Search

The article presents an interview with computer scientist Donald Knuth. The emergence of structured programming in the 1970s awakened in him an intense desire to program, rather than just teach programming. A programming assignment he gave to graduate students increased his appreciation for the ambiguities and vagueness that can arise in trying to program a specification. He considers algorithm analysis to be his primary life's work.

The article describes an object-oriented language model of concurrency that deals with reusability and methodology. The model presented in this article makes extensive use of object-oriented techniques such as inheritance and polymorphism. This article has presented a model of concurrent programming using several unification's of concepts. It allows design of a concurrent system as a collection of abstract data types. The exception-based communication reconciles asynchrony and efficiency. Another point on which to focus is the asynchronous routine call combined with the sole and systematic data-driven synchronization, wait-by- necessity. This feature brings together the safety of structured programming with the power of asynchrony. A technique has been developed on top of the model to permit the reuse of the process live routines. This is achieved by defining the process body in a more declarative manner. The method first decomposes the system into a structured collection of sequential classes. The real time specifications of the system are left aside during this decisive stage of the system development.

The article presents the author views regarding the use of Go TO statements while generating a source code. In 1968 Edsgar Dijkstra, published his comments in the "Communications," GO TO Statements Considered Harmful. One GO TO was relegated to the status of a software obscenity, some programming managers began to dictate rules such as, "No GO TO statements are allowed in new program development." The goals behind eliminating the GO TO were to enhance source code clarity, reduce maintenance costs, and improve algorithmic reliability. Cobol, as originally designed, was not well-suited to "structured programming" because it lacked the more important language constructs required to properly support that style of programming. Use of PERFORM is suggested while creating the source code, instead of GO TO. GO TO statement is adequate for very small programs but "harmful" in larger more complex programs. Similarly, PERFORM is useful in small to medium programs but "harmful" in very large programs.

The article presents a discussion on systems programming with Jackson Structured Programming (JSP). In JSP and Jackson Structured Design (JSD), a computer program or process always models a data stream or a process in the outside world. Since most external processes and data are conveniently described as regular expressions with sequences, selections, and iterations, which are also the classical program structuring elements, this is indeed possible. In the article, the basic principles of JSP and JSD have been illustrated by means of an embedded micro-program controlling a video display unit. JSP represents a systematic method for designing a correct and efficient program from the elements of any language. The JSP designer starts by analyzing the relevant data streams. When their structures have been found, the fundamental similarity between data and programs is applied. In this way, the program becomes a model of its data. JSD applies the same idea to entire systems; it encompasses the steps from a preliminary system concept to functioning programs.

Several letters to the editor are presented in response to the article "A Generalized Control Structure and Its Formal Definition," by David Parnas in the August 1983 issue.

Presents letters to the editor referencing articles and topics published in earlier issues of the journal "Communications of the ACM." Focuses on issues relating updating of master files; Information on algorithms used in updating process; Introduction on structured programming in everyday activities; Alternatives used in processing the different file records.

Criteria are suggested for determining if a statement is a folk theorem. The ideas are then illustrated with a detailed example from the theory of programming. [ABSTRACT FROM AUTHOR]

Analyzes the algorithms for the Dutch National Flag Problem program. Observation of swap complexity of proposed algorithms; Assumption on the initial configurations of the problem; Application of Pascal procedure.

Discusses the implications of structured programming for machine architecture. Description of a GOTO-less language; Analysis of procedures written in GOTO language; Proposal for a machine architecture.

Application development today is too labor-intensive. In recent years, very high-level languages have been increasingly explored as a solution to this problem. The Business Definition Language (BDL) is such a language, one aimed at business data processing problems. The concepts in BDL mimic those which have evolved through the years in businesses using manual methods. This results in three different sublanguages or components: one for defining the business forms, one for describing the business organization, and one for writing calculations. [ABSTRACT FROM AUTHOR]

The basic ideas of nondeterministic programming are critically reconsidered to single out a proper attitude and programming style for languages allowing direct control of nondeterministic features. The proposed attitude aims at retaining the purity of the nondeterministic formulation of search processes on one level (the attempt level), deferring the coordination of problem solving efforts to another (the choice level). The feasibility of recognizing these two levels is discussed, stressing that the structure to be managed at the choice level is a tree of contexts. The leaves are computational environments, each holding an alternative under inspection, while the other nodes are associated with choice points. According to the proposed programming style, a generative function is associated with each choice point, which expresses the desired choice strategy. The main advantage of this approach is the localization of the search strategies: Each nonterminal node of the tree keeps track of the state of the computation as it was when the choice point was last interrogated, holding at the same time the strategy to coordinate the available alternatives. Examples are given in term of ND-Lisp, an extension of Lisp designed and implemented according to these guidelines. [ABSTRACT FROM AUTHOR]

In contrast to the predominant use of low-level intermediate text, high-level data flow analysis deals with programs essentially at source level and exploits the control flow information implicit in the parse tree. The need for high-level flow analysis arises from several aspects of recent work on advanced methods of program certification and optimization. This paper proposes a simple general method of high-level data flow analysis that allows free use of escape and jump statements, avoids large graphs when compiling large programs, facilitates updating of data flow information to reflect program changes, and derives new global information helpful in solving many familiar global flow analysis problems. An illustrative application to live variable analysis is presented. Many of the graphs involved are constructed and analyzed before any programs are compiled, thus avoiding certain costs that low-level methods incur repeatedly at compile time. [ABSTRACT FROM AUTHOR]

Presents several letters to the editor referencing the topics and the articles published in the previous issues of the journal "Communications of the ACM." Comments on the article "Structured Programming in Cobol," which focused on the benefits of using structured programming; Information on different aspects of structured programming; Remarks on the publications from (Association for Computing Machinery).

abstract data types can play a significant role in the development of software that is reliable, efficient, and flexible. This paper presents and discusses the application of an algebraic technique for the specification of abstract data types. Among the examples presented is a top-down development of a symbol table for a block structured language; a discussion of the proof of its correctness is given. The paper also contains a brief discussion of the problems involved in constructing algebraic specifications that are both consistent and complete. [ABSTRACT FROM AUTHOR]

Presents the method for describing and structuring programs simplifying proofs of its correctness. Application of the method to semi-automatic and automatic proofs; Concept of abstract programs; Description of abstract machine; Definition of the concept of correctness of a program P of Floyd.

Presents techniques for designing and writing Cobol programs. Definition of a well formed Cobol program; Implementation of top-down program design flowcharts, specifications and verification; Discussion of the proper use of the GO TO and other Cobol coding practices.

A sample set of Cobol programming standards is offered. These standards constrain code to be developed in a "structured" form for both data and control structures. They do not require syntax beyond the existing Cobol language and in fact utilize a typical limited subset of the 1974 ANS Cobol standard. These standards have proved extremely valuable in practice and have reduced the cost and time to produce and maintain large software systems that have been deployed in live multiple customer environments. [ABSTRACT FROM AUTHOR]

This paper defines exception conditions, discusses the requirements exception handling language features must satisfy, and proposes some new language features for dealing, with exceptions in an orderly and reliable way. The proposed language features serve to highlight exception handling issues by showing how deficiencies in current approaches can be remedied. [ABSTRACT FROM AUTHOR]

The issue of program control structures has had a history of heated controversy. To put this issue on a solid footing, this paper reviews numerous theoretical results on control structures and explores their practical implications. The classic result of Böhm and Jacopini on the theoretical completeness of if-then-else and while-do is discussed. Several recent ideas on control structures are then explored. These include a review of various other control structures, results on time/space limitations, and theorems relating the relative power of control structures under several notions of equivalence. In conclusion, the impact of theoretical results on the practicing programmer and the importance of one-in, one-out control structures as operational abstractions are discussed. It is argued further that there is insufficient evidence to warrant more than if-then-else, while-do, and their variants. [ABSTRACT FROM AUTHOR]

Reports on the design of hierarchically structured programming systems based on the concept of transparency. Use of the 'outside in' approach in determining solution to software design problems; Development of a method for evaluating the cost of requiring programmers to work with an abstraction of a real machine; Illustration of the method through examples from hardware and software.

The paper describes an efficient method for reduction of the binary matrices which arise in some school time-tabling problems. It is a development of that described by John Lions. It has been generalized and adapted to fit into the complete timetabling process; to use a more compact data representation and more efficient processing techniques; to take fuller advantage of possible available previous knowledge about the matrix. And it is designed as a structured program, which can readily be coded by the reader in the high level or low level programming language of his choice. Practical tests of the method have shown it to be a good basis for a realistic timetabling algorithm. [ABSTRACT FROM AUTHOR]

Discusses the importance of structured programming as a methodology for systematic program design and development. Identification of structured programs as compound function expressions in the algebra of functions; Characterization of structured programming in terms of the selection and solution of certain elementary equations defined in the algebra of functions.

Programming systems which permit arbitrary linked list structures enable the user to create complicated structures without sufficient protection. Deletions can result in unreachable data elements, and there is no guarantee that additions will be performed properly. To remedy this situation, this paper proposes a Data Structure Description and Manipulation Language which provides for the creation of a restricted class of data structures but ensures the correctness of the program. This is accomplished by an explicit structure declaration facility, a restriction on the permissible operations, and execution-time checks. [ABSTRACT FROM AUTHOR]

Several letters to the editor are presented in response to articles and topics discussed in previous issues, including an article about structured programming, about computer education for management and the case of physicist Andrei Sakharov.

The article presents author's opinion on the computer programming language C. According to the author C language should not be taught to students of a decent university. Students can gain knowledge by going through books available in any campus bookstore. Students should be taught a wide variety of programming languages, Pascal, Ada, Modula 2, Scheme, even the sadly neglected Algol 60 and then let them loose on the C book of their choice. If they can read it and understand it without any help from teachers, they deserve to be computer scientists. The article also presents response from some authors who do not support the author's idea of teaching other languages before teaching the programming language C. Some claim that the position of C in industry is the very reason why they should not adopt C as a student's first language. If structured programming is accepted as a principle, then C is a bad language as structured programming limits variation.