Last-write-wins sync or manual JSON reconciliation for Hoppscotch offline edits

Confirmed Behavior

In v2024.x, Hoppscotch resolves sync conflicts with last-write-wins based on the updatedAt field, overwriting the entire collection or environment document. This is a whole-document replacement, not a field-level merge.

Implications for Concurrent Edits

When two clients modify different fields in the same collection or environment, the client with the newer updatedAt timestamp wins. The losing client's changes are discarded entirely, even if those changes affect unrelated fields. This is a known limitation of the current sync design.

Is Last-Write-Wins Acceptable?

No, if your team requires preservation of all edits across different fields. Even non-overlapping changes will be lost when sync occurs. If any field-level change must survive, last-write-wins is not safe.

Recommended Strategy: Manual JSON Reconciliation

1. Export both versions: From the UI, export the local collection/environment to JSON. Note the remote version's updatedAt value if visible.
2. Diff the files: Use a JSON diff tool or text comparison to identify field-level differences. Focus on request definitions, variables, headers, and auth settings.
3. Manually merge: Combine the desired fields from both versions into a single JSON document. Ensure the merged file retains all unique changes.
4. Import the merged version: Use the Hoppscotch UI to import the reconciled JSON. This becomes the new authoritative version.
5. Sync downstream: Pull from the server to propagate the merged version to other clients.

Schema-Version Parity

Enforcing client-server parity is only relevant if the sync endpoint validates schema versions and rejects mismatched payloads. Check your self-hosted server's API responses for version headers or error codes. Parity alone does not prevent last-write-wins overwrites; it only guards against potential field drops during deserialization.

CRDT Support Status

Field-level merging via Yjs or Automerge has been proposed but not accepted for v2024.x. Until a merge flag or CRDT milestone is released, treat field-level merging as unsupported. Do not rely on unverified CRDT integration in production workflows.

Verification Steps

• Check your Hoppscotch version against the official changelog for sync-related changes.
• Perform a controlled test: edit different fields in the same collection on two clients, sync, and observe which changes persist.
• Review self-hosted server logs for schema validation errors during sync operations.

Key Risks

• updatedAt is unreliable if client clocks are unsynchronized.
• Whole-document overwrite can remove auth settings, variables, or metadata added by the losing client.
• Manual reconciliation is error-prone and does not scale for large or frequently updated collections.