SOFTWARE ENGINEERING Report on a conference sponsored by the NATO SCIENCE COMMITTEE Garmisch, Germany, 7th to 11th October 1968 Chairman: Professor Dr. F. L. Bauer Co-chairmen: Professor L. Bolliet, Dr. H. J. Helms Editors: Peter Naur and Brian Randell January 1969 2 2 The present report is available from: Scientific Affairs Division NATO Brussels 39 Belgium Note for the current edition: The version of this report that you are reading was prepared by scan- ning the original edition, conversion to text through OCR, and then reformatting. Every effort has been made to do this as accurately as possible. However, it is almost certain that some errors have crept in despite best efforts. One of the problems was that the OCR software used kept trying to convert the original British spellings of words like ‘realise’ to the American spelling ‘realize’ and made other stupid mistakes. Whenever the OCR program was unsure of a reading, it called it to the attention of the operator, but there were a number of occasions in which it was sure, but wrong. Not all of these instances are guaranteed to have been caught. Although the editor tried to conform to the original presentation, certain changes were necessary, such as pagination. In order that the original Table of Contents and Indices would not have to be recalculated, an artifice was used. That is the original page breaks are indicated in the text thusly: 49 indicates that this is the point at which page 49 began in the original edition. If two such indicators appear together, this shows that there was a blank page in the original. The figures have been redrawn to improve legibility. The original size and scale was not changed. In order to accommodate the new pagination, the figures may have been shifted slightly from their position in the original document. Finally, it should be noted that the effort required to produce the current edition was a tiny fraction of the effort required for the original. The current editor therefore wants to express his apprecia- tion to the original editors, Peter Naur and Brian Randell, for producing what was clearly a land- mark effort in the Software Engineering field. Robert M. McClure Arizona 2001 NATO SOFTWARE ENGINEERING CONFERENCE 1968 3 3 HIGHLIGHTS The present report is concerned with a problem crucial to the use of computers, viz. the so-called software, or pro- grams, developed to control their action. The report summarises the discussions at a Working Conference on Software Engineering, sponsored by the NATO Science Committee. The Conference was attended by more than fifty people, from eleven different countries, all concerned professionally with software, either as users, manufacturers, or teachers at universities. The discussions cover all aspects of software including • relation of software to the hardware of computers • design of software • production, or implementation of software • distribution of software • service on software. By including many direct quotations and exchanges of opinion, the report reflects the lively controversies of the original discussion. Although much of the discussions were of a detailed technical nature, the report also contains sections reporting on discussions which will be of interest to a much wider audience. This holds for subjects like • the problems of achieving sufficient reliability in the data systems which are becoming increasingly integrated into the central activities of modern society • the difficulties of meeting schedules and specifications on large software projects • the education of software (or data systems) engineers • the highly controversial question of whether software should be priced separately from hardware. Thus, while the report is of particular concern to the immediate users of computers and to computer manufacturers, many points may serve to enlighten and warn policy makers at all levels. Readers from the wider audience should note, however, that the conference was concentrating on the basic issues and key problems in the critical areas of software engineering. It therefore did not attempt to provide a balanced review of the total state of software, and tends to understress the achievements of the field. NATO SOFTWARE ENGINEERING CONFERENCE 1968 4 4 5 CONTENTS HIGHLIGHTS . 3 CONTENTS . 5 PREFACE . .9 1. BACKGROUND OF CONFERENCE . 13 2. SOFTWARE ENGINEERING AND SOCIETY . 19 3.1. THE NATURE OF SOFTWARE ENGINEERING . 19 3.2. SOFTWARE ENGINEERING MANAGEMENT AND METHODOLOGY . 24 3.3. DESIGN AND PRODUCTION IN SOFTWARE ENGINEERING . 32 4. DESIGN . 35 4.1. INTRODUCTION . 35 4.1.1. Sources of techniques . 35 4.1.2. Need for hardware based on program structure . 35 4.1.3. Relation to mathematics . 37 4.2. DESIGN CRITERIA . 38 4.2.1. Genera design criteria . 38 4.2.2. User requirements . 40 4.2.3. Reliability and design . 44 4.2.4. Logical completeness . 44 4.3. DESIGN STRATEGIES AND TECHNIQUES . 45 4.3.1. Sequencing the design process . 45 4.3.2. Structuring the design . 50 4.3.3. Feedback through monitoring and simulation . 53 4.3.4. High-level languages . 55 4.4. COMMUNICATION AND MANAGEMENT IN DESIGN . 59 5. PRODUCTION . 65 5.1. INTRODUCTION . 65 5.1.1. The problems of scale . 65 5.1.2. The problems of reliability . 70 5.2 PRODUCTION — MANAGEMENT ASPECTS . 72 5.2.1. Production planning . 72 5.2.2. Personnel factors . 83 5.2.3. Production control . 86 5.2.4. Internal communication . 89 6 5.3. PRODUCTION — TECHNICAL ASPECTS . 93 5.3.1. Tools . .93 5.3.2. Concepts . 96 5.3.3. Performance monitoring . 99 6. SERVICE . 103 6.1. INTRODUCTION . 103 6.1.1. The virtue of realistic goals . 103 6.1.2. Initial system release . 103 6.1.3. Frequency of releases . 104 6.1.4. Responsibility for modified systems . 106 6.2. REPLICATION, DISTRIBUTION AND MAINTENANCE . 107 6.2.1. Replication . 107 NATO SOFTWARE ENGINEERING CONFERENCE 1968 5 6.2.2. Distribution . 109 6.2.3. Maintenance . 110 6.3. SYSTEM EVALUATION . 112 6.3.1. Acceptance testing . 113 6.3.2. Performance monitoring . 114 6.4. FEEDBACK TO MANUFACTURERS FROM USERS . 115 6.5. DOCUMENTATION . 116 6.6. REPROGRAMMING . 117 7. SPECIAL TOPICS . 119 7.1. SOFTWARE: THE STATE OF THE ART . 119 7.1.1. Introduction . 119 7.1.2. Problem areas . ..