docs(testing): log the 1.9.5-rc.2 regression pass - #369
Conversation
The checklist run this records covers the 65 commits since v1.9.2 rather than the rc.2 delta, which is what a release candidate actually needs. Four recordings; no defect found. The numbers that matter are in the row. The AGENTS.md addition is the one thing this pass got wrong about itself. The staleness warning I wrote yesterday said to date "the binary" — so I refreshed the capture helper and nothing else, and an export then died on `open_input: -22 (Invalid argument)` out of `compositor.exportMulti`. It reads exactly like a product bug, and I nearly filed it as one. The file was fine: `ffmpeg` opened it from the command line without complaint. The compositor addon was four days older than the av* DLLs it was built against. A full hash diff of the directory found sixteen files differing and two missing outright. So the unit is the directory, not the binary. Copy all of it and diff by hash, or a mismatched set will hand you a failure that looks like the thing you came to test.
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (2)
📝 WalkthroughWalkthroughThe pull request updates native-binary refresh guidance and adds a Windows 11 release regression-test result for recording, recovery, synchronization, rendering, persistence, refocus, and export. ChangesRelease documentation
Estimated code review effort: 1 (Trivial) | ~3 minutes Merge Risk: ⚪ Minimal · up to This PR only records manual testing results and clarifies documentation for refreshing the native-file directory; no actionable merge-blocking risk remains beyond normal checks and review. Possibly related PRs
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
What
The manual e2e record for the 1.9.5-rc.2 candidate, plus the one correction this pass made to its own instructions.
Scope was deliberately the 65 commits since v1.9.2, not the rc.2 delta — a release candidate is judged against the last thing users actually installed. ~20 of those commits are in the Windows capture helper, so the net is weighted there.
Run against
release/v1.9.5@b1b81de5with the CI-built rc.1 native payload. That payload is the rc.2 payload: nothing underelectron/native/orcrates/changed since the rc.1 tag, verified by diffing the file list.Result: pass, no defect
Four recordings. Highlights, all measured rather than eyeballed:
mvexpresent, ~1moof/s,mfraonly after a clean stop.videoInput: cpu-rgb32, so the GPU DXGI path is still opt-in.c9ff08ac) — HUD00:59→ 60.067 s at 60/1.9e00d11a) — absent with no audio track, rendered with one.9c6f2e1f,462b5572) — preview renders and export completes.moof, nomfra), all three sidecars written.Not covered, and why
DPI scaling (
60bb6d7c,71cc88d6) — not re-run here; already validated when those two landed. My first draft of this row said "this machine is at 100 %, those bugs are structurally invisible here", which is wrong: the display scale is a setting, changing it takes about two minutes, and doing so has been the documented procedure since #346. Skipping was a choice, and the record now says so rather than dressing it up as a constraint. Webcam PiP and the export webcam fixes (056cfc39,26ad478c) — no camera in the loop. Microphone, GIF, macOS/Linux, AI sections, packaging.The AGENTS.md change
Yesterday's staleness warning said to date "the binary". So I refreshed the capture helper and nothing else — and the export then died on:
That reads exactly like a product regression, and I nearly reported it as one. The recording was fine:
ffmpeg -f null -opened the same file from the command line and decoded it end to end. The compositor addon was four days older than theav*DLLs it was built against. A full hash diff ofelectron/native/bin/<platform>/turned up sixteen files differing and two missing outright.Copying the whole directory and re-running the identical export made it pass. So the unit of staleness is the directory, not the binary, and the doc now says so — with the failure mode named, because "mismatched native set" presents as a bug in whatever you happened to be testing.
Checks
node scripts/check-docs.mjs→ OK (31 files). Docs only; no code touched.