Open Source Software Development in Developing Countries
The HISP Case in Ethiopia

Nils Fredrik Gjerull

September 11, 2006


This thesis investigates free and open source software (FLOSS), and FLOSS in the context of developing countries. The research is based on two action research case studies. Both case studies are done within the Health Information Systems Programme (HISP) network. HISP is a research and development network focusing on promoting effective use of information in the health systems of developing countries.

The first case study was conducted in the Tigray region of Ethiopia. In this case study a team of researchers used action research to configure and adapt District Health Information Software (DHIS) to the local context of the Tigray health system. DHIS is a flexible health information system used to collect routine data from health systems. DHIS is distributed under a FLOSS license. I participated in this team as a software developer.

In the second case study I participated in the development of DHIS 2 which is a total reimplementation of DHIS based on a platform of FLOSS technologies. DHIS 2 is developed using distributed voluntary development and licensed under a FLOSS license. In other words DHIS 2 is developed using a community model commonly used by large scale FLOSS projects. I participated in this project as a FLOSS developer, and I focused on the extensibility of DHIS 2. I created a throwaway prototype of a plug-in framework.

Through this two case studies I investigate FLOSS and how FLOSS can benefit and are benefiting both Ethiopia and the HISP network. I argue that the access to source code facilitates technology transfer/translation of information and communication technologies (ICT). Context dependent software like health information systems need to be adapted to each local context in order to be useful and effective, having access to source code is a life saver in this process.

1 Introduction
 1.1 Two action research case studies
 1.2 Motivation
 1.3 Research domains and objectives
 1.4 Chapter presentation
I  Theory
2 Information Systems Theory
 2.1 Perspectives
 2.2 Participatory design
 2.3 Information Systems and developing countries
II  Method
3 Methods
 3.1 Research methodology
 3.2 My research approach
III  Background
4 Short History of Open Source
 4.1 The early start of programming
 4.2 The three strains of hackerdom
 4.3 Multics, Unix and AT&T
 4.4 The rise of the Internet
 4.5 Free Software Foundation
 4.6 Minix, Linux and Hurd
 4.7 The rise of Open Source into the main stream
5 FLOSS - How does it work?
 5.1 Philosophy and values
 5.2 Development practices
 5.3 Motivation
 5.4 Governance
 5.5 Property, Copyright and Licenses
 5.6 Challenges and constrains of FLOSS
 5.7 FLOSS in developing countries
6 Health Information Systems Programme (HISP)
 6.1 HISP history
 6.2 HISP philosophy, methods and processes
 6.3 Inscription of the HISP approach into DHIS
IV  Empirical Study
7 The Ethiopian Context
 7.1 Demographics
 7.2 Ethiopia, a Land of History
 7.3 Politics
 7.4 ICT in Ethiopia
8 The HISP project in Ethiopia
 8.1 Tigray Demographics
 8.2 The Tigray team
 8.3 EPI-info
 8.4 Adapting DHIS for Tigray
 8.5 Problems with the DHIS software
 8.6 Getting support from the HISP community
 8.7 Being the farench/faranji
9 FLOSS in Ethiopia
 9.1 Economic argument for FLOSS
 9.2 The TRIPS agreement
 9.3 Political support for FLOSS
 9.4 FLOSS usage
 9.5 Ethiopian FLOSS organisations
 9.6 Participation of Ethiopia in FLOSS
 9.7 Analysis of the network in Ethiopia
10 Development of a Plug-in Framework for DHIS 2
 10.1 Why the need for a reimplementation?
 10.2 The community model in DHIS 2
 10.3 My role in the project
 10.4 What motivated me to participate
 10.5 Making the application extensible
 10.6 Interaction with other DHIS 2 developers
 10.7 Interaction with projects we depended on
V  Discussion and Conclusion
11 Discussion
 11.1 FLOSS and Ethiopia
 11.2 FLOSS and HISP
 11.3 Theoretical considerations
12 Conclusion
 12.1 Validity of the research
 12.2 Concluding remarks
 12.3 Possible future research
VI  Appendixes
A Lists
 A.1 List of Acronyms
 A.2 List of Figures
 A.3 List of Tables