Repository navigation
docs: support the uniform 15-case official benchmark campaign - #614
Mohamed Mansour (mohamedmansour) wants to merge 2 commits into
Conversation
Refresh the fifteen-case campaign snapshot, validate source provenance and disclosed browser sample availability, and update the explorer and homepage previews while keeping docs-side preparation. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
|
Companion campaign exporter and synchronization: microsoft/webui-benchmarks#49. |
mddinbox
left a comment
mddinbox
left a comment
There was a problem hiding this comment.
Target: 2dab85c...255597e | Coverage: complete 12-file three-dot diff, including structured comparison of the large JSON data change; CI: Docs, Test, Build (Linux/Windows/macOS), Lint, FFI, WASM, E2E, CodeQL passed.
Challenge: 1 confirmed, 0 revised | Compatibility: displayed measurements are not bound to the claimed provenance digest.
Action: COMMENT | The campaign validator does not tie the published measurements to their source digest.
Validate canonical measurements and provenance against an independent build/test pin before replacing generated benchmark evidence. Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com> Copilot-Session: 844f2168-295c-4641-974d-8e90468b3568
mddinbox
left a comment
mddinbox
left a comment
There was a problem hiding this comment.
Reviewed the updated 15-case campaign and evidence validation. The pinned canonical digest and tamper tests address the prior provenance concern; no blocking issue found in the PR. Linux/Windows build checks currently fail in unrelated native IPC acceptance / external download steps and should be rerun or resolved before merge.
Summary
prepare:benchmarksas the docs validation/adaptation step, validating campaign provenance, both SSR response modes, and disclosed unavailable browser-memory samples.Validation
Companion exporter branch: https://github.com/microsoft/webui-benchmarks/tree/update-benchmarks