Label lifecycle

Labels, fields, and choices go through specific states throughout their lives. Additionally, labels might have different revisions. The following diagram shows the label lifecycle, including revisioning:

Labels lifecycle diagram.

1. Create a label (create())—The label is created and stored in a database as revision_id=1. The label has the state of UNPUBLISHED_DRAFT. In this state:
   • Users can't view the label
   • Users can't apply the label to Drive items.
2. (optional) Update a label, field, or choice (delta())— Every update, even before it's published, is stored in a database, and the label's revision is incremented.
3. Publish a label (publish())—The label has the state of PUBLISHED and users can apply the label. Publishing the label increments its revision.
4. (optional) Update a label, field, or choice (delta())— The label, field, or choice is updated and stored in a database as a draft label. The label has the state of PUBLISHED with hasUnpublishedChanges=true meaning there are draft changes, but they aren't available to users. Each update increments the label's revision.
5. (optional) Publish a label (publish())—If available, the most-current draft is published. The label has the state of PUBLISHED and users can apply the label. Publishing the label increments its version.
6. Disable a label (disable())—The label has the state of DISABLED though users can apply the label through the API. However, a disabled label isn't shown in a UI unless configured to be shown. Deprecating the label increments its revision.
7. Enable a label (enable())—The label is returned to a PUBLISHED state and users can apply the label. Publishing the label increments its revision.
8. Delete a label (delete())—The label has a state of DELETED and can't be applied. Deleted labels are eventually purged.

It's important to emphasize that every update to a label increments the label's revision. And, if the label has already been published, publishing it again after n updates means that its published revision number is revision + n + 1 number of successive updates.