The open source movement is something completely new. It doesn't fit neatly into any accepted economic or business theory. For that reason a lot of traditional thinkers, e.g., economists, business people, and investors, are struggling to figure out what it means. Is it the beginning of a new economic order or is it a flash in the pan?

The idea that 400,000 software programmers around the world contribute their knowledge and energy for free doesn't make sense to people schooled in Milton Friedman and Karl Marx. Traditional thinkers keep coming back to the same question: "How do these people expect to make a living?"

It's a very legitimate question, one that all of us have pondered at one time or another. Is the open source movement a quaint romantic notion that will fade as it grows old? Is it the high-tech version of the Summer of Love? Are the SuSe Linuxs of the world the equivalent to the rock'n'roll promoters and record companies that made money off the hippies and freaks?

To answer that question - if answer it we can - it's necessary to make a detour through the unexpected economic event that made the open source movement possible - the Internet and the decline in communications cost to near zero. For at the core of the open source movement is the fact that connecting with people all over the world is now virtually free. All of the open source business models exploit this fact and the ones that leverage the ability of the Internet to reach millions of people for "free" are the most successful. In that sense, the open source movement is a mother lode of free publicity. The mindshare that can be gained by an open source project would cost a traditional company millions of dollar to achieve. For example, Openexchange Inc. (www.openexchange.com) was able to generate 10,000 software downloads per month for its Open-Xchange collaboration server with a public relations budget of less than $150,000 per year.

One of my favorite visualizations of the power of free communications is the Debian world map (www.debian.org/devel/developers.loc) (see Figure 1). I have not counted the dots on the map but apparently they denote the locations of the volunteers who have created the 8,710 packages that make up the Debian project.

When I think about the scale of the various open source projects, I'm overwhelmed and astonished that so much can be done with so little money. For example, the Apache Software Foundation (ASF) lists its income as $2,875 - that's right, no zeros are missing - and its expenses at $5,793 for fiscal 2003. ASF boasts assets of $112,187, representing the donations of equipment from IBM, Hewlett-Packard, Sun, and others (www.guidestar.org/controller/
searchResults.gs?action_gsReport=1&npoId=1166096 ). In comparison, in less than three seconds, Microsoft earns the same amount of revenue. With $37.6 billion in cash, Microsoft obviously has more resources (Source: Yahoo! Finance, May 28, 2005). ASF lists 29 projects on its Web site that cost it on average $199 per year to maintain and they are each supported by only $3,860 of assets. That is as close to free as something can get.

Yet ASF is an amazing success story. As shown in Figure 2, Apache accounts for almost 70% of the Web servers used by the 63,532,742 Internet hosts surveyed by Netcraft in May 2005 (http://news.netcraft.com/archives/web_server_survey.html).

ASF is obviously getting a little help from its powerful friends at Hewlett-Packard and IBM, not to detract from its incredible accomplishment though. IBM gave up on selling proprietary Web server software in the late 1990s when it realized that it had fallen hopelessly behind. Hewlett-Packard never really tried.

Now, these two giants focus on selling hardware and services and this strategy has worked well for them. According to IDC, in the first quarter of 2005, HP held 27.7% of the $1.2 billion Linux hardware market ($4.8 billion on an annual run rate), while the number-two vendor, IBM, held 19.8%. That translates into revenue of $1.3 billion per year for HP and $950 million for IBM. Clearly the folks at HP and IBM are making a living. So what about the leadership of ASF - the people who are really driving the Apache phenomenon? What and how are they doing?

ASF Leadership:

• Brian Behlendorf, CTO of CollabNet, a subsidiary of O'Reilly Media (http://en.wikipedia.org/wiki/Brian_Behlendorf)
• Ken Coar, public speaker, author, software developer (http://ken.coar.org)
• Dirk-Willem van Gulik, partner at Tribal Knowledge Group consulting firm in Alta, Wyoming, former VP of engineering with venture capital-backed Covalent Technolgies (www.tribalknowledgegroup.com/company)
• Jim Jagielski, founder of jaguNET Access Services, ISP and Web Hosting company, senior engineering consultant for Covalent Technologies (www.jimjag.com)
• Geir Magnusson Jr., vice president of products at Gluecode Software (acquired by IBM in May 2005) (www.infoworld.com/article/05/01/19/HNopensourcepanel_1.html)
• Stefano Mazzocchi, research scientist, MIT Libraries, freelance consultant (www.betaversion.org/~stefano )
• Sam Ruby, senior technical staff member in the Emerging Technologies group at IBM (http://conferences.oreillynet.com/cs/et2005/view/e_spkr/367)
• Greg Stein (chairman), engineering manager, Blogger Group, Google (www.lyra.org/greg/#bio )
• Sander Striker, freelance consultant (www.ukuug.org/bios+profiles/SStriker.shtml)

A quick scan of this list of bios indicates that the leadership of Apache is making a living the old fashioned way - working like the rest of us - whether for a traditional corporation like Google or IBM, a venture capital investor, or any client who will pay a bill. Geir Magnusson may have hit the jackpot in May when Gluecode was sold to IBM, but the rest of the Apache leadership needs to labor on to pay their utility bills.

The obvious benefit that these individuals and the other contributors to ASF get is membership in a community - note that many of the ASF leadership have ties to two companies, CollabNet and Covalent Technologies - and access to information that can be used to fuel consulting and promotional businesses run "on the side."

This notion of generating business on the side from involvement in an open source project is the second main driver of the industry. All of the open source business models rely on free, global communications and a business model on the side, which is referred to by more sophisticated commentators as "conversion" - wherein conversion means using the mindshare and reach of the free stuff to generate some revenue on the side whether it's through advertising, selling "enterprise" maintenance, selling premium versions or add-ons, or providing professional services.

See Sidebar

Next month I'll take a look at the advertising model and suggest that the open source projects would do themselves a favor if they were to adopt some of its techniques.