B74-14 Data Base Systems

D. F. Hartley. Concentrates on functional specifications of operating systems. "Front End Systems," by P. C. Poole. Describes various frontend architectures for handling interaction, job control, and job preparation. "Building on to an Operating System," by J. I. Strauss. Brief discussion of the why and how of adding function to an existing operating system. "The Use of Simulation in the Design of a Multi-Computer Operating System," by G. K. Hutchinson. The particular system designed drives machine tools and interactive design terminals satellited to an IBM 1800. "The Interaction Between Operating Systems and Machine Architecture," by D. Howarth. Explores some of the influences of these interactions on system structure and on the functions available to the user. "Operating Systems for a Range of Computers," by B. J. Moore. Describes the operating system family designed for the ICL series of computers. "Multiprocessor Operating System Design," by L. L. Walker. Describes problems, design principles, and implementation factors for multiprocessor systems. "Operating Systems for Real Time," by P. R. Cox. Brief overview of some of the characteristics of real-time systems. "Experience of Operating Systems Performance Measurement," by J. E. Hailstone. Describes some of the measurement studies performed on the Atlas installation at Chilton. "The Development of a Special Purpose Operating System," by D. W. Barron. Describes a system for an ICL 1900 that is special purpose in the sense that it was designed to exploit a particular machine with a particular profile of user jobs. "Protection and Access Control in Op@rating Systems," by B. W. Lampson. Describes some of the problems involved with providing protection. This is one of the more technical presentations. The invited papers consist of the following: "Discussion of LP 70 as a Language for Systems Implementation," R. Aslanian and R. Recio. Describes a particular language. Would have been much better had they discussed some of the general problems dealing with implementation languages. "Deadlocks in Computer Systems," by E. G. Coffinan, Jr. This is a cut down version of his paper in Computing Surveys 3, 2. It is a rather good survey. "Thie Developing Theory of Operating Systems," by P. J. Denning. A high-level, but interesting, overview of the topics involved in current operating systems technology. "Architectural Features of System 250," by D. M. England. Describes a particular system. No general lessons are drawn. "GEORGE 3, the Compleat Operating System," by G. B. Newell. Description of another system. "Features of FACOM 230-60/75 and Its Operating System," by H. Okada and Y. Tsushima. Yet another. "System Monitoring," by L. Presser and D. N. Dumont. Short review of the techniques of monitoring an operating system's performance. "Foundation Software for Real Time," by D. E. Rimmer. Foundation software shows characteristics of the underlying hardware rather than of the functions performed. Not particularly enlightening. "Job Control Languages: What the User Really Requires," by H. J. Weegenaar. Discussion of JCL characteristics and a plea for more rationalized forms. The bibliography, prepared by Peter Denning, is annotated and is useful. Overall, this volume is a reasonably good survey of the current state of the art of operating systems at the design and architectural level. It does not contain much detailed technical information (e.g., scheduling algorithms) and is especially lacking in the areas of memory management and file systems. The expense of the volume and the lack of much hard technical information makes it a volume which is suitable for (and probably will only be used by) managers and others concerned with operating systems at a high level. For that purpose it is probably as good a volume as can be put together in a short time. Now some detailed comments: the analysis section is really not analysis at all since much of it consists of author's explanations and advertisements for their presentations later in the volume. Most of these would be better placed at the end of their respective papers. Issues are raised and discussed, but sometimes only through the medium of text justaposition by the editor. Some questions are raised but not answered. This section does serve to give one a glib introduction to the rest of the volume and provides a number of pointers. While the editing is rather skillful, a true analysis of the material in the rest of the volume would be more useful. The presentations, by and large, are rather good. They are mostly overviews and introductions, are not too technical, and are presented at a consistent level both within and between papers. In most cases their organization is rather good and easy to follow. The invited papers are more detailed for the most part. Most present details on particular systems rather than investigating particular topic areas, such as scheduling or memory management. They are not as good as the presentations, although some of them are certainly worthwhile. The fact that all of the attendees at the symposium and the majority of the authors and lecturers were European and British reduces the usefulness to an American audience a little. In particular, the references to systems and machines familiar to those present but not to Americans makes some of the detailed comments less than useful on this side of the Atlantic. The index is very thin, only 14 pages with generally only one reference for each item. Otherwise, the volume as a whole is a wellorganized document. There is a section detailing editorial conventions and how best to use the volume. In summary, this volume is probably useful to high-level technical managers and other decision makers who wish to obtain a broad overview of the current architectural state of the art of operating systems. It will not be very useful to individuals who wish to survey the current detailed technological state of the art.