Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
skip to main content
Skip header Section
Network computing system reference manualJanuary 1990
Publisher:
  • Prentice-Hall, Inc.
  • Division of Simon and Schuster One Lake Street Upper Saddle River, NJ
  • United States
ISBN:978-0-13-617085-3
Published:03 January 1990
Pages:
396
Skip Bibliometrics Section
Reflects downloads up to 06 Oct 2024Bibliometrics
Contributors
  • Apollo Technologies, Inc.
  • Apollo Technologies, Inc.
  • Microsoft Research
  • Apollo Technologies, Inc.
  • Apollo Technologies, Inc.
  • Massachusetts Institute of Technology
  • Apollo Technologies, Inc.

Reviews

Donald W. Gotterbarn

The Network Computing System (NCS) is an implementation to manage distributed computing on homogeneous and heterogeneous computer systems. NCS is a set of tools consisting of a Remote Procedure Call runtime library, a Network Interface Definition Language Compiler, and a Location Broker, all developed by Apollo Computer. The authors have written more than a reference manual for a particular implementation. The book describes an approach to distributed computing that can run on a variety of systems such as MS-DOS, VMS, ULTRIX, SunOS, and Apollo. The book introduces distributed computing concepts including remote procedure calls, client-server communications, replicated files, and concurrent processing. In discussing NCS concepts it also touches upon several other distributed computing concepts. The book would therefore be good reference reading for a distributed computing systems course, but because it only deals with NCS and does not give any reasons for the design decisions, it is not as useful as it might have been for students of distributed systems. A short section on developing distributed applications is not very useful for either the application developer or the student of distributed systems. It does not discuss design issues or error handling. Although NCS includes a facility for the implementation of replicated files at discrete nodes, the authors never mention update techniques. They do not even discuss timestamps in a single server multiclient environment. In fact, the only design recommendations in the book are a general preference for the use of opaque ports over “well-known ports” (ports which have been assigned by protocol administrators) and a preference for automatic binding over manual binding. The nature of a system militates against some designs, but the authors make no attempt to inform the reader about design decisions. The last half of the book deals explicitly with the syntax of the Network Interface Description Language (NIDL) and NCS. The chapters defining a network interface would be interesting for data communications students to look at for the overall structure of an interface definition. The book is intended as a reference manual for the application developer, but cannot be used for that purpose. The introductory sections contain nothing new to people skilled enough to develop distributed applications. The material that deals specifically with syntax and programming does not provide all the information needed by the developer. Many sections say a particular step, such as programming the server, is not included because it is covered in some other book. When talking about NIDL, the authors say, “we present NIDL syntax informally, sacrificing precision for readability.” These ellipses leave the application developer in limbo. The book does not contain enough information to be useful as a reference manual.

Access critical reviews of Computing literature here

Become a reviewer for Computing Reviews.

Recommendations