Happy birthday, Lisa: Apple's slow but heavy workhorse turns 30
The story of the machine the Mac maker would rather you forgot
Read a press release from Apple in the 1990s and it'll end with something along the lines of:
“Apple ignited the personal computer revolution in the 1970s with the Apple II and reinvented the personal computer in the 1980s with the Macintosh.”
All of which is true up to a point, but the statement does overlook the product that Steve Jobs declared “will make a dent in the universe”. At the time he was referring to the Apple Lisa, the company’s first desktop computer to feature a mouse and an icon driven interface. As far as personal computing goes, it was a world first too. Tomorrow, the Lisa hits 30.
Jobs’ excited remarks about the computer were made following a visit to the Xerox PARC facility  in 1979. He was convinced that the graphical user interface he saw at Xerox, along with the mouse, had the potential to shape the future of computing. He was right, but his handling of the Lisa project led to his departure from the team in September 1980.
With the PARC influences in place, the Lisa project continued without Jobs. The universe would have to wait as he dealt with the dent in his ego. To pursue his vision, Jobs latched on to Jef Raskin’s computers-by-the-millions Macintosh concept. More on that around this time next year when the Mac itself turns 30.
Proposed in 1978 and in development since 1979, the Apple Lisa went on sale on 19 January 1983 for $9,995. To say dealers were nervous about the price is an understatement. The Apple II computer had made the company a fortune but was six years old and the pricey Apple III had problems that, despite being fixed, damaged confidence in the product. IBM had also made in-roads with its own PC hardware platform running DOS and CP/M-86.
Another unsettling issue was that the low-cost, easy-to-use $2,000 Macintosh was waiting in the wings. The two products were to be pitched at different markets yet beyond its GUI, Lisa’s hardware sophistication was an indulgence many would appreciate but few could afford. Back then, even the memory in the Lisa would have cost a few thousand dollars – a time when most would have been grateful for 128KB.
Based around a Motorola MC68000 processor, the Lisa offered 32-bit computing and supported up to 2MB of RAM. The 7.9MHz CPU was clocked at 5MHz and original units featured 1MB RAM with two Apple FileWare 5.25in doubled sided floppy drives. In-house, they were dubbed Twiggy drives after the thin British fashion model, and each diskette could store about 871KB. Despite Apple’s best efforts to minimise wear on the drive head, the Twiggy’s reliability left a lot to be desired. It was also a non-standard media – it had additional cut-outs on the sleeve – and was widely criticised. It even started a lawsuit, which was finally settled in 1992 and cost Apple $16m.
A year later these FileWare drives would be replaced by a single Sony 3.5in 400KB floppy drive, which were also used in the Lisa 2. Available as a free upgrade for all original Lisa owners, the process also involved updating the boot ROM and I/O ROM and changing the front enclosure to a single slot. The whole set-up took about 30 minutes.
Typically, the Lisa was sold with a 5MB ProFile hard drive although a 10MB disk was available later. This storage was designed for the Apple III in 1981 and plugged into a 25-pin parallel port at the back of the Lisa. Models supplied with hard disks became known as the Lisa 2/5 and 2/10 depending on the capacity. The 2/10 actually featured an internal 10MB Widget drive and was the last of the Lisa line – that particular model was renamed the Mac XL in January 1985. Original Lisa owners wanting to upgrade to an internal drive would also need a parallel port card along with a replacement motherboard and about $2,500 in spare change.
There were also two serial ports supporting full-duplex asynchronous-mode RS-232C or RS-422 protocols at rates from 300bps to 57.6kbps. Again, the Lisa used 25-pin interfacing – and with only minor differences in speed – these serial ports functioned almost identically to the DE-9 plugs found on the Mac. AppleBus, later rebranded as AppleTalk, was also available through these ports. As for the keyboard – featuring function keys and a numeric pad – it used a 0.25in three-pole jack at the front and the mouse hooked up at the back with a 9-pin D-type connector.
Internally, the Lisa had three expansions slots but apart from the dual parallel port card, for use with ProFile drives, the range of options was fairly limited. However, a triple port parallel card existed and SCSI interfacing appeared too. If you had $9,000 to spare, the Priam DataTower 86MB HDD and 60MB tape drive combo could be added that relied on its own interface card.
A Tecmar four-port serial card existed for Unix networking as users could also run a port of Xenix based on AT&T’s Unix System III. There wasn’t much else out there and despite hopes for a colour monitor card, it never materialised.
There are two kinds of people... apparently Kevin Costner is one of them
Yup, the Lisa’s 12in 720 x 364 dots screen was monochrome and it used rectangular dots, aligned to be tall. The idea was to ensure its resolution delivered 80 columns of text. This proved to be an issue when running Mac software as the Mac’s 9in 512 x 342 screen used square dots, causing image distortion. All in all, the Lisa 48lbs of hardware amounted to what we’d call a workstation today and bordered on minicomputer status back then. Users of microcomputers were more likely to find an Intel 8088 or 8080 chip running the show. The Lisa could process twice as much data as the former and four times that of the latter. You could also anticipate the Lisa working up to 60 per cent faster than a PC with an 8088 CPU, faster still on 8080 models.
Some Apple history books give the impression that Steve Jobs just waltzed into Xerox Parc, had a chat with the guys there, and plagiarised the company’s research efforts. The fact is, at some point, a deal was struck where Xerox could acquire 100,000 pre-IPO stock from Apple for $1m and, in return, Jobs and a handful of Apple engineers were granted access to Xerox's non-public technology for a few days.
Evidently, Xerox was focussed on maintaining existing revenue streams from its profitable photocopier business and thought its internal research efforts were not ready for primetime. The business believed that its labs would only develop specialised technology for institutions and corporations that could afford it, and the costs would be prohibitive for the mainstream. So the pre-IPO shares package seemed like a good deal to Xerox and if Apple could make something of these ideas, well, the Cupertino kids could knock themselves out. Xerox didn't hold its breath.
Raskin takes credit for suggesting the PARC visit – he’d been a visiting scholar while at Stanford – but his intention was to get Jobs to see the technologies being demonstrated there to clarify similar work already going on at Apple, most notably with his Macintosh project. Indeed, the work at PARC wasn’t especially secret, as the Xerox Alto computer – the first to incorporate a mouse and graphical user interface (GUI) – had been demonstrated numerous times in the 1970s; about 2,000 people got a viewing in 1975.
And being a research facility, it published papers covering its work that would have been widely read throughout the industry. Byte magazine even ran an in-depth piece on the labs' Smalltalk System-80 in 1981.
Smalltalk-80 graphics: Points, Rectangles, Forms, Pens, and Text are five kinds of objects used to create a wide range of imagery
Certainly Raskin’s plan worked as Jobs did see technologies including networking, object-oriented programming and large portrait bitmapped displays running what-you-see-is-what-you-get (WYSIWYG) editing of text and graphics among other things. But it was the GUI on the bitmapped display that stuck in Jobs’ mind. Although Raskin had been emphatic about bitmapped screens as the future and as part of his Macintosh project, up until that point, the Apple Lisa’s early development proposal had been for a machine that churned out characters on a green screen.
The Xerox Star 8010 was described as an “an executive workstation” and was the commercial culmination of the Alto computer. A seat at the machine cost around $16,000. Yet this workstation was part of an office system that included Ethernet networking, print servers and laser printing with a total cost of about $75,000; additional machines cost $16k. While Xerox did manage to produce about 25,000 systems, the concept hadn’t been intended as a commercial product.
Apple also welcomed aboard 15 Xerox staffers after the visit. Among them was Larry Tesler who jumped ship from Xerox in July 1980, seven months after the PARC visit where he had demonstrated the mouse – a device borrowed by the work of Douglas Engelbart and Bill English at the Stanford Research Institute. In fact, at some point just about everybody in the industry borrowed from Engelbart following his showing of his NLS (oN-Line System) work in 1968, later dubbed the Mother of all Demos .
The mouse was used to operate Smalltalk, an object-oriented software programming environment, devised chiefly by Alan Kay. For many, Kay’s work is the stuff of legend and, given the questions asked by Apple engineers during the PARC visits, it was obvious they had studied Smalltalk before showing up. Xerox staffers even remarked that the experience was like talking to members of their own team.
In Smalltalk-80: The Language and its Implementation (the Bluebook) by Adele Goldberg and David Robson, it is described as:
…a language of description (a programming language) which serves as an interface between the models in the human mind and those in computing hardware, and a language of interaction (a user interface) which matches the human communication system to that of the computer.
Programming the easy way, courtesy of Smalltalk
Here, the Xerox focus is on making programming easier, much easier. Yet Jobs was inspired to work on making the whole computing experience easier, and the differences are vast. With input from Jobs and help from Glenn Edens, engineer Trip Hawkins – who would later go on to found games giant EA – redefined the spec for the Apple Lisa to combine the mouse and the GUI as key features for direct user manipulation of the objects on-screen. The target audience of business customers would remain, but the Lisa had a renewed purpose and it was to render alternative business computing platforms obsolete.
Kay eventually left Xerox for Atari and later became an Apple Fellow in 1984. Bruce Horn was largely responsible for the Smalltalk microcode used on the Xerox Dorado – a Xerox Star prototype – and he joined Apple in 1981.
The Lisa Project
The Lisa project was originally headed by Ken Rothmuller who didn't take to the new ideas being introduced. He was replaced by John Couch, who’d spent seven years at HP. Couch steered the development of the windows-and-mouse approach on the platform and saw the endeavour through beyond its launch. He was also focussed on the idea of "datagramming" – a concept inspired by his father’s difficulties with computers and lack of programming skills. His thinking was to shield everyday users from arduous programming tasks and instead allow them to input data as needed for their required routines. With the simplification of computing for users high on his agenda, Couch was an ideal choice as leader.
Apple Lisa key personnel, left to right: Paul Baker, Bruce Daniels, Chris Franklin, Rich Page, John Couch and Larry Tesler
Hardware engineer Rich Page built the Lisa prototype featuring a sample 68000 from Motorola and the team began to take shape with Wayne Rosing as director of engineering. Bill Atkinson was the main graphics programmer and Larry Tesler (now in charge application software) defined the interface protocols in 1980. Along with input from Bruce Daniels, a designer in charge of system software, drop-down menus and the one-button mouse emerged following numerous trials with users to find the most effective methods.
According to Tesler, at various stages of application and interface design the results would be tested on groups of people with varying levels of computing experience including outright novices. This process ran into hundreds of hours and it’s claimed that most of the observations were caught on video with psychologists on board to analyse the responses.
Likewise, mouse development had received intensive revisions. Following permission from Xerox, Apple went to Jack Hawley, an ex-PARC man whose Mouse House made these sought-after mechanical pointing devices. Bill Atkinson knocked out a driver hack and Apple designer Jerry Manock went through over 100 different variations of computer rodent.
This attention to detail was all part of the Lisa ethic to keep it simple and deliver a product that wouldn't take people days to understand. This was going to be key for Lisa’s acceptance as an essential office platform that could be operated by secretaries, among others.
One button to rule them all
How many buttons for the mouse is a question that still rages today, but Apple’s testing on computer novices found that one button reduced confusion and eliminated the occasional glance to check the mouse buttons being used. The extended methodology behind single, double and even treble clicks was thought out at Apple too. Xerox had utilised mouse clicking for just making menu choices or highlighting words, but on the Lisa these actions went beyond simple selection tasks and provided the means for direct control of the user interface.
Overlapping windows was the brainchild of Alan Kay with programming work from Dan Ingalls realising this concept along with pop-up windows and BitBLIT – a key technology for bitmap graphics. Yet dragging and resizing windows, and launching applications with a double click were among the innovative mouse operations devised at Apple. Even seemingly superfluous touches, such as the animated window zoom from the launched application icon, was an interface refinement that not only looked cool, but informed the user that the app was running and loading.
Tasked with delivering images to the screen, Bill Atkinson had some seriously difficult challenges ahead. Overlapping windows as a concept was all well and good, but what happened in the obscured ‘clipping’ areas? Were those unseen elements computed anyway or was there a more efficient workaround? He’d assumed Xerox had cracked it when visiting PARC, but Dan Ingall’s BitBLT (bit block transfer) approach on the Alto and Dorado machines didn’t utilise clipping, as Atkinson had imagined.
Still, initially armed with that assumption, he set about the task and over several months and an awful lot of algebra, devised routines to swiftly draw text and graphics that focussed on "regions" – the areas of the on-screen interface that need to be redrawn or kept.
He dubbed his work LisaGraf that later became known as QuickDraw, a technology that would later be considered an Apple “crown jewel”. If you take a look at the Prior Art section of the QuickDraw 1986 US patent (4,622,545)  you get an excellent backgrounder on this breakthrough achievement.
Apple QuickDraw patent drawings highlight the display functions
At one point, Atkinson decided to demo a Lisa screen rapidly drawing circles and ovals to Jobs, now encamped with the Mac team in "Texaco Towers", an Apple building near a petrol station. He was taken aback by Jobs’ underwhelming response who insisted Atkinson needed to deliver rectangles with rounded corners as they were “everywhere”. Jobs even took him out for an educational stroll to prove his point. Atkinson went back to the drawing board and came back the next day with RoundRects – which was added to LisaGraf and became a fundamental part of the interface and would naturally migrate to the Macintosh. Even today, choosing rounded corner rectangles could put you in the dock  if you mess with Apple.
Another one of Atkinson’s major contributions was the Desktop Manager. This was a collaborative effort by stealth with Dan Smith and Frank Ludolph. Smith had been working on Lisa Filer which handled file management and launched other applications. Among Filer’s features was a dialogue window that had so many prompts for tasks such as Open, Copy and Discard that it became known as the 20 Questions Interface. Unhappy with its design, Smith tested it out on his wife, who couldn’t follow its logic, which led him to discuss his misgivings with Atkinson and Rudolph.
Desktop Manager combines overlapping windows, icons and drop down menus
Recalling the Dataland prototype he’d seen at MIT, Atkinson shifted the emphasis to an icon-based approach enabling files to be seen and moved around with the mouse. The trash can icon made its debut too and the ease of use of the design began to shine through. Everyday office stationery and objects appeared graphically on this virtual desktop which could be tidied up in a stroke with icons automatically aligning. The trash would hold discarded files and you’d have to empty it to delete or simply drag out anything that was to be kept.
It was the spring of 1982 and Lisa was slated to ship for the autumn. The idea of a rethink this late in the day wasn’t likely to go down too well and so they worked on getting a mock-up together before showing it to Wayne Rosing, Lisa’s engineering manager. Rosing could see these ideas had promise, so minded of the schedule, he turned a blind eye to them working on the project for a couple of weeks to get a stable version together. If that didn’t happen, then it was back to Filer.
Another condition was not to show this early work to Steve Jobs. If they couldn’t deliver, Jobs would no doubt disrupt the schedule pursuing it. They were also sneakily trading ideas with Bruce Horn on the Mac team and his Finder concept. Near the end of the two week period, Jobs got wind of the work and the game was up. Fortunately, the Desktop Manager was in good shape by then, Jobs loved it and Rosing was happy to adopt this new modus operandi for Lisa. When presenting the new direction to the rest of the team, the covert ops Desktop Manager developers appeared in newly printed 'Rosing's Rascals' T-shirts.
Familiarity breeds contentment
The hiring at Apple continued and, depending on where you look, estimates range from 90 to 200 people on board for Lisa. Tesler’s Applications Software Group had 20 personnel working on Window manager, QuickDraw, the Desktop Manager (previously Filer) and a suite of Lisa apps. With some input from Apple’s marketing bods, these would become LisaCalc, LisaDraw, LisaGraph, LisaList, and LisaProject, LisaTerminalReceive and LisaWrite.
LisaProject enabled PERT charts to be devised swiftly for planning and scheduling
These seven applications, later called Lisa 7/7 were essential in making the Lisa attractive to business. A key feature of the software was its capacity to share information between applications with its copy and paste commands – a level of intuitive interactivity that aided ease of use enormously that also fitted with John Couch’s datagramming vision. They also featured stationery pads for creating new files; the stationery would also launch the correct application when double-clicked.
One application, LisaProject, actually came about to meet the needs of the team for tracking schedules and would later be marketed for a wider audience with customers including NASA. It was devised by Debbie Willrett, wife of one of the Lisa team. The first of its kind, LisaProject enabled PERT (Program Evaluation and Review Technique) charts to be shown on-screen showing all the interdependent relationships and the critical path method (CPM) structure. The easy updating of PERT chart activity and the schedule output as a Gantt chart was a significant enticement to business. Indeed, it was the dependence on LisaProject that worried NASA most after Lisa's eventual demise.
Back in the office though, Lisa’s simplified printing dialogue functions were another breakthrough that, together with the ImageWriter dot matrix printer, enabled an uncomplicated workflow for printing jobs.
Easy-to-use WYSIWYG applications aside, the Lisa had innovations as a workstation environment too, with advanced features including protected memory, virtual memory and cooperative multitasking. Beyond the GUI that was named the Lisa Office System there was the Workshop program development mode which was almost entirely text-based. One issue with having these two machine environments was that ideally you’d need two Lisas to develop software on – one for coding in Pascal, the other for testing on the GUI, although you could always dual boot from one into the other. Also, the Mac owed an enormous debt to Lisa, given its hardware muscle; it was the primary tool for writing apps for the fledgling platform.
With Lisa’s primary software challenges resolved, work began in earnest on debugging and performance enhancements, a rather less glamorous endeavour than reinventing the wheel. Adding to the strain was the poor reception of the Apple III. Launched in May 1980, it was received as an incremental upgrade, rather than an innovation. Worse still, the machine was so buggy the first 14,000 units were recalled. Despite fixes, Apple’s seemingly unblemished reputation had taken a big hit and the Lisa team felt an increased responsibility to deliver a reliable product – one that could put a dent in the universe.
Another long shadow hanging over the endeavour was the schedule. The Apple Lisa was supposed to ship in March 1981. When quizzed on this timescale the previous summer, Rosing diplomatically remarked that it was “optimistic”. Indeed, following the January 1983 launch, the Lisa didn’t make it into the shops until May. The delays also undermined Apple’s standing and speculation was rife that the company couldn’t deliver. As the project’s existence was widely known, industry pundits were already armed with an opinion without even having seen one.
Although the proposal for Lisa appeared in 1978, Tesler’s view is that it only took three years to get it into production given the direction change in March 1980 – a few months after the PARC visit – that redefined Lisa. Part of Apple’s marketing strategy was that it had taken “200 person-years to develop programs you can learn in 20 minutes” and few could argue that what they saw on the screen was intuitive and instantly recognisable. As the Apple sales blurb declared, “Now you know where the world’s going, consider the advantages of getting there first”.
If you read through some of the interviews of the Lisa team post-Macintosh, you soon get a sense of deflation. The Mac eclipsed their efforts and the advanced features of the Lisa went unnoticed beyond computing professionals and Mac software programmers. While a sense of pride in their achievements is undimmed, the Lisa’s shortcomings in terms of price and performance, and its early retirement, must surely have stung quite a bit. They never expected the dent in the universe to be the landfill site where the unsold Lisa inventory was dumped to facilitate a tax depreciation write-off.
On reflection, Rosing remarked that it was the “largest project ever undertaken for a microprocessor”. Fewer than 100,000 Lisas were sold but according to Trip Hawkins, Apple still raked in $100m. As for that microprocessor choice, Alan Kay had his own view of this modern marvel: "The Lisa failed because it was very underpowered, and so, while it did beautiful things, it did them very slowly."
Direct manipulation of the on-screen interface was certainly the future, but in giving the world an office environment on the desktop, Apple had underestimated users' expectations once they became familiar with the machine. Using a Lisa might only take 20 minutes to learn, but getting on with the job could become frustratingly slow especially when using virtual memory on the sluggish hard drives of the day.
It was this fatal combination of poor performance and high price that was to dog broader acceptance in the business marketplace that Apple had always aimed for. The delays had allowed the IBM PC to get a foothold, too. At the time Apple had arrogantly mused on the Big Blue’s arrival with its 1981 WSJ ad: "Welcome, IBM. Seriously." Undoubtedly, a 64KB IBM PC was inelegant and technically inferior, but its price point made it a hit.
The Mac would also prove that, regardless of hardware prowess, price was an all-important factor for gaining acceptance. Indeed, despite heavy discounting a year on – the Lisa 2 floppy-only was $3,495, Lisa 2/5 $4,495, and Lisa 2/10 $5,495 – to coincide with the Mac’s launch, the Lisa’s star appeared to be fading with sales of a few hundred a month by the end of 1984.
By April of that year, Apple had also devised MacWorks, a Macintosh environment for the Lisa. This, combined with the announcement of Macintosh Office in January 1985 and the Lisa rebranded as the Mac XL with its hard disk advantage, saw a resurgence in sales orders. The pitch was that Macintosh Office would deliver networking featuring a laser printer with the
Lisa 2/10 Mac XL functioning as the file server. Even though Macintosh Office never actually materialised, it didn't stop Apple running a TV ad campaign called Lemmings.
Yet, the plan was to only continue the Mac XL from the remaining inventory of parts while essential networking and hard drive additions were engineered for Macintosh itself. With a lack of parts for the rise in new orders, and concerns that they couldn’t be acquired in time, Apple announced the Mac XL was to be discontinued - while Mac engineering manager Bob Belleville and Apple co-founder Steve Jobs were visiting Japan.
Apple had finally found the price point and – with the aid of third-party networking applications – realised the functionality that the corporates wanted, yet missed the opportunity to deliver. The Mac XL and Macintosh Office were essential pieces of the puzzle but the sudden decision to cancel the Lisa 2 three months down the line had industry analysts questioning Apple’s competence.
Even the notion that the Mac would be Lisa's younger brother and the two could work in tandem, couldn't be so. There was no way the Mac could run Lisa's apps, so it was, in effect, a one-way street if you had both machines. Sure, you could easily swap floppies between them, but you'd need to be using software that ran on the Macintosh to make use of the files. Another issue was that the Mac relied heavily on its 64KB ROM. In its Mac XL form, the Lisa acquired a Mac ROM as its original ROM was only 16KB and performed hardware tests and boot instructions.
With its initial release, Lisa did find a place in the sun for a time. As a technical achievement, it couldn't help but wow those that came across it – the press included, price misgivings notwithstanding. After booting up, the "paper-like" black-on-white screen displayed a completely new environment and its applications changed working methods for good.
Spreadsheets would never be the same again and neither would the way we relate to computers. This world of icons, folders and office stationery remains to this day and likewise the impression that these places exist on the computer continues to shape our thinking as we engage daily in direct manipulation of computational tasks.
Oh, and the name? It's claimed that Lisa was named after the daughter of an Apple engineer. Lisa is Jobs' daughter's name too, so go figure. Later, Apple invented acronyms to suit this moniker, but the one that sticks in the minds of disaffected users of the time is: Lost Interest Start Again.
If you want to grab a piece of history without grubbing around a landfill, there's a Lisa 2/5 with all the extras in museum piece condition on eBay  that will set you back about £16,000. When it comes to return on investment, not even Apple's marketing would have conceived such optimism.
LisaEm provides a different kind of Apple time machine
Another way to sample a slice of history is to run an emulator and if you don't have the necessary ROMs you don't need to search very hard to find them. I tried the free LisaEm , which is available for Mac OS X or Windows and hasn't been updated in five years. Still, it works and given the installation quirks and the overall speed, it certainly seemed authentic. If you give it a spin you'll find that even 30 years on, the interface remains familiar and intuitive.
Happy Birthday, Apple Lisa. ®
A hat tip to the folks behind the ftp.apple.asimov.net site, a huge resource of information related to early Apple products.