Francisco J. Proenza Food and Agriculture Organisation (FAO), United Nations, Italy Francisco.Proenza@fao.org

In most developing countries, the adoption of a national programme can prevent technological lock-in through selective, judicious and cost-effective use of open source software.

Software is subject to network economies that makes an application rise in value rapidly as the number of users increase. This leads to winner-take-most markets, where a single enterprise achieves overwhelming dominance. Consumers become captive or ‘locked’ into a single technology because everyone uses it and because the costs of shifting and learning to use alternative products are high.

The role of government is to promote competition in all markets . The leading software vendor often bundles features in its software that increase consumer dependence and thwart competitors that directly threaten its dominance. Regulators have sought to level the playing field by imposing fines for anti-competitive behaviour and by requiring the sharing of technical information. It will make easier for the competitors to connect their desktop applications to the dominant vendor’s servers. But often, it is not the effort of regulators but rather the maturing of technology and innovation in business models that gnaw away the leading software vendor’s dominance.

Network effects and technological lock-in are highest where a significant investment in a proprietary technology is already in place. This is hardly the case in most developing countries where e-Government and computerisation is only beginning. The re-training and other transitional costs of moving from proprietary technology to a low-cost open source technology are much higher in the US, Australia, Sweden, Korea or Singapore, where there are over 60 computers for every 100 people. In Asia, very few countries have even 3 computers per 100 people. In most developing countries, the adoption of a national programme can prevent technological lock-in through selective, judicious and cost-effective use of open source software.

Analytical work has firmly established the reliability of open source software as practical and robust technological platforms supported by sensible business models. Many large companies like Sun Microsystems, IBM, Novell, etc. who are direct competitors of the leading proprietary vendor have participated in the development of open source software. These companies have provided the coordination and investment resources needed to ensure that some widely used applications developed under an open source are reliable, sustainable and available across several technological platforms. The companies have done well because their support of open source serves as a viable business model. They are regarded as ‘community friendly’ (a powerful form of advertising), and they can make profit on services (e.g. training, technical support) or by selling enhanced software products.

Some governments are making large scale all-purpose migrations from proprietary to open source software. The small Municipality of Extremadura, Spain was perhaps the first to make the move and cities like Bergen, Barcelona and Munich followed suit. Paris considered migration, but given its strong dependence on proprietary software, opted for a gradual shift to open source. After many years of recommending migration to open source systems in government desktops, the Government of Brazil appears to be on the verge of issuing a presidential decree mandating migration of all the computers of its 22 federal ministries to open source (Linux and FreeBSD) operating systems.

Challenges by type of application
The choices of the governments of developing countries regarding software may be broadly categorised into three major groups:

• e-Government portals and service delivery systems;
• Desktop office applications;
• Community networking and online collaboration software.

A distinction between user requirements is necessary. The most successful open source systems - Perl, Linux, Apache and PHP are used primarily by information technology specialists who value the ability to make changes in the code to suit specialised needs. Many e-Government applications fall in this category. The possibility of modifying code is valuable to public agencies developing their online service applications. It can enable an agency to share code and coordinate developments with other agencies, without having to reinvent the wheel or pay hefty proprietary fees.

In contrast, the much larger market for desktop applications – spreadsheets, word processing, presentation, and publishing is made up of people interested in ease of use and the standard features of an application. Their desire or technical capability to alter code is mostly negligible. Also the costs of shifting from one technological platform to another are generally high for users of desktop applications.

Networking and online collaboration software are in a separate class. This kind of software is important in developing countries because it is through the interaction of people and communities with shared interest that the truly empowering effect of computers and the Internet will be realised. Most community group members are not expert users. They use mailing lists and interact with others to achieve social and economic objectives, and rely on administrators or webmasters to manage the software. The costs of shifting technologies is not an overriding concern to members of the community, but the availability of a system that meets the needs of many countries and multiple language requirements and that can be continuously upgraded at low cost should be of concern to governments, donors and development agencies.