Showing papers by "Michael Merritt published in 2003"

Objects shared by Byzantine processes

Abstract: Work to date on algorithms for message-passing systems has explored a wide variety of types of faults, but corresponding work on shared memory systems has usually assumed that only crash faults are possible. In this work, we explore situations in which processes accessing shared objects can fail arbitrarily (Byzantine faults).

Appraising two decades of distributed computing theory research

Abstract: The field of distributed computing started around 1970 when people began to imagine a future world of multiple interconnected computers operating collectively. The theoretical challenge was to define what a computational problem would be in such a setting and to explore what could and could not be accomplished in a realistic setting in which the different computers fell under different administrative structures, operated at different speeds under the control of uncoordinated clocks, and sometimes failed in unpredictable ways. Meanwhile, the practical problem was to turn the vision into reality by building networks and networking equipment, communication protocols, and useful distributed applications. The theory of distributed computing became recognized as a distinct discipline with the holding of the first ACM Principles of Distributed Computing conference in 1982. This paper reviews some of the accomplishments of the theoretical community during the past two decades, notes an apparent disconnect between theoretical and practical concerns, and speculates on future synergy between the two.

Resilient consensus for infinitely many processes

Abstract: We provide results for implementing resilient consensus for a (countably) infinite collection of processes.