CSV export: relation columns should export the related record's name, not only its id #24542
Replies: 2 comments
|
@Bonapara we've talked about it long time ago 😅 |
|
Kind: Probably a feature request. CSV export intentionally emits relation UUIDs, while the request is to include the related record’s human-readable label. |
Uh oh!
There was an error while loading. Please reload this page.
(from Val's Claude)
Feature request
When exporting a view to CSV, columns of type RELATION currently export the related record's id (e.g.
studentId: f7de0f2a-ee6b-4dfd-bdff-1513426db31f). In the table UI the same column renders the related record's label as a chip ("Jane Doe"), so the export is a downgrade from what the operator sees: a spreadsheet full of UUIDs that mean nothing to the people the export is usually for (management, teachers, external partners).Proposal: export the related record's label-identifier value (its
name) for relation columns — either instead of the id, or as an additional column next to it (Student= "Jane Doe",Student Id= uuid) so programmatic round-trips keep working. This would also pair naturally with #14693 (importing relations by name), making name-based CSV round-trips symmetric.Current workaround: we maintain read-only TEXT "projection" fields (e.g.
ownerNameProjection) on each object, synced from the related record by an external service listening to*.updateddatabase events, purely so exports carry human-readable names. It works, but it duplicates data Twenty already has and requires external infrastructure for what the export could resolve in one join.Context: self-hosted v2.25, CRM used by non-technical operators who export views for selection committees and reporting.
Related: #12496 (relation ids missing from export — fixed), #14693 (import-side name resolution).
All reactions