinism; 138 spec files / 13 projects accounted for. Product fixes kept from the investigations (reachable on real GPUs too): wx 7799fd1be5 — paint flags clear before dispatch + Invalidate always propagates; kicad 3dcfea5e45 — SwiftShader pass-boundary flush + per-instance font texture + first-frame GL-error drain (GAL recovery recovers instead of falling back to Cairo) + the user-facing eeschema switch navigates again under __EMSCRIPTEN__ (project-sync's FaceRegistered gate had rerouted it into the hidden sync player; caught by the newly-gated web suite). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018eUxiPApHgGiu9NFyQfhAq
2026-07-17 12:10:40 +02:00
const saveSingle = (img: PNG, key: string, status: 'added' | 'removed'): void => {
const suffix = status === 'added' ? 'added' : 'removed';
e2e/CI: dual-engine suites, per-engine screenshots, SwiftShader retired, prod web suite, CI-coverage gate Squash of experiment/ff-big-modules vs main. Big-module routing removed: native-EH shrank kicad_editor below SpiderMonkey's x86-64 code budget (runs 29355049705/29356152413 green on stock Firefox), so BIG_MODULE_SPECS routing and the baseline-only-JIT crutch are gone — kicad-firefox and kicad-chromium both run the full suite, with the module compiled the way real users' browsers compile it. Per-engine screenshots end to end: stableShot/shotPath write test-results/<engine>/<name>.png; baselines move to baseline-screenshots/{chromium,firefox}/ and the whole tools/screenshots pipeline (compare/promote/manifest/spec-map/changelog/Discord) keys on <engine>/<name>. Previously Firefox and Chromium renders of one spec overwrote each other and Firefox renders were never actually gated. Seeded from CI run 29421380806 (92 new firefox baselines, +24 chromium web-suite shots); manifest generated from the baseline tree. One merged playwright.config.ts (kicad/asyncify/coroutine/perf as projects); ~25 dead npm scripts dropped. The web suite is gated in CI for the first time ever (4 rotted specs fixed, 5 broken lib-bridge specs triaged as fixme in docs/features/web-e2e-rot/); cheap lint step after npm ci; last 26 blind-sleep violations fixed. SwiftShader retired: CI Chromium renders WebGL on ANGLE → Mesa llvmpipe (--use-gl=angle --use-angle=gl --ignore-gpu-blocklist; the blocklist flag is mandatory — llvmpipe is blocklisted and WebGL is silently unavailable without it) in BOTH configs. Under WORKERS=4 congestion SwiftShader transiently failed the first post-board-load draw and the recovery cascade ended in a silent permanent Cairo fallback — that engine flip was the "~1.2% changedRatio both directions" occ-export baseline flake. Validated 160/160 across two 80-repeat rigs; full analysis in docs/features/wx-parity-bugs/occ-export-context-eviction.md. Chromium baselines shift slightly on llvmpipe — promote once from the first green run. Deflakes the new coverage exposed: presence baselines settle before capture; presence fixtures declare current file formats; perf gets its own outputDir so CI evidence survives; occ-export settles the board paint before the export dialog; menu-item waits (waitForRenderedByLabel before clickMenuItem) in 4 specs + the TESTING.md rule. Web suite runs the PROD build, in parallel: webServer becomes backend `start` + the standalone's e2e:preview (build-preview.mjs: link-wasm → stash the public/wasm symlink aside during vite build, build-demo.mjs's move — then vite preview as the persistent server). The wasm middleware serves /wasm/* in preview and emits COOP/COEP/CORP itself (a pthread worker script's own response must carry COEP or Chrome kills it with ERR_BLOCKED_BY_RESPONSE). VITE_* flags bake at build time; VITE_ALLOW_USER_OVERRIDE joins turbo globalEnv. fullyParallel + default workers: 5.2m → 1.4m. Determinism fixes the parallel run exposed: shared-page specs become serial groups; locks.spec grabs alice's exact item via the new kicadCollabTestSelectByUuid hook (cross-tab "first footprint" order is not a ysync invariant); quit specs poll page.url() (quit supersedes its own navigation — NS_BINDING_ABORTED on Firefox). Suite: 51 passed / 12 skipped / 0 failed in 1.6m. CI-coverage gate (lint:ci-coverage): every tests/**/*.spec.ts must be reachable from the npm scripts the workflows invoke — scraped from .github/workflows/, resolved through package.json, coverage asked from playwright --list itself. Rules: uncovered-spec + orphan-project (with a documented LOCAL_ONLY_PROJECTS allowlist). Gating next to lint:determinism; 138 spec files / 13 projects accounted for. Product fixes kept from the investigations (reachable on real GPUs too): wx 7799fd1be5 — paint flags clear before dispatch + Invalidate always propagates; kicad 3dcfea5e45 — SwiftShader pass-boundary flush + per-instance font texture + first-frame GL-error drain (GAL recovery recovers instead of falling back to Cairo) + the user-facing eeschema switch navigates again under __EMSCRIPTEN__ (project-sync's FaceRegistered gate had rerouted it into the hidden sync player; caught by the newly-gated web suite). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018eUxiPApHgGiu9NFyQfhAq
2026-07-17 12:10:40 +02:00
const rel = path.join(DIFF_OUT_DIR, `${key.replace('/', '_')}.${suffix}.png`);
savePng(path.join(cwd, rel), withBottomLabel(img, labelText(status, key, specFor(splitKey(key).name)), LABEL.colors[status]));
report[status].push({ name: key, image: rel });
};
feat(tests): screenshot regression + Discord review, perf-tracked New tooling in tests/tools/screenshots/ (TypeScript via tsx): - compare.ts: one pixelmatch engine (AA-excluded), connected-component "where to look" boxes, old|new+boxes|heatmap triptych, per-engine floors. - promote.ts: churn-free updater — overwrite a baseline only when decoded pixels differ beyond the floor, copying CI bytes verbatim (no re-encode churn); pulls a CI run via `gh run download` or a local --from dir. - post-discord.ts: always-on CI-on-main report (SHA + e2e status + the track-only runtime-perf table), then screenshot triptychs, batched + size-capped + flood-collapsed + 429-aware. - perf-report.ts: perf table with Δ vs the previous main run (via gh). - changelog.ts: no-build git-history baseline differ (Discord trigger B). - noise.ts / gen-manifest.ts: calibration + manifest generation. CI wiring: - wasm-build.yml: post-test step runs the gate + report on the already- produced test-results (no extra build); report-only (continue-on-error), posts only on push to main, inert without DISCORD_WEBHOOK_URL. - ci-ubicloud.yml: secrets: inherit (pass the webhook through). - screenshot-changelog.yml: ~30s no-build changelog on baseline changes. screenshot-manifest.json: canonical 354-name set + best-effort engine tags (313 chromium-swiftshader / 41 firefox-llvmpipe). Normalize scale:'device'->'css' across 18 spec files (no-op at CI DSF=1) so committed baselines are uniformly css-scaled. Design: CI's Linux render is the single source of truth; no pinned container (accept rare env drift -> re-promote); dev commits via promote. Replaces the byte-cmp compare-screenshots.sh + file-size-proxy update-baseline-screenshots.sh (kept for now until the first re-baseline). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-01 18:01:15 +02:00
for (const line of diff.split('\n')) {
const parts = line.split('\t');
const status = parts[0][0]; // M/A/D/R
const repoPath = parts[status === 'R' ? 2 : 1];
const oldPath = status === 'R' ? parts[1] : repoPath;
e2e/CI: dual-engine suites, per-engine screenshots, SwiftShader retired, prod web suite, CI-coverage gate Squash of experiment/ff-big-modules vs main. Big-module routing removed: native-EH shrank kicad_editor below SpiderMonkey's x86-64 code budget (runs 29355049705/29356152413 green on stock Firefox), so BIG_MODULE_SPECS routing and the baseline-only-JIT crutch are gone — kicad-firefox and kicad-chromium both run the full suite, with the module compiled the way real users' browsers compile it. Per-engine screenshots end to end: stableShot/shotPath write test-results/<engine>/<name>.png; baselines move to baseline-screenshots/{chromium,firefox}/ and the whole tools/screenshots pipeline (compare/promote/manifest/spec-map/changelog/Discord) keys on <engine>/<name>. Previously Firefox and Chromium renders of one spec overwrote each other and Firefox renders were never actually gated. Seeded from CI run 29421380806 (92 new firefox baselines, +24 chromium web-suite shots); manifest generated from the baseline tree. One merged playwright.config.ts (kicad/asyncify/coroutine/perf as projects); ~25 dead npm scripts dropped. The web suite is gated in CI for the first time ever (4 rotted specs fixed, 5 broken lib-bridge specs triaged as fixme in docs/features/web-e2e-rot/); cheap lint step after npm ci; last 26 blind-sleep violations fixed. SwiftShader retired: CI Chromium renders WebGL on ANGLE → Mesa llvmpipe (--use-gl=angle --use-angle=gl --ignore-gpu-blocklist; the blocklist flag is mandatory — llvmpipe is blocklisted and WebGL is silently unavailable without it) in BOTH configs. Under WORKERS=4 congestion SwiftShader transiently failed the first post-board-load draw and the recovery cascade ended in a silent permanent Cairo fallback — that engine flip was the "~1.2% changedRatio both directions" occ-export baseline flake. Validated 160/160 across two 80-repeat rigs; full analysis in docs/features/wx-parity-bugs/occ-export-context-eviction.md. Chromium baselines shift slightly on llvmpipe — promote once from the first green run. Deflakes the new coverage exposed: presence baselines settle before capture; presence fixtures declare current file formats; perf gets its own outputDir so CI evidence survives; occ-export settles the board paint before the export dialog; menu-item waits (waitForRenderedByLabel before clickMenuItem) in 4 specs + the TESTING.md rule. Web suite runs the PROD build, in parallel: webServer becomes backend `start` + the standalone's e2e:preview (build-preview.mjs: link-wasm → stash the public/wasm symlink aside during vite build, build-demo.mjs's move — then vite preview as the persistent server). The wasm middleware serves /wasm/* in preview and emits COOP/COEP/CORP itself (a pthread worker script's own response must carry COEP or Chrome kills it with ERR_BLOCKED_BY_RESPONSE). VITE_* flags bake at build time; VITE_ALLOW_USER_OVERRIDE joins turbo globalEnv. fullyParallel + default workers: 5.2m → 1.4m. Determinism fixes the parallel run exposed: shared-page specs become serial groups; locks.spec grabs alice's exact item via the new kicadCollabTestSelectByUuid hook (cross-tab "first footprint" order is not a ysync invariant); quit specs poll page.url() (quit supersedes its own navigation — NS_BINDING_ABORTED on Firefox). Suite: 51 passed / 12 skipped / 0 failed in 1.6m. CI-coverage gate (lint:ci-coverage): every tests/**/*.spec.ts must be reachable from the npm scripts the workflows invoke — scraped from .github/workflows/, resolved through package.json, coverage asked from playwright --list itself. Rules: uncovered-spec + orphan-project (with a documented LOCAL_ONLY_PROJECTS allowlist). Gating next to lint:determinism; 138 spec files / 13 projects accounted for. Product fixes kept from the investigations (reachable on real GPUs too): wx 7799fd1be5 — paint flags clear before dispatch + Invalidate always propagates; kicad 3dcfea5e45 — SwiftShader pass-boundary flush + per-instance font texture + first-frame GL-error drain (GAL recovery recovers instead of falling back to Cairo) + the user-facing eeschema switch navigates again under __EMSCRIPTEN__ (project-sync's FaceRegistered gate had rerouted it into the hidden sync player; caught by the newly-gated web suite). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018eUxiPApHgGiu9NFyQfhAq
2026-07-17 12:10:40 +02:00
const key = keyOf(repoPath);
feat(tests): screenshot regression + Discord review, perf-tracked New tooling in tests/tools/screenshots/ (TypeScript via tsx): - compare.ts: one pixelmatch engine (AA-excluded), connected-component "where to look" boxes, old|new+boxes|heatmap triptych, per-engine floors. - promote.ts: churn-free updater — overwrite a baseline only when decoded pixels differ beyond the floor, copying CI bytes verbatim (no re-encode churn); pulls a CI run via `gh run download` or a local --from dir. - post-discord.ts: always-on CI-on-main report (SHA + e2e status + the track-only runtime-perf table), then screenshot triptychs, batched + size-capped + flood-collapsed + 429-aware. - perf-report.ts: perf table with Δ vs the previous main run (via gh). - changelog.ts: no-build git-history baseline differ (Discord trigger B). - noise.ts / gen-manifest.ts: calibration + manifest generation. CI wiring: - wasm-build.yml: post-test step runs the gate + report on the already- produced test-results (no extra build); report-only (continue-on-error), posts only on push to main, inert without DISCORD_WEBHOOK_URL. - ci-ubicloud.yml: secrets: inherit (pass the webhook through). - screenshot-changelog.yml: ~30s no-build changelog on baseline changes. screenshot-manifest.json: canonical 354-name set + best-effort engine tags (313 chromium-swiftshader / 41 firefox-llvmpipe). Normalize scale:'device'->'css' across 18 spec files (no-op at CI DSF=1) so committed baselines are uniformly css-scaled. Design: CI's Linux render is the single source of truth; no pinned container (accept rare env drift -> re-promote); dev commits via promote. Replaces the byte-cmp compare-screenshots.sh + file-size-proxy update-baseline-screenshots.sh (kept for now until the first re-baseline). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-01 18:01:15 +02:00
if (status === 'A' || status === 'R') {
const img = blobPng(head, repoPath);
e2e/CI: dual-engine suites, per-engine screenshots, SwiftShader retired, prod web suite, CI-coverage gate Squash of experiment/ff-big-modules vs main. Big-module routing removed: native-EH shrank kicad_editor below SpiderMonkey's x86-64 code budget (runs 29355049705/29356152413 green on stock Firefox), so BIG_MODULE_SPECS routing and the baseline-only-JIT crutch are gone — kicad-firefox and kicad-chromium both run the full suite, with the module compiled the way real users' browsers compile it. Per-engine screenshots end to end: stableShot/shotPath write test-results/<engine>/<name>.png; baselines move to baseline-screenshots/{chromium,firefox}/ and the whole tools/screenshots pipeline (compare/promote/manifest/spec-map/changelog/Discord) keys on <engine>/<name>. Previously Firefox and Chromium renders of one spec overwrote each other and Firefox renders were never actually gated. Seeded from CI run 29421380806 (92 new firefox baselines, +24 chromium web-suite shots); manifest generated from the baseline tree. One merged playwright.config.ts (kicad/asyncify/coroutine/perf as projects); ~25 dead npm scripts dropped. The web suite is gated in CI for the first time ever (4 rotted specs fixed, 5 broken lib-bridge specs triaged as fixme in docs/features/web-e2e-rot/); cheap lint step after npm ci; last 26 blind-sleep violations fixed. SwiftShader retired: CI Chromium renders WebGL on ANGLE → Mesa llvmpipe (--use-gl=angle --use-angle=gl --ignore-gpu-blocklist; the blocklist flag is mandatory — llvmpipe is blocklisted and WebGL is silently unavailable without it) in BOTH configs. Under WORKERS=4 congestion SwiftShader transiently failed the first post-board-load draw and the recovery cascade ended in a silent permanent Cairo fallback — that engine flip was the "~1.2% changedRatio both directions" occ-export baseline flake. Validated 160/160 across two 80-repeat rigs; full analysis in docs/features/wx-parity-bugs/occ-export-context-eviction.md. Chromium baselines shift slightly on llvmpipe — promote once from the first green run. Deflakes the new coverage exposed: presence baselines settle before capture; presence fixtures declare current file formats; perf gets its own outputDir so CI evidence survives; occ-export settles the board paint before the export dialog; menu-item waits (waitForRenderedByLabel before clickMenuItem) in 4 specs + the TESTING.md rule. Web suite runs the PROD build, in parallel: webServer becomes backend `start` + the standalone's e2e:preview (build-preview.mjs: link-wasm → stash the public/wasm symlink aside during vite build, build-demo.mjs's move — then vite preview as the persistent server). The wasm middleware serves /wasm/* in preview and emits COOP/COEP/CORP itself (a pthread worker script's own response must carry COEP or Chrome kills it with ERR_BLOCKED_BY_RESPONSE). VITE_* flags bake at build time; VITE_ALLOW_USER_OVERRIDE joins turbo globalEnv. fullyParallel + default workers: 5.2m → 1.4m. Determinism fixes the parallel run exposed: shared-page specs become serial groups; locks.spec grabs alice's exact item via the new kicadCollabTestSelectByUuid hook (cross-tab "first footprint" order is not a ysync invariant); quit specs poll page.url() (quit supersedes its own navigation — NS_BINDING_ABORTED on Firefox). Suite: 51 passed / 12 skipped / 0 failed in 1.6m. CI-coverage gate (lint:ci-coverage): every tests/**/*.spec.ts must be reachable from the npm scripts the workflows invoke — scraped from .github/workflows/, resolved through package.json, coverage asked from playwright --list itself. Rules: uncovered-spec + orphan-project (with a documented LOCAL_ONLY_PROJECTS allowlist). Gating next to lint:determinism; 138 spec files / 13 projects accounted for. Product fixes kept from the investigations (reachable on real GPUs too): wx 7799fd1be5 — paint flags clear before dispatch + Invalidate always propagates; kicad 3dcfea5e45 — SwiftShader pass-boundary flush + per-instance font texture + first-frame GL-error drain (GAL recovery recovers instead of falling back to Cairo) + the user-facing eeschema switch navigates again under __EMSCRIPTEN__ (project-sync's FaceRegistered gate had rerouted it into the hidden sync player; caught by the newly-gated web suite). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018eUxiPApHgGiu9NFyQfhAq
2026-07-17 12:10:40 +02:00
if (img) saveSingle(img, key, 'added');
if (status === 'R') {
const oldImg = blobPng(base, oldPath);
e2e/CI: dual-engine suites, per-engine screenshots, SwiftShader retired, prod web suite, CI-coverage gate Squash of experiment/ff-big-modules vs main. Big-module routing removed: native-EH shrank kicad_editor below SpiderMonkey's x86-64 code budget (runs 29355049705/29356152413 green on stock Firefox), so BIG_MODULE_SPECS routing and the baseline-only-JIT crutch are gone — kicad-firefox and kicad-chromium both run the full suite, with the module compiled the way real users' browsers compile it. Per-engine screenshots end to end: stableShot/shotPath write test-results/<engine>/<name>.png; baselines move to baseline-screenshots/{chromium,firefox}/ and the whole tools/screenshots pipeline (compare/promote/manifest/spec-map/changelog/Discord) keys on <engine>/<name>. Previously Firefox and Chromium renders of one spec overwrote each other and Firefox renders were never actually gated. Seeded from CI run 29421380806 (92 new firefox baselines, +24 chromium web-suite shots); manifest generated from the baseline tree. One merged playwright.config.ts (kicad/asyncify/coroutine/perf as projects); ~25 dead npm scripts dropped. The web suite is gated in CI for the first time ever (4 rotted specs fixed, 5 broken lib-bridge specs triaged as fixme in docs/features/web-e2e-rot/); cheap lint step after npm ci; last 26 blind-sleep violations fixed. SwiftShader retired: CI Chromium renders WebGL on ANGLE → Mesa llvmpipe (--use-gl=angle --use-angle=gl --ignore-gpu-blocklist; the blocklist flag is mandatory — llvmpipe is blocklisted and WebGL is silently unavailable without it) in BOTH configs. Under WORKERS=4 congestion SwiftShader transiently failed the first post-board-load draw and the recovery cascade ended in a silent permanent Cairo fallback — that engine flip was the "~1.2% changedRatio both directions" occ-export baseline flake. Validated 160/160 across two 80-repeat rigs; full analysis in docs/features/wx-parity-bugs/occ-export-context-eviction.md. Chromium baselines shift slightly on llvmpipe — promote once from the first green run. Deflakes the new coverage exposed: presence baselines settle before capture; presence fixtures declare current file formats; perf gets its own outputDir so CI evidence survives; occ-export settles the board paint before the export dialog; menu-item waits (waitForRenderedByLabel before clickMenuItem) in 4 specs + the TESTING.md rule. Web suite runs the PROD build, in parallel: webServer becomes backend `start` + the standalone's e2e:preview (build-preview.mjs: link-wasm → stash the public/wasm symlink aside during vite build, build-demo.mjs's move — then vite preview as the persistent server). The wasm middleware serves /wasm/* in preview and emits COOP/COEP/CORP itself (a pthread worker script's own response must carry COEP or Chrome kills it with ERR_BLOCKED_BY_RESPONSE). VITE_* flags bake at build time; VITE_ALLOW_USER_OVERRIDE joins turbo globalEnv. fullyParallel + default workers: 5.2m → 1.4m. Determinism fixes the parallel run exposed: shared-page specs become serial groups; locks.spec grabs alice's exact item via the new kicadCollabTestSelectByUuid hook (cross-tab "first footprint" order is not a ysync invariant); quit specs poll page.url() (quit supersedes its own navigation — NS_BINDING_ABORTED on Firefox). Suite: 51 passed / 12 skipped / 0 failed in 1.6m. CI-coverage gate (lint:ci-coverage): every tests/**/*.spec.ts must be reachable from the npm scripts the workflows invoke — scraped from .github/workflows/, resolved through package.json, coverage asked from playwright --list itself. Rules: uncovered-spec + orphan-project (with a documented LOCAL_ONLY_PROJECTS allowlist). Gating next to lint:determinism; 138 spec files / 13 projects accounted for. Product fixes kept from the investigations (reachable on real GPUs too): wx 7799fd1be5 — paint flags clear before dispatch + Invalidate always propagates; kicad 3dcfea5e45 — SwiftShader pass-boundary flush + per-instance font texture + first-frame GL-error drain (GAL recovery recovers instead of falling back to Cairo) + the user-facing eeschema switch navigates again under __EMSCRIPTEN__ (project-sync's FaceRegistered gate had rerouted it into the hidden sync player; caught by the newly-gated web suite). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018eUxiPApHgGiu9NFyQfhAq
2026-07-17 12:10:40 +02:00
if (oldImg) saveSingle(oldImg, keyOf(oldPath), 'removed');
feat(tests): screenshot regression + Discord review, perf-tracked New tooling in tests/tools/screenshots/ (TypeScript via tsx): - compare.ts: one pixelmatch engine (AA-excluded), connected-component "where to look" boxes, old|new+boxes|heatmap triptych, per-engine floors. - promote.ts: churn-free updater — overwrite a baseline only when decoded pixels differ beyond the floor, copying CI bytes verbatim (no re-encode churn); pulls a CI run via `gh run download` or a local --from dir. - post-discord.ts: always-on CI-on-main report (SHA + e2e status + the track-only runtime-perf table), then screenshot triptychs, batched + size-capped + flood-collapsed + 429-aware. - perf-report.ts: perf table with Δ vs the previous main run (via gh). - changelog.ts: no-build git-history baseline differ (Discord trigger B). - noise.ts / gen-manifest.ts: calibration + manifest generation. CI wiring: - wasm-build.yml: post-test step runs the gate + report on the already- produced test-results (no extra build); report-only (continue-on-error), posts only on push to main, inert without DISCORD_WEBHOOK_URL. - ci-ubicloud.yml: secrets: inherit (pass the webhook through). - screenshot-changelog.yml: ~30s no-build changelog on baseline changes. screenshot-manifest.json: canonical 354-name set + best-effort engine tags (313 chromium-swiftshader / 41 firefox-llvmpipe). Normalize scale:'device'->'css' across 18 spec files (no-op at CI DSF=1) so committed baselines are uniformly css-scaled. Design: CI's Linux render is the single source of truth; no pinned container (accept rare env drift -> re-promote); dev commits via promote. Replaces the byte-cmp compare-screenshots.sh + file-size-proxy update-baseline-screenshots.sh (kept for now until the first re-baseline). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-01 18:01:15 +02:00
}
} else if (status === 'D') {
const img = blobPng(base, repoPath);
e2e/CI: dual-engine suites, per-engine screenshots, SwiftShader retired, prod web suite, CI-coverage gate Squash of experiment/ff-big-modules vs main. Big-module routing removed: native-EH shrank kicad_editor below SpiderMonkey's x86-64 code budget (runs 29355049705/29356152413 green on stock Firefox), so BIG_MODULE_SPECS routing and the baseline-only-JIT crutch are gone — kicad-firefox and kicad-chromium both run the full suite, with the module compiled the way real users' browsers compile it. Per-engine screenshots end to end: stableShot/shotPath write test-results/<engine>/<name>.png; baselines move to baseline-screenshots/{chromium,firefox}/ and the whole tools/screenshots pipeline (compare/promote/manifest/spec-map/changelog/Discord) keys on <engine>/<name>. Previously Firefox and Chromium renders of one spec overwrote each other and Firefox renders were never actually gated. Seeded from CI run 29421380806 (92 new firefox baselines, +24 chromium web-suite shots); manifest generated from the baseline tree. One merged playwright.config.ts (kicad/asyncify/coroutine/perf as projects); ~25 dead npm scripts dropped. The web suite is gated in CI for the first time ever (4 rotted specs fixed, 5 broken lib-bridge specs triaged as fixme in docs/features/web-e2e-rot/); cheap lint step after npm ci; last 26 blind-sleep violations fixed. SwiftShader retired: CI Chromium renders WebGL on ANGLE → Mesa llvmpipe (--use-gl=angle --use-angle=gl --ignore-gpu-blocklist; the blocklist flag is mandatory — llvmpipe is blocklisted and WebGL is silently unavailable without it) in BOTH configs. Under WORKERS=4 congestion SwiftShader transiently failed the first post-board-load draw and the recovery cascade ended in a silent permanent Cairo fallback — that engine flip was the "~1.2% changedRatio both directions" occ-export baseline flake. Validated 160/160 across two 80-repeat rigs; full analysis in docs/features/wx-parity-bugs/occ-export-context-eviction.md. Chromium baselines shift slightly on llvmpipe — promote once from the first green run. Deflakes the new coverage exposed: presence baselines settle before capture; presence fixtures declare current file formats; perf gets its own outputDir so CI evidence survives; occ-export settles the board paint before the export dialog; menu-item waits (waitForRenderedByLabel before clickMenuItem) in 4 specs + the TESTING.md rule. Web suite runs the PROD build, in parallel: webServer becomes backend `start` + the standalone's e2e:preview (build-preview.mjs: link-wasm → stash the public/wasm symlink aside during vite build, build-demo.mjs's move — then vite preview as the persistent server). The wasm middleware serves /wasm/* in preview and emits COOP/COEP/CORP itself (a pthread worker script's own response must carry COEP or Chrome kills it with ERR_BLOCKED_BY_RESPONSE). VITE_* flags bake at build time; VITE_ALLOW_USER_OVERRIDE joins turbo globalEnv. fullyParallel + default workers: 5.2m → 1.4m. Determinism fixes the parallel run exposed: shared-page specs become serial groups; locks.spec grabs alice's exact item via the new kicadCollabTestSelectByUuid hook (cross-tab "first footprint" order is not a ysync invariant); quit specs poll page.url() (quit supersedes its own navigation — NS_BINDING_ABORTED on Firefox). Suite: 51 passed / 12 skipped / 0 failed in 1.6m. CI-coverage gate (lint:ci-coverage): every tests/**/*.spec.ts must be reachable from the npm scripts the workflows invoke — scraped from .github/workflows/, resolved through package.json, coverage asked from playwright --list itself. Rules: uncovered-spec + orphan-project (with a documented LOCAL_ONLY_PROJECTS allowlist). Gating next to lint:determinism; 138 spec files / 13 projects accounted for. Product fixes kept from the investigations (reachable on real GPUs too): wx 7799fd1be5 — paint flags clear before dispatch + Invalidate always propagates; kicad 3dcfea5e45 — SwiftShader pass-boundary flush + per-instance font texture + first-frame GL-error drain (GAL recovery recovers instead of falling back to Cairo) + the user-facing eeschema switch navigates again under __EMSCRIPTEN__ (project-sync's FaceRegistered gate had rerouted it into the hidden sync player; caught by the newly-gated web suite). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018eUxiPApHgGiu9NFyQfhAq
2026-07-17 12:10:40 +02:00
if (img) saveSingle(img, key, 'removed');
feat(tests): screenshot regression + Discord review, perf-tracked New tooling in tests/tools/screenshots/ (TypeScript via tsx): - compare.ts: one pixelmatch engine (AA-excluded), connected-component "where to look" boxes, old|new+boxes|heatmap triptych, per-engine floors. - promote.ts: churn-free updater — overwrite a baseline only when decoded pixels differ beyond the floor, copying CI bytes verbatim (no re-encode churn); pulls a CI run via `gh run download` or a local --from dir. - post-discord.ts: always-on CI-on-main report (SHA + e2e status + the track-only runtime-perf table), then screenshot triptychs, batched + size-capped + flood-collapsed + 429-aware. - perf-report.ts: perf table with Δ vs the previous main run (via gh). - changelog.ts: no-build git-history baseline differ (Discord trigger B). - noise.ts / gen-manifest.ts: calibration + manifest generation. CI wiring: - wasm-build.yml: post-test step runs the gate + report on the already- produced test-results (no extra build); report-only (continue-on-error), posts only on push to main, inert without DISCORD_WEBHOOK_URL. - ci-ubicloud.yml: secrets: inherit (pass the webhook through). - screenshot-changelog.yml: ~30s no-build changelog on baseline changes. screenshot-manifest.json: canonical 354-name set + best-effort engine tags (313 chromium-swiftshader / 41 firefox-llvmpipe). Normalize scale:'device'->'css' across 18 spec files (no-op at CI DSF=1) so committed baselines are uniformly css-scaled. Design: CI's Linux render is the single source of truth; no pinned container (accept rare env drift -> re-promote); dev commits via promote. Replaces the byte-cmp compare-screenshots.sh + file-size-proxy update-baseline-screenshots.sh (kept for now until the first re-baseline). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-01 18:01:15 +02:00
} else {
// Modified: captioned triptych old vs new.
const oldImg = blobPng(base, repoPath);
const newImg = blobPng(head, repoPath);
if (!oldImg || !newImg) continue;
e2e/CI: dual-engine suites, per-engine screenshots, SwiftShader retired, prod web suite, CI-coverage gate Squash of experiment/ff-big-modules vs main. Big-module routing removed: native-EH shrank kicad_editor below SpiderMonkey's x86-64 code budget (runs 29355049705/29356152413 green on stock Firefox), so BIG_MODULE_SPECS routing and the baseline-only-JIT crutch are gone — kicad-firefox and kicad-chromium both run the full suite, with the module compiled the way real users' browsers compile it. Per-engine screenshots end to end: stableShot/shotPath write test-results/<engine>/<name>.png; baselines move to baseline-screenshots/{chromium,firefox}/ and the whole tools/screenshots pipeline (compare/promote/manifest/spec-map/changelog/Discord) keys on <engine>/<name>. Previously Firefox and Chromium renders of one spec overwrote each other and Firefox renders were never actually gated. Seeded from CI run 29421380806 (92 new firefox baselines, +24 chromium web-suite shots); manifest generated from the baseline tree. One merged playwright.config.ts (kicad/asyncify/coroutine/perf as projects); ~25 dead npm scripts dropped. The web suite is gated in CI for the first time ever (4 rotted specs fixed, 5 broken lib-bridge specs triaged as fixme in docs/features/web-e2e-rot/); cheap lint step after npm ci; last 26 blind-sleep violations fixed. SwiftShader retired: CI Chromium renders WebGL on ANGLE → Mesa llvmpipe (--use-gl=angle --use-angle=gl --ignore-gpu-blocklist; the blocklist flag is mandatory — llvmpipe is blocklisted and WebGL is silently unavailable without it) in BOTH configs. Under WORKERS=4 congestion SwiftShader transiently failed the first post-board-load draw and the recovery cascade ended in a silent permanent Cairo fallback — that engine flip was the "~1.2% changedRatio both directions" occ-export baseline flake. Validated 160/160 across two 80-repeat rigs; full analysis in docs/features/wx-parity-bugs/occ-export-context-eviction.md. Chromium baselines shift slightly on llvmpipe — promote once from the first green run. Deflakes the new coverage exposed: presence baselines settle before capture; presence fixtures declare current file formats; perf gets its own outputDir so CI evidence survives; occ-export settles the board paint before the export dialog; menu-item waits (waitForRenderedByLabel before clickMenuItem) in 4 specs + the TESTING.md rule. Web suite runs the PROD build, in parallel: webServer becomes backend `start` + the standalone's e2e:preview (build-preview.mjs: link-wasm → stash the public/wasm symlink aside during vite build, build-demo.mjs's move — then vite preview as the persistent server). The wasm middleware serves /wasm/* in preview and emits COOP/COEP/CORP itself (a pthread worker script's own response must carry COEP or Chrome kills it with ERR_BLOCKED_BY_RESPONSE). VITE_* flags bake at build time; VITE_ALLOW_USER_OVERRIDE joins turbo globalEnv. fullyParallel + default workers: 5.2m → 1.4m. Determinism fixes the parallel run exposed: shared-page specs become serial groups; locks.spec grabs alice's exact item via the new kicadCollabTestSelectByUuid hook (cross-tab "first footprint" order is not a ysync invariant); quit specs poll page.url() (quit supersedes its own navigation — NS_BINDING_ABORTED on Firefox). Suite: 51 passed / 12 skipped / 0 failed in 1.6m. CI-coverage gate (lint:ci-coverage): every tests/**/*.spec.ts must be reachable from the npm scripts the workflows invoke — scraped from .github/workflows/, resolved through package.json, coverage asked from playwright --list itself. Rules: uncovered-spec + orphan-project (with a documented LOCAL_ONLY_PROJECTS allowlist). Gating next to lint:determinism; 138 spec files / 13 projects accounted for. Product fixes kept from the investigations (reachable on real GPUs too): wx 7799fd1be5 — paint flags clear before dispatch + Invalidate always propagates; kicad 3dcfea5e45 — SwiftShader pass-boundary flush + per-instance font texture + first-frame GL-error drain (GAL recovery recovers instead of falling back to Cairo) + the user-facing eeschema switch navigates again under __EMSCRIPTEN__ (project-sync's FaceRegistered gate had rerouted it into the hidden sync player; caught by the newly-gated web suite). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018eUxiPApHgGiu9NFyQfhAq
2026-07-17 12:10:40 +02:00
const { result, heatmap, triptych } = comparePair(oldImg, newImg, key, floorFor(key));
const triptychRel = path.join(DIFF_OUT_DIR, `${key.replace('/', '_')}.triptych.png`);
const heatmapRel = path.join(DIFF_OUT_DIR, `${key.replace('/', '_')}.heatmap.png`);
savePng(path.join(cwd, triptychRel), withBottomLabel(triptych, labelText('changed', key, specFor(splitKey(key).name)), LABEL.colors.changed));
feat(tests): screenshot regression + Discord review, perf-tracked New tooling in tests/tools/screenshots/ (TypeScript via tsx): - compare.ts: one pixelmatch engine (AA-excluded), connected-component "where to look" boxes, old|new+boxes|heatmap triptych, per-engine floors. - promote.ts: churn-free updater — overwrite a baseline only when decoded pixels differ beyond the floor, copying CI bytes verbatim (no re-encode churn); pulls a CI run via `gh run download` or a local --from dir. - post-discord.ts: always-on CI-on-main report (SHA + e2e status + the track-only runtime-perf table), then screenshot triptychs, batched + size-capped + flood-collapsed + 429-aware. - perf-report.ts: perf table with Δ vs the previous main run (via gh). - changelog.ts: no-build git-history baseline differ (Discord trigger B). - noise.ts / gen-manifest.ts: calibration + manifest generation. CI wiring: - wasm-build.yml: post-test step runs the gate + report on the already- produced test-results (no extra build); report-only (continue-on-error), posts only on push to main, inert without DISCORD_WEBHOOK_URL. - ci-ubicloud.yml: secrets: inherit (pass the webhook through). - screenshot-changelog.yml: ~30s no-build changelog on baseline changes. screenshot-manifest.json: canonical 354-name set + best-effort engine tags (313 chromium-swiftshader / 41 firefox-llvmpipe). Normalize scale:'device'->'css' across 18 spec files (no-op at CI DSF=1) so committed baselines are uniformly css-scaled. Design: CI's Linux render is the single source of truth; no pinned container (accept rare env drift -> re-promote); dev commits via promote. Replaces the byte-cmp compare-screenshots.sh + file-size-proxy update-baseline-screenshots.sh (kept for now until the first re-baseline). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-01 18:01:15 +02:00
savePng(path.join(cwd, heatmapRel), heatmap);
report.changed.push({ ...result, triptych: triptychRel, heatmap: heatmapRel });
}
}
report.changed.sort((a, b) => b.changedRatio - a.changedRatio);
const sha7 = (process.env.GITHUB_SHA || head).slice(0, 7);
let subject = '';
try {
subject = git(cwd, ['log', '-1', '--pretty=%s', head]);
} catch { /* ignore */ }
const header =
`🗂️ **Baseline changelog** · \`${sha7}\`` +
(subject ? `\n> ${subject}` : '') +
`\n${report.changed.length} changed, ${report.added.length} added, ${report.removed.length} removed` +
(report.removed.length ? '\n➖ REMOVED: ' + report.removed.map((r) => `\`${r.name}\``).join(', ') : '');
const { files, notes } = buildAttachments(cwd, report);
const messages = paginate(header + (notes.length ? '\n' + notes.join('\n') : ''), files);
const isMainPush = process.env.GITHUB_REF === 'refs/heads/main' && process.env.GITHUB_EVENT_NAME === 'push';
if (args.dryRun || (!args.force && !isMainPush)) {
for (const [i, m] of messages.entries()) {
console.log(`--- message ${i + 1}/${messages.length} (${m.files.length} files) ---`);
if (m.content) console.log(m.content);
for (const f of m.files) console.log(` [attach] ${f.name} (${f.buffer.length} bytes)`);
}
if (!args.dryRun) console.log('[changelog] not a push to main — not posting');
return;
}
const webhook = process.env.DISCORD_WEBHOOK_URL;
if (!webhook) {
console.log('[changelog] DISCORD_WEBHOOK_URL unset — skipping');
return;
}
for (const m of messages) await postMessage(webhook, m);
console.log(`[changelog] posted ${messages.length} message(s)`);
}
if (require.main === module) {
main().catch((e) => {
console.error(`[changelog] ${e.message}`);
process.exitCode = 1;
});
}