Open-source software is customizable, expandable and available at low or no cost. Closed-source software is difficult to customize or expand and often is exorbitantly expensive. As a result, businesses and governments worldwide are beginning to make extensive use of open-source technologies, in some cases going as far as introducing legislation to mandate its use. You'd expect developing nations to be leading this charge for free, high-quality software, but you'd be wrong.

In Geekcorps' experience supporting over a hundred IT projects in developing nations, open source is a surprisingly difficult sell. The chief mistake in marketing to the developing world is an overemphasis on “free beer”. The expression comes from Richard Stallman's demand that software be “free as in 'free speech', not free as in 'free beer”'. Emphasizing the low cost of open-source software often backfires. The history of technology transfer and appropriate technology projects includes efforts that dumped obsolete technology on developing nations when it was no longer marketable in developed nations. As a result, inexpensive and inferior are often falsely equated.

In fact, free software frequently costs users in developing nations more than proprietary software. Widespread copyright infringement means software often is available for the price of the media on which it's delivered. In Yerevan, Armenia, I recently found Microsoft Windows XP and Red Hat 8 shelved side by side, both selling for less than $5 US. For users familiar with Windows, Linux has an incremental cost—the cost of the manuals necessary to use the software. And, at copyright infringement prices, manuals can cost 20 times as much as software.

Although on-line documentation can address some of these problems, it can be expensive and intimidating to access this information. Downloading the 3MB Emacs manual gets expensive when you're paying by the minute for connectivity at a cyber café where 20 machines share a single 28.8kbps modem. Asking questions of experts in another country, often in an unfamiliar language, becomes even more intimidating when new users encounter netiquette for the first time. A well-meaning response of “RTFM” is likely to be misinterpreted by someone from a culture that places a high value on politeness or formality. Almost every culture in the world is more polite than global geek culture.

Although open-source advocates need to consider the linguistic, cultural and bandwidth barriers that affect adoption in developing nations, our chief focus must be on demonstrating that open-source software can be more advanced and powerful than the alternatives. This requires a focus on the free speech benefits of open source: expandability, transparency and resulting high performance. A great example of “free speech” application development is the translate.org.za project, dedicated to providing operating systems and critical applications in all 11 of South Africa's official languages. Thanks to their work—and the availability of source code for Mozilla—a browser now exists in Zulu, Xhosa and four other languages.

IDN, one of Ghana's leading ISPs (www.idngh.com), is refocusing its business around its SAVA wireless access points. Manufactured in Ghana, SAVA boxes are Linux servers that let businesses and cyber cafés offer connectivity using WiFi, avoiding Ghana's inadequate local phone system. ISPs around the continent are looking at IDN's success, and some are migrating their architectures to Linux servers.

Before open source can have a true worldwide audience, developers need to focus on IT problems specific to developing nations. Efforts such as Simputer (www.simputer.org)--a low-cost Linux-based handheld for India—suggest one model for these efforts: a cadre of talented engineers from a developing nation supported by a global community. MIT Media Lab's ThinkCycle Project (www.thinkcycle.org) demonstrates another model, matching design students in universities around the world with the engineering challenges of nonprofit organizations in developing nations. The US government's new Digital Freedom Initiative (www.digitalfreedom.gov) is trying a third model, creating solution teams of Senegalese developers, business and IT expert volunteers from US and Senegalese end users, to create new IT applications for use in Senegal and throughout Africa.

Political developments also may advance the global adoption of open source. An increasing number of politicians, like Peru's Edgar Villanueva Nuñez, are pushing legislation to mandate the use of free software to protect the “integrity, confidentiality and accessibility” of information.

Advocates in the developed world can push in a different direction. Demand that money spent by your nation's foreign aid agency on development of new IT products in developing nations goes toward the creation of open code. The result likely would be a boom of developers in poor nations and a wealth of code that could be used in other developing nations—not to mention a few more Xhosa-speaking geeks.