Dean Hovey on Mouse Innovations

Source: Interview with Dean Hovey, 22 June 2000.

A-Ha! Moments

(Brainstorming session notes)

Hovey: [Looks in box of drawings and mouse prototypes, pulls out and unfolds papers] I'm looking at a couple large sheets that look like they were from a brainstorming session. It looks like we were trying to figure out how we were going to get around the fundamental problems we had seen in the limited production-run mice we saw from Xerox.

Problems with Xerox Mouse

One of the key elements of the Xerox mouse was a large, stainless-steel bearing, that they encoded off of, just like normal mice. But the bearing was also a support, so just like a ball bearing supports load, they used that ball to support the load of the case. So there were pressure points on the ball pushing down. What happened then was as a result of that extra friction, the ball would slip on the table surface, so they had to require a special pad. Well, one of the things Steve Jobs said in my conversations with him was, "I don't want to have to use a pad; I want to be able to do my mousing on my Levis."

See also Jim Sachs, Jim Yurchenco, and Jim Yurchenco and Rickson Sun on the Xerox mouse.

So we were struggling with this because the ball continued to slip, and when it slips, you would then not get encoding from the ball that would indicate its motion. So we were trying to brainstorm around that. One idea was, why don't we have a lead-filled rubber ball? Rubber would track better on a surface, and lead would give it weight. Well, that would work fine, but now it couldn't be load-bearing.

Eliminating the Bearing

I remember there were a couple "A-has!" when I was designing the mouse.

(Brainstorming session notes)

We were located in a building on 514 Bryant Street in Palo Alto, which was above a woman's apparel store called Kitty O'Hare, and Jim's Coffee Shop. Jim's is still there, but I think a Starbuck's is beside it. We had our first office there, and there were a number of other painters, writers, etc.; there was a linoleum floor, and the building was quite tipsy. I remember having balls of various size on my desk, and after you'd leave them they'd roll off the floor, following the slant of the table. And I said, "That's exactly what I want it to do: I want it to roll without slipping." And I remember grabbing the ball that was with me-- we had gone with steel balls with some sandblasting to get some surface friction [pulls a sandblasted, steel ball bearing out of box]-- and I remember saying, "You know, that's what I want it to do. How do I get it to continue to roll without slipping?" And I put my fingers around it like a little cage.

See also Sachs and Yurchenco and Sun on ball choice.

When you do that, you immediately realize-- if you think about your physics-- that you can touch it at the point of rotation, there's no lever-arm to introduce friction. So at some point on the ball, which is at that center, you can touch it and you won't add any friction. So we said, "If we took encoding off that point, and also guided at that point, we could actually minimally affect the friction on the ball." So that was the big "A-ha!" The ball was no longer being pushed on as a bearing support, it was actually free to roll, and we'd barely touch it to get the information about where it was moving. And we redesigned it as a result of thinking of that.

Pang: Bill Lapson tells the story of you demonstrating this at a meeting with you and several other people--

Hovey: I don't remember that, but it's possible.

Pang: So this was something that happened first in your office.

Hovey: Yes.

Pang: Is it possible that Lapson was there? I'm trying to line these stories up.

Hovey: Maybe. I remember where I was when I had that "A-ha." Immediately I got up and grabbed some people to tell them. It was either David Kelley, or Jim Sachs, or maybe Jim Yurchenco. It was just one of those things were you go, "That's it."

Cleaning the Mouse

(Mouse with ring)

The other thing that was causing significant problems on the Xerox mouse was that it would collect dirt, which would gum things up. And while this may just seem obvious, I said to myself, "Well, we're not going to change the world, there's going to be dust and grit, so let's just make the mouse easy to clean." So this little thing on the bottom [removes ring] makes it possible to take the ball out, blow on it [demonstrates], wipe it off, and put it back in, because we're not going to be able to get around the problem.

See also Sachs and Yurchenco and Sun on adding the cleaning ring.

Pang: How did you clean the PARC mouse?

Hovey: You literally had to take it apart with a screwdriver. And because it was a bearing surface, that just complicated things dramatically. As soon as you get grit in there, it gummed up even faster than this mouse [holds up Apple mouse]. So I remember those being the two significant "A-has."

The thing I think Hovey-Kelley did was take a neat idea and executed on it in a way that would inexpensive, reliable, manufacturable, and usable by people. That was the value-add. Xerox had obviously come up with the idea of the mouse, and had implemented various versions of it, but the execution around it was still flawed. It wasn't something that would scale such that it could change the world as it had the potential to.


Another one of fun parts about the project was that all the part drawings were anatomically named for mice. There was the ribcage, the outside cable was the tail, the cover was the hide, and a guy named Douglas Dayton did some wonderful sketches animating the mouse character. We definitely had fun on it.

I believe it was one of the things that was actually under budget and on time on the Mac project. We had working prototypes in the late 1980, early 1981 time frame, and Mac didn't actually make it out the door until 1984. I think the first mice actually came out on the Lisa, rather than the Macintosh.


Document created on 10 July 2000;