The Shape of Days

In January 1978, a UC Berkeley graduate student in computer science named Andy Hertzfeld spent a significant chunk of his life savings on an Apple II personal computer. By the fall of 1979, Andy had dropped out of school and taken a full-time job as a systems programmer at Apple’s corporate headquarters in nearby Cupertino. In February 1981, Apple co-founder Steve Jobs invited — though that may be too delicate a word for it — Andy to join a new project, a project that took him out of the Apple II group and set him on the road toward building something entirely new.

Between 1981 and 1984, Andy, along with a small team of bright, creative, dedicated people, redefined the personal computer and changed the world forever.

The result of that project, of course, was the Apple Macintosh, surely among the most influential products of the computer age. Even before the Macintosh debuted in January 1984, the ideas that lay at its heart were starting to resonate throughout the industry. Microsoft, for example, announced their Windows desktop environment at the Comdex trade show in November 1983, three full months before the Macintosh was introduced. If you think that was just a coincidence, you obviously haven’t read Andy’s recently published memoir, Revolution in the Valley. (A review of Andy’s book can be found here.)

Ironically, if the Mac had been a flop, Andy might have enjoyed a life-long career at Apple. As it was, the success of the Macintosh — and the relative disappointment of the Mac’s big sister, Lisa — led Apple’s management to make some sweeping staffing adjustments. Andy saw the transition as an opportunity to make a change. On March 1, 1984, Andy started what was officially a six-month leave of absence from Apple. It turned out to be just a formality. In September 1984, Andy tendered his official resignation from Apple Computer and went on to … well, you name it.

I had the chance to interview Andy recently. We talked about what he’s been doing for the past 20 years: Radius, General Magic, Eazel, his sincere belief in the open-source software movement and what’s on his iPod.

Jeff Harrell: So I guess being one of the Founding Fathers of the Macintosh, a co-founder of Radius, a co-founder of Eazel and a prominent open-source software advocate wasn’t enough. You’ve now added published author to your résumé. How’s that working out for you?

Andy Hertzfeld: It’s been great so far. I’m very pleased with how the book turned out, and it’s been very well received, as far as I can tell.

Jeff Harrell: What was it like working with the folks at O’Reilly?

Andy Hertzfeld: O’Reilly is a superb publisher, and they’ve been a dream to work with. I think they understood and appreciated what I was trying to accomplish with both the website and the book, and they were completely committed to making the book as high-quality as possible. Plus, they’re all good people and fun to work with.

Jeff Harrell: Revolution in the Valley is the story of how the Macintosh came to be, and as such mostly takes place between 1981 and your departure from Apple in 1984. Where did you go from there?

Andy Hertzfeld: At first, I did what I probably would have done if I had stayed at Apple, continuing to evolve and improve the Mac’s system software, only I did it as an independent developer instead of as an Apple employee, coming up with my own ideas and selling them back to Apple. During this period, I wrote Switcher, Servant and QuickerDraw, all of which I sold to Apple, as well as developing Mac peripherals at Radius.

Starting in 1989, I began to move beyond the Mac platform, with limited success. My first post-Mac project was a computerized television set and stereo system that I worked on with Hartmut Esslinger in 1989 called Frox, which was probably at least ten years ahead of its time.

Jeff Harrell: Odds are good that many of my readers have never heard of Frox. Can you describe it for us?

Andy Hertzfeld: Frox was the brain child of Hartmut Esslinger, the principal behind Frog Design and one of the world’s most accomplished industrial designers, who designed the Apple //c, the NeXT Cube and lots of consumer electronics — the first time he came to my house, he pointed out that he had designed over half the items in my living room! His dream was to combine his computer and consumer electronics experience by developing a computer-based home entertainment system that used signal processing to achieve the maximum possible quality. I developed a very interesting prototype of the system using a Macintosh II and jury-rigged hardware.

Jeff Harrell: Frox had an impressive debut at CES; it was even reported at the time that Microsoft’s Bill Gates was an early adopter, though I don’t know whether that was true or just a show-floor rumor. What happened? Where did all that momentum go?

Andy Hertzfeld: Hartmut was a genius in many ways, but he wasn’t a very careful businessman, and unfortunately he hired a poor management team in Germany that spent the venture capital on expensive cars and a fancy office. Eventually, the investors decided to blame Hartmut and fire him, which caused me to leave the project, too. But it was probably doomed anyway, because it was way too expensive — more than $15,000 as I recall.

Jeff Harrell: Going from helping to write the Macintosh Toolbox to founding Radius — a company best remembered for its professional computer monitors — seems like quite a change. What drew you to Radius?

Andy Hertzfeld: It really wasn’t that much of a change — Radius was founded to help the Macintosh transcend its limitations at the time. We saw the desktop-publishing revolution on the horizon and realized that the Mac needed a much bigger screen to live up to its potential, as well as a faster processor.

But I have to admit that the main reason that I got involved with Radius was to help my friend Burrell [Smith]. I was skeptical about starting a new company in those days, but Burrell was so committed to it that he kind of dragged me along. And I did get to write some cool software for Radius, patching QuickDraw to support multiple displays simultaneously, for example. We take that for granted nowadays, but it was quite jaw-dropping to drag windows between screens when we first came up with it in 1986.

Jeff Harrell: In 1990, you co-founded a company called General Magic with fellow Apple veteran Bill Atkinson. You left General Magic in 1996, but while you were there you worked on two projects called Magic Cap and Telescript. What can you tell us about those?

Andy Hertzfeld: General Magic was founded to pioneer a new product category that we called “personal intelligent communicators.” We developed hand-held devices that are sort of like what PDAs are today, but with a stronger emphasis on communication. Bill Atkinson was my mentor at Apple, and the person from whom I had learned the most, so I became involved with General Magic mainly to get a chance to work with him again.

Magic Cap was a highly graphical, object-oriented device platform that included a personal information manager and email application. You could send and receive little electronic postcards called “Telecards” that were annotated with scribbles and graphics.

Telescript was an interpreted programming language intended to run in the network, enabling smart, customizable communication services by allowing new code to be injected into the network in a safe and secure fashion, a bit similar to what Java or Jini is today, but even more ambitious. Telescript enabled little programs called “agents” to roam the network on behalf of a user. Even though it’s more than ten years old now, Telescript still seems ahead of its time to me in 2005.

General Magic didn’t come close to living up to its goals, even though we did ship products with Sony, Motorola and AT&T. There’s lots of reasons for that, which I don’t have the space to discuss here, but the biggest one was probably that we were overly ambitious.

Jeff Harrell: Let’s talk about Eazel. Your work on the Nautilus desktop environment for Linux could be seen as a natural extension of what you started at Apple. Do you see it that way?

Andy Hertzfeld: Eazel’s mission was to make free software easier to use, so in that sense it was an extension of our work on the Mac. Nautilus wasn’t a “desktop environment” by itself, but it was an important part of the desktop environment called “GNOME.” Eazel worked closely with the GNOME community to improve GNOME.

Jeff Harrell: Eazel had to close its doors in May 2001. What’s the status of Nautilus now? Is it just gathering dust somewhere?

Andy Hertzfeld: Nautilus is still the GNOME file manager, and consequently is installed on millions of systems today. If it had been proprietary, it would have gone down with Eazel, but its open source nature allowed it to continue to be maintained by the GNOME community.

I have to say that Nautilus didn’t live up to our vision for it, though. It was really only half completed when Eazel folded, and it improved in some ways but sort of regressed in other ways thereafter. I don’t want to knock the community of volunteers that continued to work on it, but they didn’t understand what we were aiming for and made it more like Mac OS 9 or Windows instead of moving it forward like I had hoped. Some of that was my fault, as I didn’t stick with it enough after Eazel closed down.

Jeff Harrell: At the risk of asking a ridiculously open-ended question, what would you have done differently if you’d stayed involved?

Andy Hertzfeld: We were trying to make Nautilus a graphical shell and universal viewer, using the Bonobo component system to provide customized viewing components for each file and directory type. We were driving the component model down to the icon level, so video file icons could actually play in place, for example, or any file type could be portrayed in its icon. I already had code checked in on a branch that implemented icon components.

The most important thing Eazel was trying to do with Nautilus that never got accomplished was to offer network-based system-management services, like automatic backup or software updating; since so much of system management is manipulating files, we thought the file manager was the ideal place to integrate system management services.

Jeff Harrell: I think it’s fair to say that you’ve become quite a fan of Linux. How did you first get involved with that whole scene?

Andy Hertzfeld: For years in the 1990s, I despaired that the software industry had fallen into a quagmire called the Microsoft monopoly, where innovation was intentionally throttled and thwarted. The open sourcing of Mozilla in early 1998 drew my attention to the free software community, and Eric Raymond’s stupendous essay, “The Cathedral and the Bazaar,” made me see that open source software had the potential to fix the structural problems plaguing the industry, creating a level playing field that was open to innovation at all levels. Once I realized that, I decided to devote myself to helping to make it happen sooner, which eventually led to me start Eazel in the summer of 1999.

Jeff Harrell: What did you do after Eazel shut down?

Andy Hertzfeld: On the very day that Eazel closed its doors, I met with Mitch Kapor, who told me about his ideas for establishing an non-profit organization to develop free software. I began to volunteer with his “Open Source Applications Foundation,” where I worked until I started my Folklore project in the summer of 2003.

Jeff Harrell: Tell us about Folklore. What I find fascinating there is the way you took your own personal itch — the desire to collect your stories about the early history of the Macintosh — and created a solution to a much more general problem.

Andy Hertzfeld: I wanted to pursue a project that combined writing prose with writing code, and came up with an idea to build a website for what I called “collective historical storytelling,” that allowed a group of participants to collaborate to recollect an important event. I initially started writing the Mac anecdotes to provide fodder for developing the site, but eventually I saw that people were more interested in the stories than the code.

I still plan on releasing the Folklore Python scripts as free software, but I’ve been procrastinating, since there’s only boring grunt work left to do and people don’t seem to be clamoring for it. So far, I’ve given a pre-release out to around a half dozen people, but as far as I know, no one is really using it.

Jeff Harrell: The conversation turns, as it must, back to Apple. First things first: What’s on your iPod?

Andy Hertzfeld: I’m a huge Bob Dylan fan, so it’s filled with lots of Dylan shows. The best new album that I’ve been listening to recently is Bright Eyes’ “I’m Wide Awake, It’s Morning,” which hasn’t been officially released yet. Conor Oberst (the main guy in Bright Eyes) is an amazing songwriter.

I maintain my own music server using a Linux box and software that I wrote myself in 1998; it now has around 7,000 albums and concerts on it, occupying a terabyte or so. So much new music is available from the net these days that I don’t have time to listen to everything that I download, which is kind of pathological.

Jeff Harrell: Would you like to comment on the recent back-and-forth over music piracy? That’s such an incredibly sticky issue — and basically one that’s off the point of this interview — that I won’t be offended at all if you choose not to go on the record. But you sort of raised the issue in a sideways fashion, and I’m sure there are lots of people out there who would be interested in hearing your opinion.

Andy Hertzfeld: I could talk about that subject for days. Basically, I’m totally on the side of the users — if I’m your customer, I want to be respected, and not have my fair-use rights stolen by awkward DRM schemes that aren’t effective at stopping real piracy anyway. Apple says that people want to own their music instead of renting it, but I think that you don’t own it if it’s encrypted and you don’t have the keys. There are many useful features that should be in your DVD player but aren’t because it’s controlled by a cartel that caters to the content providers, not the users.

I am infuriated that the DMCA is the law of the land — it’s horrific to make it illegal for me to take something apart to figure out how it works. We have an appallingly bad government in the United States right now. I better shut up before I get arrested.

Jeff Harrell: You helped define the original Macintosh desktop experience, and you helped to shape the desktop experience on Linux, and now of course you’re familiar with Mac OS X. How do you see the lay of the land? How do Linux and the Mac compare today in terms of user experience?

Andy Hertzfeld: These days, I use all three platforms, Mac, Linux and Windows, and I still think that the Macintosh has the best user experience by far. But OS X still has lots of flaws — I often despair that Apple values appearance more than usability these days. It seems sad to me that in the last 20 years the hardware has improved by three or four orders of magnitude, while the software has only improved incrementally. The industry still has lots of work to do to make the software realize the potential of the hardware.

Jeff Harrell: What’s your biggest usability gripe with Mac OS X? From your point of view, what’s the most egregious shortcoming?

Andy Hertzfeld: The way that the Finder hangs for tens of seconds at a time when accessing a powered-down disk, or the way that directories are updated at exactly the wrong times — not until the user clicks on it, which is exactly when you shouldn’t be changing it.

My favorite trivial “appearance over usability” bug is the way a scroll bar elevator doesn’t look like it’s at the top, even when it is. It’s a gratuitous visual flourish that interferes with usability. I also think the traffic light window buttons are stupid — sometimes I click on a plus to make things smaller, for example.

Jeff Harrell: A huge part of Apple’s overall strategy since Steve Jobs returned has been the idea of the “digital hub.” The company no longer uses that term as a buzzword the way it once did, but it’s obviously an idea that still dominates their long-term plans. What’s your opinion of the digital hub idea?

Andy Hertzfeld: The tremendous success of the iPod has certainly validated the “digital hub” strategy, which is good as far as it goes, but I think PCs are inherently too complex for most consumers to deal with satisfactorily. I suspect that the winning “digital hub” will eventually be a networked service that provides intelligence to more appliance-like devices in the home.

Jeff Harrell: A sort of “divergence not convergence” idea?

Andy Hertzfeld: I wouldn’t say that. I want the computer to more or less disappear, instead of being at the center of the user’s consciousness. The best way to do that is to have it out in the network. Let them focus on the spokes, not the hub.

Jeff Harrell: From a user-interface programmer’s perspective, what’s your opinion of Apple’s Cocoa toolkit? It’s obviously very different from the classic Macintosh Toolbox that you helped create.

Andy Hertzfeld: I think Cocoa is an excellent object-oriented UI toolkit. I’m not that a big fan of Objective-C, though. Interface Builder was great in its day, but it’s showing its age; I think it might have fallen behind newer approaches like XAML. I’ve dabbled a bit with Cocoa, but I’ve never written a substantial program using it, and probably never will, so I’m not in a great position to judge. These days, I prefer to use a cross-platform toolkit like wxWidgets whenever possible.

[At this point, to my eternal shame, I had to confess to Andy that I’d never heard of XAML. He explained to me that it’s Microsoft’s technology for building user interfaces. At that point, my eyes glazed over — I’ve never been much for acronyms that I can’t pronounce and that require a detailed design document to explain — and I decided to return to more comfortable territory.]

Jeff Harrell: For a while now, you’ve been concentrating on Folklore, and later on Revolution in the Valley. With the Web site established and the book on shelves everywhere, what’s next for Andy Hertzfeld?

Andy Hertzfeld: I’m not sure. My main interest is still the user interface, and I have lots of ideas about innovative user interfaces that I’d like to explore. I’m also very interested in media browsing and playing, as well as various network applications. I might also try to write a novel, but I’m not sure I can pull it off.

• PSA: The next book review
• ‘Revolution in the Valley’
• ‘Revolution in the Valley’