The Apple Lisa

Apple found massive success on the back of the Apple II. They went public like many of the late 70s computer companies and the story could have ended there, as it did for many computer companies of the era who were potentially bigger, had better technology, better go to market strategies, and/or even some who were far more innovative.  But it didn’t. The journey to the next stage began with the Apple IIc, Apple IIgs, and other incrementally better, faster, or smaller models. Those funded the research and development of a number of projects. One was a new computer: the Lisa. I bet you thought we were jumping into the Mac next. Getting there. But twists and turns, as the title suggests.  The success of the Apple II led to many of the best and brightest minds in computers wanting to go work at Apple. Jobs came to be considered a visionary. The pressure to actually become one has been the fall of many a leader. And Jobs almost succumbed to it as well.  Some go down due to a lack of vision, others because they don’t have the capacity for executional excellence. Some lack lieutenants they can trust. The story isn’t clear with Jobs. He famously sought perfection. And sometimes he got close.  The Xerox Palo Alto Research Center, or PARC for short, had been a focal point of raw research and development, since 1970. They inherited many great innovations, outlandish ideas, amazing talent, and decades of research from academia and Cold War-inspired government grants. Ever since Sputnik, the National Science Foundation and the US Advanced Research Projects Agency had funded raw research. During Vietnam, that funding dried up and private industry moved in to take products to market.  Arthur Rock had come into Xerox in 1969, on the back of an investment into Scientific Data Systems. While on the board of Xerox, he got to see the advancements being made at PARC. PARC hired some of the oNLine System (NLS) team who worked to help ship the Xerox Alto in 1973, shipping a couple thousand computers. They followed that up with the Xerox Star in 1981, selling about 20,000. But PARC had been at it the whole time, inventing all kinds of goodness.  And so always thinking of the next computer, Apple started the Lisa project in 1978, the year after the release of the Apple II, when profits were just starting to roll in.  Story has it that Steve Jobs secured a visit to PARC and made out the back with the idea for a windowing personal computer GUI complete with a desktop metaphor. But not so fast. Apple had already begun the Lisa and Macintosh projects before Jobs visited Xerox. And after the Alto was shown off internally at Xerox in 1977, complete with Mother of All Demo-esque theatrics on stages using remote computers. They had the GUI, the mouse, and networking - while the other computers released that year, the Apple II, Commodore, and TRS-80 were still doing what Dartmouth, the University of Illinois, and others had been doing since the 60s - just at home instead of on time sharing computers.  In other words, enough people in computing had seen the oNLine System from Stanford. The graphical interface was coming and wouldn’t be stopped. The mouse had been written about in scholarly journals. But it was all pretty expensive. The visits to PARC, and hiring some of the engineers, helped the teams at Apple figure out some of the problems they didn’t even know they had. They helped make things better and they helped the team get there a little quicker. But by then the coming evolution in computing was inevitable.  Still, the Xerox Star was considered a failure. But Apple said “hold my beer” and got to work on a project that would become the Lisa. It started off simply enough: some ideas from Apple executives like Steve Jobs and then 10 people, led by Ken Rothmuller, to develop a system with windows and a mouse. Rothmuller got replaced with John Couch, Apple’s 54th employee. Trip Hawkins got a great education in marketing on that team. He would later found Electronic Arts, one of the biggest video game publishers in the world. Larry Tesler from the Stanford AI Lab and then Xerox PARC joined the team to run the system software team. He’d been on ARPANet since writing Pub an early markup language and was instrumental in the Gypsy Word Processor, Smalltalk, and inventing copy and paste. Makes you feel small to think of some of this stuff.  Bruce Daniels, one of the Zork creators from MIT, joined the team from HP as the software manager.  Wayne Rosing, formerly of Digital and Data General, was brought in to design the hardware. He’d later lead the Sparc team and then become a VP of Engineering at Google.   The team grew. They brought in Bill Dresselhaus as a principal product designer for the look and use and design and even packaging. They started with a user interface and then created the hardware and applications.  Eventually there would be nearly 100 people working on the Lisa project and it would run over $150 million in R&D. After 4 years, they were still facing delays and while Jobs had been becoming more and more involved, he was removed from the project. The personal accounts I’ve heard seem to be closer to other large out of control projects at companies that I’ve seen though.  The Apple II used that MOS 6502 chip. And life was good. The Lisa used the Motorola 68000 at 5 MHz. This was a new architecture to replace the 6800. It was time to go 32-bit.  The Lisa was supposed to ship with between 1 and 2 megabytes of RAM. It had a built-in 12 inch screen that was 720 x 364.  They got to work building applications, releasing LisaWrite, LisaCalc, LisaDraw, LisaGraph, LisaGuide, LisaList, LisaProject, and LisaTerminal. They translated it to British English, French, German, Italian, and Spanish.  All the pieces were starting to fall into place. But the project kept growing. And delays. Jobs got booted from the Lisa project amidst concerns it was bloated, behind schedule, wasting company resources, and that Jobs’ perfectionism was going to result in a product that could never ship. The cost of the machine was over $10,000.  Thing is, as we’ll get into later, every project went over budget and ran into delays for the next decade. Great ideas could then be capitalized on by others - even if a bit watered down. Some projects need to teach us how not to do projects - improve our institutional knowledge about the project or product discipline. That didn’t exactly happen with Lisa.  We see times in the history of computing and technology for that matter, when a product is just too far advanced for its time. That would be the Xerox Alto. As costs come down, we can then bring ideas to a larger market. That should have been the Lisa. But it wasn’t. While nearly half the cost of a Xerox Star, less than half the number of units were sold. Following the release of the Lisa, we got other desktop metaphors and graphical interfaces. Agat out of the Soviet Union, SGI, Visi (makers of Visicalc), GEM from Digital Research, DeskMate from Tandy, Amiga Intuition, Acorn Master Compact, the Arthur for the ARM, and the initial releases of Microsoft Windows. By the late 1980s the graphical interface was ubiquitous and computers were the easiest to use for the novice than they’d ever been before.  But developers didn’t flock to the system as they’d done with the Apple II. You needed a specialized development workstation so why would they? People didn’t understand the menuing system yet. As someone who’s written command line tools, sometimes they’re just easier than burying buttons in complicated graphical interfaces.  “I’m not dead yet… just… badly burned. Or sick, as it were.” Apple released the Lisa 2 in 1984. It went for about half the price and was a little more stable. One reason was that the Twiggy disk drives Apple built for the Lisa were replaced with Sony microfloppy drives. This looked much more like what we’d get with the Mac, only with expansion slots.  The end of the Lisa project was more of a fizzle. After the original Mac was released, Lisa shipped as the Macintosh XL, for $4,000. Sun Remarketing built MacWorks to emulate the Macintosh environment and that became the main application of the Macintosh XL.  Sun Remarketing bought 5,000 of the Mac XLs and improved them somewhat. The last of the 2,700 Lisa computers were buried in a landfill in Utah in 1989. As the whole project had been, they ended up being a write-off. Apple traded them out for a deep discount on the Macintosh Plus. By then, Steve Jobs was long gone, Apple was all about the Mac and the next year General Magic would begin ushering in the era of mobile devices.  The Lisa was a technical marvel at the time and a critical step in the evolution of the desktop metaphor, then nearly twenty years old, beginning at Stanford on NASA and ARPA grants, evolving further at PARC when members of the team went there, and continuing on at Apple. The lessons learned in the Lisa project were immense and helped inform the evolution of the next project, the Mac. But might the product have actually gained traction in the market if Steve Jobs had not been telling people within Apple and outside that the Mac was the next thing, while the Apple II line was still accounting for most of the revenue of the company? There’s really no way to tell. The Mac used a newer Motorola 68000 at nearly 8 megahertz so was faster, the OS was cleaner, the machine was prettier. It was smaller, boxier like the newer Japanese cars at the time. It was just better. But it probably couldn’t have been if not for the Lisa. Lisa was slower than it was supposed to be. The operating system tended to be fragile. There were recalls. Steve Jobs was never afraid to cannibalize a product to make the next awesome thing. He did so with Lisa. If we step back and look at the Lisa as an R&D project, it was a resounding success. But as a public company, the shareholders didn’t see it that way at the time.  So next time there’s an R&D project running amuck, think about this. The Lisa changed the world, ushering in the era of the graphical interface. All for the low cost of $50 million after sales of the device are taken out of it. But they had to start anew with the Mac and only bring in the parts that worked. They built out too much technical debt while developing the product to do anything else. While it can be painful - sometimes it’s best to start with a fresh circuit board and a blank command line editor. Then we can truly step back and figure out how we want to change the world.