Two main issues need to be covered when dealing with the dependability of component-based systems: quality assurance of reusable software components and quality assurance of the assembled component-based system. By focussing on the assembly, a software architecture specification of a component-based system allows to explicitly model the structure and required system behavior by specifying how components and connectors are intended to interact. Software architecture-based conformance testing techniques can yield confidence on the implementation conformance to expected structural and behavioral properties as specified in the architectural models. In this paper we explore software architecture-based regression testing methods that enable reuse of earlier saved results to test if a different assembly of components conforms to the evolved software architecture. The approach is presented through a running example.

During software maintenance, regression testing is an expensive but critical activity to ensure that modified versions of the system do not "regress". As software evolves, chances are that not only the implementation changes, but that the specification of the system changes too. We argue that guiding regression testing by the system specifications generated early in the software lifecycle will be more accurate than extracting specification from source code, and more cost effective. Model checking is used to reason about the truth of statements about the system specification. We use a model checker as part of a highly automated regression test generation tool, providing a practical approach to specification based regression testing. Features of our approach that support different specification description languages and different test coverage criteria are also presented.

Concurrent event-based components present characteristics that impose difficulties in understanding their dynamic behavior, mainly for interpreting the cause and effect relations between input and output events in component interactions. In this paper, we propose a technique to help in the process of understanding the dynamic behavior of concurrent event-based components. It checks the event trace (generated by monitoring the component execution) against a specification of the component communication protocol (even with a possibly incomplete or incorrect specification). The technique identifies and presents the more probable cause and effect relations between the component events, providing also a measurement related to this probability.

This formal research demonstration attempts to present an approach to develop and assess architecture and component-based systems based on specifying software architecture augmented by statecharts representing component behavioral specifications [1]. The approach is applied for the C2 style [2] and associated ADL and is supported within a quality-focussed environment, called Argus-I, which assists specification-based analysis and testing at both the component and architecture levels.