Workflows are defined by creating subtasks of a parent task, and linking these subtasks via shared data fields. Data fields within the scope of a parent and its subtasks are identified by their name only. For example, when a subtask input field is defined as 'a' and 'a' is also a parent input data field, then subtask and parent task share data field 'a'. Sharing a data field means not only sharing the name, but also the value when valid. This has the consequence that the value of data field 'a' is copied from parent task to subtask when it becomes available, in this case when the parent task is requested. The value of the data field 'a' cannot be filled in in the subtask independently of the parent task. Figure 6.3.4-1 shows an example where e.g. subtask Task T1 takes its input from the parent task Task T and feeds its output into peer tasks Task T2 and Task T3.