The most popular programming languages (Java, C++, Objective-C, Python, etc.) and toolkit (Qt, JavaFX, etc.), commonly used for building interactive systems, are still based on the paradigm of sequential programming originally built for computation. In this tutorial we present djnn, a new way of programming interactive systems that takes as its primary concept coupling between processes. We will take as an example the building of a well-known component of an aircraft cockpit, the primary flight display. The first part of this tutorial will be dedicated to the presentation of the basic principles of djnn, its main concepts and how to build a simple program. Then we will introduce the various control structures and provide an overview of the existing libraries of djnn components. We will pursue by showing how to import graphical components from a SVG file, and how to connect various input modalities. Finally, we will demonstrate a rapid prototyping process through the integration of increasingly realist graphical components.

Interaction design is considered important for achieving usable user interfaces. In this tutorial, we present our discourse-based approach for the specification of (classes of) dialogues. We also explain a previously identified and published duality of discourse- and task-based interaction design according to these approaches to high-level modeling for GUI generation. In addition, this tutorial demonstrates how graphical user interfaces (GUIs) can be automatically generated from such discourse-based models specifying a high-level interaction design. This generation approach is especially useful when user interfaces tailored for different devices are needed. It is based on model-transformation rules according to the model-driven architecture. This tutorial also shows how customization can be included into the generation itself. It presents our approach for managing model transformation rules for customization of GUIs in the context of their automated generation. Finally, this tutorial presents our approach for including custom widgets already during the automated generation, in order to make the result persistent also in case of re-generation.