Each dataset based on an embedded data model in the TIBCO EBX® repository is associated with a publication of a data model, rather than directly to the data model itself. The first time you publish a data model using the Publish button in the navigation pane, a publication is created. Once the publication exists, datasets can be created based upon it.
The Publish button is only displayed to users who have permission to publish the data model. See Data model permissions for more information.
As datasets are based on publications, any modifications you make to a data model will only take effect on existing datasets when you republish to the publication associated with those datasets. When you republish a data model to an existing publication, all existing datasets associated with that particular publication are updated.
You can publish a data model using either 'Embedded' mode or 'In module' mode. The 'Embedded' publication mode generates a publication that is managed and persisted within the EBX® repository and thus has associated versioning and rollback functionality. The 'In module' publication mode creates an XML Schema Document contained in a module that is not managed or versioned within the repository.
Depending on the configuration of the data model, EBX® automatically determines the publication process to use when you click the Publish button in the navigation pane. When a data model specifies the publication mode 'In module' and provides a target XSD to be generated, the publication process generates an XSD file contained in the module defined in the configuration.
The first time you publish a given embedded data model, a new publication with the same name as your data model is automatically created in the repository. If more than one publication has already been created for this model, you will need to select a target publication for this process.
See Viewing and creating publications for more information on the use of different publications.
During the publication process, you have the opportunity to review the structural differences being introduced by the current publication in a side-by-side comparison view, if the data model has previously been published.
The publication process also offers the option to create a read-only snapshot of the current state of the data model for reference purposes. This snapshot can be useful if the data model ever needs to be rolled back to the current state after other modifications have been made.
Snapshots, which are static archives of the state of the data model, must not be confused with data model versions, which act instead as parallel evolving branches of the data model. See Versioning embedded data models for more information on data model versions.
To access the publications that exist for the current data model, select 'Manage publications' from its data model 'Actions' menu in the navigation pane. From there, you can view the details of the publications and create new publications.
In certain cases, it may be necessary to employ several publications of the same data model, in order to allow datasets to be based on different states of that data model. Multiple publications must be handled carefully, as users will be asked to select an available publications to target when publishing if more than one exists. The action to create a new publication is only available to users who belong to the 'Administrator' role.
To create a new publication, select 'Manage publications' from the data model 'Actions' menu of the data model in the navigation pane, then click the Create publication button. The name you give to the publication must unique in the repository.
User guide table of contents