A Conversation with General Magic’s Andy Hertzfeld

This month, we had the pleasure to speak with General Magic software wizard, Andy Hertzfeld. We spent the same amount of time with Andy as with other people we feature in this section, however Andy simply speaks twice as quickly.

Since Andy had so many interesting things to say, covering the time from the start of his professional career, we decided to break this article into two parts. Please enjoy part one of this interview presented below, and look forward to the conclusion in the December issue of Pen-Based Computing.

In the world of microcomputers, Andy Hertzfeld is a legend. For his work on the Apple II, Macintosh, Radius, and now General Magic, Hertzfeld receives from software engineers the kind of adulation usually reserved for celebrities. And after all this, Hertzfeld can truly be said to have joined the ranks of his early hero, Steve Wozniak.

Hertzfeld started: “I’m an example of an individual whose life was changed by a machine, by an Apple II that I bought in January 1978 when I was a grad student at UC Berkeley. It really did transform my life.”

Pull-quote: “I want to create an ecstatic experience as well as making it very useful. But more important than that, to capture this thing that I didn’t invent by myself but saw inside the Apple II, which is the spirit of fun that Woz put into his machine.”

Apple II Forever

Bitten by the microcomputing bug, Hertzfeld succumbed to the fact that: “the Apple II quickly became about a hundred times more interesting than my classes.” Driven by an active curiosity, Hertzfeld delved deeper and deeper. “I was driven as a systems programmer to get to the bottom of it and the more I learned, the more I appreciated it. It’s just an amazing piece of work, full of passion, spirit, and excellence in so many dimensions.”

After taking the early route of publishing programs in magazines, a friend convinced him to try selling them. Hertzfeld recalled: “So I started to take them around to various places in the industry. When a company said that they wanted to sell my program, I immediately knew that they weren’t big enough.”

Enter Steve Jobs

Hertzfeld continued: “So I would go up the next level until finally, in December 1978, I ended up at Apple talking to Steve Jobs about Apple selling one the programs I wrote. Meanwhile, I had just built up this complete idolization of Steve Wozniak. I sort of deduced his personality from his work on the Apple II and I was pretty correct.”

Hertzfeld started working at Apple in the summer of 1979 and, after working on several projects for the Apple II, got his big break when he joined the Macintosh team. Hertzfeld recounted: “They had a trauma at Apple on February 26, 1981 called ‘Black Wednesday’ where 40% of the engineering team on the mainstream Apple II product was fired in one day.”

Hertzfeld wanted to work on the Mac and made it clear to Mike Scott, president of Apple at the time. In addition, Hertzfeld described that: “I had become friends with Burrell Smith and actually—on the very, very first Macintosh prototype which was 6809-based—I had written the first graphics demonstration software in my free time.”

An Insanely Great Computer

“That afternoon, Steve Jobs came bopping into my office asking in his inimical way: ‘Are you any good?’ He had already decided that I was good enough to come work with him so he said: ‘OK, you’re working on the Mac now.'”

Hertzfeld amusingly recalled: “I said that I had to finish up this thing. I was designing a new operating system for the Apple II at that time and since I had already worked on it for two months, I had a fair amount of work done. I said give me two weeks to get this in good order so that someone else can pick it up and Jobs replied: ‘Oh, that’s shit, you don’t want to work on that.’ And, believe it or not, he unplugged my Apple II—even though it had live code on the screen that I was working on—powered it off and started walking away with it.”

He continued: “And what could I do but follow him. He dumped it into the trunk of his car, drove me over to this remote building, plopped my Apple II down on the desk and said ‘you’re working on the Mac now.’ And that was how it started.”

Hertzfeld’s first job was helping Bud Tribble, the lone Mac programmer at the time. Their goal, in some sense, was to refine the work on the Lisa and put it in a smaller, more affordable package. “The vision for the Mac, at least at the time I started, was very, very clear, and it was to take the Lisa and make it into the people’s Lisa—that’s a three word way of saying it.”

Hertzfeld described the mission behind the Macintosh: “We were going to make something for the mainstream—not for the elite people in offices—that was easy to use. In fact, I think that’s been the direct straight line of my career including Magic Cap. It has always been to make computers more accessible to ordinary individuals.”

The Ghost in the Machine

In talking with Hertzfeld, you quickly realize that he extends his passion into his work. Hertzfeld described: “[Computers] are so magical to me, they’re just so expanding of the human spirit and just the sense of joy, ecstasy even, that I get from creating things on computers, I want to bring that to people with no technical experience.”

He continued: “I want to create an ecstatic experience as well as making it very useful. But more important than that, to capture this thing that I didn’t invent by myself but saw inside the Apple II, which is the spirit of fun that Woz put into his machine.”

Hertzfeld feels that they made important progress with the Macintosh, however much of the work remains unfinished. In fact, this is much of his drive behind his work on Magic Cap. Hertzfeld stated: “It just shocks me that people to this day think that with the Windows or Mac user interface we’ve solved the user interface problem. You know, it’s done, let’s not change it.”

In Hertzfeld’s opinion, however, “it’s about one percent done compared to how it will be even ten years from now. Maybe not one percent, but the lack of progress in user interface design is astonishing to me, because we’re nowhere close to making it good enough for everyone.”

Hertzfeld summarized: “Part of the indefinable charisma of the machines is, in my opinion, the love that the designers pour into it. In the case of the Apple II, Woz working from his heart, working completely out of his passion. That’s the key trick, choose the projects that you care about. Create the things that don’t exist in the world that you’d like to have in the world.”

Time to Move On

Hertzfeld left Apple in March, 1984, staying on to ship the Mac. According to Hertzfeld, Apple was building a bureaucracy and losing some of its original values. Hertzfeld noted: “More specifically, it was because I had a boss who thought that an engineering group should be run like the marine corp—and I wasn’t ready to salute to people.”

Following Apple, Hertzfeld worked independently on a range of projects. He recalled that: “The two most significant ones I did were Thunderscan—that I did during the summer of 1984—which transformed the ImageWriter printer into a scanner.

“Right after that—actually something I started before Thunderscan was finished—was a program called Switcher which was the first multitasking environment for the Macintosh. I was able to sell that to Apple and get that adopted as the Macintosh system software standard.”

The Switcher application, in addition to being useful, also featured another Hertzfeld trademark: a pleasing animation that caused one application to slide from the screen and another to slide into its place. Hertzfeld summed this up by saying: “I always like to have the user be exhilarated and astonished by the work I do and so I’m always looking for those little tricks that will create that feeling of excitement.”

Innovating at Radius

At around this time, Hertzfeld also performed some work at the start-up monitor and accelerator manufacturer Radius. While never a formal employee of the company, Hertzfeld played a key role in its product development. However, his personal goals still called for him to remain somewhat independent, allowing him to pursue the projects he deemed important.

Hertzfeld described his relationship with Radius: “I was what’s called a founding shareholder, which means I put up ten percent of the original seed money and I software engineered their first two products—the accelerator and the full page display.”

“I didn’t want to become an employee because I was working on Servant at that point, which didn’t fit into the context of Radius.” Servant was an evolution of Switcher, and according to Hertzfeld, more ambitious in many ways. He sold Servant to Apple, but was disappointed when Apple failed to put their support behind it.

Hertzfeld recalled: “By that time Apple had gotten Jean Louis Gassee and built up this huge ‘not invented here’ syndrome. In the course of the reevaluation, I realized that I had to get off the Mac.”

Discovering Object-Orientation

Around the fall of 1987, Hertzfeld became engrossed in a relatively new model of programming known as object-oriented. Hertzfeld quickly realized that he could build “systems that were an order of magnitude beyond the Mac both in user experience and flexibility for developers.”

Finding a project into which he could channel his efforts, he began in earnest. Hertzfeld recalled: “It was an intelligent television set for a company called Frox. It was kind of a nascent version of Magic Cap, and in some sense it was my first system that was completely object-oriented and completely customize-able.”

Hertzfeld purposefully wrote the prototype in assembly language, an act that may appear a bit strange to most programmers. However, Hertzfeld’s experience made the selection obvious. “I had experienced before how prototypes become products. Also, I was most comfortable writing in assembly language and got the best results that way.”

That’s not a claim too many people can make, especially in today’s world of high-level languages and tools. Hertzfeld conceded however that: “It’s less true today. I write almost entirely in C today because the processors are fast enough now. [However, going back to the old days,] my head is still full of Apple II ROM addresses, even though I haven’t used them for 13 or 14 years. FDDA to print out a byte in hex, FB1E to poll the paddles; I know the hook addresses.”

Pull-quote: “It just shocks me that people to this day think that with the Windows or Mac user interface we’ve solved the user interface problem. You know, it’s done, let’s not change it… we’re nowhere close to making it good enough for everyone.”

Enter Marc Porat

Though Hertzfeld completed much of the technical work to launch Frox, the management of the company fell into disarray—a factor outside of Hertzfeld’s control. One short month after deciding to drop the Frox project, he received a call from his old friend from Apple, Bill Atkinson.

Atkinson was incredibly excited about a set of ideas he had heard at Apple from a guy called Marc Porat. As Hertzfeld described it: “Marc was espousing working in the area of consumer electronics, and I had already dove into that pool by trying to work on Frox.”

Listening to Hertzfeld, it’s clear that he has incredible respect for Atkinson. “Bill was sort of my mentor in a way. I learned more about how to be a great programmer from Bill than anyone else while working at Apple.”

Porat had initiated a project within Apple that was known as Pocket Crystal. Hertzfeld noted that: “Actually, at the time I became involved with it, it wasn’t called Pocket Crystal any more, it was called Paradigm. It was a further evolution of his Pocket Crystal ideas.”

Strangely, Paradigm represented a convergence of sorts for Hertzfeld. He recounted: “I had been cooking up a project with Burrell [Smith] that I called the ‘scan computer’ which was sort of similar to this and was essentially making a small hand-held computer with a scanner in it.”

“We thought of all kinds of great things that you could do with it. Maybe that’s still something I’d want to pursue. You could scan it over sheet music and have it play the music. Or you could scan over engineering documents and have it analyze them, or over maps while computing distances, etc.”

Founding General Magic

Sensing Atkinson’s excitement, along with a snug fit with his own ideas, Hertzfeld was interested. He noted: “I was a co-founder of General Magic, but the way it happened was this: Bill called me up on a Saturday toward the end of February 1990. I was interested and the next day Marc was over to my house brainwashing me.”

Anyone who’s heard Marc Porat speak immediately realizes the clarity with which he’s able to express his vision. Hertzfeld was no exception. He recalled that: “I very quickly made the decision that this was really interesting and exciting, even though I had a great reluctance to work in the context of Apple because I essentially hated the VP’s at Apple at that point. I thought that all they cared about was their own compensation and they were destroying the company that I loved.”

Early on, the new team consisting of Porat, Atkinson, and Hertzfeld recognized the importance of gaining autonomy from the burgeoning Apple. Hertzfeld remembered that: “We made a case at the April 4, 1990, board of directors meeting for doing that and they turned us down. They felt that what they wanted to do instead was to fold us into the Newton project, which at that time had been going for over three years and had no results whatsoever. They were, at the time, designing an $8000 computer.”

However, history has recorded that General Magic was indeed able to escape the grasp of Apple. Hertzfeld credited his partners for this early victory: “It was really a testament to Marc Porat’s skills of persuasion and Bill Atkinson’s clout with the board at Apple that allowed us to slip through.”

The Sculley Connection

The clout with Apple paid off in other ways as well. Hertzfeld described that: “Sculley was very enthusiastic about us for the first year of the company, and really helped us infinitely to get a relationship with Sony and Motorola. We wouldn’t be here today if it wasn’t for him, and he liked us so much that sometime in early 1991 he decided that he wanted to be us!”

Hertzfeld continued: “There couldn’t be two us-es so he decided to destroy us, or try to destroy us, by telling Motorola to drop their relationship with us and by pretending that he was working on, Newton, was a personal communicator when in fact it was nothing like that.”

Hertzfeld evaluated the situation by saying: “The Newton project had to be restarted—all the technical work was thrown out multiple times. So really one of the reasons it wasn’t so well received was they hardly had time to work on it, whereas we were working steadily on a very difficult thing for over four years. And I think it shows in the depth of it.”

— End of Part I —