Versioning Policy

Data Share Versioning Policy

This page defines how Optimove Data Share schema versions are named, released, supported, and retired — so schema evolution is always predictable for your data team.

How versions coexist

Data Share view generations live side by side in the share: when a new generation is released (e.g. _V1_5), the previous generation (_V1_3) remains fully queryable. You migrate on your own schedule, within the support window, by changing the view suffix in your queries.

  • Exactly one generation is marked ✅ Recommended — use it for all new integrations.
  • Older generations remain accessible but unsearchable: they are removed from this site's navigation and search, but the views stay live in your share.
  • Never join views from different generations in a single query — key columns may differ.

Naming convention

Views follow VW_<DATASET>_V<major>_<minor> (e.g. VW_SCHEDULED_CAMPAIGNS_V1_5). A new generation always ships as new views — live views never receive breaking in-place changes. Version numbers may skip (e.g. V1_4 was never shipped); the changelog always states this.

What counts as a breaking change

ChangeClassification
Column addedNon-breaking
New view addedNon-breaking
Value domain extended (new code/value)Non-breaking — always documented
Column renamedBreaking
Column removedBreaking
Column type changedBreaking
Column meaning changed (same name)Breaking
Row grain / granularity changedBreaking
Primary key changedBreaking
View removedBreaking

In-place changes to a live generation are limited to non-breaking changes, and are always recorded in the changelog.

Version lifecycle

StatusMeaningDuration
✅ RecommendedLatest generation; use for new integrationsUntil the next release
SupportedPrevious generation; fully functional≥ 12 months after its successor ships
DeprecatedRemoval date announced≥ 6 months notice before removal
RemovedViews dropped from the shareSchema preserved permanently in the ERD archive

Removal dates are never moved earlier; delays are announced in the changelog.

What ships with every release

Every version release includes: a dated changelog entry (with total and breaking change counts), updated dataset pages, an updated Entity Relation Diagram (the previous generation's ERD moves to the archive with its full schema), and a migration guide whenever a breaking change exists.