Dependable computing and dependable software architectures and infrastructures has been a topic of research for many years. Due to the extensive research and development of techniques, that can help in the development and maintenance of computer systems, we are now able to develop reasonable dependable distributed systems in which people and firms can put their trust. This is at least true for almost static deployments, in which one can control the environment of the running system.
On the other hand, mobility and pervasiveness seems to be a trend entering the lives of more individuals each day. While the infrastructure becomes more end more heterogeneous and complex, due to the mobility and pervasiveness, people are increasingly demanding dependable systems. This imbalance, in the demand for dependable systems on one hand, and the degradation of dependability due to increased complexity and transient infrastructures, has recently been termed the dependability gab by Laprie in "Resilience for the Scalability of Dependability". It shows that research is needed within the field of dependable software architectures for pervasive computing.
Pervasive applications are currently being deployed to support areas such as clinical work, emergency situations, education, ad hoc collaboration, and many areas involving information sharing and collaboration. These environments are all characterized by their transient properties, either due to the underlying infrastructure changing or the nomadic work habits of users. Existing methods for developing distributed applications are ill-suited to meet this challenge in that they rely on existing infrastructure and demand configurations and reconfigurations when the environment changes, as for example when utilizing a client-server architecture pattern. A more suitable method for developing distributed applications for pervasive computing would demand no existing infrastructure and would automatically adapt to the environment.