Visual Rule-Based Programming
Visual Rule-Based Programming
Frank Drewes (Umeå University, Schweden)
Berthold Hoffmann (Universität Bremen)
Mark Minas
Many problems in computer science or other application domains are based on graphs or are easily modelled by graphs. Computing by graph transformation, therefore, is natural when dealing with such graphs. However, users do not accept graphs as (the only) data structure if they are visible on the user interface. This is one of the reasons why programming by graph transformation does not have the success that might be expected. But usability of programming languages using the rule-based approach of graph transformations is expected to be improved by domain-specific visualizations of graphs (e.g., as diagrams). Hence, the objective of this research project is a visual programming language that makes use of diagrams as data structures which are internally represented by graphs and which are manipulated by graph transformations in a rule-based way.
The cooperation has started with a computational model that uses graph transformations on hierarchical hypergraphs. In order to create domain-specific visualizations, we use the diagram specification and generation tool DiaGen which will be advanced in the course of this cooperation.
The long term goal of this project is to make available domain-specific programming languages with customizable visualizations of data structures and user interfaces. Such languages would thus be adaptable to different application domains, and they would be more usable than (textual) programming languages as of today. Graphical approaches are currently omnipresent in analysis and design (e.g., UML). Programs, however, are still written as text, even for inherently visual problems or problems that use visual representations. Instead, future approaches should allow for visual modelling and visual programming without such a large and error-prone gap between analysis and design on the one hand, and programming on the other hand. Programs would then be easier to write, easier to maintain, and easier to understand.
