Sellout or Savior?

There’s a sense of dissonance in the office of Miguel de Icaza. On one hand, here is the celebrated hacker – as in programming whiz, not virtual trespasser – wearing a T-shirt, looking boyish and rail-thin, and resembling an impoverished graduate student who has been living on coffee. But here also is the vice president of product technology for staid software giant Novell, entirely at ease as he takes command of a plush corporate conference room in Cambridge, MA, with a view of the Charles River and the Boston skyline. It’s a dissonance, however, that de Icaza is quick to wave away. “There are a lot of motivations in the open-source community, like the freedom to choose software platforms and the chance to innovate,” he says, referring to the global community of programmers who write software that others are free to download and modify. “Now one of my motivations is that I’m being paid to do this, and I have to deliver products.”

Both determinedly idealistic and stubbornly pragmatic, de Icaza is in many ways the new face of the open-source software movement. A programming firebrand, de Icaza has rocketed in stature in just a few years from an unknown student at a Mexico City university to one of the leaders of the increasingly successful challenge to Microsoft’s hegemonic grip on computing. And though he remains deeply connected with the community of idealistic programmers who make open source possible, his meteoric rise is fueled in large part by a keen marketing sense. De Icaza recognized early on that to be truly popular with everyday users, the Linux operating system – the freely available operating system that serves as open source’s alternative to Windows – needed the same icon-based bells and whistles familiar from the Windows desktop and access to applications of the same variety and quality as those that run on Windows machines. Open source might offer cheaper and better software, but de Icaza instinctively recognized that it would only change the world if people actually used it.

Indeed, nothing better illustrates the surprising evolution of open source than de Icaza’s trajectory over the last eight years. Starting as an unpaid programmer who contributed minor software programs that supplemented Linux, de Icaza cofounded a startup company called Ximian in 1999 to bring Linux-based desktop tools to market. The company quickly became a major player in the open-source movement, and last year Novell, a financially troubled software vendor best known for its corporate networking products, bought the startup.

Novell hopes to use open source to stop the hemorrhaging of its customer base. Novell’s once vaunted Windows-oriented corporate networking software had been waging an increasingly vain battle against Microsoft’s own products. “We were losing 10 to 12 percent of our base every year,” says Hal Bennett, Novell vice president for business development. “We needed a bright new place to go.” Novell hitched itself to the one approach making headway against Microsoft: open-source software. “Open source is starting to look like the future to a lot of people, and Novell desperately needed a future,” says Rob Enderle of the market research firm Enderle Group in San Jose, CA. “Open source and Linux gives Novell a road map that appears credible.”

What’s most significant about de Icaza’s move to the corporate world is that it’s not unique. “In some ways Miguel has been a totem for those of us in the open-source community,” says Michael Shaver, an open-source project leader who recently joined Oracle Software. “He’s gone from bits of hacking to dealing at a high level with both corporate and community interests. And that’s similar to the direction the entire open-source movement has followed.”

Smashing Windows

De Icaza’s story, like those of most programmers active in the open-source community, has been closely entwined with the story of Linux. Linux appeared seemingly out of nowhere in 1991 when a Finnish student named Linus Torvalds posted the first version of it on the Web, offering PC users a way to escape what many saw as Microsoft’s tyranny. Up until that point, the open-source movement had been relatively tiny, struggling to put an operating system together piece by piece. With the appearance of Linux, the community had a viable, working alternative to Windows.

But there was a catch: using Linux required typing out arcane commands. Eliminating these commands is where de Icaza would make his mark. In 1991, he was an 18-year-old mathematics student at the National Autonomous University of Mexico – Mexico’s Harvard. De Icaza couldn’t afford a PC, but he had access to the university’s computers and soon stumbled onto the suddenly thriving online world of the open-source community – a series of bulletin boards on which a sort of rebel alliance of hackers posted programs built on the freshly posted Linux platform and swapped extended barbs on everything from code to politics.

De Icaza soon made a name for himself writing a file management program for Linux, and by 1997 he had come to the attention of Microsoft, which flew him to its headquarters in Redmond, WA, for a job interview. According to Nat Friedman, a then intern at Microsoft and later a cofounder of Ximian, de Icaza took the interview as an opportunity to lecture managers on why Microsoft should abandon its multibillion-dollar business model and embrace open-source programming. Not surprisingly, de Icaza wasn’t hired. Instead, he returned to Mexico City to launch the project that would strike at the heart of Microsoft’s business.

The project was called GNOME, and it set the audacious goal of giving Linux a graphical interface as easy to use as the Windows desktop, thereby rendering Linux a legitimate alternative for everyday users in businesses and the home (see “The Linux Revolution, Part II,” p. 50). In 1997, a few hundred open-source programmers around the world flocked online to join de Icaza. By 1999 an initial version of GNOME was complete, and it was soon adopted by tens of thousands of users, as well as by Hewlett-Packard, Novell, Red Hat, Sun Microsystems, and other companies. (De Icaza was named TR100 Innovator of the Year by Technology Review in 1999.)

GNOME had established de Icaza as not only a top-notch coder but also as someone with a keen eye for the big project and a knack for getting people on board. “I enjoy working on open-source software, but I particularly enjoy working with Miguel,” says Todd Berman, a software developer at Medsphere Systems, a health-care infotech company in Aliso Viejo, CA. “He’s a visionary, and he has a way of keeping things fun.” De Icaza, says Oracle’s Shaver, has a “terrifying talent for finding the right people and getting a project to critical mass.”

Later in 1999, Friedman suggested to de Icaza that the two of them start a software company. Ximian was aimed at providing Linux users with the sort of application software that Windows users take for granted – tools like e-mail programs and online calendars, as well as software that helps information systems managers at companies keep employees’ computers running smoothly. For most people, getting a fledgling company up and running would have been enough. But about a year later, de Icaza was already on to his next big thing. Tapping many of the GNOME programmers and using Ximian as a base, de Icaza plotted the launch of open source’s ultimate assault on Microsoft’s dominance.

Wider Net

De Icaza explains the project with a story about a cornflakes box. The box advertised a computer game inside, written, of course, to run only on computers that used the Windows operating system. “I realized no one’s going to write a game in a cornflake box for Windows, Mac, and Linux,” says de Icaza. “And that’s what keeps people using Windows today – a large selection of applications that aren’t there for Linux.” Popular programs like Word, Excel, PowerPoint, and Photoshop, for example, as well as the vast majority of computer games, are not designed for Linux. For typical PC users, Linux places favorite programs out of reach.

That problem was likely to get far worse, since Microsoft had designed a complex set of software building blocks called .Net, with which the majority of new software designed to run on networks is being written. Keeping Linux users from being shut out of this new universe required a set of programming tools that in essence would translate software written for Linux into .Net and vice versa. Such tools would let Linux users take advantage of new software written for Windows, while Linux programmers could sell their new software – for example, next-generation open-source versions of Photoshop and PowerPoint – to the giant Windows market. This proliferation of choices would inevitably drive down the price of software. De Icaza launched a project to come up with these new tools. He called it Mono, Spanish for “monkey.”

Ximian raised $15 million in venture capital by the beginning of 2001 to pursue both the new tools and Linux application software and was bringing in steady revenues by helping companies develop and use open-source software. A handful of larger companies had approached Ximian about a possible acquisition, and that had gotten de Icaza thinking: maybe he could reach his goals faster as part of a big company. “When you’re a startup, you don’t have the resources to get software deployed quickly and to support it,” he says.

But de Icaza inevitably ended up disappointed with the companies that came calling, mostly because of what he perceived as their lack of commitment to open-source software. Then, early last year, on the day de Icaza was starting a much needed vacation in Brazil, Friedman called him to say that Novell executives were coming to Cambridge the next day. “I said, ‘God, not another meeting that doesn’t go anywhere,’” de Icaza recalls. But he flew back and gave a presentation about Mono.

Novell had long ago ruled the lucrative world of networking software – the programs that help computers talk to one another, now ubiquitous at companies and becoming common in homes – until Microsoft usurped the throne. Novell’s executives sensed a way out of their predicament in remaking the company’s product lines as open-source programs. Mono would then be a jewel in its crown, allowing Novell’s customers to move freely between open-source and Microsoft-based applications. It would make even Microsoft-dependent companies free to use Novell’s open-source products.

The acquisition was announced in August of last year. But since Mono remains an open-source program that anyone can work on or modify, there are still a hundred or so programmers around the world who continue to work on the project on a volunteer basis. In some cases the volunteers do it because it gives them a chance to garner major-project experience that enhances their résumés, and in other cases they do it because they want to make sure the software meets their needs. Novell will profit from the results, not by selling the basic software, but by charging for embellished and integrated versions with service and support.

Reconciling Ximian’s and Novell’s very different cultures, meanwhile, wouldn’t be a slam dunk. “To integrate the companies, we had to let go of our culture of independence,” says de Icaza. “We didn’t want to be considered a small research facility in Novell. If Ximian had a cool-hacker reputation, then our job was to help make Novell cool.” Matt Asay, Novell’s Linux business office director, concedes that it took Novell employees a while to get used to “this Bohemian invader,” as he calls de Icaza. “People were uncomfortable at first,” he says. “But you walk around now, and the pendulum has swung the other direction. People are giddy with the prospects for open source. I would say that the greatest benefit that Novell got from Ximian was not their technology; it was their DNA.”

Rebel Sells

To some in the open-source software community, de Icaza is a fallen angel – a legendary hacker who has strayed from the good and pure. One gripe within the community: Mono enlists code that appears dangerously similar to .Net code heavily patented by Microsoft. More generally, some worry that de Icaza represents the breakdown of the once strong barrier between open source and the corporate world. In fact, Microsoft has apparently come to deeply regret spurning de Icaza back in 1997. Microsoft software architect Don Box even wrote a song imploring de Icaza to join the company and sang it to him in front of a large audience at a party late last year. But even though he didn’t heed Box’s siren song, de Icaza has essentially been accused of selling out to the corporate world.

The notion of being a sellout doesn’t precisely amuse de Icaza, but he doesn’t seem put out by it, either. “There’s always some anticorporate fanaticism, but it’s a tiny minority,” he says. “Even back when I was working on GNOME, I knew companies would have to get involved. If you want to get all the benefits of the software, get rid of bugs, deploy it in real solutions, and bring it to a wide market, then you need big companies.”

There have, of course, been a few drawbacks to going corporate. De Icaza sometimes has to stagger in for early-morning marketing meetings, though for the most part he has been able to maintain his customary work schedule of noon to 2:00 a.m. More importantly, says de Icaza, his new job has given him the best of both worlds: the freedom and shared passion of the open-source world, backed by the deep pockets and long-term strategy of a large corporation. “We were at the mercy of VCs at Ximian,” he says. “We were constantly adjusting to whatever their latest idea was. But at Novell I’m thinking about what’s going to happen over the next seven years.”

If de Icaza seems to have left behind open source’s rebel personality, maybe it’s because he had a different revolution in mind all along. Having himself grown up in a country where most people can’t afford computers, de Icaza has long championed the open-source movement as a means of bringing affordable computing to poor communities. “I’ve got a global goal,” he says. “I want to make Linux successful on the desktop for countries where people can’t afford computers with proprietary software.” And while GNOME is already helping make Linux usable to less affluent computer users, Mono aims to ensure that those users will also have access to new generations of Web-enabled software.

If Mono achieves its goal, then Linux and other open-source programs are likely to continue to gain favor at the expense of Windows. Microsoft might then very well lose its long-standing domination of the computing world altogether, which would likely lower the cost of computing and, theoretically, provide more and better choice in software. That, at least, is the open-source vision. But revolutions seldom go according to plan. To know how the odyssey of de Icaza, and his fellow open-source programmers, will turn out, you’ll just have to stay tuned.

David H. Freedman is a freelance writer based in Needham, MA.