On a Method of Multiprogramming - Practical Formal Development

In this review of On a Method of Multiprogramming the author examines a focused, formal approach to designing parallel and multiprogram systems from specification. The bottom line: this monograph is best for readers who need a rigorous method for moving from specification to correct multiprogram implementations, and its greatest value is the clear connection it makes between the Owicki/Gries theory and practical program development. It is a specialist work rather than an introductory textbook, and those seeking broad coverage of parallel hardware or high-level concurrency libraries may want a different resource.

Key Features

• Formal foundation: The book ties multiprogram design directly to the Owicki/Gries theory so readers gain a principled method for reasoning about correctness.
• Specification to implementation: The text demonstrates how to develop parallel programs from formal specifications, helping engineers keep correctness central during design.
• Applicability: Techniques are presented for multiprograms whether they are distributed or run on shared resources, making the method broadly useful.
• Focused scope: The monograph concentrates on program design issues rather than hardware details, so practitioners get concentrated guidance on algorithm correctness.
• Academic rigor: The presentation preserves theoretical rigor while emphasizing how the approach can be put to practical use in software engineering.

Who It's For

The book is aimed at advanced students, researchers, and software engineers who work with concurrent or parallel algorithms and who require a formal basis for correctness. Those familiar with program verification or formal methods will extract the most value because the material builds on the Owicki/Gries framework and expects some prior exposure to formal reasoning.

Readers looking for an introductory guide to parallel programming, language-specific concurrency tutorials, or hands-on coverage of parallel hardware should look elsewhere. This monograph is not a general-purpose how-to for multithreaded application APIs but a concise treatment of formal design techniques for multiprograms.

Pros & Cons

Pros

• Provides a direct link between specification and correct multiprogram design, useful for building reliable concurrent systems.
• Maintains theoretical rigor while showing practical application, which benefits researchers and methodical engineers.
• Applies to both distributed and non-distributed multiprograms, offering flexibility across deployment models.

Cons

• The focused, formal approach makes it less suitable for beginners or those seeking broad pragmatic tutorials on parallel APIs.

Specifications

Title	On a Method of Multiprogramming (Monographs in Computer Science)
Author	W. H. J. Feijen
Subject	Formal development of multiprograms from specification
Theoretical basis	Owicki/Gries theory
Scope	Design of parallel programs, distributed or not
Intended audience	Advanced students, researchers, software engineers

Our Verdict

On a Method of Multiprogramming is a concise, rigorous monograph that rewards readers who need a formal method to derive correct multiprograms from specifications. It is good value for researchers and engineers who prioritize correctness and formal development, though it is not a beginner's companion to parallel programming.

Frequently Asked Questions

Is this book practical for real-world projects?
The book emphasizes methods that can be applied in practice, particularly for projects where formal correctness is a priority, but it assumes familiarity with formal reasoning.

Do I need prior knowledge of Owicki/Gries theory?
Familiarity with Owicki/Gries or program verification concepts helps; readers without that background may find the text dense.

Does it cover distributed systems specifically?
Yes, the techniques are presented so they apply to both distributed and non-distributed multiprograms.