Walter R. Bischofberger

Computer supported cooperative software engineering with Beyond-Sniff

Teamwork is a prerequisite for the development of large complex software systems. In conventional software engineering coordination of teamwork is achieved by exchanging formal documents and by providing support for keeping these documents consistent even while several developers are evolving them. In order to support teamwork more effectively it is important to move the focus beyond coordination towards cooperation. The goal of the Beyond-Sniff project is to provide cooperation support in three ways. First, by providing a terminological and conceptional foundation for the field of cooperative software engineering. Second, by developing tools for computer supported cooperative software engineering. Third, by developing a platform for control and data integration, a technical prerequisite for computer supported cooperative software engineering.<<ETX>>

A Generic Multicast Transport Service to Support Disconnected Operation

Many mobile computing applications can profit from process groups and reliable multicast communication to maintain replicated data, but most operating systems available today fail in providing the primitive operations needed by such applications. In this paper we describe a highly configurable, Generic Multicast Transport Service (GTS), which supports the implementation of group-based applications in wide-area settings. GTS is unique in that it offers fault-tolerant, order-preserving multicast on arbitrary communication protocols, including e-mail. As another distinguishing mark, messages can be sent to processes even when they are temporarily unavailable, which permits disconnected operation and mobility. We further propose an object-oriented system design consisting ofadaptor objects interconnected to form aprotocol tree. Adaptor objects offer a common interface to dissimilar communication protocols, and make it easy to incorporate new protocols into GTS. Currently, GTS is being used in a cooperative software engineering environment and in other projects. GTS is available for anonymous ftp.

Designing object systems for evolution

First, we give an overview of several approaches to the evolution of object systems and discuss their potential to make a system better prepared for anticipated evolution. Then, we place our emphasis on unanticipated evolution where we find the flexibility of object coupling to be a crucial system property. We present the major concerns of object coupling and a set of flexible, design-oriented object coupling mechanisms. Each one of these mechanisms exhibits considerable flexibility regarding at least one of the concerns of object coupling. Flexible object coupling mechanisms raise the issue of run-time failures. Thus, these mechanisms should not be used extensively throughout the entire system, but only where they can be most effective. We introduce Design for Slippage, a small conceptual framework providing decision support for the application of flexible object coupling mechanisms.

Architecture support for global business objects: requirements and solutions

The development of world-wide distributed object-oriented systems poses a c onsiderable number of hard questions. In this paper, we summarize these questions as a set of requirements which we consider to be important for a software architecture t o be successful, and we present our solution for such a software architecture. Our main conclusion is that such a software architecture must be reflective in all its key abstractions in order t o allow analyzing and operationalizing its properties. A mainstream banking project is on its way which conforms to this architecture. At Ubilab, we are focusing on the research aspects of the project like enabling smooth evolution and explicitly modeling and operationalizing the software architecture at runtime. 1 Introduction and Motivation The Union Bank of Switzerland (UBS) is a large globally operating bank the operations of which require more and more world-wide distributed applications. New applications must integrate with old applications and must be prevented from turning into legacy applications themselves. To address these problems, UBS is developing a homogenous object-oriented software architecture for both wrapping and integrating legacy systems and providing a common base for new applications. In this paper, we present the main requirements for such an architecture as well as our solutions. Key requirements for this architecture are that it must explicitly support evolution from the very first day, must flexibly utilize and integrate existing and new middleware, and must provide information to analyze the architecture and to operationalize it, for example to guarantee pre-specified runtime behavior. We approach this goal with a distributed object-oriented virtual machine based on a small number of reflective key abstractions. We use this virtual machine to provide flexible evolution support and an operationalized software architecture model that lets components analyze and control the system. UBS is undertaking this effort in form of a regular banking project, for which Genesis Development Corporation is consulting. The authors of this paper are three of the four authors of the key software architecture specification document [1]. Earlier this year, a first prototype has shown the feasibility of