René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"random seed: a declaration fault: brokenOwner: a drawn history does not enter 'blocked by brokenOwner' (seed -1269403528), on main too","articleBody":"## Summary\n\nThe random unit `random seed: a declaration fault: brokenOwner` found a counterexample in CI on PR #90 (head cd8e0f9, job `Gap report end-to-end oracle (random seed, 3/3)`, run 43701076084): the drawn history did not enter the state `blocked by brokenOwner`. The constructed history must reach every state it names.\n\nIt reproduces with the same seed and path on `origin/main` (4d3c8a9, twice) and on the PR head (twice), so the base has it. Under the M phase-1 rule it is filed, not blocking. Filed under P4; not rerun away.\n\n## Replay\n\n    cd packages/oracles && ORACLE_SEED=-1269403528 ORACLE_PATH=0 ORACLE_PROPERTY=oracle-gaps.report ORACLE_UNIT='random seed: a declaration fault: brokenOwner' pnpm exec vitest --run --config oracle-gaps/vitest.config.mjs gap-report-oracle -t \"^oracle-gaps\\.report(?: | > )random seed: a declaration fault: brokenOwner$\"\n\nThe Replay line the framework prints omits `--config oracle-gaps/vitest.config.mjs gap-report-oracle`, so it does not run here either (#102).\n\n## Open questions\n\nWhether the state should be reached by every drawn `brokenOwner` history, or only by the constructed one. The drawn history may not reach the blocking step under this seed (the blocked state needs an owner that the drawn history leaves out). A decision on the law, not only on the drawn data, is needed.\n","author":{"url":"https://github.com/KyleAMathews","@type":"Person","name":"KyleAMathews"},"datePublished":"2026-11-15T10:17:00Z","url":"https://github.com/TanStack/oracles/issues/107"}