Technical Details:
Bill didn’t like how we made space on the stack for a function result. He was used to doing it with CLR.W -(SP), while Larry and I preferred SUBQ #2, SP. Bill’s version required a memory access while ours didn’t. Bill didn’t like our technique because it essentially pushed a random, unknown value into the stack, which shouldn’t have mattered, but he thought it introduced unnecessary indeterminism.
“Hey, that’s not the right way to code. What are you guys, a bunch of hackers? I’m not sure that I want to work with a bunch of hackers.”
Both Larry and I cared more about pleasing Bill than saving every possible byte or cycle, so we changed our fix to use the slower, more conservative, Bill-approved technique. We also added a comment to the instruction in the source code to remind us why we did it the slower way in this circumstance. The comment said “We’re Not Hackers!”
A Rich Neighbor Named Xerox
November 1983
Steve confronts Bill Gates about copying the Mac
When Steve Jobs recruited Microsoft to be the first third-party applications software developer for the Macintosh, he was already concerned that they might try to copy our ideas into a PC-based user interface. As a condition of getting an early start at Macintosh development, Steve made Microsoft promise not to ship any software that used a mouse until at least one year after the first shipment of the Macintosh.
Microsoft’s main systems programmer assigned to the Mac project was Neil Konzen, a brilliant young Apple II hacker who had grown up in their backyard, in the suburbs of Seattle. Neil started working at Microsoft while he was still a high school student and single-handedly implemented the system software for their hit Z80 card, which allowed the Apple II to run CP/M software.
Neil loved Apple, so it was natural for Microsoft to assign him to their new, top-secret Macintosh project. He was responsible for integrating Microsoft’s byte code–based interpreted environment (which was actually a copy of a system used at Xerox that favored memory efficiency over execution speed and was appropriate for the Mac’s limited memory) with the rapidly evolving Macintosh OS, so he quickly became Microsoft’s expert in the technical details of the Mac system.
By the middle of 1983, Microsoft was far enough along to show us working prototypes of their spreadsheet and business graphics programs, Multiplan and Chart (they were also working on a word processor, but they neglected to mention it because it would compete with MacWrite). I would usually talk with Neil on the phone a couple times a week. He would sometimes request features that I would implement for him, or he would perhaps complain about the way something was done. But most of the time I answered his various questions about the intricacies of the still evolving Application Programming Interface (API).
I gradually began to notice that Neil often asked questions about implementation details he didn’t really need to know about. In particular, he was really curious about how regions were represented and implemented, and would often detail his theories about them to me, hoping for confirmation.
Aside from intellectual curiosity, there was no reason to care about the system internals unless you were trying to implement your own version of it. I told Steve that I suspected Microsoft was going to clone the Mac, but he wasn’t worried because he didn’t think they were capable of doing a decent implementation, even with the Mac as an example.
Then, in November 1983, Microsoft made a surprising announcement at Comdex, the industry’s premier trade show, which was then held twice a year in Las Vegas. Microsoft announced a new, mouse-based system graphical user interface environment called Windows, competing directly with an earlier environment announced by Personal Software called Vision. They also announced a mouse-based option for Microsoft Word. When Steve Jobs found out about Windows, he went ballistic.
“Get Gates down here immediately,” he fumed to Mike Boich, Mac’s original evangelist who was in charge of our relationships with third-party developers. “He needs to explain this, and it better be good. I want him in this room by tomorrow afternoon, or else!”
To my surprise, I was invited to a meeting in that conference room the next afternoon. Bill Gates had somehow manifested, alone, surrounded by 10 Apple employees. I think Steve wanted me there because I had evidence of Neil asking about the internals. But that never came up. I was just a fascinated observer as Steve started yelling at Bill about violating their agreement.
“You’re ripping us off!” Steve shouted. “I trusted you, and now you’re stealing from us!”
But Bill Gates just stood there coolly, looking Steve directly in the eye, before starting to speak in his squeaky voice.
“Well, Steve, I think there’s more than one way of looking at it. I think it’s more like we both had this rich neighbor named Xerox and I broke into his house to steal the TV set only to find that you had already stolen it.”
Unfortunately, it turned out, what we had agreed to was only that Microsoft not ship mouse-based software until a year after the Mac introduction, a date defined in the contract as September 1983. Foolishly, we hadn’t allowed for a floating ship date, so Microsoft was within their rights to announce Windows when they did. And since Apple still needed Microsoft’s apps for the Macintosh, Steve really couldn’t cut them off.
In fact it took Microsoft two more years to ship Windows 1.0—in the fall of 1985. It was pretty crude, just as Steve had predicted, with little of the Mac’s thoughtful elegance. It didn’t even have overlapping windows, preferring a simpler technique called “tiling.” When its utter rejection became apparent a few months later, Bill Gates fired the implementation team and started a new version from scratch, led by none other than Neil Konzen.
Neil’s version of Windows, released a couple of years later, was good enough that Apple filed a monumental copyright lawsuit against Microsoft in 1988, but they eventually lost on a technicality; the judge ruled that Apple inadvertently gave Microsoft a perpetual license to the Mac user interface in November 1985.
“Well, Steve, I think there’s more than one way of looking at it. I think it’s more like we both had this rich neighbor named Xerox and I broke into his house to steal the TV set only to find that you had already stolen it.”
PRICE FIGHT
October 1983
We feel betrayed by the unexpectedly high price
The plan had always been to have the Macintosh be a very low-cost, high-volume personal computer. We wanted to keep the price as low as possible so ordinary individuals would be able to afford them. The initial target price was $500, less than half the price of an Apple II at the time, but it quickly rose to $1000 after the design team added up the cost of various components.
In early 1981, after switching from the 6809 to the more expensive 68000 microprocessor and doubling the RAM to 128K bytes, we realized we’d have to raise the retail price to $1500 in order for Apple to make its standard profit margin. $1500 was approximately the original price of the Apple II, and that was about as high as we could go while still being affordable to individuals. We worked hard to keep the price from rising further and were able to hold it at $1500 for most of the time the product was under development.
Pricing a brand new computer is tricky because costs are highly dependent on volume: the more units of a component you are willing to order, the lower the price per unit. But how can you predict how well a new type of computer will sell? It’s literally a confidence game, and we had no shortage of that. Steve Jobs knew we were going to sell Macintoshes by the millions, and he was good at convincing our suppliers to share some of the risk with us via lower initial prices, which would be rewarded as volumes soared in the years ahead. For example, Steve was able to get Motorola to commit to a price of $9.00 for the 68000 microprocessor, which was less than a quarter of what they were quoting at the time.
By the summer of 1983, it was becoming clear that the disk division’s Twiggy floppy disk drive wasn’t going to make it, and if we weren’t careful, it could drag down t
he Macintosh with it. We had to scurry (see “Quick, Hide In This Closet!” on page 158), but we were able to replace Twiggy with the Sony 3.5“ drive without slipping the schedule, which was better in every way except one: it cost us an extra $50 or so. When combined with a few other recent splurges, it pushed us over the top, so we grudgingly accepted that the Macintosh would have to debut for $1995.
Meanwhile, Apple hired a new CEO, John Sculley, in April 1983. John was the former CEO of Pepsi and a world-class marketing whiz, having invented the concept of the “Pepsi Generation” and other successful promotions. He was hired by Apple mainly to apply his marketing skills to the personal computer market, and the Macintosh in particular. But big-time marketing costs big-time money.
As plans for the Macintosh launch were being finalized in October 1983, and we were frantically trying to finish the software, Steve Jobs strode into the software area one evening. He looked angry. “You’re not going to like this,” he told us, “but Sculley is insisting that we charge $2495 for the Mac instead of $1995, and use the extra money for a bigger marketing budget. He figures the early adopters will buy it no matter what the price. He also wants more of a cushion to protect Apple II sales. But don’t worry, I’m not going to let him get away with it!”
The design team was horrified. One of the main reasons we were so passionate about the Macintosh was that we thought we were working on something we would use ourselves, along with our friends and relatives. It was crucial that it be affordable to ordinary people. $2500 felt like a betrayal of everything we were trying to accomplish. We had worked so hard to keep the price down in every aspect of the design, and we resented the idea that it was being artificially inflated to cover a glitzy ad campaign. But we believed Steve would be able to convince John that we’d do better at the lower price.
After a week or so of wrangling, much to our surprise and dismay, Steve was the one who gave in. The Mac was launched at $2495, a thousand dollars more than our target. It sold quickly at first, but sales soon bogged down. This was partially due to the lack of available software, but also to the high price. Even after sales picked up in 1986, with the Mac Plus and the proliferation of desktop publishing, Apple continued to overcharge for the Macintosh, preferring huge profit margins to growing their market share, which eventually caused major problems when this practice caught up with them in the 90s.
One of the main reasons we were so passionate about the Macintosh was that we thought we were working on something we would use ourselves, along with our friends and relatives. It was crucial that it be affordable to ordinary people. $2500 felt like a betrayal of everything we were trying to accomplish.
90 Hours a Week and loving it
October 1983
Burrell modifies his sweatshirt
Most Macintosh software team members were between 20 and 30 years old, and with few family obligations to distract us, we were used to working long hours. We were passionate about the project and willing to more or less subordinate to it the rest of our lives, at least for a while. By the fall of 1983, as pressure mounted to meet our January 1984 deadline, we began to work longer and longer hours. It wasn’t unusual to find most of the software team in their cubicles on any given evening—weekday or not—still tapping away at their keyboards at 11 P.M. or even later.
The rest of the Macintosh team, which had now swelled to almost a hundred people (nearing the limit Steve Jobs swore we would never exceed), tended to work more traditional hours. But as our deadline loomed, many of them began to stay late as well to help us out. Dinner was brought in for those who stayed late, and we would all put the software through its paces, competing to see who could find the most bugs, of which there were still plenty, even as the weeks wore on.
Burrell is wearing the sweatshirt in this picture taken at Larry Tesler’s birthday party at Caroline Rose’s house in Palo Alto in April 1983. Steve Capps is up top, left to right is Burrell Smith, Caroline Rose, Marge Boots (Capps’ wife), Andy Hertzfeld, and Larry Tesler.
Earlier in the year, Debi Coleman’s finance team decided to commemorate the team’s effort in the traditional Silicon Valley manner: they made a T-shirt. Actually, to make it a little more special, they chose a high-quality, gray hooded sweatshirt. Steve Jobs had bragged to Time magazine that the Macintosh team was working “90 hours a week.” In honor of this exaggerated assertion, they chose the tag line, “90 Hours a Week and Loving It.”
The sweatshirt featured the Macintosh name in red letters, purposefully misspelled as “Mackintosh,” as it had been in the Time article, with a black squiggle crossing out the errant “k.” The “90 Hours” tagline was emblazoned in black across the back. The software team wasn’t all that pleased with the whole sweatshirt thing because while we really were working that hard, and most of the other sweatshirt recipients weren’t even coming close to 90-hour work weeks. But it was a pretty nice sweatshirt, and lots of the engineers wore them frequently, including Burrell Smith.
When Burrell finally quit Apple in February 1985, he continued to wear the sweatshirt. But, as soon as he returned home following his resignation, he immediately took some masking tape and made a big “X” across the “9,” virtually obliterating it from view. From then on, he wore it nearly every day, proudly displaying the updated motto, which reflected exactly how he felt. It now read “0 Hours A Week And Loving It.”
Burrell at home in March 1985.
MacPaint Gallery
October 1983
A gallery of Susan Kare’s MacPaint art from 1983
Susan Kare
Bill Atkinson began writing MacPaint in February 1983, just after Susan Kare joined the Mac team to design bitmaps for fonts and icons. Susan became one of the first and most accomplished users of MacPaint, trying out new features as they were developed and using it for a wide range of practical applications.
As the Mac team struggled to finish in time for the launch, Susan kept a notebook of many of the MacPaint documents she created. They provide an interesting glimpse of the daily life of the Mac team during that period.
The image on the left is an announcement of a combined birthday party for seven Mac team members whose birthdays fell in early April, including me. I remember how terrified I was of turning 30. As a birthday gift, Susan made me a jersey with a large hexadecimal number “1E” (which is 30 in base 16) on it, so I could still say I was a teenager, at least in hexadecimal.
The image on the right is for a picnic held in July 1983 to celebrate the wedding of two members of the software team: programmer Larry Kenyon and librarian Patti King. Larry and Patti actually eloped in June; the picnic happened after they returned from their honeymoon.
During the last couple of months before shipping, we held some testing marathons where the software team enticed employees from the rest of the division to stay late and help test the software by bribing them with dinner (see “90 Hours A Week And Loving It” on page 196). The image on the right is an announcement of another bug hunt, accompanied by the bug report form that was used during the testing.
Below these is an invitation to celebrate a ROM freeze at software manager Jerome Coonen’s house. This wasn’t the final ROM freeze, which took place in September, but the first of a series that led up to it. We had a party celebrating the final ROM freeze at Woz’s house in Scott’s Valley.
The Mac group worked hard, but Apple would occasionally throw parties for the team, sometimes with unusual themes. The poster on the far left promoted our “Punk Party.” Everyone was supposed to attend in “punk” attire and a local band, “The Medflys,” was hired to play. The other poster advertised a dance that featured a DJ. Some of us were so disappointed with the music he played that we left and drove to a nearby record store to buy some better music for him to play.
Steve Jobs is very visually oriented and everyone knew he would react more favorably to material presented with nice graphics and high production values. Because of this, the Finance team recruited Susan to use MacPaint to
make attractive covers for the monthly financial reports. Here are some covers she made for the June and July 1983 reports.
As we began preparing for the launch in the fall of 1983, Susan was asked to produce art for various marketing materials, to show off the kinds of things you could do with MacPaint. The frog on the left demonstrates how you might use MacPaint in a biology class. Susan also had a fondness for whimsical applications, as demonstrated by the picture on the right.
Here are some examples that Susan created for marketing. The “Japanese Lady” on the left is pretty famous because it was used in the original brochure. She started with a scan of a fine Japanese woodcut Steve had procured. I think the hilarious detective agency letter on the right was also used in an advertisement.
Sports were another outlet to blow off steam between longs hours of working. In the spring of 1983, we managed to get facilities to install a basketball hoop in the back of Bandley 4 and usually played a half-court game every afternoon. We also had co-ed softball games once per month, which could sometimes get pretty intense.
Apple gave out brand new Macintoshes to everyone on the team right after the launch (see “The Times They Are A-Changin’” on page 217). On the left is a certificate that accompanied each machine. On the right is a bottle of champagne, in honor of the launch, to complete the celebration.
Steve Capps Day
December 1983
We all dress up in honor of Steve Capps
Revolution in The Valley [Paperback] Page 19