There are hundreds of Linux distributions. Some are great, some are not. But each has its own special focus and constituency. In this cornucopia of distros, a growing focus on national distributions has emerged, especially in developing countries. China has had a well-known national distribution, Red Flag Linux, for over 5 years. Today Spain has at least 5 regional distributions. India now has one distribution called Bharat Operating System Solutions (BOSS). Most of these distributions have been built in-house, based on non-commercial sources, or have been selected from a single vendor or non-competitive consortium.

“It is better to offer a reference implementation instead of a government sponsored point-solution.”

As developing countries race to close the digital divide, open source has become an attractive way to jump start their local ICT industries. So why not take the bull by the horns and develop and control a country’s or region’s own Linux distribution? Such a solution can be localized in as many languages as needed and is free from payments for royalties, upgrades and unneeded support. However, government sponsorship of a particular Linux distribution may have unintended consequences. First, it may stifle the development of local industry by discouraging competition from alternative providers. The chosen implementer, whether in-house or a vendor, may become a favored provider with advantages of both precedent and access. Second, selections made for a single distribution may work against the natural evolutionary benefits of wider choice, whereby multiple solutions continually compete.

Instead, distributions that purport to represent national interests have special responsibilities.

Most national distributions have been built to solve particular problems. However, a better approach may be to use the considerable resources of government to encourage the ICT industry through promotion of open standards and broad competitive activity. Such a recommendation should not be mistaken as an unqualified encouragement to add standards-based but proprietary products to the mix of ICT solutions. Indeed proprietary vendors typically are not honest brokers of common standards, but instead seek to embrace and extend them through additions and vendor lock-ins.

It is better to offer a reference implementation instead of a government sponsored point-solution. A national open source distribution should serve primarily to validate requirements, explicate relevant standards and provide guidance for solving the practical ICT challenges that a developing country faces. In other words it should provide a working profile of the solutions desired that can be fulfilled sustainably by an industry encouraged to grow in capability and responsiveness. Therefore, most effort should be addressed toward the development of a framework of national and international standards and a profile of specific features derived from the standards that are selected. Furthermore, national distributions should encourage open source practice and methods. They should promote, by sponsorship or through the bidding process, the progress of existing upstream open source projects and the formation of new projects as needed.

So instead of building solutions that just address the immediate problem of one country or another, national open source efforts should provide a framework which encourages collaborative activity and the growth of an open source economy as widely as possible. Only then will a healthy open source ecosystem be in the position to promote a more comprehensive and robust national ICT industry and begin to bridge the digital divide.

Some of the most popular open source projects are Apache, Mozilla, Ubuntu, Asterisk and Wikipedia. Today, they are at the top of the popularity curve. Many other projects, of course, are not so well known and live farther down the tail of the curve. While they may be less successful today, they nonetheless represent the future of the open source world. Every project which is now at the head of the curve was itself in the long tail when it began.

What is the Long Tail?

The Long Tail is a term popularized by Chris Anderson of Wired magazine. The idea is that most things are of interest to only a few people at a time. Only a small percentage of things are hits with everyone. Everything else prospers in its own niche. But taken together, the aggregate bulk of niche items far exceeds the weight of the relatively small number of hits.

Think of the Long Tail as the engine for the continued success of open source software.

A great example is seen at Amazon.com which sells millions of books. Most of Amazon’s sales involves a few copies of specialized titles to a few people at a time, over and over again. When tallied up, the bestsellers make up only a small fraction of Amazon’s total volume of sales.

A study by the MIT Sloan school of management called “From Niches to Riches: Anatomy of the Long Tail” explains the dynamics of the Long Tail phenomenon. On the demand side, lots of people with their individual interests make up a huge market for new ideas and products. On the supply side, digital forms of communications like the Internet offer an almost free way to reach these niche markets. Before the Internet, the cost of physical storage and handling, such as advertising, logistics and shelf space, prevented less popular items from being available. E-commerce drives down the cost of handling and distribution. It promotes centralized logistics, customized advertising (like Google’s AdSense), and virtually unlimited shelf space. There is no difference in cost between handling popular or obscure items.

Why is it cool?

The Long Tail model of commerce is revolutionary. Hit and non-hit products get equal treatment in an on-line storefront. No longer does an idea or product have to climb to the top to be noticed. It just has to be relevant to someone somewhere.

The Long Tail model is inclusive. It gets rid of the middleman. It offers an instantaneous, direct and personalized connection between supply and demand. As the case of Amazon demonstrates, it can be extremely successful for both supplier and customer. Still, the Long Tail can fail if the supply side is not setup properly. For example, if licensing or distribution agreements are not worked out, the demand side may be stunted as individual users face unattractive complications. Exchanges in the Long Tail market must be smooth and transparent.

So why is the Long Tail cool for open source?

Open source has all the ingredients to benefit from the dynamics of the Long Tail. It is on-line and digital. It has lots of itches that need to be scratched. The Long Tail is an infinite pipeline of projects and solutions for the ever changing IT marketplace. By self-selection, the best projects from the Long Tail repeatedly move to the head in a process that fuels innovation. Think of the Long Tail as the engine for the continued success of open source software.