# Reproduce the date measurements Original code is MIT licensed; original fixtures and observations are CC BY 4.0. No third-party libraries are required. Install Node.js 24.19.0 and CPython 3.12.14 separately, with an IANA time-zone database available to Python. Runtime binaries are not shipped. From the package root, run exactly: ```sh TZ=America/Los_Angeles node data/examples/task-date-portability/measure-node.mjs > data/task-date-portability/node-results.json python3 data/examples/task-date-portability/measure-python.py > data/task-date-portability/python-results.json ``` The first command sets the zone only for that invocation. Python uses its explicit `ZoneInfo` configuration and attaches `fold=0` to naive values. Check runtime versions in each output before comparing rows. The shipped run used Node.js 24.19.0 and CPython 3.12.14. Zone data and runtime executable hashes are in the output; the Python zone-data hash identifies the installed file without publishing its filesystem location. No user data, network request, account or platform import is involved. The output timestamps will differ on a rerun. Compare `records` after checking versions and zone data. `index.json` joins matching record IDs from the two outputs; CSV is a display export. After changing any input or runner, regenerate the outputs, joined index and hashes rather than retaining stale results.