National Linux Distributions

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.