Advent of Computing - Episode 18 - Evolution of the Mouse

Episode Date: December 2, 2019

The computer mouse is a ubiquitous device, it's also one of the least changed devices we use with a computer. The mice we use today have only seen small incremental improvements since the first mouse was developed. So how did such a long lasting design take shape, and how did it travel the decades up to now? Like the show? Then why not head over and support me on Patreon. Perks include early access to future episodes, and stickers: https://www.patreon.com/adventofcomputing Important dates in this episode: 1961: First Mouse Developed at Engelbart's ARC Lab 1972: Xerox Develops Rollerball Mouse for Alto 1979: Apple LISA Mouse Designed

Let's face it, we all use computers. If not on a daily basis, then pretty close to that. But how much thought do you give to how exactly you use a computer? If this was an earlier time, then you may have used switches, wires, and punch cards to load in data. But today we're a lot luckier. You're more likely to be using a keyboard and a mouse for some type of input. The keyboard here is undoubtedly the older of the two devices. Its design would grow and morph from older typewriters. In fact, some of the earliest keyboards were just modified electric

typewriters. The mouse, on the other hand, is a much more recent development. Unlike the keyboard, there was never a pre-computer analog to the now ubiquitous mouse, and even then it took a while for there to really be a need for the mouse once computers developed. Despite being a late comer, the overall design of the mouse has remained surprisingly consistent. You have a movable puck that fits in the hand with two, maybe three buttons along the top edge. Computers themselves have totally transformed since the creation of the mouse, transitioning from mainframes to mini computers and eventually into the personal computers that we're familiar with today. But in that time, the mouse has seen surprisingly little change. That being said though, there have been small updates to what's under the hood.

So let's take a look at the most ubiquitous and enduring piece of computer hardware. Just how old is its design? And who exactly should take the credit for such a long legacy? Welcome back to Advent of Computing. I'm your host, Sean Haas, and this is episode 18, Evolution of the Mouse. Today, we're going to be diving headfirst into the long and sometimes controversial history of the computer mouse. And yes, you did hear right. Even something as mundane and everyday as the computer mouse has a complicated and not entirely agreed upon history. As near as we can tell, the first mouse appeared sometime around 1964.

At least, that would be the date for the first pointing device that's close to a modern mouse. However, there are earlier computer pointers. One example that predates the mouse by a good deal of years is a device called a light gun. And despite not shooting anything, it's still shaped and functions pretty similarly to a gun. In modern day, you're much more likely to see these in arcades, but in the 1950s they were more commonly used with military radar systems. One example was SAGE, a computerized US defense system built in 1958. At a SAGE terminal, light guns were used to point and click on targets displayed on large

radar screens. Even earlier than that was the trackball. It may come as a surprise, I know it did for me, but the trackball actually predates the mouse by nearly two decades. The first prototype trackball was built in 1946 by Ralph Benjamin. It was intended to be used for the British Royal Navy's Comprehensive Display System, a radar system similar to America's later SAGE project. Benjamin's device, called the Rollerball, definitely has a strong claim for the first mouse.

But there are some complicating factors here. The Rollerball only ever existed as a prototype. It wasn't mass-produ produced, and it wasn't even used with the final display system. And due to the secrecy around the development of the comprehensive display system, the design for the rollerball was kept classified, only being made public some 20 years later. And in a lot of cases, I think that would make for the exact kind of story I like to cover. It's an unused military prototype that was a state secret for decades.

That's a winning combo to be sure. But when you consider the larger context of the history of the mouse, we run into some issues. As we'll see in this episode, there's a direct lineage from early mice to the modern mouse. The fact that the rollerball was a state secret during the development of the earliest mice means that it's totally cut off from that lineage. I think it's better to think of the rollerball as more of a predecessor to the computer mouse than a direct ancestor.

So with that aside, let's start looking at some of the first computer mice, how the device developed into what we use today, and the strange and very tight chain of custody that the invention traveled down over the years. The seed of the idea that we create the modern computer mouse as we know it dates all the way back to Augmenting Human Intellect. That's a paper published by Doug Engelbart in 1962. Now, Augmenting Human Intellect is a definitive and seminal work in the field of computing. The core of the paper outlines a major problem in the sciences.

To quote from its introduction, Man's population and gross product are increasing at a considerable rate, but the complexity of his problems grow still faster, and the urgency with which solutions must be found becomes steadily greater in response to the increased rate of activity and the increasingly global nature of that activity. End quote. There is just too much information for any one person or even team of people to sort through. And unless something changed, then that inability to process mass data could stymie progress. Luckily, a solution was already on the horizon. The way Doug saw it, computers represented a huge untapped potential that could easily fix the information overload issue.

But that potential could only be realized if computers could be used more easily and by more people. To simplify things considerably, Augmenting Human Intellect is all about how to make a better computer user interface. In order to continue his research and work towards that goal, Engelbart founded the Augmentation Research Center, much more easily just called ARC, at Stanford. Over the course of the 1960s, he and a group of other computer scientists would design a totally new style of computer interface, the Graphical interface, or GUI. First demonstrated in 1968, their new interface originated a lot of what we see in modern graphical interfaces. It had Windows, hypertext, video conferencing, collaborative editing, the list goes on. The entire system was controlled by a combination of a traditional text input and an on-screen pointer,

which Doug's team called a bug. Text was sent into the computer via, of course, a keyboard. But the pointer was a totally different story. You can't really control a cursor with a keyboard alone. You need a new kind of device. really control a cursor with a keyboard alone. You need a new kind of device. The idea of some kind of pointer had been part of Engelbart's plans from the very beginning. It's a pretty natural way to interact with a computer if you think about it.

It's a lot easier to point at something that you want to use instead of describing where it is. There were already systems that used a pointing device of some kind or another, but only in a very limited capacity, as a kind of special purpose input. In Engelbart's new interface, a pointer would be used in conjunction with a keyboard. That may seem like a small change, but it has a lot of implications for the system's design. Like I said, there were some options available. The light guns from Sage are a good example. They were meant to only be used to tag items on a radar screen, an occasional task. Operators weren't expected to hold the light gun at the ready pointed at the screen for hours on end. Simply pick it up when they see a suspicious blip on the radar, mark that blip, and then put the light gun down.

Overall, a user wouldn't really use a light gun for that much of the time. In Engelbart's conception, he'd be using a pointing device much more often, and there wasn't yet a well-established device that could do that. So with that, the search was on. well-established device that could do that. So with that, the search was on. Possibly as early as 1962, Doug and his team at ARK started to vet their options. And when looking at new contenders, they had four main criteria in mind. The first was access time. How long it takes to go from actively typing to grabbing the pointing device. The second was motion time. How long do you have to be moving the device to get to where you need to be on screen? The third is just how easy is it

to learn the device? If you have a great pointing device, that's great, but if no one can learn to use it, then that's garbage. The fourth was error rate, or just how often does a user miss the target on screen. Those were all measurable criteria, and over the course of their research, ARC devised an entire test program based around these metrics. A test subject would have to go from typing to clicking on either a word or a letter on the screen and then back to typing, while the computer timed their movement and measured their accuracy. The whole thing has a fantastic scientific approach. ARC actually published multiple papers just on the data findings from these pointing device experiments.

It was also decided to run these tests on two separate groups, one of experienced users and one of inexperienced computer users, in order to just get a better spread of test subjects. ARC would end up testing a handful of different options, some more recognizable than others. One was a light pen, essentially the same device used at Sage but shaped like a pen instead of a ray gun. They also tested using a joystick, something more at home in an arcade than a research lab, really.

Another device tested was a thing called a Graphicon. And I don't know how well I can describe it over audio, but here goes. It basically consists of a stylus that's mounted on a long metal arm. It basically consists of a stylus that's mounted on a long metal arm. The arms attach to a base by this compound bracket so it's free to rotate, but also move in and out, inwards and outwards from the base. Still with me? Anyway, those three devices were all pre-built. Like pins, joysticks, and the bizarre Graficon were already in production and being

used for special purpose short time frame applications. ARK also designed a few of its own pointing devices to try out. And my favorite device, and one that actually turns out to be fast to use and relatively easy to learn, is what Engelbart called the knee control. relatively easy to learn is what Engelbart called the knee control. This is another one that you kind of have to see and even then I still don't fully get how you'd use it. Basically, it's a lever arm that's mounted under the desk. It rests atop your knee and can be moved up, down, and side to side. So to point to the screen, you have to rock and pivot your leg.

It does have the advantage that you can keep both hands on the keyboard while moving the pointer, but I don't know, personally I'm glad it didn't win out, it seems a little bit too out there for my tastes. The other device that Ark purpose built for these tests was the main attraction, the computer mouse. Engelbart had the idea for a mouse as early as 1961. He envisioned a device that would fit in the hand and be moved across the desk to position an on-screen pointer. The trick to measuring movement was to use two perpendicular wheels mounted below the mouse. One was for the x direction and another

for the y. As you move the mouse across the desk, each wheel measures how much the pointer needs to move in each direction. Plop a button on the top and you have yourself a mouse. But Doug wouldn't be the one to build the mouse. That task fell to Bill English, ARC's chief engineer. And if Engelbart was on the theory and research side, then English was the hardware guy. He was the first employee hired for ARC, and would end up being responsible for building a lot of the hardware that the lab used. And of course, that includes the mouse. Sometime in 1963, English would put together the first prototype, and like I mentioned earlier, it was very close to modern mice, but with a few key differences. The most visible was the casing.

It was built using wood. It was a prototype after all, and wood was easy to pick up and deal with. There was initially only one button, but pretty quickly that would be stepped up to three separate switches. But the largest difference was the operating principle. It used two perpendicular metal wheels, each turned variable resistors called potentiometers. As the mouse wheels moved, the resistance would change. That resistance value was then converted from an analog voltage to a digital number using a circuit called an analog-to-digital converter.

And finally, after all that complication, it was fed into the computer. It's pretty convoluted, but it is effective. When this new mouse was put head-to to head with other pointing devices, something really interesting happened. ARC would publish their findings in a 1965 paper. And they're somewhat contradictory at first look. For experienced users, the mouse was a clear winner. It was by far faster and more accurate than all other options.

But it turned out that for new users, the mouse was actually third place on almost all metrics measured. For inexperienced users, the light pin was the fastest and most accurate option, followed by the strange knee control. If the lab's conclusions had been based purely on math, then the mouse may have died off then and there. But the statistics couldn't take into account the human factor. To quote from the ending of they liked the various devices, reveals that the like pin, while operating in a natural way, does tend to be fatiguing, and that the mouse, though it requires some practice, seems to be a satisfying device to use. End quote. When interviewing users

in the test group, it became apparent that the like pin wasn't a good fit. The ergonomics were just all wrong. Using it for more than short bursts was tiring since you had to hold it perpendicular to the screen. The like pin felt natural to use, but it wasn't usable for long periods of time. The knee control had its own issues. It turned out that mechanically it just wasn't very reliable. And beyond all of that, people seemed to just plain like using the mouse. Over the next few years, the mouse would become the pointing device of choice at ARK, but it wouldn't stay in that one lab forever. Eventually, the mouse would escape into the wider world.

It was only a matter of time before an idea that good got away. It got its chance in 1971 when a group of ARC employees moved from their Stanford lab to work at Xerox Palo Alto Research Center, usually just called Xerox PARC. That team of defectors included, among others, Bill English himself. The loss of so many core researchers set ARK back considerably, but at the same time, it gave Xerox a miraculous injection of institutional know-how. At the time, Xerox was more than just the printer-copier company that we think of today. A large part of their operation was centered around computing in general. The company even had its own research lab in PARC, and with help from

English and his ARC team, Xerox would quickly start to develop its own graphical system. Polos, or the PARC Online Office System, would be an early and short-lived attempt at this. It was essentially a recreation of Engelbart's graphical system, complete with a mouse, made by some of the same people who worked on that original system. But eventually, Polos was scrapped. The team inside Xerox started to realize that while Ark's work was groundbreaking, they needed to find a way to move forward and into the future. And the future for Xerox was a machine called Alto.

If we think of Engelbart's earlier systems as an experiment in how to make better use of computers, then we can think of Alto as an experiment in how to make computer use better. The computer was built from the ground up to be a user-friendly graphical system, so much so that a lot of users just described it as a graphical display and a mouse. Art's work brought us a lot of the ideas that would go into future GUI systems, but the Alto would bring us the look and feel of modern computers. Alto's interface was packed full of things like icons, drop down menus, even had windows

with frames and a little bar at the top, and buttons. And all of this was controlled with a keyboard and a very modern looking mouse. The Alto project started in 1972, with early models using a mouse that's nearly identical to the ones made by English and Engelbart at ARC. It's no surprise since they did have English from that lab. The first Alto mice used two perpendicular wheels inside a metal and plastic housing with three buttons. But it became apparent that this older design

would need to have an update to go with the new shiny machine. And this is where sources start to conflict. Sometime near 1973, a new mouse that used a ball instead of the two wheels would be developed at Xerox. And I've seen this invention attributed to Bill English directly in some sources. In others, it's a team effort between English and Ronald E. Ryder, another engineer that was working at Xerox. This of course gets more complicated because Ryder is the name on the patent for the ball mouse, but the two worked together while at Xerox. So it may be a case that Ryder was the one who submitted the final patent, but English

was involved in the development. There's also a third option, that another engineer named Jack Hawley actually created the mouse for Xerox while working outside the company. A couple years later, Hawley would open his own company selling mice. So, the true creator of the mouse may not be clear cut, but however it was designed, in 1973, a new mouse would evolve. Like I alluded to earlier, the mouse developed at ARC was good, but it could be made a lot better. For one, the wheelbase design restricted motion. But more importantly, it was an analog device. And there's a lot of implications

that are caused by having an analog device hooked into a digital one. But the big issue is that analog circuits tend to be more complicated. They also, on the mean, require more power to work. So the new mouse would have to do away with the perpendicular wheels while also using purely digital components. The solution that came out of Xerox is essentially a totally new device housed in the same shell. Gone were the two separate wheels, instead they were replaced with a single large metal ball bearing. That ball sat against two perpendicular rollers, one to measure the X and the Y. When the mouse was moved across the desk, the ball rotated each roller proportional to their respective axes displacement. Basically, you can think of this as a mechanical way to break down the

movement of the mouse into a single x and y value. The ball turned out to be a lot less restrictive than the older dual wheel mechanism. You can slide it anywhere on the desk and it's not going to stop your motion at all. But that was only one of the requirements. There's still the matter of a way to somehow digitally detect how the mouse moves. That ends up being accomplished in a pretty ingenious way. Both of the rollers inside the mouse were connected to their own small metal drum. These drums were striped in conductive and non-conductive material, alternating. And then a metal brush would rest on the top of each drum.

If the brush was resting on a part of the drum that was conductive, it would complete a circuit, so it would register an on state. And if it rested on a non-conductive part, then the drum would just read an off. So as the drum spun, the mouse would register a series of on-off signals, or if you like, a bunch of ones and zeros, a digital data stream. From there, some fancy digitized circuits can make short work of interpreting those signals and figuring out how and which direction the mouse is moving. It's a pretty smart way to entirely remove analog circuits from the equation. The new mouse was dealing with digital data as soon as the ball rolled. And what I think is really cool about it is it's a physical device that's producing a purely digital

signal. That's just something that I think is neat. Now this mouse, and especially the new digital ball mouse, was what made the Xerox Alto possible. And while the Alto is a milestone for computing, it was actually a pretty big commercial failure. From its creation in 1972 all the way to the end of its lifespan in 1979, only 2,000 machines were ever produced, and of those, only a handful would ever make it outside of Xerox. The Alto just never got a foothold in the larger computer community. Between the researchers at ARK, Xerox, and a few outsiders that have seen their systems, it's possible that by 1979, only a few thousand people on the planet had ever

seen a computer mouse face to face. And once again, the mouse could have been trapped then and there. But as fate would have it, another player would pick up the mouse, and it would be one of those few thousand people that would get to see a demo. The next leg of the mouse's journey takes place in 1979 at another Silicon Valley giant. Apple had been working on a new product, tentatively called the Lisa. The Apple II computer had been their best-selling product line for years, and the Lisa was originally slated to be an update to that machine. But that would all change when in 1979, Steve Jobs took a trip to the park, and when he

was there, he came face to face with a Xerox Alto computer. The Xerox team had been giving demos to outside companies in the computer industry for a couple years, so the idea of the mouse had been starting to spread around. But by far, out of everyone who saw the demos, Jobs would take what he saw the furthest. As soon as Jobs returned to Apple, he decided to change the entire Lisa project. Seeing a mouse face to face had turned him into somewhat of a devotee. Instead of an update to the Apple II, it would be an entirely new computer modeled after the Xerox Alto.

And Jobs decided that this new computer, well, it would have to have everything. It would have windows, it would have menus, it would have buttons, and even a desktop. And all of it would be driven by a mouse. You gotta remember here that the Alto was a total commercial flop. Most people outside Xerox had never seen or even heard of the technology at play here. Steve knew, but none of his employees had the slightest idea what he was talking about. So over the course of 1979, he worked out a deal with Xerox. Jobs would trade pre-public offering Apple stock for two more demos of the Alto. And this time, he sent a team

of Apple engineers to learn everything that they could during these demos. From there, Apple set about cloning the Alto, with some alterations. The biggest hurdle was that the Xerox machine just wasn't very sellable. It was big, expensive, and not all that easy for an everyday consumer to use. So Apple's challenge was to take the core of the Alto and shape it into something that could win in the computer market. And they had some experience doing this from their Apple II computers. But at the core of this new project was something new, something that made up the core of the earlier Alto.

And that was a mouse. Jobs was exacting about the entire project, and especially the mouse. He knew that it would be the primary way that a user would interact with the computer, so it had to be perfect inside and out. The actual implementation of the mouse would come down to an outside contractor, Hovey Kelly Design. This was a startup that had worked with Apple in the past, mainly doing the design of cases and keyboards. mainly doing the design of cases and keyboards. In early 1980, Jobs brought the proposal to Dean Hovey,

one of the founders of the design firm. To quote from Hovey's recollection, the conversation went something like this. Jobs, What you guys need to do, what we need to do together, is build a mouse. Hovey, A what? And I think that this is something important mouse. Hovi. A what? And I think that this is something important to keep in mind.

At this point in 1980, maybe a few thousand people knew what a mouse was. Apple was in a really powerful position if they could bring a workable system with a mouse to market. And they had to get it right on the first shot. system with a mouse to market, and they had to get it right on the first shot. As Jobs saw it, the new mouse had to cost under $35 to manufacture. That was down from the $400 price tag of an Alto mouse. It also had to work reliably for at least two years of constant use. And it should work on any surface. According to Jobs, that included rolling on his own jeans. While the Lisa team worked away at how to make a graphical computer, Dean Hovey and his team would work away at how to make a mouse. And over 1980, a prototype started to take shape.

The basics of the Alto Mouse were kept intact. It would be shaped roughly like a puck, and it would roll on a large ball bearing. But a lot of the specifics, especially internally, had to be changed. The entire housing and all internal components had to be injection molded, since that would save a lot of cost. After some back and forth with Apple, it was also decided that the mouse would only have one button to simplify things for the end user. They found in tests that users got confused if there was more than one button. The new mouse was also designed with a removable hatch that held the ball in place.

That made it so you could disassemble some of it and clean it easily if any gunk got up around the ball. But the largest change was the actual mechanism inside. The encoder drums used in the Alto Mouse were more expensive to produce, and since they had brushes in contact with them, they'd wear down over time. Jim Sachs, one of Hovi's employees, redesigned the whole mechanism in a very clever way. Saks' new device used an optical encoding disk. This replaced the striped drums and brushes by using a plastic disk with a pattern of slots cut on the outer edge. The disk would spin between a set of infrared emitters and receivers, alternating between blocking and letting light

shine through. It did the same as the earlier drum encoders, but without anything touching the discs. The new optical pickups were way cheaper to produce, and they would last longer than the Xerox method, mainly because they couldn't wear down, there was nothing touching the encoding disc. The design process would take a lot of back and forth with jobs, but was nothing touching the encoding disk. The design process would take a lot of back and forth with Jobs, but by 1980, the Apple Lisa had a new mouse. It was cheap, it was reliable, and the thing just plain looks cool. But there's still one twist in the story. In 1982, Steve Jobs would get kicked off of the Lisa project. By all accounts, he had been constantly suggesting tweaks and changes to the project to the point that the Lisa got pushed behind schedule.

So the rest of the team decided that it would be best for everyone if he went and bothered someone else at the company. Angry and probably a little dejected, he decided to find the closest team working on a computer and shape it up into what he thought Lisa could be. And after some changes, that project would turn into what's now known as the Macintosh, another graphical system based around a mouse. And after some changes to the outside look, the Mac project would use an almost identical mouse to Lisa. The Apple Lisa would be released in 1983 to some mixed reviews. It was expensive. It initially cost just under $10,000 in 1983 money. It was also underpowered for the time.

Overall, it was another huge flop in the market, like the Alto before it. But it brought the mouse to the attention of the tech world. This was the first time that the mouse would appear in a large way with much marketing force behind it. The following year was 1984, and that's the release of the Macintosh with its own, internally similar, mouse. And the Mac, well, that system would fare much better in the market. Alright, so that's the story of the mouse. How it went from a niche device in Engelbart's lab, to another lab at Xerox, and finally to the market with Apple.

There are two big takeaways from this story for me. First, while the insides of the mouse have changed a lot over the years, the overall design has remained almost constant since the 1960s. That blows me away. There are very few things, especially in the realm of computing, that are that long lived. Secondly, you can easily trace where the idea changes hands. From Arc to Xerox to Apple, each has a single point of contact from where the mouse spreads. It's almost like a contagion in that sense. I guess an intellectual contagion if we want to talk about it in those terms. So the next time you find yourself clicking a mouse,

remember that you're holding nearly 60 years of heritage in your hand. Thanks for listening to Advent of Computing. I'll be back in two weeks time with a new episode on the strange history of computers. Until then, if you like the show, then please take a second to share it with a new episode on the strange history of computers. Until then, if you like the show, then please take a second to share it with a friend. You can also rate and review on Apple Podcasts. If you have any comments or suggestions for a future topic, go ahead and shoot me a tweet.

I'm at Advent of Comp on Twitter. If you want to be a super fan, you can also buy some merch from my TeePublic shop. I'll post a link in the description. And as always, have a great rest of your day.