ENIAC: The Triumphs and Tragedies of the World's First Computer.

ENIAC: The Triumphs and Tragedies of the World's First Computer by Scott McCartney Walker & Company, $23

JOHN MAUCHLY WAS OBSESSED with weather prediction, not with computing. But his meteorological obsession led Mauchly, a distracted visionary, to conceive of the first general-purpose computer--the machine with a mouthful of a name, Electronic Numerical Integrator and Computer, but a snappy evocative acronym, Eniac.

His partner, the engineering ramrod for the Eniac project, John Presper Eckert, once explained that Mauchly's "goal was to forecast the weather, and it was only incidental that he found out there was just no computing machine that existed that would handle all this voluminous stuff."

This is but one of the intriguing details in Scott McCartney's book that constantly remind the reader how little has changed, in some ways, since the start of the computer age. The Internet, like the Eniac, was the result of government sponsorship and serendipitous invention. At every turn, we read stories in newspapers, magazines and books that assume we live in a unique period of technological innovation and round-the-clock work regimens. The Eniac team in the 1940s may have dressed differently, but their seven-days-a-week ways would have been familiar to the tattooed employees of any Internet start-up in Silicon Valley.

Like so many modern computer development efforts, the Eniac, designed for calculating ballistics trajectories in World War II, was late. It was ready to work in the fall of 1945, just as the war had ended. And weather prediction--Mauchly's motivation at the outset--remains one of the most ambitious chores in computing, requiring the most powerful super computers.

McCartney, a reporter for The Wall Street Journal, has written a little gem of a book, a resonant work of technological history strengthened by ample helpings of fresh research. And it is a refreshing change at a time when so many books on technology are hasty efforts to chronicle the Internet gold rush and cash in on it.

The Eniac story has been told many times before. But McCartney deftly places the Eniac project into historical perspective, sketches out the principal characters, and makes the strongest argument yet that Eckert and Mauchly did indeed invent the world's first real computer. That is a conclusion that remains a subject of debate, but McCartney's book is an explanation of why he comes down firmly on the side of the Eniac and the two men behind it.

McCartney traces the 300-year genealogy of the computer, dating back to an eight-digit calculator fashioned by a 19-year-old named Blaise Pascal in 1642, the year Galileo died. His Pascaline, McCartney writes, was before its time and proved "two axioms of computing history." The first, the author states, is that "the young will lead," adding that Pascal was "perhaps the first teenage whiz kid in computer history." The second law of computing, McCartney asserts, is that "new technologies will catch on slowly unless there is a clear and urgent use for them."

The author's brief summation of the march of computing history includes Charles Babbage, the 19th-century astronomer and economist who sought to automate mathematics, just as the machines of the Industrial Age were converting factories to mass production; Herman Hollerith, an engineer at the Massachusetts Institute of Technology, who built a punch-card tabulator for the national census in 1890; and Vannevar Bush, also of M.I.T., who developed a machine for calculating differential equations in 1930.

Eckert and Mauchly are cast in Eniac as a complementary duo of contrasting personalities and strengths, a pattern of binary human teamwork that later proved so effective in the personal computer industry (e.g., Steve Jobs and Steve Wozniak at Apple Computer and Bill Gates and Paul Allen at Microsoft). Mauchly contributed the key concept of building a general-purpose computer that could be programmed for many tasks. And by McCartney's account, Mauchly was "a calming influence" and "a lovable intellectual who always kept his eye on the big picture."

Eckert, in contrast, was the cando team leader with a fiery temper who, as one Eniac engineer observed, was "always pressing everything to the limit." Eckert possessed a laser-like attention to detail, which was necessary if the complex machine was to work reliably. He was very concerned, McCartney recounts in one example, that mice would gnaw on the wiring, causing malfunctions. So Eckert got some mice and starved them for a few days. Then, he put various kinds of wire in their cages. "The least appetizing brand was used in Eniac," McCartney notes dryly.

John von Neumann was a latecomer to the project, but he has often been given credit as the "father of the computer" because of a paper he wrote in June 1945. Communicating ideas can be as important as developing them, and von Neumann was a brilliant communicator. His 1945 paper, McCartney writes, "poetically equated the structure of the computer with the structure of the human brain."

The most original part of McCartney's book is his revisiting of a 1970s court case and the enduring and controversial assertion that an Iowa engineer, John V. Atanasoff, had the rightful claim to be called the inventor of the computer. The court case pitted Honeywell against Sperry Rand, which had purchased Eckert and Mauchly's company in 1950, but a key element in the case was the validity of the Eniac patent. Mauchly came off poorly as a witness, evasive and misleading about a visit he made to Atanasoff in Iowa in 1941. His witness-stand claim that Atanasoff's work was "irrelevant" to his own was undercut by laudatory letters he had written to Atanasoff 30 years earlier--a use of documentary evidence that was intriguingly reminiscent of the Justice Department's use of Microsoft's email in its antitrust trial.

Still, McCartney is unconvinced. He first points out the basic differences in the two designs--the Eniac was general-purpose and programmable, while the Atanasoff machine was single-purpose and not programmable. And McCartney then uses recently uncovered records to contradict Atanasoff's assertion that he only came forward in the '70s to make his claim to invention because, after his early work, he had dropped out of computing and was unaware of how much the Eniac mimicked his design.

In the '40s Atanasoff was named to lead a computer-building program at the Naval Ordinance Laboratory and given a $300,000 budget. But his effort went nowhere and was quickly abandoned. And McCartney says that records at the Library of Congress and the Smithsonian Institution show that Atanasoff had "intimate access to Eniac, and he even hired yon Neumann as a consultant to help with his computer-building program for the Navy."

On the evidence in his book, McCartney makes a convincing case for his conclusion: "Many played a role in the development of the computer, but Eckert and Mauchly were the ones who ultimately put it all together."

Steve Lohr is a technology reporter for The New York Times.