The man who built a better mouse trap

COMMENTARY--Doug Engelbart doesn't come across as a "mad scientist" type. The 78-year-old engineer seems more like the type of person who votes in every election and would walk three blocks rather than park close and partially obstruct someone's driveway. But for more than two decades, the man who invented the computer mouse was often confined to the scientific fringe.

In 1951, Engelbart first came up with the idea of using a manual device to manipulate data on a computer. During World War II, he learned that the scale and other aspects of radar displays could be controlled by hand, and he figured the same techniques could be applied to computer displays. He suggested pursuing research on the topic while working at Ames Aeronautical Laboratory (the NASA Ames Research Center) and later, as a professor at the University of California at Berkeley.

"Everyone--for at least 10 years--thought I was totally crazy," Engelbart said in an interview last week, during a press event held by Logitech to mark the Swiss company's shipment of its 500 millionth mouse.

No one remembers who suggested the name, but we all started calling it the 'mouse. --Doug Engelbart

Subsequently, Engelbart took a position at SRI International (then known as the Stanford Research Institute). "I was told not to talk about it," he said in the interview. Instead, he worked on a computer made from magnetic cores, a project that has vanished in the mists of time.

Finally, his research group at SRI obtained a grant from NASA in 1963 to study interactive devices for displays. Of the eight or nine different devices the group studied--pointers, joysticks, trackballs--a brown, wooden box with two rolling wheels and a red push button on top achieved the best results.

"No one remembers who suggested the name, but we all started calling it the 'mouse,'" Engelbart recalled.

Still, NASA never really ran with the idea, even after a successful public debut of the mouse in December 1968 at the Joint Computer Conference.

Bill English, who built the first mouse based on Engelbart's designs, moved from SRI to Xerox's Palo Alto Research Center (PARC) and began to work on input devices with Stuart Card, an expert on human-machine interaction. At Xerox, the mouse remained a lab-only phenomenon until Apple Computer copied the idea in the early 1980s--three decades after it was first mooted.

Interestingly enough, the fact that mice are associated with simplicity is a disappointment to Engelbart.

Most people look upon their computer mouse as something that fell out of the sky. It's always been there and has pretty much always functioned the same way. The devices, of course, are ubiquitous. According to Logitech, 500 million mice laid end to end, without cords, would circle the Earth 1.6 times.

Yet most experts failed to realize the potential of the idea. Even after the mouse was proven to work, many companies resisted it.

"IBM was dead set against it for a long time," Card said.

Even after IBM decided to pick up the mouse, Logitech didn't get the contract. "We didn't have a document shredder," which IBM required, Logitech's Chairman Daniel Borel recalled.

To a large degree, the mouse technology has survived because it harmonizes almost perfectly with human hand-eye coordination, Card said. If you drew two targets on a screen and bounced between them with different devices, you would perform the task faster and with more accuracy with a mouse than with a track-pointer, a joystick or a touchpad--even if the mechanisms all ran at the same speed.

The only device that might work better is a pen, Card said. Mice, however, stand out better than pens in peripheral vision. The extra time it takes to find and pick up a pen relegates it to second place.

Invite Michael Kanellos into your in-box Senior department editor Michael Kanellos scrutinizes the hardware industry in a weekly column that ranges from chips to servers and other critical business systems. Enterprise Hardware every Wednesday.

An explanation for the phenomenon can be had through Fitts' Law, which states that the time to home in on a target is a function of the distance to, and size of, the target. Personally, I think the answer is more Neolithic: We Homo sapiens have been throwing rocks a lot longer than we've been playing computer games, so the sweeping motion is natural.

Because of this biological magic, the mouse still functions pretty much the way it always has--unlike microprocessors or graphics chips. Two of the biggest changes have been that the cord was moved from the back of the mouse to the front, and the main control button has shifted from the right side to the left.

Interestingly enough, the fact that mice are associated with simplicity is a disappointment to Engelbart. He believes that better, more-elaborate input-output devices could increase collaboration and cooperation. Over the years, he has developed control devices for both hands--a mouse for one hand and a five-key keyboard for the other--a setup that can increase the speed of performing certain tasks, experimental results suggest.

The primary knock against these devices is that they require training. But how many of us stopped at the tricycle as kids? As Engelbart noted, we all managed to move up to the more-complex two-wheeler.

biography
Department Editor Michael Kanellos will chronicle the fortunes of promising start-ups in this regular column. If you've heard of a young company with a hot technology, he wants to know about it. Kanellos also runs the Enterprise Computing and Personal Technology sections at CNET News.com.