René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"determinism: the hang-replaced-context history records different traces across two runs on main (observation 18, reach)","articleBody":"## Summary\n\nOn origin/main (`35422e3`), the determinism check fails on a constructed history that replaces a hung context: two\nruns of the same history record different traces. The failure is at observation 18 (\"reach\"). It predates #61.\n\n## Replay\n\n```sh\nvitest --run --config oracle-gaps/vitest.config.mjs gap-report-oracle -t \"an edit that changed the cold result after a hang replaced the only context\"\n```\n\nRun at `35422e3` (origin/main), `packages/oracles`. The check fails with:\n\n```\noracle-gaps.report: two runs of \"constructed history: an edit that changed the cold result after a hang replaced the only context\" recorded different traces at observation 18 (\"reach\").\n```\n\nLog: `/tmp/oom/hang-main.log` on the reporting machine.\n\n## Related\n\nThe same history on the #61 branch (`11105c5`) also fails the determinism check, at observation 0 (\"cli\"). Its\noutput is truncated (see #81), so the difference is not visible. The two failures are at different observations, so\nthey may not share a cause. A wall-clock cut is one candidate for both; the deterministic-cut follow-up is the fix.\n\n## Classification\n\nPredates #61 (it fails on main). Per the maintainer's phase 1 decision (2026-11-12), it does not block #61.\n","author":{"url":"https://github.com/KyleAMathews","@type":"Person","name":"KyleAMathews"},"datePublished":"2026-11-14T18:00:14Z","url":"https://github.com/TanStack/oracles/issues/82"}