From Airline Reservations to Sonic the Hedgehog: A History of the Software Industry

by Martin Campbell-Kelly
The MIT Press, Cambridge, MA, 2003
392 pp., illus. 24 b/w. Trade, $16.95
ISBN: 0-262-53262-X.

Reviewed by John Knight
User-Lab, Birmingham Institute of Art and Design
Gosta Green, UK

John.knight@uce.ac.uk

Martin Campbell-Kelly was co-author of the brilliant Computer: A History of the Information Machine. The book, published in 1997, traces the origins of the computer. The authors uncovered a fascinating, but hidden story, and brought key figures (e.g. Ada Lovelace and Grace Hopper) to a wider audience. This new book takes a similar tack. It is crammed with information corroborated with personal reminiscences.

From Airline Reservations to Sonic the Hedgehog: A History of the Software Industry describes the commercial exploitation of software. It covers 43 years from 1952 to1995. The first computers, such as ENIAC, had to be programmed in binary code. Writing machine code is a formidable task, and programming languages were developed to increase productivity. Produced between the 1957 and 1964, the most influential of these, FORTRAN and BASIC, enabled programmers to work more efficiently and to develop useful functionality for a broad range of applications.

Large-scale military projects such as the SAGE air defence project bootstrapped the nascent industry and created a market for programmers and suppliers. This research enabled companies such as International Business Machines (IBM) to move from office machinery to computing.

However, it took over 10 years for demand to develop. Increased demand for computers occurred with hardware reliability, reduced costs, and the development of useful business applications. Business-based computer systems such as SABRE, the airline reservation system, often involved reshaping military applications for civilian use.

Campbell-Kelly charts the growth of the computer industry during the "go-go" years of the sixties. This era heralded a US dominated industry with international reach. The decade saw the release of groundbreaking products including IBM’s 360 and the first use of the term "software product." Despite these advances, the industry faced a "software crisis" that was characterised by late delivery and spiraling costs:

"Today we tend to go on for years, with tremendous investment to find that the system, which is not well understood to start with, does not work as anticipated. We build systems like the Wright brothers built airplanes . . . build the whole thing, push it off a cliff, let it crash, and start over again." (Graham cited in Naur and Randell, 1969:10)

In order to promote best practice, NATO convened the first software engineering conference in 1968. The 1970s began with IBM ‘unbundling’ its software. This process meant that third party vendors could sell software to IBM’s customers. This strategy led to a growth in the number of software providers and stock market speculation that foreshadows the dot.com boom:

"In the 1970s, the positive effects of IBM’s unbundling decision on the software industry were overshadowed by the effects of the crash in computer stocks and the computer recession of 1970-71. At the end of the decade, the market for software products was less than $2 billion. Nonetheless, the 1970s was the decade in which the industry was fully shaped and in which its numerous submarkets were defined." (p. 161)

As well as a maturing market, the 1970s crystallised research in the software user-interface. The personal computer provided a platform for innovations such as graphical user interfaces and windowed applications and promised software that could be used by anyone. The growth of the home and small business market resulted in a gold rush for PC products in the 1980s and a number of technical innovations, including Object-Oriented Programming. Through a combination of guile and luck Microsoft was able to exploit these advances and established an early industry standard operating system and office software. However, Campbell-Kelly points out that, "As this book has been at pains to point out, Microsoft still constitutes only about one-tenth of this extremely fragmented industry" (p. 264).

The latter part of the book deals with the struggle to monopolise markets. While grabbing headlines in anti-trust cases, some of the most successful and lucrative software products are relatively obscure. These include IBM’s CICS data communications programme that processes "20 billion transactions everyday" (p. 152).

Outsourcing and open source suggest a more turbulence for the traditional industry shapers. These factors come into focus in the last section of the book, which deals with computer games and has, perhaps, more in common with the music and film industries than with the origins of the software industry in office machinery.

Campbell-Kelly has written a fascinating book on an engaging but hidden subject. He is impartial and enthusiastic. There are some disappointments caused by scoping such a huge topic. Thus, we are left to wonder what changes have occurred since 1995 with the growth of Internet and mobile applications.