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.

e-Government systems
Public intervention in support of e-Government under an open source platform may be justified on social welfare grounds. The open source movement has often promoted ‘viral’ licenses that discourage innovation by preventing subsequent developers from making a profit. Software developed under the GPL license, for example, requires that any future developments built from the original software must be distributed freely with full access to the code. Governments, however, need not follow a restrictive license regime. Some licenses enable government agencies to make the software developments they sponsor freely available, but also allow private entrepreneurs to use the code and sell improvements under a proprietary license.

Use of open source need not be an all or none proposition. When a wholesale shift in software technology is not practicable, significant economies may still be achieved by sharing selected open source applications. In the US, the States of Massachusetts, Rhode Island, Pennsylvania, Utah, Kansas, Missouri, West Virginia, and the cities of Glaucester, Massachusetts, and Newport News, Virginia, have formed a Government Open Code Collaborative Repository to enable open source software code sharing by government agencies. The municipalities of Extremadura and Barcelona, in Spain, and Porto Alegre in Brasil, have established a similar network to exchange experiences and software developments.


http://www.itu.int/ITU-D/ict/
A similar initiative has been proposed for the European Commission (EU). e-Government efforts have often resulted in ‘data dungeons’ that do not interact with each other. These disparate systems reside in different agencies and become outdated rapidly. They tend to rely on proprietary software that does not conform to open standards; interconnection may be achieved but often at a high cost.

To solve this problem, Great Britain has adopted a flexible open source policy, developed through a public consultation process. The proposed architecture mandates the use of on open standards, e.g. XML, by all government agencies. Proprietary software is not excluded, provided that it meets the open standards. Similarly, Brasil’s interoperability architecture (e-ping), envisages the occasional need to use proprietary software, but will rely mainly on open source solutions and open standards.

Desktop applications
Technological lock-in in desktop applications started in high income countries when the software industry was still in its infancy. Desktop systems have since become quite sophisticated in functionality and interoperability. Businesses are resisting expensive changeovers to new versions that exhibit only minor changes in functionality. This is especially true of the standard office desktop applications – spreadsheet, word processing, presentation and desktop publishing for which robust free downloads or inexpensive alternatives are available.

The leading open source office suite OpenOffice (www.openoffice.org) is available in more than 70 languages.

The leading open source office suite OpenOffice (www. openoffice.org) may be downloaded for free. It is rich in features and its files are readable by other leading vendor office suites. The new version’s (Beta 2.0) interface is practically the same as that of the leading vendor. OpenOffice is available in more than 70 languages, a feature made easier by free access to the source code.

Because OpenOffice is available in common operating system platforms, i.e. Windows, Macintosh, Solaris, Linux, and FreeBSD, the decision to migrate to OpenOffice may be considered apart from the decision to change the operating systems. This is important for governments considering migrations to open source. The applications on offer are still larger for Windows than for Linux and problems of compatibility are more serious across operating systems. For novice developing country users who only need to type letters, write emails, work on spreadsheets, the basic functionality of low cost open source solutions that can operate in different operating systems is usually adequate.

Since 2001, the city of Sao Paulo, Brazil has run a telecentre programme under an open source environment including operating system (Linux) and desktop applications. The city sponsors 110 telecentres, all located in the most impoverished parts of the municipality, showing that inexperienced users can perform well in an open source environment. After an initial trial with dual operating systems, the city decided to exclusively install the open source Linux operating system on all of its telecentre computers. This was a crucial strategic choice which aimed purposefully to avoid technological lock-in to the dominant technology by new users. Sao Paulo’s programme served as a model for Brazil’s nationwide 3,200 telecentre programme, presently under implementation.

Pirate software has been and continues to be commonplace in developing countries (Table 2). The basic desktop applications make a very large portion of pirated software. Policing small shopkeepers, cybercafe operators and low-income users using pirated software is impracticable and politically untenable. Governments are inevitably forced to adopt lenient enforcement policies that lead to public contempt and lax attitudes towards intellectual property rights that run counter to international commercial obligations and end up hurting budding local proprietary software industries.

If governments were to encourage their citizens through widespread governmental use in its own offices and projects to use a low cost desktop office application software alternative, the common excuse of high cost of software could no longer be regarded as compelling. The real costs of policing intellectual property rights violations would be lower and could therefore be enforced more strictly and effectively by focusing on fewer violators.

Community networking
Software to establish mailing lists, web pages and enable resource sharing is a most valuable tool for empowering rural communities and encouraging collaboration online.

There are powerful proprietary software options (e.g. First Class, Lyris), some highly specialised (e.g. Blackboard for education content management). There are also open source list servers (PHPList), web page creation programmes (Postnuke) and online collaboration tools (e.g. http://wikipedia.sourceforge.net, http://www. kolabora.com, Basic Support for Cooperative Work http://bscw.fit.fraunhofer.de, etc.). Existing online services like Yahoo Groups (http://groups.yahoo.com) and Dgroups (www.dgroups.org) have limited functionality but are presently available for free.

What appears to be missing is an integrated low-cost system available or easily translated into local languages that is suited to facilitate information exchange and user friendly web page creation by small community groups in developing countries. Such a system would have the following features:

• The software developed should use open source software, under a non-restrictive license regime.
• The software should enable the use of special fonts (e.g. Sinhala, Tamil) as well as the corresponding standard fonts to allow a variety of optional languages of communication between users.
• The software should make it possible for community groups, schools, small businesses to have their own distinct unique portal shell with its own logo and banner.
• The software developed should be easy to use and run directly and independently by individual user groups, requiring no intervention of any external institution.
• The software developed should have a separate section for simultaneous chatting by registered group members, through a Web page interface within the system.

Conclusion
A rapid expansion in e-Government applications is imminent in many developing countries. The present setting offers unique opportunities to expand systems rapidly and to avoid duplications and locking the country’s e-Government services into proprietary technologies that could prove costly. It is an opportunity to be purposefully seized and planned for.

Most present users of computers in developing countries are English speakers and use proprietary software, but they represent a very small proportion of developing country populations (Table 1). For the few well off individuals who use computers, the costs of shifting to another software technology are high. Yet their views carry weight because most decision-makers and government officials fall in this category.

A dependency on proprietary software in desktop systems should not be forced upon the vast majority of people who do not speak English, have no vested interest or training in the dominant technologies, have limited income, and will be the ones to pay the most if an expensive proprietary software platform is adopted either by design or by default by their country’s leaders.

Ultimately, the decision regarding what type of software a government agency implements should be dictated by a sober analysis of economic and social considerations. Such a choice should be technology neutral, as advocated by the leading vendor and the Business Software Alliance, but should not be blind to future costs and benefits and should seek to avoid costly technological lock-in on proprietary software technology.

Recommendations

• Software developed under developing country government sponsorship should consider open source solutions, particularly if these developments are potentially useful to other members of society or to government dependencies. These developments should be subsequently made available for use by third parties (e.g. through an online software code sharing repository), under a license that enables further development and reasonable commercial exploitation.
• The formation of national Task Forces on open source and interoperability in e-Government applications bringing together senior IT officers from ministries and agencies planning e-Government systems is recommended. The first order of business should be the drafting of guidelines for the development, use, and sharing of low cost interoperable applications across public agencies.
• Most people do not need sophisticated office suites, and the freely available and low-cost desktop alternatives are sturdy enough for widespread country-wide adoption at significant savings and without having to change operating system. Government sponsored tele-centre, school computerisation programmes should consider widespread migration to low-cost open source software.
• The decision to migrate government ICT development programmes to open source operating systems is more complex. In a developing country, the proposed migration should involve a large number of PCs. Also, because of network effects, if only a few computers are involved in the change over, the end effect will be to punish the new users who sooner or later have to migrate to proprietary standard operating system.
  
  Piracy rate: Number of pirated software units divided by total number of units put into use.
  Source: BSA-IDC [2005]
  
  
• Seemingly generous offers to supply software to a few government sponsored tele-centres or school initiatives usually ignore the large mass of small cybercafe owners and low-income computer users. These offers should in general be resisted by government decision-makers. They are part of a concerted effort from the dominant vendor to retain market dominance. The risks are high that the temporary free provision of the dominant desktop applications to new computer users will end up locking the country into a high cost proprietary software spiral from which it will be increasingly difficult to extricate itself.
• Developing country governments would do well in supporting international cooperating bodies made up of academic and civil society developing country representatives that set standards for open source operating systems, as a means of ensuring their special language and font requirements are met and to encourage development country software development capabilities.
• Government and donors would do well to contribute and support the design and implementation of an open source community portal and networking software.

Works Cited

British Educational Communications and Technology Agency (BECTA), “Open Source Software in Schools: A study of the spectrum of use and related ICT infrastructure costs”, 13 May 2005.
(www.becta.org.uk/corporate/press_out.cfm?id=4681)

Blitz, Roger, “Savings gain led council to reject Microsoft”, The Financial Times, 9 May 2005.

Blitz, Roger, and Richard Waters, “Threat to Microsoft as local authorities plan to use free open source software”, The Financial Times, 9 May 2009.

Business News Americas, “Government Plans Decree for Compulsory Open-Source Migration”, 25 April 2005.
(www.technewsworld.com/story/42584.html)

Business Software Alliance and International Data Corporation (BSA-IDC), Second Annual BSA and IDC Global Software Alliance Piracy Study, May 2005.
(www.bsa.org/globalstudy/loader.cfm?url=/commonspot/security/getfile.cfm&pageid=24342&hitboxdone=yes

Cabinet Office, Open Source Software: Use within UK Government, 28 October 2004.
(www.govtalk.gov.uk/documents/oss_policy_version2.pdf))

Cabinet Office, e-Government Interoperability Framework, Version 6.1, 18 March 2005. (www.govtalk.gov.uk/schemasstandards/egif_document.asp?docnum=949)

Cobo, Matías, “Linux, la alternativa del software libre”, El Ojo Crítico. (www.la-planeta.com/elojocritico/n1-junio04.html#apoyo)

Comino, Stefano, and Manenti, Fabio M., “Open Source vs. Closed Source Software: Public Policies in the Software Market”, June 2003.
(http://papers.ssrn.com/sol3/papers.cfm?abstract_id=469741)

Economides, “United States v. Microsoft: A Failure of Antitrust in the New Economy”, 2001.
(http://arxiv.org/ftp/cs/papers/0109/0109069.pdf)

Emerging Media Communications, City of Munich selects Novell Open Enterprise Server, TWeb, 13 May 2005.
(www.itweb.co.za/sections/computing/2005/0505130853.asp?A=SRW&S=Servers%20and%20Workstations&O=FPIN)

European Commission, Commission concludes on Microsoft investigation, imposes conduct remedies and a fine24 March 2004.
(http://europa.eu.int/rapid/pressReleasesAction.do?reference=IP/04/382&format=HTML&aged=0&language=en&guiLanguage=en)

Delio, Michelle, “Munich Open Source Plows Ahead” Wired, (http://wired-vig.wired.com/news/infostructure/0,1377,62236,00.html?tw=wn_story_related)

Dravis, Paul, “Open Source: Perspectives for Development”, Infodev, 2004.
(www.infodev.org/symp2003/publications/OpenSourceSoftware.pdf)

Dravis, Paul, Open Source Software: Case Studies Examing its Use”, April 2003.
(www.dravis.net/documents/OSSCaseStudies-Dravis.pdf)

The Economist, “Open Source’s Local Heroes”, 13 December 2003.
(www.economist.com/science/tq/displaystory.cfm?story_id=2246308)

Evans, David S. and Bernard Reddy, Government Preferences for Promoting Open- Source Software: A Solution in Search of a Problem, National Economic Research Associates, 21 May 2002.
(http://papers.ssrn.com/sol3/papers.cfm?abstract_id=313202).

Franke Nikolaus, and von Hippel, Eric, “Satisfying Heterogeneous User Needs via Innovation Toolkits: The Case of Apache Security Software”, MIT Sloan School of Management Working Paper 4341-02, January 2002.
(http://papers.ssrn.com/sol3/papers.cfm?abstract_id=299419)

Government Open Code Collaborative Repository (GOCC), website launched June 28 2004
(www.gocc.gov/docs/about_GOCC).

Governo Brasileiro, “e-PING Padrões de Interoperabilidade de Governo Eletrônico”. Comitê Executivo de Governo Eletrônico, May 2004.
(www.governoeletronico.e.gov.br/governoeletronico/publicacao/down_anexo.wsp?tmp.arquivo=E15_241e-PING%20minuta%20v1.0%2020%2012%2020041.pdf)

Hahn, Robert W., ed. Government Policy Toward Open Source Software, AEI-Brookings Joint Center for Regulatory Studies, Washington, DC. 2002.
(www.aei.brookings.org/publications/abstract.php?pid=296)

Hazlet, Thomas Rivalrous Telecommunications Networks With and Without Mandatory Sharing”, AEI-Brookings Joint Center Working Paper No. 05-07, March 2005.
(http://papers.ssrn.com/sol3/papers.cfm?abstract_id=707633)

Interoperable Delivery of European eGovernment Services to public Administrations (IADBC), “Paris migration to Open Source: evolution, not revolution”, E-Government News, 13 October 2004.
(http://europa.eu.int/idabc/en/document/3382/5640)

InformaticaPublica “L’Ajuntament de Barcelona comença la migració al software Lliure en català”, 21 July 2004. (www.informaticapublica.com/notis/noticia.php?id=169&show_comments=true)

Kriss, Eric, “Open Mind on Open Source”, Power Point Presentation, 23 June 2004.
(www.mass.gov/eoaf/docs/OpenMindonOpenSource_June-23-2004.pdf)

Lemer, Josh, and Jean Tirole, “The Economics of Technology Sharing: Open Source and Beyond”, December 2004.
(http://papers.ssrn.com/sol3/papers.cfm?abstract_id=620904)

Lettice, John, “MS offers 57% price cut as Paris tilts to open source”, The Register, 2 July 2004.
(www.theregister.co.uk/2004/07/02/ms_fights_paris_open_source)

Libbenga, Jan, “Munich embraces the penguin”, The Register, 17 June 2004.
(www.theregister.co.uk/2004/06/17/munich_embraces_penguin)

Miyajima, André, “Brazilian Government to Adopt Linux”, 4 July 2003.
(www.yankeegroup.com/public/home/daily_viewpoint.jsp?ID=10190)

Moglen, Eben, “Questioning SCO: A Hard Look at Nebulous Claims”, 2003.
(www.osdl.org/newsroom/articles/osdl_eben_moglen_position_paper.html)

Naq, Aditya, “OpenOffice.org Beta 2.0: Trouble for Microsoft”, 9 March 2005.
(www.cooltechzone.com/index.php?option=content&task=view&id=1073)

Oksa, Jukka, and Turunen, Jarno, “Local Community Net: Evaluation Study of the Learning Upper Karelia Project”, 2000.
(www.joensuu.fi/ktl/projsoc/infosoc/upperkar.pdf)

Open Source Risk Management (OSRM), “OSRM Position Paper: Mitigating Linux Patent Risk”, Release Version 1.1, August 2, 2004.
(http://www.osriskmanagement.com/pdf_articles/linuxpatentpaper.pdf)

Porto Alegre, “Porto Alegre integra Rede de Cooperação de Governos Locais pelo Software Livre” 3 June 2004. (www.portoalegre.rs.gov.br/noticias/Default.asp?proj=81&secao=369&m1=21862)

Rosenberg, Ron, “Great Britain and Open Source Software”, January 2002.
(www.stromian.com/Corner/Jan2002.html)

Schmitz, Patrice-Emmanuel and Castiaux, Sébastien, “Pooling Open Source Software: An IDA Feasibility Study – Interchange of Data Between Administrations”, UNISYS report prepared for the European Commission, 2002
(http://europa.eu.int/ISPO/ida/export/files/en/1115.pdf).

Shapiro, Carl, and Hal R. Varian, Information Rules: A Strategic Guide to the Network Economy, Harvard Business School Press, 1999.

UN, UN Global e-Government Readiness Report: Towards Access for Opportunity, Department of Economic and Social Affairs, November 2004.
(www.unpan.org/egovernment4.asp)

Varian, Hal R. and Carl Shapiro, Linux Adoption in the Public Sector: An Economic Analysis, 1 December 2003.
(www.sims.berkeley.edu/~hal/Papers/2004/linux-adoption-in-the-public-sector.pdf)

Wall Street Journal, “Microsoft is Dealt Blow by EU Judge”, Wall Street Journal, 23 December 2004.


Wall Street Journal, “Microsoft Faces EU Deadline: Software Giant has Until 1 June to Resolve Disputes or Face Sanctions”, 23 May 2005.

Wheeler, David A., “Why Open Source Software / Free Software (OSS/FS)? Look at the Numbers!”, 9 May 2005,
(www.dwheeler.com/oss_fs_why.html)

Wildstrom, Stephen H., “A Big Fly in the Open Source Soup”, Business Week Online, August 13, 2004.
(www.businessweek.com/technology/content/aug2004/tc20040813_1107_tc120.htm

Znet, “Norwegian city government switches to open source”, June 15, 2004.
(http://hardware.silicon.com/desktops/0,39024645,39121367,00.htm)