Efficient C/C++ Programming: Smaller, Faster, Better/Book and Disk - Softcover

Heller, Steve

Softcover

ISBN 10: 0123390958 ISBN 13: 9780123390950

Publisher: Morgan Kaufmann Pub, 1994

Synopsis

This book provides C and C++ programmers with invaluable tools for optimizing their code to run "large problems" on "small machines." Optimization is presented at the algorithmic level, and explanations are provided for how to re-think and re-cast applications to make them work more efficiently. Includes a disk containing source code for all examples found in the book.

About this title

Review

This book provides C and C++ programmers with invaluable tools for optimizing their code. Optimization is presented at the algorithmic level and explanations are provided for how to rethink and recast applications to make them work more efficiently. Numerous programming techniques are covered, including sorting, hashing, data compression, and the quantum file access method, which allows efficient access to records whos sizes vary dynamically.

Excerpt. © Reprinted by permission. All rights reserved.

You'll find that attitude again and again throughout this book: What's good is what works, and works efficiently; that is, without spending either space or cycles to no clear gain. Steve is not afraid to create a data structure that bows to the inevitable peccadilloes of the underlying hardware, if doing so will make it a more efficient data structure. Steve's totally awesome quantum file access algorithm (see Chapter 7) is a superb example. It's an algorithm, yes, but it was brought to life with the express intent that it run efficiently on late-build Intel processors-- primarily the 486.

Hey, if you want portability, be prepared to pay for it.

So let me make a few suggestions: Read this book, from start to finish, and while doing so set aside any preconceived notions you may have about "beautiful" code, or portable code, or lord knows, fast code. (Recursion, for example, is "beautiful." It's also dangerous, and slow.) Also, don't just look at the code. Load it, compile it, run it, understand it--not just as code, but as an element of a larger system that includes--rather than willfully ignores--the underlying hardware.

Once you do that, you will have learned a lot.

Not just theoretically. For real--and for all time.

Edition details

• Publisher: Morgan Kaufmann Pub
• Publication date: 1994
• ISBN 10: 0123390958
• ISBN 13: 9780123390950
• Binding: Paperback
• Language: English
• Number of pages: 500