Yurchenco: Engelbart of course deserves credit for the concept of a hand-held pointing device. The credit our company gets is a bit misplaced: nobody should assume we invented the mouse. What we did was take a laboratory concept and turn it into a manufacturable product at a reasonable price. It's an important step: commercialization is really the "D" part of R&D, which Xerox Parc wasn't doing, and SRI wasn't doing at that time. But the intellectual idea of the mouse, certainly we do not deserve any credit for.
Pang: When you started your work, you had seen the Xerox mouse, correct? [Nods] Had you seen the Engelbart mouse, the great big one?
Yurchenco: The only mouse I had seen was the Xerox PARC mouse, which had commutators for doing the encoding, a small ball, and an elaborate gimbaling system.
Yurchenco: It was immediately obvious that this was not a viable product, and you couldn't build them in any quantity. Using optical electrical encoders is something that 99% of mice in the world still do today.
Sun: The forces [in the mouse] are very low that are available to move an encoder to pick up information, so anything with little wires rubbing against spinning disks was going to take more force out of the system than interrupting light beams would. So the early direction was to use light beams rather than rubbing insulators and contacts past a little contacting finger.
Yurchenco: The people at Hovey-Kelley who were involved, of course, were to begin with, myself and Sachs, and Dean Hovey, who was one of the founders of Hovey-Kelley and David's partner. Jim had responsibility for electrical engineering, and Dean and I had responsibility for product design and mechanical engineering.
The first things were did were to go out and see what ideas we could steal, which is how engineers work-- why reinvent the wheel?-- and one of the things we came up with was a trackball module, a very large trackball used in Atari game machines. We looked at it, and what they were doing was interrupting beams with slotted wheels-- a different orientation, and everything was at a much larger scale, and of course the ball was supported by the structure and was depending just on gravity, but it sure seemed like a real promising approach....
So very quickly we built some prototypes, scaling it down to a size to at least [gets clear prototype]-- I don't know what generation this was find out, can be made as something that can be hand-held
Pang: Now, this does not have a button--
Yurchenco: No, because the button was not a real issue: it was just a switch with a cosmetic cover, so there's no real technology involved there at all. The technology was here on the inside, and there were a couple things I think were key that we brought to the table.
Sun: The first was going in the direction of optical encoding. It was the answer then, and as you can see, it's still the way mice are made. Now, the optical encoding in those days was more complex, because you had to use off-the-shelf phototransistors and off-the-shelf LEDs, and everything now is custom to mice, and they're packaged in dual packages, the tolerances are better, and you get more dots per inch and fewer parts, but those are just incremental improvements.
The second things was allowing the ball to float. In all the mice before-- in the Xerox PARC mice, Engelbart's mouse didn't even house a ball, but two encoder disks-- you're actually forcing the ball on the table, and the Xerox PARC mouse had a gimbal on top of it that tried to force the ball down. We very quickly realized that you didn't need to do that, that you could get rid of all those parts that were fussy and tended to get dirty and so forth, and that gravity alone would suffice to allow you to do that.
The third innovation was coming up with an idler here [points to idler in ribcage], so the ball is actually suspended between two orthogonal shafts, and is pressed against those shafts by a little spring-loaded wheel of some kind. In this case it's a wire spring, and there are a lot of variations now, but every mouse you open now will have that third point of contact on the ball, and allow the ball to be forced against the shaft, so there's a little bit of friction between the ball surface and the shaft surface. So in terms of technology innovations, those were probably the key elements in developing this thing.
Pang: Now this [prototype] has two of these three, right?
Yurchenco: This doesn't have the idler; this was built before we realized the need for it. This was adjusted so you could try to push it [the ball] against these foam rubber rollers, but foam rubber rollers would never have the lifetime you need for a commercial product, and they take up the tolerance, so we knew we'd have to go with some kind of hard roller. After a while it became obvious that you'd need some kind of compliant member on this side [the side with the idler], since you couldn't have one over there [with the rollers]. And so that's how that evolved.
The other things we brought to it was the actual product design, which was miniaturizing and cost-reducing this thing, which first of all takes you into the world of injection molding, which is the lowest-cost way to make complex parts. Then it was a matter of basically squeezing the geometry down. That was a fun exercise, and that was primarily my responsibility, designing this [unrolls drawing]-- what came to be known as the ribcage. There were a whole bunch of little pieces I designed separately, in different sections, and then pieced together.
This was in the waning days before CAD took over the design world, when paper and pencil was still the operative mode. There were primitive CAD machines available at that time-- primitive by today's standards-- but they were relatively expensive and Hovey-Kelly design didn't have the wherwithal to buy them. So we were still operating in a paper and pencil world.
And so behind this drawing was a lot of chickenscratching on paper, and a lot of calculations in terms of spring forces and snap forces and so forth, just the-- I guess it's just the magic of visulaization in terms of how are things going to fit together in a 3D way.