Pang: Between the time that Steve Jobs gives you the assignment to produce the mouse for under $10, and you deliver the model that makes it into the Lisa and is redesigned for the Macintosh, how much involvement did Apple people have in the design process?
Sachs: The Apple involvement was transitional. Hovey-Kelley was actually were responsible for getting the mouse into production in Apple's own factory. They took it over from there, there was minor involvement on Hovey-Kelley's part on a consulting basis after that. It had been documented well enough that Apple could take it over, with one exception.
One of the interesting production stories is about my color-blindness. I had built-- myself, by hand-- all of the mice until mass-production. There were a few dozen, and I actually soldered the wires. When we went into production and built the first hundred, none of them worked. They quickly found out that there were wires crossed, and it turns out that because I'm red-green colorblind, I'd misidentified the wires-- but I could consistently misidentify the wires and make them work! I documented it as such. When a non-color blind person tried to build the production mice, they didn't work. I never admitted why: we just said, "It's a documentation error," and I saved the document that showed that somehow the red and green wires ought to be reversed. So that was the transition.
The patent application didn't come until long enough after Hovey-Kelley's involvement that nobody at Hovey-Kelley was given credit for the design: the people at Apple who were processing the paperwork simply put the names of the manufacturing engineers involved. There was no continuity, which was startling, and a patent was actually issued without the names of the people who actually did the design. When we found this out, we immediately called Apple and said, "We think maybe you forgot the names of the inventors!" And so the patent was actually reissued with the inventors' names on it. The original patent was a combination patent that included both the mechanical mouse, and the notion of pull-down menus, which was Bill Atkinson's work. When they reissued the patent, they elected to split it up into separate pieces, and have the mechanical design of the mouse separate from the pull-down menus.
So an interesting bit of lore is that there are two patents on file, one without the inventors and with pull-down menus, the other with the inventors and without pull-down menus.
Pang: How difficult was moving the mouse into production?
Sachs: It actually was quite a non-event. Not for the design team to take complete credit, but the whole point was to have the mouse be easily produced. I've never heard any stories other than, how fast could they get the components, and how fast could they assemble them. It has been essentially a trouble-free operation. The design that we created, because it inherently and fundamentally eliminated design tolerance issues, and precision issues, and alignment issues-- there weren't even any adjustments in the product-- allowed Apple simply to build them and they worked. So it might have been a pretty big non-event.
Probably most of the changes had to do with subtle aesthetic things like the feel of the switch. We spent a lot of time figuring out how the switch should feel. Most people never realize that any switch they touch has had a designer agonize over things like, What does it feel like? What does it sound like? How far does it go down? Is it springy or mushy after it goes down? Does it make a sound when it clicks? and How much force is required to press it? Now, all of those things are controllable, and all of those things result in a very different experience, whether it's a key on a keyboard, or the clicker on a ballpoint pen that has a long throw and a "ker-chunk" sound, or the mouse button on a mouse, or a membrane keyboard on a microwave oven. All of them have different characteristics.
And again, we were designing this in advance of knowing really what people were going to do with a mouse. But we felt it was important that a user be able to rest their finger on the button and not activate it, because you didn't want to have to hold your finger up in the air all day. So the original mouse, and I think mice today, had an activation force that won't trigger if you just rest your finger there. But we also didn't want you to have to press too far, and spend a lot of energy, and tensing a lot of muscles in your hand when you use the mouse. We felt that a good tactile feedback was necessary, so you could feel it click and not only rely on what you saw on the screen. It also had to have enough audible feedback so you could hear it click, but we didn't want it to be annoyingly loud.
So we balanced all those, plus wanting the reliability to be very high, meant using a relatively expensive microswitch in the mouse. With a microswitch, the manufacturer of the switch can tightly control the tolerance, instead of the manufacturer of the mouse having to have a high-tolerance switch design. The one area of very high tolerance and precision in the mouse was in the microswitch, and microswitch manufacturers are set up to do that in mass-production, and to screen them. We chose a switch that we could simply insert into the mouse assembly, and reliably would get this feel that was subjective and appealed to users. And I think that changed over time: I think it was a little stiff at first.
But the issues of mass-production were essentially using a lot of product in the hands of a lot of people to get feedback. Whether it was noise of the mouse, which resulted in a rubber ball; the weight of the steel ball, which resulted in a solid rubber ball, not a rubber-coated steel ball; the activation of the switch; and the shape of the mouse-- what would make your hand cramp and what wouldn't make your hand cramp-- those things then got cycled into mass-production. They were all compatible with one another, except for the connector.
Although the mouse cord was always an issue. The mouse cord on the Xerox laboratory instrument was very, very flexible; you almost couldn't sense it was there. But as a result, it was prone to breakage. It also failed what we called the "yo-yo" test, which was, if the mouse was dropped off a table and the cord saves it, did that stretch the cord or break it? But because the cord is coming out the of the mouse and somewhat constraining it, it needed to be very flexible, and not feel like you had a garden hose attached to the mouse.
That required technology that cable manufacturers weren't used to having to care about. They generally didn't have to worry about something being flexible and strong. If it really needed to be flexible, because it was going to be inside a moving component in some mechanism, the mechanism would never yank on the cord; or, it would be very sturdy because you would yank on it, but you didn't need it to be flexible. The intersection of requiring it to be a cord you could yank on and be flexible was something new. It took some time, and I think over time cable manufacturers were able to customize them more, and mouse manufacturers learned what was strong enough and what was flexible enough. Now there are wireless mice, which eliminate the cord completely-- at a price. Back then, it would have been unconscionably expensive to have a cordless mouse, but now it's not so. So [in news announcer voice] one of the great advancements in mice technology is eliminating the cord. [Pang laughs]