Customer Reviews

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

5 star:		(4)
4 star:		(5)
3 star:		(1)
2 star:		(1)
1 star:		(1)

 

 

I wasnt sure how I would like Martin Campbell-Kellys new book since I have grown up with the software industry--I started as a programmer in 1973and thought I had a pretty good understanding of it. But on the contrary I found his book fascinating. Over all I thought Campbell-Kelly weaved together many hard-to-find facts and statistics into a comprehensive, well-researched, and well-written story of how the software industry evolved

Basically, it reviews the development of the various software industry segments; what we now refer to as professional services companies, enterprise product software companies, personal computer software companies and game companies. For each, it describes the major events that created these industry segmentse.g. in 1970 IBM unbundled its software pricing from its hardware pricing ushering in age of product software companies; the major companies who dominated these segments; and the business models.

A few things that were of particular interest to me.

When I grew up in the software business, we used lines of code as a way to estimate the work it would take to create an application. I have overseen the development of systems with millions of lines of code. So I was surprised to learn how few lines of code were in DOS 2.0, just 20,000+, and even more surprised to learn that the venerable early business language Fortran, in its first incarnation, was only 18,000 lines of code.

And it helped put into context what I have lived through. For example, in 1973, straight out of C.C.N.Y., I first entered the job market as young Cobol programmer trainee at Royal-Insurance Company in New York. As a youngster I had no awareness that Thomas Watson Jr. bet his business on the IBM360, or that it was the reason for IBMs unprecedented industry dominance. I just knew that I worked on IBM 370-145, an updated, at that time, version of the IBM 360.

And does this sound familiar? Never before has the stock market shown quite so much enthusiasm about an industry as it has lately about the computer industry. Recent prices of computer stocks represent some of the highest prices-earning ratios ever recorded. Even shares of giant I.B.M, which increased six fold between 1957 and late 1966, have doubled since. In July the market valued I.B.M., whose physical assets amount to less than $6 billion , at more than $40 billionmore than any other company in the word, actually as much as gross national product of Italy. And the market value of smaller and newer companies in the industry has gone up even more steeply than I.B.M.s. In less than three years the price of University Computing Co. of Dallas rose form $1.50 a share(adjusted for splits) to $155. The stock market valued this newcomers, who sales last year were than $17 million, at more than $600 million.

This quote that Cambell-Kelly provides is about the timeframe 1964-1968, which was another time period of irrational exuberance, like our recent dot.com debacle. Only I hope that our recovery will be sooner than that one. It evidently, took almost ten years for the market to again buy software company stock.

I do have one quibble, though. Campbelly-Kelly challenges the common perception that Microsoft dominates the technology industry. He quite correctly cites statistics which show that, even if Microsoft is a giant, IBM is a titan, towering over Microsoft; where Microsoft has at most owned 10% of the software industry, IBM owned fully 75% of the entire IT industry consisting of computer hardware, software and services. To put this into perspective, the total of computer hardware, software and services revenues world-wide today is thought to be something like $1 trillion. If IBM had maintained its earlier level of control it would now be worth $750 billion while Microsoft today has revenues of roughly $28 billion. His point is that while IBM was almost the entire industry by itself, Microsoft controls just a piece of the software industry.

While this is true, I dont believe the issue of Microsofts dominance is about revenues. I believe it is about cultural impact--which Campbell-Kelly does allude to. Microsoft, almost single-handed, created a cultural revolution, making it fashionable to be a nerd, to dress casually in business, to work insane hours, to believe as a young twenty-something you can be an over-night millionaire. Whereas IBM made an impact on businesses--certainly being viewed for a long time as one of, if not, the best-run business on the planet--Bill Gates and company, captured the publics imagination and changed our society in a way IBM never did. And why not, when Bill Gates and Paul Allen are the first and third richest men in the world.

All in all though, this book was an enjoyable and insightful read.

This is a history of the Software Industry. "Software" was coined to distinguish it from hardware; it describes the spirit that activates electronic machines. There are three sectors: software contracting, corporate software products, and mass-market software products (pp.3-8). The book covers events from around 1950 to 1995 in the USA. Chapter 1 gives an overview of the sources available. Chapter 2 tells of the origins of software writing, and its need for high-maintenance. Could errors arise from "one minor change"? Early users cooperated in sharing software. FORTRAN and COBOL became the first standard programming languages. But high costs and slipped schedules became typical. Government support for SAGE helped establish US dominance of the computer industry (p.48). The "Great Society" led to investments in non-defense projects.

Chapter 3 discusses "Programming Services". The established techniques of engineering management filtered into programming projects. Program flowcharts became institutionalized, then flushed away by the "fad for 'structured programming'" (p.69). The boom for software companies in the late 1960s reminds me of the dot-com fever in the late 1990s. All fueled from government spending (p.75, P.80). The arrival of minicomputers around 1970 allowed middling companies to own a computer. Chapter 4 tells about the change to "Software Products". Computers were more plentiful and more powerful (pp.90-91), programmers didn't keep up. Lines of code used increased 1000% every 5 years, the cost of developing quadrupled by 1965. Page 100 discusses flowcharting, whose purpose was to graphically represent a program's operations. Sort of like a condensed slide presentation of a topic. Page 102 tells of a secret machine instruction used to improve sorting speed (what was it?).

Chapter 5 tells how the software industry acquired its current shape, and gives an overview. Software products was a capital goods business. Industry specific software requires in-depth knowledge; in systems software programming skills are critical. The success of CICS can be compared to a system of roads where applications can freely travel (p.151). Chapter 6 discusses the maturing of corporate software packages, and growth through acquisition. It focuses on three large firms that became prominent in the 1990s. Some grew by acquiring smaller firms for their products (diversification). The rise of the relational database had an adverse affect on older database technologies. The use of fully integrated business application software (ERP) created new companies. Pages 182-4 overviews the successes of Computer Associates. A relational database did not require knowledge of the internal structure of the database; ever faster computers masked its relative inefficiency. Sales of SAP R/3 benefited from the "fad for business re-engineering" (p.195). Page 197 explains why SAP is more important that Microsoft.

There are strong parallels with other historical systems, such as railroads to airlines. If the database was bundled with the operating system there would be no independent vendors. European firms were able to pioneer ERP because they not not been locked into "legacy software" (p.199). The remaining chapters discuss the history of the personal computer.

From Airline Reservations to Sonic the Hedgehog may sound like a mystifying title, but this book provides a reasonable overview of the history of the software industry. At times, given the ups and downs in the industry, it can't avoid sounding like a catalog of defunct firms and obsolete software. However, this chronology is quite useful for anyone who wants to come up to speed very quickly and very generally on the main trends in the industry. Author Martin Campbell-Kelly covers some of the industry's seminal events and the main categories of software. Vexingly or refreshingly, he takes pains to say as little about Microsoft as possible, making it clear that others have written enough on that subject. So, with that absence duly noted, we recommend this book to those who want an inside history of the software industry, from massive mainframes to little blue cartoon porcupines.

This history of the Software Industry covers personal computers in the last three chapters. The "Acknowledgments" lists his sources and references. Chapter 7 reviews the early development of microcomputers. The invention of the microprocessor in 1971 made microcomputers possible (p.201). The Apple II was the transforming event of April 1977. The fall 1979 release of VisiCalc transformed "toy computers" into a useful machine for businesses. Digital Research's CP/M allowed any application to run on any computer that used CP/M; this allowed program vendors to access a larger market. Microsoft eclipsed DR by providing DOS for the IBM PC, and its games and programming languages. PC software was usually sold by mail, then at stores. The invention of VisiCalc is credited as boosting the market for personal computers. Productivity applications drove the software industry in the early 1980s (p.215). Word processing was aimed at home computing; Word Star was the most successful. Most computer games were produced by sole authors, lasted a few months, and made little money.

Chapter 8 discusses the now mature PC industry. Why did a few companies succeed where many failed? "The Autodesk File" says: product improvements, complementary products, training networks (p.243). Technical competence does not guarantee success unless it meets user needs (p.244). The need to work with two or more applications simultaneously led to "windowing" (p.247); but this required more time and money than first estimated (p.251). Page 253 tells of the big mistake by Lotus' management in rewriting the program. A similar mistake doomed Word Star (p.255). Ashton-Tate's demise is described on page 257. These were one-product companies. Page 259 explains Microsoft's winning strategy for its Office Suite. Page 264 tells of Symantec's strategy for success.

Chapter 9 describes software used for entertainment, and looks at videogames, CD-ROM encyclopedias, and personal finance software. Arcade games replaced older pinball machines during the 1970s. Videogame consoles for the home allowed playing many games. Home computers had a keyboard and secondary storage, and could be programmed by the user. Videogames are similar to recorded music's stream of new titles, and relatively short life. The purpose of a CD-ROM with an encyclopedia was to justify the cost of a computer (p.289). Microsoft's Encarta broke into the 1993 consumer market with multimedia. This coincided with the falling price for CD-ROM drives (p.292), and lowered prices for CD-ROM software. By the early 1990s Quicken was the best selling consumer software product of all time. Its founder entered a crowded field with no track record, an untried product developed by a single programmer (p.295). It was designed to be easy to use, and continually improved.

Chapter 10 discusses the success of Silicon Valley, and the economic and physical environment that created its culture (p.303). Hardware companies tended towards success, software companies less so (p.304). The great number of computers in the US created a market for software companies. The prices for their mature products ruled out competitors. This pattern continued to the personal computer age. One effect of manpower training is to create off-shore body shops to benefit US multi-national corporations. Clustering firms in a small geographic area helps, as does Government subsidies (like the Internet). But misdirecting support can hurt rather than help (p.311). [I found Robert X. Cringely's book to be better.]

The emphasis of the book is on the economic history of the software industry. There is a very small number of references to the technological developments that enabled the progress that the author outlines. Among others, he leaves out the one significant event that explain, for example, the mushrooming of Microsoft. The author glosses over the importance of the scientific and technological aspects of software development and their influence on the history of software. Also, in the exposition of his methodology (he is a "certified" historian) he says that most of the software history writings are anecdotal -- but that his isn't. Then throughout the book we find lots of quotes of peoples opinions; of tables published in magazines whose origins are not questioned or cross-validated. This does not ring of rigorous historical research and the double and triple typical crosschecks. In a nutshell, if what one is looking for financial "facts" regarding the software industry history -- which company was the first, second, etc., money wise, during a given decade, then you will enjoy the book. For me, it wasn't what I expected.

From Airline Reservations to Sonic the Hedgehog may sound like a mystifying title, but this book provides a reasonable overview of the history of the software industry. At times, given the ups and downs in the industry, it can't avoid sounding like a catalog of defunct firms and obsolete software. However, this chronology is quite useful for anyone who wants to come up to speed very quickly and very generally on the main trends in the industry. Author Martin Campbell-Kelly covers some of the industry's seminal events and the main categories of software. Vexingly or refreshingly, he takes pains to say as little about Microsoft as possible, making it clear that others have written enough on that subject. So, with that absence duly noted, we recommend this book to those who want an inside history of the software industry, from massive mainframes to little blue cartoon porcupines.

I wasn't sure how I would like Martin Campbell-Kelly's new book since I have grown up with the software industry--I started as a programmer in 1973-and thought I had a pretty good understanding of it. But on the contrary I found his book fascinating. Over all I thought Campbell-Kelly weaved together many hard-to-find facts and statistics into a comprehensive, well-researched, and well-written story of how the software industry evolved

Basically, it reviews the development of the various software industry segments; what we now refer to as professional services companies, enterprise product software companies, personal computer software companies and game companies. For each, it describes the major events that created these industry segments-e.g. in 1970 IBM unbundled its software pricing from its hardware pricing ushering in age of product software companies; the major companies who dominated these segments; and the business models.

A few things that were of particular interest to me.

When I grew up in the software business, we used lines of code as a way to estimate the work it would take to create an application. I have overseen the development of systems with millions of lines of code. So I was surprised to learn how few lines of code were in DOS 2.0, just 20,000+, and even more surprised to learn that the venerable early business language Fortran, in its first incarnation, was only 18,000 lines of code.

And it helped put into context what I have lived through. For example, in 1973, straight out of C.C.N.Y., I first entered the job market as young Cobol programmer trainee at Royal-Insurance Company in New York. As a youngster I had no awareness that Thomas Watson Jr. bet his business on the IBM360, or that it was the reason for IBM's unprecedented industry dominance. I just knew that I worked on IBM 370-145, an updated, at that time, version of the IBM 360.

And does this sound familiar? "Never before has the stock market shown quite so much enthusiasm about an industry as it has lately about the computer industry. Recent prices of computer stocks represent some of the highest prices-earning ratios ever recorded. Even shares of giant I.B.M, which increased six fold between 1957 and late 1966, have doubled since. In July the market valued I.B.M., whose physical assets amount to less than $6 billion , at more than $40 billion-more than any other company in the word, actually as much as gross national product of Italy. And the market value of smaller and newer companies in the industry has gone up even more steeply than I.B.M.'s. In less than three years the price of University Computing Co. of Dallas rose form $1.50 a share(adjusted for splits) to $155. The stock market valued this newcomers, who sales last year were than $17 million, at more than $600 million."

This quote that Cambell-Kelly provides is about the timeframe 1964-1968, which was another time period of "irrational exuberance," like our recent dot.com debacle. Only I hope that our recovery will be sooner than that one. It evidently, took almost ten years for the market to again buy software company stock.

I do have one quibble, though. Campbelly-Kelly challenges the common perception that Microsoft dominates the technology industry. He quite correctly cites statistics which show that, even if Microsoft is a giant, IBM is a titan, towering over Microsoft; where Microsoft has at most owned 10% of the software industry, IBM owned fully 75% of the entire IT industry consisting of computer hardware, software and services. To put this into perspective, the total of computer hardware, software and services revenues world-wide today is thought to be something like $1 trillion. If IBM had maintained its earlier level of control it would now be worth $750 billion while Microsoft today has revenues of roughly $28 billion. His point is that while IBM was almost the entire industry by itself, Microsoft controls just a piece of the software industry.

While this is true, I don't believe the issue of Microsoft's dominance is about revenues. I believe it is about cultural impact--which Campbell-Kelly does allude to. Microsoft, almost single-handed, created a cultural revolution, making it fashionable to be a nerd, to dress casually in business, to work insane hours, to believe as a young twenty-something you can be an over-night millionaire. Whereas IBM made an impact on businesses--certainly being viewed for a long time as one of, if not, the best-run business on the planet--Bill Gates and company, captured the public's imagination and changed our society in a way IBM never did. And why not, when Bill Gates and Paul Allen are the first and third richest men in the world.

All in all though, this book was an enjoyable and insightful read.

From Airline Reservations to Sonic the Hedgehog: A History of the Software Industry - by the time of its publication and writing of this comment - is the only book that is solely on history of software industry. As an academician, Mr. Campbell-Kelly builds a historical view using carefully collected data. The scope of the book is mostly software industry in USA and this point is explicitly phrased at the beginning of the book, with plausible reasons. Among with absence of non-USA part of the history, history of military software is also missing (except SAGE). But this should also be an expected result, since it is very hard to collect data about military systems.

Because of the scope issues, the book may not satisfy all expectations (which is the reason for 4 stars), but from the opposite point of view, I believe that a better book can hardly be composed in the same conditions. As Mr. Campbell-Kelly pointed out, there are a lot of data about success stories or the firms that managed to live until today, but a lot more data have been lost to history. In this regard, "From Airline Reservations to Sonic the Hedgehog: A History of the Software Industry" is among the most valuable source about the history of software industry. Everybody, who works in software industry should read this book.

A last note: the book really makes to think about history. The interested reader may also read books about meaning of history, like E. Carr's "What is History".

There has been a need for an authoritative history of the software industry, and this book fills the gap. It is entertaining and gives an excellent account of the origins and development of the various sectors of the industry. Its focus is wholly on the United States, a decision for which the author gives full justification, but despite my interest in the UK software industry I still found the book fascinating and illuminating. It provides an important corrective to a number of myths, from the Software Crisis to the dominance of Microsoft, and offers plenty of evidence in support of its analysis. There are entertaining illustrations at the beginning of each chapter. As a university lecturer, I shall certainly be using this book extensively and referring my students to it.

This book is scholarly, but accessible to the ordinary public, albeit people who are keenly interested in tables of rates of growth. The author describes the sources of his research humorously and in a most interesting manner. I give him very high marks for that, for letting the general public appreciate how hard it is to do research of this kind, sifting through all the junk for the few gems. I thought the author's technical competence was excellent, too, and his insights into such subtleties as operating systems was great.

I have only two complaints. He too often says "The story of ... is well documented elsewhere." A brief recap would have helped for those who don't have time to read the other sources.

Second, there is exactly one picture in the book, and while it is great (a stack of 60,000 punch cards for the SAGE system), I wish there would have been a few more, especially of earlier systems.

But overall a fabulous book, very neatly wedged between business history and computer history!