Large Open Systems Research Articles

The programming language PS440 as a tool for implementing a time-sharing system

The use of assembly language for the programming of basic software of a large operating system often proves inflexible and impractical. Writing the programs in assembly language is time-consuming and expensive. Furthermore, the programs are hard to read and difficult to modify. Nevertheless, assembly languages have facilities which must be preserved when higher level programming languages are to be used for basic software. A programming language, “PS440”, is described which provides many of the advantages of higher level languages while at the same time maintaining some needed facilities of assembly language.

Right and Wrong

I think some clarity is needed in the flak flying in the REVIEW between what Engel represents (see PAR Communications July/August 1969) and what Hamilton represents (ibid.). Both are partially right and partially wrong in their conclusions. Engel is correct in stating that the methodologies gathered under the heading of operations can be helpful. That is, if: (1) the analyst realizes that the methodology that is used with in one type of system (e.g., physical) may necessitate reformulation for use in another system type (e.g., social systems in the civil sector), and (2) that the content of the analysis must include the political variables of attitudes and value structures, spectrum positions, and intensity levels of particular publics and policy formulators. If Mr. Engel inferred such when he stated using the combined knowledge of hard and soft scientist alike, it is not clear. Hamilton is correct in his disdain for the lack of success to date. A number of the RAND efforts in New York have been performed on the basis of collecting measurable data, running such through packaged computer analysis, and sitting in one's office or at hectic meetings drawing conclusions from the output. Much of this social system has been carried out by men trained only in physical systems analysis such as electrical engineering. The fireman's friend statement, by example, should not mean that a study not be made, but that operational purpose models necessitate the combined efforts of the theoretician trained in the behavioral sciences and OR methodologies and the administrator, politician, or low-level public servant engaged in the day-to-day activity. Much of the research by systems-thinking organizations has instead used a canned approach they have refined as a result of weapons and communication systems analysis. Although the OR analyst does not hold that facts . . . speak for themselves, he however is not accustomed to the amount of interpolation necessary in the modelling of large open systems with such a high degree of human activity. The firms referred to above are basically guilty of a lack of imagination and humility to the degree that they have produced much research but little that is relevant for implementation. If Mr. Hamilton inferred such with his statement bent on systematizing that which is not and cannot be a system within their rubric, it is not clear. STEPHEN D. SLINGSBY Public Administration and Management Science Faculties, and Director, Politicometrics Program Ohio State University

An Engineering Approach to Software Configuration Management

Increasingly, digital computers are being incorporated as major hard-ware subsystems in today's large support and operational systems. As a result, computer programs and complex operational procedures, software, are also becoming major system elements. Contrary to much current practice, software must be managed, engineered, and controlled in the same manner as hardware if past pitfalls are to be avoided. NASA's NPC 500-1 and the USAF 375 series establish a basis for an effective hardware/software systems development methodology that must be adopted by both hardware and software engineers and managers.

The Vertical-Flow Interrupter and Its Application to ``Oil-Poor'' Circuit Breakers

The historical background of the development of the vertical-flow interrupter for ``oil-poor'' circuit breakers is reviewed. A description of the interrupter and its theory of operation is given. The application of two of these interrupters to a 138-kv 1½-million-kva ``oil-poor'' porcelain-clad circuit breaker is shown, together with laboratory test results. A review is given of important field tests on a 138-kv single-pole unit, tested line to ground on a large 220-kv operating system up to approximately 2¼ million three-phase kva. Magnetic and cathode-ray oscillograms of these tests are shown.

Progress in the Study of System Stability

In the second part of this paper and in the appendices attention is given to simplified methods of treating the problem of system stability. Methods are recorded which have been found useful in making many system studies. These methods have had considerable verification by tests both on a model system and on large operating systems and have been simplified to such an extent that many operating companies are now finding it to their advantage to undertake the work of making careful studies of their own systems, just as they now make short circuit studies which formerly were thought too difficult and too highly theoretical. The first part of the paper gives some comments and conclusions of the authors and their colleagues as a result of many such system studies as well as studies involving more detailed methods, and also as the result of practical experience through contact with various operating companies. Some of these comments are based directly on an example which is given in detail in Appendix I.

Abridgment of progress in the study of system stability

In the second part of this paper and in the appendices, attention is given to simplified methods of treating the problem of system stability. Methods which have been found useful in making many system studies are recorded. These methods have had considerable verification by tests both on a model system and on large operating systems and have been simplified to such an extent that many operating companies are now finding it to their advantage to undertake the work of making careful studies of their own systems, just as they now make short-circuit studies which formerly were thought too difficult and too highly theoretical. The first part of the paper gives some comments and conclusions of the authors and their colleagues as a result of many such system studies as well as studies involving more detailed methods, and also as the result of practical experience through contact with various operating companies. Some of these comments are based directly upon an example which is given in detail in Appendix I.