Skip to main content

Merging sources

The merge engine is the heart of Cairn. When several sources describe the same workout, Cairn produces one activity by choosing the best source per field group — not by picking a single "winner" source.

Field groups​

Fields are grouped (classification, time, duration, title, distance, elevation, speed, hr, power, …, gps_track, laps). For each group, a PolicyResolver(ActivityType) priority list decides which source wins. So a ride might take its GPS track from your bike computer, power from the trainer, and title from Strava — automatically.

Your edits are protected​

  • preserveUserEdits — title, description, tags, privacy and gear are user-owned; re-merging never clobbers them.
  • Field-source override — pin any field group to a specific source; applied on every recompute.
  • Classification override — a per-activity overlay for type/discipline/flags and summary metrics (distance/elevation/moving time), which survives re-import.

Streams stay per-source​

Streams are never merged — each source keeps its own. A primary stream is selected for rendering. You can even graft "source data from another activity" (a donor's selected groups + its stream, rebased in time) onto an activity.

Configure the policy​

Set your merge priorities in the web UI (the policy editor) per activity type. Manual entry is just a manual source fed through the same engine.