Revert "feat(tracing): allow configuring the screenshots size"

[Yury Semikhatsky (yury-s)]

Reverts #42283

Reference #42274

[github-actions]

Test results for "tests 1"

1 failed
❌ [playwright-test] › global-setup.spec.ts:313 › globalSetup should work for auth @windows-latest-node22

4 flaky

⚠️ [chromium-library] › library/chromium/chromium.spec.ts:301 › should report intercepted service worker requests in HAR `@chromium-ubuntu-22.04-node24`
⚠️ [chromium-library] › library/beforeunload.spec.ts:130 › should support dismissing the dialog multiple times `@chromium-ubuntu-22.04-node20`
⚠️ [firefox-page] › page/page-emulate-media.spec.ts:144 › should keep reduced motion and color emulation after reload `@firefox-ubuntu-22.04-node20`
⚠️ [webkit-library] › library/browsercontext-add-cookies.spec.ts:426 › should allow unnamed cookies `@webkit-ubuntu-22.04-node20`

51260 passed, 1237 skipped

---

Merge workflow run.

[github-actions]

Test results for "MCP"

8259 passed, 1361 skipped

---

Merge workflow run.

[github-actions]

Hi, I'm the Playwright bot and I took a first look at the CI failures.

🟡 One failure, most likely a Windows timeout flake — but I can't prove it

[playwright-test] › global-setup.spec.ts:313 › globalSetup should work for auth timed out on windows-latest-node22. This is a pure revert of the tracing screenshot-size option, which never touches global setup or auth — so I don't think the PR caused it. But the only failure of this test in the results DB is this run, so I can't confirm it's a pre-existing flake either. Calling it uncertain.

Details

Overall: The PR reverts #42283 — it only removes the configurable tracing screenshot size (docs, channels, tracing.ts, testTracing.ts, videoRecorder.ts, protocol). None of that reaches the global-setup / auth path the failing test exercises, so a real regression here would be surprising.

Uncertain — global-setup.spec.ts:313 › globalSetup should work for auth (windows-latest-node22)

• Signature: Test timeout of 30000ms exceeded with expect(exitCode).toBe(0) receiving 1 — a Windows-only timeout, a classic infra signature.
• Base rate: this test is passed 4541 / 4542 runs (99.98%) across the results DB; on windows-latest-node22 specifically 642 / 643. It's not a known flake.
• The one and only failure on record is this PR's run (194bef2, run 32911875350). Per the triage hard rule, a flake needs the same test failing somewhere the PR can't be responsible for — and there's no such run — so I can't call it a confirmed flake. Nothing in the diff plausibly reaches it, so I don't think it's caused by the PR either.

To be sure, a re-run of the tests 1 job on Windows would settle it — a green re-run would confirm the timeout was noise.

The 4 flaky results (chromium HAR service worker, beforeunload dialog, firefox emulate-media reload, webkit unnamed cookies) all passed on retry and are unrelated to tracing config — nothing to triage there.

_{Triaged by the Playwright bot - agent run}