Distributed systems substantially improve our ability to compute and exchange information, as is evidenced by the dramatic success of the so-called World Wide Web. At the same time, distributed systems --and computer networks in particular-- are hard to design, control, and maintain, as they consist of a variety of complex hardware and software components that are subject to faults and dynamic changes.

Self-stabilisation has emerged as a promising paradigm for the design, control, and maintenance of fault-tolerant distributed systems. As its name suggests, self-stabilisation enables systems to automatically recover from the occurrence of faults. Its essential idea is this: Regardless of what state a system is placed in, by virtue of being self-stabilising, the system converges to desired behaviour. Thus, even if faults cause the system to be placed in an arbitrary state, the system can eventually resume its desired behaviour.

The field of self-stabilisation is young and rapidly growing. To facilitate research in this field, experts in this field and in allied fields were invited to share their research interests and work with each other. The Dagstuhl Seminar on "Self-Stabilisation" brought together thirty five researchers from seven different countries. The opening talk was given by Edsger W. Dijkstra, then an overview of the state-of-the-art and future directions was provided by Shmuel Katz. The talks that followed presented new results and directions:

Formal methods for verification and specification of self-stabilising algorithms,

Use of the self-stabilisation concept in the context of security and privacy,

Integration with other fault models,

Transient fault detectors,

Design frameworks for achieving self-stabilisation and other fault tolerances,

Self-stabilising algorithms and their time/space efficiency; impossibility results

The pleasant atmosphere of Dagstuhl was an important incentive for the lively interaction between the participants. The success of the seminar in stimulating new ideas and dialogue has led us to start planning the next seminar two years hence. We would like to thank all who contributed to the seminar, and in particular the encouragement we received from Professor Dr. Reinhard Wilhelm. The support of the TMR Program of the European Community is gratefully acknowledged.