David J. Farber

Recoverability of Communication Protocols--Implications of a Theoretical Study

A study is presented which permits the formal analysis and synthesis of recoverable computer communication protocols. This study is based on a formal representation of processes by a model of computation, the Petri nets (PNs). The PN model is generalized to include a representation of the possible failures, and then, the concept of recoverability is formally defined. A set of necessary and sufficient conditions which a process must satisfy in order to be recoverable is derived. In the PN model, the processes that satisfy these conditions are shown to have some practical limitations. A new model, the time-Petri net (TPN), is introduced to remove these limitations. This new model allows the introduction of constraints in the execution times of its part. As shown in this paper, the TPN appears to be a suitable model for the study of practical recoverable processes. Several practical communication protocols are formally designed and analyzed using this new model, and some interesting properties of these protocols are formally derived.

A Parallel Mechanism for Detecting Curves in Pictures

In Hough [1], Duda and Hart [2], and Griffith [3] procedures were proposed for detecting lines in pictures and in [2] Duda and Hart extended their method for more general algebraic curve, fitting. This correspondence shows how this method can be used to detect any given curve in a specific orientation. The procedure presented here con be easily implemented and can be efficiently implemented in a parallel machine.

The convergence of computing and telecommunications systems.

Since we do not know of any completely workable solutions to the problems we have considered, we shall settle for a few general observations: 1) The public availability of socially useful computer communications services is and has been held back by legal battles that are now under way between the potential suppliers. 2) No simple resolution of these issues in the near future seems likely in view of the past conceptual separation of computers and communications doctrines. 3) The current policy is to determine whether the nation shall or shall not have certain computer communications services, by the adversary process. In this process, often only the voices of the loudest adversary suppliers are heard. 4) Although there can be no certainty that better alternatives cannot be devised, we believe that such a possibility assumes a higher probability if the key actors come from the technical community sectors more representative of the future consumers. 5) If we are to have the new services that are possible, we need an approach that makes better use of the technologists dreams and goals rather than have future prospects excessively bound by lawyers paid to preserve the interests of their clients, irrespective of any secondary consequences. 6) We cannot be sanguine about this possibility, as technological statesmanship is too easily corrupted by the same forces that have placed us in this predicament. Furthermore, even if not corrupted, beneficial cooperation can too readily be regarded as simply collusion. 7) Although we do not have any clear answer, we do know that present approaches are not taking us where we want to go very rapidly and that alternative approaches should at least be considered.

Software considerations in distributed architectures

This presentation is an outgrowth of a workshop session held at Lake Arrowhead in 1973 on Distributed Software. The participants in the session were Dave Farber of UC Irvine (Chairman). Don Bennett of Sperrty Rand. Bob Bressler of Bolt. Beranek and Newman, Larry Rowe of UC Irvine, Bob Metcalfe of XEROX PARC, and Marty Graham of UC Berkeley.

THE IMPACT OF OFFICE AUTOMATION

This chapter discusses the strategies for successful implementation of office of the future, methods by which impacts are studied, impacts as they have been observed to date, and how these impacts relate to organizational productivity. The office and organization are complex communication systems. The chapter further discusses the importance of communication in the implementation of office automation systems and the realization of beneficial impacts. The communication system concept has enabled the understanding of far deeper changes than the scattered efficiencies obtained from data and word processing support. There is a need for mature office automation technologies and also for mature installations. Most implementations have been rather immature because the knowledge of how to implement a system of the needed size and capability has not been applied. The strategies have not been specified in detail, and there have not been systems advanced enough to warrant a staged evolution. The chapter presents a step-by-step strategy to manage the human use of computer-based office systems. Commitment to evolving a mature implementation requires well-developed methods of assessing impact. The complex communication system requires an elaborate structure because of many interdependencies and variable relationships. This structure includes lists of the variables by kind and type, some of the possible values of the variables, a framework for measuring the variables, and methods of measurement.

Recoverability of modular systems

In [1] GOSTELOW-VAN WEERT, it was shown that processes can be described by Petri-nets (PN). In this note we will demonstrate a way of designing programs using Petri-nets as a way of describing the desired behavior. This approach makes it possible to utilize all the theoretic knowledge developed for PN and thus provides a powerful tool for program design. In an example of this approach we show how to design a recoverable program. This design is based on a study of recoverable PNs presented in [2], [3], [4] and [5].n In this note, we will not deal with the code itself, but with the structural organization of the programs. Our approach is based on an hierarchical, top-bottom, modular design. At each level a description of the desired system behavior is given. This description is then modeled by a PN and analyzed using the theory developed for PNs.n This analysis allows one to check for, among other things, properties of the systems such as recoverability, proper termination, etc. These properties are inherently structural and not dependent on internal module semantics.

A survey of the systematic use of macros in systems building

Assemblers with macro capabilities have been available for over ten years. There has been a limited number of mainly unpublished systematic uses of such capabilities in the construction of a variety of systems. This paper will cover a number of these cases. We will examine the features of the macro systems which allowed their usage as well as the method used in the system implementation. We will comment on the effect on efficiency and flexibility that the use of this facility has produced.

A note on recoverability of modular systems

In a paper by Gostelow-Van Weert, it was shown that any processes can be described by Petri-nets (PN). Thus PN can be used as a model in the design (or analysis) of any computer system. This approach makes it possible to utilize all the theoretic knowledge developed for PN and thus provides a powerful tool for computer systems (or program) design. In this case, the PN can be used as a framework that provides a tool for:n 1. the understanding of the design,n 2. handling the incremental changes of the design (without redesigning the entire system),n 3. the implementation of desired properties (like recoverability in case of failures, proper termination, mutual exclusion, etc.), andn 4. at least, the analysis of the program in order to check if it has the desired properties.

USES OF COMPUTERS IN THE OFFICE OF THE FUTURE

This chapter discusses the use of computers in the office of the future. Society is entering the “information age.” Computer communication is one essential ingredient in accessing information. The communication link may stretch halfway around the world, but it is equally important that it may only stretch to computers in adjacent offices, with very high communication bandwidths available. The ability to get widely distributed information is economically feasible and an important element of the office of the future. Most of the work that is conducted in business offices involves narrative information, rather than the kind of highly structured digital information stored in computers today. The office of the future will have interactive computer tools capable of helping knowledge workers in the office and handling the large volumes of narrative information in which they deal on a daily basis. The personal computer will put computer tools into the hands of knowledge workers, in the offices where they work, to aid them in doing their jobs. Because the commodity they deal in is information, these computers need to be highly interactive, and they need to function as friendly extensions to individual brain power. A major process in any office is planning. Some other major processes include resource allocation, monitoring execution of plans, and decision making.

Management impact of networks

Technical feasibility of networking is putting pressure on management in education, government and industry to deal with some new but fundamental policy issues. There are a number of important management issues to consider.