Abstract
It was the development of efficient and cost effective interconnection structures on the one hand and the price-performance revolution in microelectronics on the other hand which were prerequisites for the development of distributed communication applications. Changes in technology, however can be a driving force only because they make it possible to match user needs which tend to become more and more sophisticated. Parallel to the decline of hardware costs and the development of interconnection technologies, there is a general trend for introducing computing systems into almost every business-oriented or leisure-oriented activity, as well as improving computing system interfaces. Moreover, new communication applications became possible to meet economic or managerial user constraints. One of those new application areas is computer supported cooperative work (CSCW). This field can be seen as a conceptual shift in our understanding of computing - whereas the classical view of computing was manipulation of data by isolated individuals, the CSCW view of computing is an exchange of data and information as well as cooperation by individuals aware of their environment and of other users. Communication facilities play a key role in the design and implementation of distributed systems for cooperation purposes. Interprocess communication facilities allow processes to be spread over a distributed computer system to communicate and to access remote resources. Performance of the whole distributed system, in particular a distributed workflow and communication system, depends on this facility. Knowledge of distributed computer systems, in particular local area networks, is important, not only because distributed systems are based on these networks today, but also because many techniques used in communication systems can be adapted in cooperation environments. After many years of research and experiments, and as a result of the development of several experimental and commercial cooperation and coordination systems, we know how to build centralised systems. On the other hand, although significant progress has been made in the development of distributed systems for Computer Supported Cooperative Work (CSCW) during the last decade, they are still in an early phase of development, with many unanswered questions and relatively little agreement among workers in the field about how things should be done. There are three reasons for this situation. First, there is the problem of obtaining a full understanding of distributed applications. The second reason is that awareness information about all processes and objects of a distributed computer system is not located in one place as in a centralised cooperation system. The last reason is that synchronous and asynchronous events have an important influence on application processes and for handling those, there is a variety of different solutions. The design process for an adequate solution has to be managed very carefully. In this session we contribute towards understanding some aspects of CSCW design by presenting some results from different phases involved in system design and adaptation. The first paper "Termination Handling in Inter-Organisational Workflows - An Exception Management Approach" by H. Ludwig proposes an approach for the handling of terminations in inter-organisational business processes that are enacted using Workflow Management Systems. A model for cross-organisational process management is introduced to identify cases where human management has to be notified and involved. The second paper "Agent-based Systems Co-operation " is by B. Kumova and A. Kut. The paper proposes a system in which users incrementally define and store their knowledge about distributed data and information sources. The authors have developed a framework that provides for network connectivity at the information level and they introduce the architecture of their prototype. The implementation is a solution for cooperative work, where user-user communications are transformed into user-agent communications, in which users do not need to communicate with each other directly. The title of the third paper is "Supporting Cooperative WWW Browsing: A Proxy-based Approach," written by G. Cabri, L. Leonardi, and F. Zambonelli. The authors present a system for cooperative browsing that permits users within a Workgroup to share and cooperate towards a common goal. The system implementation is based on a Java proxy, thus achieving portability without requiring modification neither to browsers nor to servers. The last paper "Support for Tailoring CSCW Systems: Adaptation by Composition" is by M. Koch and G. Teege. It deals with flexibility and tailorability, two key requirements for CSCW systems. An important mechanism for providing tailorability is the composition of existing functionality modules to a compound functionality. Two variants of this composition approach are presented and it is shown how both variants can be combined in the same application. The aim of this session is to contribute to the discussion on design and implementation of user-centered groupware applications. Usability, however, is one of the main problems of groupware applications today. For example, tailorable solutions directed toward the specifics of group communication processes are lacking. The concepts presented in these papers show approaches for improving usability in real world applications.