In order to shield the modeler from the complexity of features like higher-order, static typing, and generics we provide a preconfiguration feature. Nearly everything (except the control-flow of a process) can be configured, including names, generics, types, context variables, the data-flow and documentation. This offers to create new activity libraries, just via configuring existing ones.


Excerpt of the preconfiguration of the Chainreaction domain
Excerpt of the preconfiguration of the Chainreaction domain

The configuration layer is realized via configuration graphs, SLGs (Service Logic Graphs) without any control-flow information, declaring activities with (partially) preconfigured data-flow and further meta-information denoted by preconfigured activity (cf. Fig. 1). A modeler may

  • simply drag-and-drop activities to the canvas of a configuration graph,
  • create context variables,
  • connect context variables to input and output parameters of the activities,
  • set type arguments of a generic activity,
  • change the icon of an activity,
  • set the label of activities, and
  • add further meta-data as, e.g., validations or temporal formulae for verification.