Reverse engineering object-oriented distributed systems

Abstract

A significant part of the modern software systems are designed and implemented as object-oriented distributed applications, addressing the needs of a globally-connected society. While they can be analyzed focusing only on their object-oriented nature, their understanding and quality assessment require very specific, technology-dependent analysis approaches. This doctoral dissertation describes a methodology for understanding object-oriented distributed systems using a process of reverse engineering driven by the assessment of their technological and domain-specific particularities. The approach provides both system-wide and class-level characterizations, capturing the architectural traits of the systems, and assessing the impact of the distribution-aware features throughout the application. The methodology describes a mostly-automated analysis process fully supported by a tools infrastructure, providing means for detailed understanding of the distribution-related traits and including basic support for the potentially consequent system restructuring.