Search

Microscopic level Job stream data obtained in a production environment by an event-driven software probe is used to drive a model of a multiprogramming computer system. The CPU scheduling algorithm of the model is systematically varied. This technique, called trace-driven modeling, provides an accurate replica of a production environment for the testing of variations in the system. At the same time alterations in scheduling methods can be easily carried out in a controlled way with cause and effects relationships being isolated. The scheduling methods tested included the best possible and worst possible methods, the traditional methods of multiprogramming theory, round-robin, first-come- first-served, etc., and dynamic predictors. The relative and absolute performances of these scheduling methods are given. It is concluded that a successful CPU scheduling method must be preemptive and must prevent a given job from holding the CPU for too long a period. [ABSTRACT FROM AUTHOR]

An overview of the current and future positions with respect to operating system performance is given. While a great deal of information and a large number of models for subsystems have been developed, gaps still exist in our knowledge. Because of the severe interactions between the various subsystems of an operating system, an overall model of the total system must be developed to be able to analyze and design the performance aspects of an operating system. Although such total system designs are exceptional today, it is projected that they will become increasingly more common and necessary in the near future. Such a design philosophy will clearly have a severe impact on the way we go about modularizing operating and computer systems. [ABSTRACT FROM AUTHOR]

Presents letters to the editor referencing articles and topics discussed in previous issues. Description of proprietary software packages; Importance of proprietary software packages in the computer field; Problems related to the access to the U.S. Government's scientific and technical information; Selection of IFIP Congresses sites.

This article presents news briefs related to the computer industry. Fred P. Brooks has been appointed Association of Computing Machinery's (ACM) representative to the National Academy of Sciences-National Research Council. Brooks is a professor in the department of information sciences at the University of North Carolina in Chapel Hill, North Carolina. The ACM Professional Development Program will continue its policy of one and two-day seminars in conjunction with national meetings at major cities with two one-day seminars to be presented on August 28, 1967 prior to the ACM 1967 National Conference in Washington D.C. A Special Interest Committee on Programming Languages of the ACM has decided to add an information bulletin on the string manipulation language SNOBOL to those on ALGOL, PL/I, COBOL and JOVIAL. William M. Waite, a scholar of the University of Colorado, has been appointed editor of the bulletin. He invites contributions concerning SNOBOL which may be of interest to users and implementers of the language or to the computing community in general.

This article focuses on the techniques for representing strings in Fortran IV. FORTRAN IV is one of the most widely used programming languages, and programs written in FORTRAN tend to be quite portable. Although the language has, at best, only primitive string-handling capabilities, many programs that deal primarily with strings are sometimes written in FORTRAN simply to achieve a degree of portability. It is the purpose of this note to illustrate a very simple technique for structuring and manipulating varying-length string data in FORTRAN IV. Two pieces of information characterize varying length strings--their location in memory and their length. In languages such as SNOBOL4 this information is usually contained in a descriptor. The location field, or first word of the complex variable, may be referenced by referencing the integer variable OFF. Likewise, the second word is referenced via the integer LEN. The complete string descriptor is referenced via the complex variable D. One advantage of the descriptor representation is that strings can overlap since there are no special markers embedded in the storage array.

Presents the author's views on the issues related to the Association for Computing Machinery (ACM). Opinion concerning what an appropriate relationship between the ACM and the computer industry should be; View that ACM must continue to be involved in cooperative ventures with the computer industry so long as these are beneficial to ACM's primary scientific-technical-educational purpose; Discussion over a recent ruling by the U.S. Supreme Court concerning the patentability of computer programs.

The article focuses on the need for the United States government to improve computer software procurement since the enactment of the Brooks Bill in 1965. The relevant part of the Brooks Bill sets up computer-buying powers under the General Services Administration and gives strong guidance on the need to save the taxpayer's money in such procurements. For its time, the Brooks Bill performed an outstanding service. Through the years, however, the procurement process has woven a web that threatens to immobilize the people who are responsible for effective and innovative applications. The bureaucratic implementation of the Brooks Bill has been exquisitely precise--and completely blind to the changing economics of computer applications. The General Accounting Office has issued a report that indicates a desire to place software procurement by the government on a commodity basis. The only way that hardware and/or software procurement for the government can be improved is by thorough attention to the skills of the information systems staff, to the total costs of the whole information system and to the performance of the resulting systems when compared to management's requirements.

The article presents information on the 1970 Fall Joint Computer Conference that will be held in Houston, Texas, on November 17-19. The conference will have a truly Texas-sized setting, the Astrohall, the world's largest convention and exhibit structure. The technical program may also appear to be Texas-sized in view of its size and diversity. However, it is not so much the program that has grown as the computer industry. The program has therefore been expanded to reflect new developments in computer usage and technology while maintaining its emphasis on various subjects of continuing importance. One of the chief areas of interest is computer software, its present problems and future promises. Two sessions will be devoted on it. Session 21, chaired by M.A. Melkanoff, University of California at Los Angeles, is devoted to the three principal and self-excluding approaches to long-range programming language development being advocated today. A distinguished panel will present a position paper on each of the three approaches and a fourth position paper on relevant research problems.

This article analyzes the pros and cons of patenting computer programs in the U.S. A proposed set of guidelines that would aid patent examiners in reviewing applications for computer program patents has raised opposition in Washington D.C. The consensus in the patent office is that the guidelines are inadequate. The House Judiciary Committee plans to hold hearings this year on an entire revamping of the patent laws in the U.S. Bell Laboratories is supporting the program patents, but most manufacturers remain solidly against patenting programs. Some software firms would like to receive program patents; however, the resistance shown by the Patent Office to patent computer programs is currently blocking the way.

This article presents information on various products related to the computer industry launched in the U.S. Sperry Rand Corp.'s Univac Division has announced the establishment of a new series of integrated computers that will span the small, medium and large-scale data processing market. The first members of the new 9000 computer series, recently unveiled, are the UNIVAC 9200 and UNIVAC 9300. A small, low-cost real-time computer for scientific, engineering and process control applications, Sigma 2, was announced on August 2, 1966 by Scientific Data Systems (SDS). Featuring multiprogramming and multiprocessing capabilities, Sigma 2 is the second real-time computer in the new SDS family of Sigma machines. A new, low-cost offline printer system, designed to relieve general purpose computers of routine printing operations, has been introduced by Potter Instrument Co. A new software package for the PDS 1020 digital computer has been developed by Pacific Data Systems Inc. of Santa Ana, California.

Information about several panel discussions held at the International Federation for Information Processing Congress 65 held at Paris Hilton Hotel in New York City in 1965 is presented. At the panels "Legal Aspects of Software," and "Content Addressable Memories," use of patent law in protecting ownership of computer programs and organizational features of associative memories respectively were discussed. "Economics of Programming" panel discussed strategies for managing programming costs.

Presents information on a letter from the ACM President about scientific activities sponsored by the North Atlantic Treaty Organization (NATO). Problems related to software engineering; Participation of NATO Science Committee in scientific activities; Discussion on matters related to software designing.

Focuses on the increase in regulation of computers by the U.S. government. Monopoly in the U.S. computer industry; Involvement of court cases in antitrust law; Difficulty in patenting in the computer field for the future availability of high quality, competitive software.

This article presents information related to the revised algorithm policy of 1961. A contribution to the algorithms department must be in the form of an algorithm, a certification, or a remark. Contributions should be sent in duplicate to the editor, type written double-spaced in capital and lower-case letters. An algorithm must be written in the ALGOL 60 Reference Language, and normally consists of a commented procedure declaration. Each algorithm must be accompanied by a complete driver program in ALGOL 60 which generates test data, calls the procedure, and outputs test answers. Moreover, selected previously obtained test answers should be given in comments in either the driver program or the algorithm. The driver program may be published with the algorithm if it would be of major assistance to a user.