RobbyDBMS

Abstract

The development of a highly configurable data management system is a challenging task, especially if it is to be implemented on an embedded system that provides limited resources. We present a case study of such a data management system, called RobbyDBMS, and give it a feature-oriented design. In our case study, we evaluate the system's efficiency and variability. We pay particular attention to the interaction between the features of the data management system and the components of the underlying embedded platform. We also propose an integrated development process covering both hardware and software.

On the Relationship between Preprocessor-Based Software Variability and Software Defects

“…Reference to the experimental method of Literature [4], select different projects range from the small project (~10KLOC) to large projects (~1000KLOC) that using C/C++ language were selected. These projects are varying in different areas, including operating system and application software.…”

“…We refer the software variability metrics mainly form literature [4]. The metrics of variability contains several sub-metrics, including program size metrics (lines of feature constants, lines of feature code and classes contain variability), program crosscutting metrics (scattering degree, tangle degree and the average nesting depth), program variability granularity metrics (coarse-grained and fine-grained).…”