The nature of programming is changing. These changes will accelerate as improved software development practices and more sophisticated development tools and environments are produced. This paper surveys the most likely changes in the programming task and in the nature of software over the short term, the medium term, and the long term.
In the short term, the focus is on gains in programmer productivity through improved tools and integrated development environments. In the medium term, programmers will be able to take advantage of libraries of software components and to make use of packages that generate programs automatically for certain kinds of common systems. Over the longer term, the nature of programming will change even more significantly as programmers become able to describe desired functions in a nonprocedural way, perhaps through a set of rules or formal specification languages. As these changes occur, the job of the application programmer will become increasingly analysis-oriented and software developers will be able to attack a large number of application areas which could not previously be addressed effectively.
Computing Applications
Advertisement
Some practical experience with a software quality assurance program
Within a production programming environment, a software quality assurance program (QAP) was instituted to produce standards, conventions, and methodologies for all phases of the software development process. Software language standards and several support processors, in turn, developed. The authors offer a plan which may help others avoid some of the pitfalls they experienced while attempting to construct a meaningful software QAP.
Control flow and data structure documentation: two experiments
Two experiments were carried out to assess the utility of external documentation aids such as macro flowcharts, pseudocode, data structure diagrams, and data structure descriptions. A 223 line Pascal program which manipulates four arrays was used. The program interactively handles commands that allow the user to manage five lists of items. A comprehension test was given to participants along with varying kinds of external documentation. The results indicate that for this program the data structure information was more helpful than the control flow information, independently of whether textual or graphic formats were used.
Generating gamma variates by a modified rejection technique
A suitable square root transformation of a gamma random variable with mean a ≥ 1 yields a probability density close to the standard normal density. A modification of the rejection technique then begins by sampling from the normal distribution, being able to accept and transform the initial normal observation quickly at least 85 percent of the time (95 percent if a ≥ 4). When used with efficient subroutines for sampling from the normal and exponential distributions, the resulting accurate method is significantly faster than competing algorithms.
DIAGRAM: a grammar for dialogues
An explanatory overview is given of DIAGRAM, a large and complex grammar used in an artificial intelligence system for interpreting English dialogue. DIAGRAM is an augmented phrase-structure grammar with rule procedures that allow phrases to inherit attributes from their constituents and to acquire attributes from the larger phrases in which they themselves are constituents. These attributes are used to set context-sensitive constraints on the acceptance of an analysis. Constraints can be imposed by conditions on dominance as well as by conditions on constituency. Rule procedures can also assign scores to an analysis to rate it as probable or unlikely. Less likely analyses can be ignored by the procedures that interpret the utterance. For every expression it analyzes, DIAGRAM provides an annotated description of the structure. The annotations supply important information for other parts of the system that interpret the expression in the context of a dialogue.
Major design decisions are explained and illustrated. Some contrasts with transformational grammars are pointed out and problems that motivate a plan to use metarules in the future are discussed. (Metarules derive new rules from a set of base rules to achieve the kind of generality previously captured by transformational grammars but without having to perform transformations on syntactic analyses.)
Reciprocal hashing: a method for generating minimal perfect hashing functions
A method is presented for building minimal perfect hash functions, i.e., functions which allow single probe retrieval from minimally sized tables of identifier sets. A proof of existence for minimal perfect hash functions of a special type (reciprocal type) is given. Two algorithms for determining hash functions of reciprocal type are presented and their practical limitations are discussed. Further, some application results are given and compared with those of earlier approaches for perfect hashing.
Analysis of future event set algorithms for discrete event simulation
New analytical and empirical results for the performance of future event set algorithms in discrete event simulation are presented. These results provide a clear insight to the factors affecting algorithm performance, permit evaluation of the hold model, and determine the best algorithm(s) to use. The analytical results include a classification of distributions for efficient insertion scanning of a linear structure. In addition, it is shown that when more than one distribution is present, there is generally an increase in the probability that new insertions will have smaller times than those in the future event set. Twelve algorithms, including most of those recently proposed, were empirically evaluated using primarily simulation models. Of the twelve tested, four performed well, three performed fairly, and five performed poorly.
An algorithm for exhaustive generation of building floor plans
The combinatorial complexity of most floor plan design problems makes it practically impossible to obtain a systematic knowledge of possible solutions using pencil and paper. The objective of this paper is to contribute to the development of computer methods providing such knowledge for the designer. The paper describes an algorithm which generates all possible rectangular plans on modular grids with congruent cells, subject to constraints on total area, room areas, wall lengths, room adjacencies, and room orientations. To make room sizes regular and limit the solution set only such grids are used which minimize the number of cells in the smallest room. The description is sufficiently detailed to serve as a basis for programming. Test results for a Pascal implementation of the algorithm are reported. Realistic problems of up to ten rooms have been solved in modest lengths of computer time. The results indicate that the approach of exhaustive generation may prove to be more fruitful than generally assumed.
A two-list synchronization procedure for discrete event simulation
The traditional mechanism for maintaining a list of pending events in a discrete event simulation is the simple linked list. However, in large scale simulations this list often becomes cumbersome to maintain since the number of pending events may become quite large. As a result, the execution time required by the simple linked list is often a significant portion of total simulation time. Several papers have been published suggesting improved synchronization procedures. The most efficient procedures reported are the time-indexed procedure and the two-level procedure. Both methodologies are much more efficient than simple linked lists; however, neither has been adopted by a general purpose simulation language. Further, both procedures require external parameter definition, which is a major handicap to their adoption by a general purpose language. This paper introduces a new sychronization procedure, the two-list procedure, which is much faster than simple linked lists for large pending event files. This procedure was designed for implementation in Fortran, and properly implemented it is transparent to the user. Thus it is ideal for adoption by general purpose simulation languages.
A generalized user interface for applications programs
A general method for using disk files (instead of a more conventional parameter-passing mechanism) to transfer control information from a user interface to a set of related applications programs is described. This technique effectively moves much of the user interface, including command decoding and limited parameter checking, from the applications programs to a table-driven executive.
Microcomputers as remote nodes of a distributed system
A time-sharing system may be extended into a general-purpose distributed computing facility through the use of remote microcomputer systems. In such an extension the microcomputer system is not merely a remote terminal but a complete computer system. It becomes part of an integrated network of computers, each with significant local computational power with access to host system data resources. The implementation described emphasizes ease-of-use by minimizing any differences between local and remote data use.
Shape the Future of Computing
ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.
Get InvolvedCommunications of the ACM (CACM) is now a fully Open Access publication.
By opening CACM to the world, we hope to increase engagement among the broader computer science community and encourage non-members to discover the rich resources ACM has to offer.
Learn More