Skip to content

fix: deterministic resource selection for decoder qualification - #89

Merged
tcconnally merged 1 commit into
mainfrom
fix/issue-78-deterministic-resource
Sep 5, 2026
Merged

tcconnally merged 1 commit into
mainfrom
fix/issue-78-deterministic-resource

Conversation

@tcconnally

Copy link
Copy Markdown
Owner

Summary

  • add deterministic --item-id selection for controlled VideoToolbox resource phases
  • apply exact selection against the full Emby response, so direct-only filtering cannot hide the requested item
  • fail closed for missing or ambiguous IDs
  • record the selected resource in the private phase manifest without credentials or source URLs
  • add launcher cleanup, focused tests, and operator documentation

Verification

  • python3 tests/run_all.py — OVERALL: PASS (all suites green)
  • candidate commit: f4132740a4b7ab1c2552614d894a8e1b1a9b1ad1
  • binary diff SHA-256 from main 074258ddb6b1b2e3e1e8bd5cbe09742310c5d99b: 0de5b16a9c020acb76ccc571f5b3941911b4a46507c788615a2e9196a0109b7f

Acceptance boundary

This resolves the deterministic harness gap for #78. It does not claim native VideoToolbox qualification: the known-good and malformed-resource Mac phases still need to run on the target Mac with current per-resource decoder telemetry and media gates. The target host was unavailable from the current Linux environment, so no native result is fabricated.

Addresses #78

@tcconnally
tcconnally merged commit a858488 into main Sep 5, 2026
2 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant