René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Self-oracle counterexample: calibration selector finds no history when history names repeat (seed 1187311827)","articleBody":"A random-seed run of the framework self-oracle on main found a counterexample in the calibration-history selector from #31 (e2af13a). The selector found no history for an enumerated source when the property declares duplicate history names.\n\n- **Run:** PR #46, run 43288589486, job 131428487356, \"Framework self-oracle (3/3)\". #46 changes no file under `packages/oracles/src` or `tests/`.\n- **Unit:** `self-oracle.framework > random seed: generated`\n- **Violation:** checkpoint \"child\", law registration.\n- **Difference:** `unexpected child load/suite errors: {\"id\":\"self-oracle.synthetic-0\",\"errors\":[\"self-oracle.synthetic-0: calibration history selector for \\\"cal-0\\\": enumerated \\\"items 0 | 1\\\" must identify exactly one history (found 0).\"]}`\n- **History shape:** `duplicateHistoryNames: true`, an enumerated source `[[0]]`, and an unbound probe.\n- **Replay:**\n  ```\n  ORACLE_SEED=1187311827 ORACLE_PATH=1:0:0:0:1:1:1:1:2:5:2:2:2:2:2:2:2:2:2:2:2:2 ORACLE_PROPERTY='self-oracle.framework' ORACLE_UNIT='random seed: generated' vitest run -t '^self-oracle\\.framework(?: | > )random seed: generated$'\n  ```\n\nEither the selector mishandles duplicate history names, or the grammar generates a property that #31's model does not cover. Fix it oracle-first: shrink it, decide whether the defect is in production or in the model, show RED, then fix. Otherwise, record it in the coverage map with this trace.\n\nThis issue was written by Isaac.\n","author":{"url":"https://github.com/KyleAMathews","@type":"Person","name":"KyleAMathews"},"datePublished":"2026-11-12T20:56:36Z","url":"https://github.com/TanStack/oracles/issues/50"}