|Contents:||Main||Chapter||See Also:||Getting Started Manual||Advanced User Manual|
Next, the init installs the data dictionaries for files sent with the init. The data dictionaries are then reindexed.
Data dictionaries are set in place node-by-node, integrating with what already exists. In other words, if a node is brought in by the DD that exists on the target system, the existing node will be replaced. However, if a node that is not included in the init exists on the target system, the init will NOT delete that node. This feature allows users to create local fields and cross-references.
However, this does mean that the developer must carefully consider what the target system's data dictionary will look like after installation. For example, if the developer in the account used to build the init changes the definition of a field or removes a cross-reference, the field or cross-reference must be deleted or otherwise cleaned up on the target account by the PRE-INIT AFTER USER COMMIT routine. This cleanup ensures that the data dictionary will not end up with an inconsistent structure after the init.
Further, each line of a word processing field resides on a separate node. Thus, a change in one of the field attributes that is a word processing field (e.g., field description or technical description) may not completely overwrite a pre-existing attribute. If the incoming value has fewer lines than the pre-existing one, the install will not delete the surplus lines automatically.
Reviewed/Updated: March 10, 2007