Paprika: Rapid UI Development of Scientific Dataset Editors for High Performance Computing

Abstract

The increasing demand for numerical simulation software leads to productivity bottlenecks in the development of scientific dataset editors. New methods must be found to improve the time to market and the maintainability of these software applications. In this paper we present Paprika, an Eclipse-based application which tackles this problem through two approaches, both of them relying on model-driven engineering techniques. The first is a generative approach where a tailored scientific dataset editor is created for each numerical simulation. The second approach provides a generic editor, which uses reflective interfaces. We present the implementation of these approaches and assess them using results from an empirical study.