← Blog

Keeping timezone data fresh

· Data Team · 5 min read

Timezone rules change more often than people expect. Governments shift daylight-saving boundaries, redefine offsets, or abolish DST entirely, sometimes with only weeks of notice. The /v1/timezones endpoint has to reflect those changes before they take effect, not after.

Our pipeline tracks the upstream IANA time zone database. When a new release is published we diff it against the version we currently serve, review the changed zones, and stage an update. Each entry exposes its current UTC offset and DST state, so a client never has to compute transitions itself.

Validation runs before anything ships: we replay known DST transitions for the affected zones and confirm the computed local times match the new rules. Only then is the dataset promoted and replicated to the edge nodes.

Responses are versioned and carry a Last-Modified date, so a client can issue a conditional request and get a cheap 304 when nothing has changed since it last synced — which, for timezone data, is most of the time.