Real Computing Made Real

In the acknowledgements to the first edition of Numerical Recipes in Fortran, Wm. Press et al write "We also wish to ackowledge two individuals we have never met: Forman Acton, whose 1970 textbook Numerical Methods That Work has surely left its stylistic mark on us...." (The second being Donald Knuth) In Real Computing, Acton expands in greater detail upon the theme of the earlier work, which is that good computing routines are problem-specific. That insight and understanding are called-for; recipes and black box routines are insufficient. In the fifties, with powerful mainframes few and far between, engineers farmed out their computations to a computing center. There you could find the guys who knew intelligent machine computing. When the IBM 370 came along, every university had its own machine, and computations became in-house. The expertise you found in the old centers wasn't there. Acton worked in one of those centers, and this book is written to guide the reader in acquiring some of the old magic.