Checking dataset history
The History tab displays the latest modifications made to the dataset. By default, your workspace keeps the 20 most recent versions of each dataset. Older versions are deleted and can’t be restored.
Each listed item gives the following information:
Name of the user who operated the actions
Date when the modification was operated
Nature of the modification
The only item without any modification description should be at the bottom of the history. It corresponds to the creation of the dataset.
What the History tab records
Each item in the History tab lists the types of changes saved in that version of the dataset.
Label in the History tab | Recorded when you… |
Changed metadata | Edit the dataset's metadata, including its visualization settings, or add or remove an attachment. |
Changed data | Add or remove a source, add, edit or remove a processor, change the field configuration, or change the parameters of a federated dataset. |
Changed scheduling | Add, edit or remove a publication schedule. |
Changed security | Add, edit or remove the access permissions of a user or a group on the dataset. |
Restored version | Revert the dataset to a former configuration from the History tab. |
An item with no label corresponds to the creation of the dataset. When you duplicate a dataset, the copy starts its own history with an unlabeled item.
Note that older items may show a Changed visualizations label. Visualization settings are now recorded under Changed metadata.
What the History tab doesn't record
The History tab only tracks changes to the dataset configuration. The following actions don't create an item:
Publishing, republishing or unpublishing the dataset, including scheduled publications, and aborting its processing.
Deleting the dataset.
Changing the dataset's technical identifier.
Editing the dataset's translations.
To track actions such as publishing or deleting a dataset, see Analyzing back office activity.
Reverting to a former configuration
It is possible to revert to a former version of the dataset by clicking on an item from the History tab.
Instead of erasing items from the list, it creates a new item indicating which modifications were operated to revert to the chosen former configuration.
Using this feature will revert your dataset back to this previous configuration. Note that this might alter the schema of the dataset, so use it with caution as it may impact existing ways the dataset is being used.
Note also that, in case that the dataset was duplicated, it isn't possible to revert to a former version if the dataset source was added by another user.