Skip to content

Refresh Android benchmark coverage with validated measurements #472

Description

@janicduplessis

Problem

The Android benchmark catalog has unavailable cells even though the follow-up measurements have now completed. The site needs the validated timings, command timelines, and matching screenshot/recording evidence.

Evidence

The current catalog lives in website/src/data/benchmarks/*-android.json. Its exporter verifies transcript, screenshot, recording, and cleanup evidence before publishing. The session-classification correction in #471 also needs an auditable publication path that preserves original coordinator records.

Cause

Published artifacts are a reviewed snapshot rather than a live view of the benchmark coordinator.

Fix idea

Refresh the Android catalog from validated measurements, retain existing valid cells, and keep iOS data unchanged. Preserve original records when applying the reviewed session-classification correction, verify evidence hashes, and export only sanitized results and proof media. Keep private attempts and machine paths out of the website and issue discussion.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions