March 9, 2003
The Ascent of the Software Civilization
By STEVE LOHR
THE NEW YORK TIMES
TECHNOLOGY euphoria was in the air and stock prices were stratospheric. The
shares of some software and computer services companies jumped a hundredfold
in a few years. "Never before has the stock market shown quite so much
enthusiasm about an industry," a writer for Fortune magazine observed.
The year was 1968.
History may provide no sure guide to the future, but it does offer context
and insight for the present. And, given the proper twist, history glimmers
with a certain knowing humor. In his incisive, panoramic book, "From Airline
Reservations to Sonic the Hedgehog: A History of the Software Industry" (MIT
Press, $29.95), Martin Campbell-Kelly delivers all three - context, insight,
even occasional humor. His treatment of the similarity in the computer mania
of the late 1960's and the Internet bubble of the late 1990's is one many
Mr. Campbell-Kelly's book is not for everyone. He is an instructor in
computer science at the University of Warwick in England and is a
professional historian. His is not a book of insider gossip or of recreated
scenes of clashing egos and executive tirades - the stuff of so many
business books. Instead, it is the product of his reading and distilling of
books, professional journals, magazine and newspaper articles and historical
archives over the last four decades.
The result is a sweeping survey of the software business since the early
1950's, its evolving structure, economics and marketing. The product and
company names are so numerous that, at times, they seem to race by in a blur
- Cincom, Cullinane, Panosophic, Cytation
nyt-com/html-companyprofile.asp&symb=CYTY>, Famicom, Xenix, Comshare
nyt-com/html-companyprofile.asp&symb=CSRE>, Syncsort, Super- Writer,
MacNeal-Schwendler and hundreds more. The book is a ready reference for any
misguided soul who wants to create a computer trivia game.
Yet, along the way, Mr. Campbell-Kelly pauses for longer passages on
subjects ranging from the role of the government's SAGE air
defense-and-surveillance system to the decision by I.B.M.
nyt-com/html-companyprofile.asp&symb=IBM> in 1969 to "unbundle" hardware and
software sales. The unbundling was "a turning point," the author says,
because it ensured a sizable software industry separate from hardware.
Mr. Campbell-Kelly gives the United States government credit for creating a
market for programmers in the 1950's and early 1960's with SAGE, a name
derived from "semiautomatic ground environment." About $150 million was
spent on its software, and thousands of programmers were trained to work on
the project - a veritable "university for programmers." Its alumni went off
to populate corporate data centers and to start companies elsewhere. SAGE
also tackled a host of technical problems in real-time transaction
processing that eased the way for I.B.M.'s development of the Sabre online
reservations system, called "the Kid's SAGE," for American Airlines.
Time and again, the book shows that what seems new in the software industry
in fact echoes the past. The engineers who used software formed user groups
starting in the 1950's with SHARE, whose members came from companies that
used I.B.M. 704 computers. The group's name, though capitalized, was not an
acronym. It was what the group did; it shared programs and expertise to help
drive down programming costs.
This was when software was treated as a marketing cost of hardware instead
of as a stand-alone product. But, as the author writes, "one of the abiding
legacies of the user groups was the perception of software as a free good."
In their attitude and behavior, the user groups were very similar to those
engaged in today's open-source software projects like GNU Linux, an
operating system that is distributed free and is steadily improved and
debugged by a network of programmers.
IN the late 1960's and early 70's, there was great excitement about - and
investment in - the notion of computing as a utility-like service that could
be delivered to offices and homes to solve all manner of problems. The idea
flopped because it proved to be too complex a challenge to write the
software needed to distribute computing as a service from a central
time-shared machine to many users. Besides, microcomputers, later called
personal computers, were about to arrive, putting affordable computing on
desktops thanks to the miracle of the microprocessor.
Today, the utility concept is making a comeback. This time, the industry is
betting that 30 years of advances in software, hardware and networking can
deliver utility computing. I.B.M. is promoting "on demand" computing, and
nyt-com/html-companyprofile.asp&symb=HPQ>, Sun Microsystems
nyt-com/html-companyprofile.asp&symb=SUNW> and others have similar
offerings. Companies like Salesforce.com are offering software as a service
in an emerging market for so-called application service providers, or
A.S.P.'s. "Because the concept was never fully tested in the 1970's," Mr.
Campbell-Kelly writes, "history has few lessons to offer the A.S.P.
industry." Well, the A.S.P. start-ups certainly hope so.
As software became a product in its own right, marketing became more
important. In the fledgling personal computer industry, some of the tactics
were particularly inventive. In 1980, George Tate, a former electronics
industry salesman, began marketing a database program he named dBase II.
There was no "I," but "II" suggested a second, improved product. This ploy
was not uncommon. But in 1983, Mr. Tate took the more imaginative leap,
renaming his company Ashton-Tate.
"There was no Ashton," the author writes, "but it was considered a
euphonious and high-sounding name."
Mr. Campbell-Kelly argues convincingly that Microsoft's
nyt-com/html-companyprofile.asp&symb=MSFT> stature in the software industry
as a whole tends to be overstated. After all, he notes, Microsoft dominates
only about 10 percent of the industry, while it has no such sway over large
portions of the software business like corporate software, software
contractors and consultants.
"If this book serves no other purpose," he writes, "I hope it will serve as
a corrective to the common misconception that Microsoft is the center of the
software universe around which all else revolves."
Mr. Campbell-Kelly is dismissive of the federal antitrust suit filed in
1998. "Exactly why the action was brought remains a mystery," he writes. He
is fastidious with footnotes, but there are none near that sentence. It's
his opinion, period.
He may not understand the Microsoft antitrust suit, but he lucidly analyzes
Microsoft's mastery of using software and business acumen to build a
lucrative technology platform. In Microsoft's case, of course, the
technology platform is the Windows operating system. The company encouraged
and helped software developers write programs that run on the Microsoft
platform, increasing the value of Windows and prompting even more
applications to be written on top of Windows. Economists call this
commercial snowball "network effects."
It is nothing new, really. I.B.M., as Mr. Campbell-Kelly explains, did the
same thing with its operating system for the I.B.M. 360 mainframe and
nyt-com/html-companyprofile.asp&symb=SNE>, in video game consoles, deftly
used software to build its technology platform and a thriving ecosystem of
game developers supporting its PlayStation platform. Once Sony mastered the
economics of software, Sega's
nyt-com/html-companyprofile.asp&symb=SEGNF> Sonic the Hedgehog never had a
Distributed through Cyber-Society-Live [CSL]: CSL is a moderated discussion
list made up of people who are interested in the interdisciplinary academic
study of Cyber Society in all its manifestations.To join the list please visit: