Cover Image
close this bookBridge Builders: African Experiences with Information & Communication (BOSTID, 1996, 304 p.)
close this folderCase studies on electronic networking
View the document(introduction...)
View the documentElectronic Networking for the Research Community in Ethiopia
View the documentNetworking in West Africa
View the documentBackground Summary: SatelLife and HealthNet
View the documentMUKLA: Evolution of a Homegrown Network in Uganda
View the documentThe UNIMA Fidonet Network: Computer Networking for Communications in Malawi
View the documentBackground Summary: A Global Computer Network for Change
View the documentBringing the Internet to Zambia


Perhaps no information technology generates as much interest as networking. Bringing electronic networking and Internet capability to a country opens avenues of communications that create hundreds of other opportunities. Computer-based communications provides a means to bridge time and distance to facilitate interpersonal communication. People who have the need or desire to communicate about a particular subject can do so without being either physically present in the same location (as in a conventional meeting), or even available at the same time (as in a telephone conference call or a video teleconference).

Electronic communication is in everyday use in many organizations in Africa; however, getting it into place in a new environment poses many challenges. Complex economic, social, political, and legal factors affect the use of the technology and present barriers to its successful implementation. The telecommunications systems in many African countries are suffering from deteriorating equipment and inadequate investment. Telecommunications costs are high in relation to other costs, and participation in some computer-based communication activities can require scarce foreign exchange.

Still progress is being made. In the last year or two several African countries have gained full Internet connectivity. Others are creating local networks and adding users one by one until they build a user base large enough to sustain an Internet connection. In the meantime, the users and the system operators are learning valuable skills.

The case studies in this section describe the difficult processes that five system operators have gone through in order to build the networking capability in their countries. These five projects are interconnected since the project managers were introduced to the technology by many of the same people and through many of the same megaprojects. Their individual stories are unique, however, because each author overcame different technological, managerial, and infrastructural constraints.

As one author wrote, networking tends to sell itself. The more users you have online' the more users you have waiting to be connected. So, almost one user at a time, these authors are helping to build Africa's information highway.

Electronic Networking for the Research Community in Ethiopia

by Lishan Adam

Lishan Adam has a B. Sc. in Electrical Engineering and a M.Sc. in Computer Information Systems. He is working towards his Ph.D at the University of Sheffield, United Kingdom. He has worked in the information field for the last six years, focusing mainly on technology for accessing information, which includes electronic communication networks. Currently, he is the coordinator for the Capacity Building for Electronic Communications in Africa (CABECA) project. His research interests include electronic communications for grassroots institutions, networking technologies, and techniques for building qualitative information systems and networks.


This case study describes the challenges faced in setting up a research communication network in Ethiopia. It covers the activities I undertook and the constraints I faced while expanding electronic connectivity to a research community. It also discusses the lessons I learned in the process.

Ethiopia is located in the horn of Africa and shares borders with Sudan, Kenya, Somalia, Eritrea, and Djibouti. Addis Ababa, the capital city, is often called "a city for Africa" because it hosts two major organizations: the United Nations Economic Commission for Africa (UNECA) and the Organization of African Unity (OAU). UNECA is the African regional arm of the United Nations Economic and Social Council. UNECA has established and sponsors a number of African institutions to promote all aspects of socioeconomic development of the region.

The Pan African Development Information System (PADIS), one such regional institution under UNECA, was created to build both information and networking capability in African countries. PADIS was instrumental in setting up a national electronic network in Ethiopia and in connecting members of the research community to each other locally and with colleagues worldwide.

There are only a few advanced research institutions and two major universities in Ethiopia. Addis Ababa University, the country's largest university, administers most of the colleges. The Agriculture University of Alemaya, in Harar Province 500 kilometers east of Addis Ababa, is the next largest research center. Health research is undertaken by the medical faculty of Addis Ababa University and at two other research colleges in Gondor and Jimma. The country has several teacher training schools and research centers.

The level of scientific and technological information (STI) and infrastructure in these research centers varies. Some colleges lack the most essential resources, such as paper and ink. Others use computers and connect to the local electronic networking host. The need for STI and communication is great, especially by colleges outside of Addis Ababa.

The Current STI Environment in Ethiopia

Ethiopia is the second poorest country in Africa, having been devastated by long-lasting war, drought, and mismanagement. The national STI environment is generally weak, reflecting the poor economic situation in the country. The transition' economy is too weak to support the population, which is growing at an annual rate of 3.1 percent, with the fertility rate mounting to 7.5 percent. The STI environment has suffered from the poor economic performance during the last twenty years.

All STI centers, including university libraries, have been unable to expand and keep up with demands. There were hardly any books coming to the country during the previous socialist government. According to the head of the Addis Ababa University libraries "most collections are those of the 1960s." Those were the "golden years" of library collection development - when resources were made available to support new acquisitions.

Most colleges are not equipped with computers and other information technology tools. The more remote the college from the capital city, the less it is exposed to computing technology. (See Box 1.) Under-use of existing equipment due to lack of training and low quality of maintenance is also common. Lack of knowledge about different hardware contributes to the under-use of computer accessories.

Computer equipment is generally expensive. A non-error correcting, 2400 bps modem that costs $40.00 in the United States costs the equivalent of $200 in Ethiopia. The local price of accessories is between 400 percent to 1500 percent more than the original cost. A bureaucracy for clearing equipment through customs exasperates most possibilities of getting it through mail order or from friends residing in foreign countries.

One additional problem, unique to the country, is related to the use of Ethiopian script. Ethiopia is one of the oldest countries using its own script. The shortage of good, easy-to-use software that is adaptable to local script is a major problem. Generic native-language software interfaces that allow easier storage and retrieval of textual information in local languages and scripts are not well developed. Modern tools, such as Windows [Byte, 1994], that have features for processing data in all languages simultaneously are not readily available in Ethiopia.

A lack of good training schools and colleges in computing technology has created a chronic shortage of trained personnel in advanced networking, although Addis Ababa University now offers courses in computer applications and system design. Due to the lack of incentives and of an enabling environment, one seldom finds dedicated computer "gurus." The culture for competition and independent effort to solve complex computing problems is not well developed.

Several small computer companies were set up during the transition period but they have not become a source of objective advice in computing technology. Most companies focused on generating quick income through computer applications training. Qualitative and advanced services in networking and local area network (LAN) management are not available at these companies. Due to the lack of a good professional background, techniques such as application development, system integration, and networking are not practiced.

The telecommunication sector is another underdeveloped area in the country. National telecommunication service in Ethiopia is owned by the government. Analog telephone2, digital leased lines, and radio links to rural areas are major services of the national operator. There is no modern communication technology. Rural telecommunications is still operator assisted. Packet switching and ISDN (Integrated Services Digital Network) are under consideration.

A three to five year queue is usually the case for individual households that wish to install a telephone. Business and international organizations may get telephone lines within one to three months when lines are available. Some institutions have to wait for months to get direct lines.

The above national STI environment dictated the type of network technologies we chose to use, as well as the pace at which we could install and apply them. Policy, regulatory, management, and cultural problems were by far the most challenging - even when compared to the technical bottlenecks faced in setting up electronic mail links to the research community. All the same, electronic networking in Ethiopia was started during one of the most rigid governments in Africa.³

BOX 1 Under-use of Equipment

A recent survey conducted prior to linking colleges outside of the capital city with the network showed that one college received a few modems in 1993 and did not know what to do with them. The modems were locked up in cupboards for two years!


Networking activity in the country began under a project funded by International Development Research Centre (IDRC) entitled "Computer Networking in Africa." The aim of the project was to assess the viability of networking between African institutions. The specific objectives of the project in Ethiopia were to:

· improve the exchange of scientific information within Ethiopia by establishing a working, efficient, and reliable electronic network that brings contact with other networks locally, regionally, and internationally;· develop human resources in electronic networking through training, transfer of skills. and university teaching;· develop corps of skilled users in the country through training, troubleshooting, and ongoing technical support;· establish a national scientific network that supports all levels of technologies (Fidonet, UUCP, and TCP/IP, which stands for transmission control protocol/Internet protocol and connotes a full, interactive Internet connection) for various categories of colleges and users under different situations; and· set up an Internet link through development of the user base that justifies the cost of a TCP/IP connection.


The first activity of the project in Ethiopia was initiated in September 1991, when the Ethiopian Science and Technology Commission, which coordinates the national STI networks, participated in the first African workshop on low cost networking. The workshop provided training on low cost communications technology and charted connectivity strategies among research institutions in Africa and those in Ethiopia. Subsequently other academic institutions (universities, colleges and faculties) began to link to the main national host.4

To date the main national host serves over 1,000 users drawn from NGOs, research institutions' government, business, and others. Researchers evolved from network-"hesitant" to power users. For example, a connection was made to the chemistry department of Addis Ababa University at the beginning of 1993. At that time, a professor said openly "this is a waste of time and of computer resources." In 1995, the department became the top user of the national hub and a center for network support to Addis Ababa University users.

Network usage and development activity at the university intensified in October 1993, immediately after a workshop was organized for participants drawn from various departments of the Addis Ababa University. The workshop created conditions for setting up a steering committee to promote networking within the university. The steering committee evolved from a few actors to a full fledged university-wide networking group. It established a networking committee to promote networking in the country and develop connectivity to the Internet.

The research community became the second largest user of the national network (See Table 1.) It constitutes 14 percent of PADIS installed sites and 26.2 percent of total local users.

TABLE 1 Composition of PADISnet Users (April5 1995)

User Category

Number of installed sites

Percent of total sites

Average number of users

Total number of users


Research and academic
















































Source: PADISnet registration forms.

In addition to developing a sound user base that would eventually justify the upgrade to a full Internet connection, one of the major activities of the project at the beginning involved experimenting with a mix of technologies to select the hardware that was most appropriate for working with poor telephone lines. We conducted the experiments using a mix of technologies to arrive at appropriate solutions. We evaluated the following four connection techniques during the process:

· long distance dial-in to UNIX hosts in North America;· UUCP link through a leased line by running a gateway software;· packet-radio connection to health institutions; and· Fidonet based mail connection to Internet using GnFido gateway in London.

Our first attempt to link researchers in the country used long distance dial-in to UNIX hosts in Canada and the United States. Access to large academic networks such as BITNET was the main interest at the time. A direct dial-in to a UNIX server at Carleton University in Canada was made every day to collect and send messages to the research community. Rudimentary terminal access programs such as Procomm were used to link to the Carleton server. This turned out to be one of the most frustrating means of making connections as the lines broke at almost every trial. Successful connections were so rare that the connection was discontinued after a few months. The cost of telecommunications mounted every month. The cost was made up of:

· long distance telephone charges;· low speed connections (successful connections were at 300 bps); and· additional procedures required as a result of long distance dialing and line breaks (i.e., transferring files several times after line breaks; login procedures to packet switching network in the U.S.).

Paying enormous bills every month for long distance connections forced PADISnet to switch to a more cost effective site. While exploring sites, a link using the United Nations Alternate Voice and Data (AVD) leased line to the Institute of Global Communications (IGC) became cheaper and more attractive. The IGC link enabled us to convert local Fidonet messages into WCP format before leaving the PADIS host via a gateway software. This created conditions for developing local expertise in running gateway software and in experimenting with WCP packets. However, since the line was devoted to UN communications (voice and fax)? it became slow and inadequate for data communications. PADISnet was given only a 15 minute window to send and receive WCP messages. This was an inconvenience and resulted in the final suspension of the connection after ten months of operation.

Another technology used for STI connectivity in Ethiopia was a HealthNet link. PADIS helped to establish this link and to secure its ground station license. The HealthNet ground station, which links to a low earth-orbiting (LEO) satellite operated by SatelLife, was licensed in April 1994 - after 18 months of negotiation with the national telecommunications operator. The ground station was installed at a teaching hospital, one of the largest in the country. This connection was not used as effectively as hoped due to the following difficulties:

· Unreliable power supply. Fluctuation in electric current at the site of the ground station led the system to crash from time to time.· Very complex software. A home-written Fidonet software made configuration difficult. Lack of documentation and troubleshooting tools heightened the problem.· Synchronization of systems. The HealthNet system is composed of a receiver, transmitter, modem, antenna, microcomputer, satellite, and software. Failure in one or more of these systems resulted in a complete crash of the node from time to time.· Lack of an overall plan for user support. The national HealthNet node did not come up with a plan for introducing the technology, connectivity, or regular user support.· Low bandwidth. The uplink and downlink speed of the modem and satellite pass time was limited. With growing user demand for band width intensive documents, such as reports with graphics or images, the system became inadequate.· Marketing. The failure to develop an expansion plan and lack of marketing of the services limited its use in the country.

A direct Fidonet connection to the Association for Progressive Communications (see description on page 189) in London using public telephone lines remained the most attractive of all the above techniques. This connection is currently our main link to the global networks and provides all store-and-forward Internet services. Due to a lack of knowledge about the potential of store-and-forward technology, researchers mainly use the email and bulletin board services of the linkage. Some students and a few researchers are exploring other potential services such as:

· searching the PADIS database using email query;· accessing the Hornet bulletin board system (BBS);· batch Internet services such as ftp, gopher, and archie through mail; and· conferences and Fidonet echo mail.

The Hornet BBS is perhaps the most unique application of low cost technology in Ethiopia for information exchange. Hornet offers an introduction to the potentials of electronic information systems end provides a reservoir of background information and discussion on the Horn of Africa. [Parker, 1994] The bulletin board offers:

· Fidonet conferences (or echoes) to as wide an audience as possible, with no on-site software set-up necessary;· a mechanism to exchange information and skills among users of the PADIS system; and· a central store of information, files, help, useful Internet addresses, and so on.

Simplicity and minimum hardware requirements on the users' side make the BBS more popular. Prompted by this offering, a number of research institutions, including the Addis Ababa University and the Ethiopian Science and Technology Commission, are setting up a scientific bulletin board system. Local bulletin board systems connected to information reservoirs, such as databases and tools on CDROM, are found to be the most useful tools for STI networking under the local infrastructure.

During the last two years, increasing user demands for real time connections, the need for joining the Global Information Infrastructure (GII), and congestion of the host telephone line forced the national node to look for a better technology. At the time of this writing, the node is exploring full Internet connectivity and/or other intermediate solutions. The effort to get a full Internet (TCP/IP) connectivity is hampered by three major problems:

· high cost of leased lines;· institutional and national constraints; and · financial resources to cover implementation and running costs.

Leased Line Costs

The national telecom operator provides leased lines at a very high tariff.
Table 2 shows approximate monthly charges for international leased lines to the United States and South Africa.8

TABLE 2 Cost of Leased Line from Ethiopia to Selected Countries

Type of circuit



19.2 kbits analog



19.2 kbits analog

South Africa


64 kbits digital



64 kbits digital

South Africa


Source: Action plan document of a committee for Bringing Internet To Ethiopia (BITE).

National Infrastructure and Political Constraints

The low level of infrastructure as well as policy-specific problems continue to hamper the progress of setting up full connectivity to global networks in Ethiopia. Changing the attitude of the government and of the PT&T (the national telecommunication provider) remains one of the most challenging tasks. Some of the national infrastructure and political constraints we face include:

· competition and unwillingness of the national telecom operator to endorse operation of national cooperative or private nodes,· low participation of government institutions in overall network implementation;· lack of knowledge of electronic communication by government officials that results in delays to approve research network (see Box 2);· education structure and low value accorded to information management, transmission, exchange and use, and· lack of a critical mass of trained UNIX experts or "gurus."

Despite the above problems, the project continued to work towards improving network access to researchers. The bottom-up approach for building a network (starting from users, soliciting institutional support, and then approaching the government) helped us bypass an often difficult bureaucratic structure in the country.

Eventually, the increase in the availability of local "hand holders" and troubleshooters encouraged expansion of the network. The local troubleshooters created institutional champions and then the institutional champions helped users get around configuration problems. In time, we motivated others to get connected.

Reliability of the network has also contributed towards the diffusion of networking technology in Ethiopia. Network reliability is a function of good systems operators (sysops) running the node, advanced computing technology tools, and reliable telecommunication infrastructure. PADIS' effort to make the system reliable by hiring additional system operators9 and acquiring the latest technologies favored the situation.

Other factors that contribute to further network expansion include:

· the availability of dedicated local champions and· the availability of simple network learning tools, such as the Hornet Bulletin Board System.

PADIS management played a major role in leading and supporting networking in the country. The availability of information10 at PADIS provided incentive for several users to seek connection to the network. PADIS made its resources available via the network to satisfy users who wanted to go beyond electronic mail. Individual willingness - especially from top government officials - to embrace existing technology and their demand for better services are perhaps the major motivating factors for expansion to a full TCP/IP. Influenced by their former contacts (usually at universities or NGOs in developed countries ) researchers who are not connected demand quicker and better connectivity. Contact requirements and realization of benefits of networking compelled several users to demand email services even under difficult circumstances. What are the major benefits that prompted researchers to look for electronic networking?

BOX 2 HealthNet in Ethiopia

Often government policies work against the introduction of new technologies. Lack of understanding can load to delays and other frustrations. For example, it was a painful process to secure a HealthNet ground station license for Ethiopia. The process took over 18 months. Simply clearing the ground station equipment from customs took over a month. An engineer from SatelLife, who could have been used for training and set-up, ended up running between the national telecom operator's office and the Ministry of Health. The license was approved with a traffic compensation fee of $350 per month. The ground station, which was donated as national asset to aid in the transmission of health information, was instead treated as a CNN hub. The word ´'ground station" was found to be misleading for technicians who did not bother to know what a packet radio connection to low earth orbit satellite was. The traffic compensation fee was waived six months later.


Reduced Communication Costs

The most outstanding benefit of electronic networking to the research community in Ethiopia was that of bringing a new tool for reducing high communications costs. Ethiopia is a country with one of the lowest pay scales in the world. The following vital figures shows the situation.

Professor or researcher salary per month$ 162.00

Cost of:

1 page fax to UK


5 page fax to UK


3 minute telephone call to UK


5 minute telephone call to UK


Electronic networking brought a considerable amount of communications cost savings compared to fax, telex, telephone, and courier service. Though email costs are complex and difficult to quantify, we do know that networking offers a far cheaper option to all existing transmission media. (See Table 3.) The following tables and computations show a considerable comparative advantage of email over a fax - without involving costs of other elements such as hardware, or training.

Using compression tools, such as PKZIP, an average document can be compressed 2.5 times. The amount of bytes per minute after compression would be:

Compression factor 2.5

Total bytes per minute 2.5 x 103342 = 258357

High speed modems, with good error correction, synchronized to the receiving modem, can also increase the amount of bytes transmitted per minute. Using 28.8 Kbps modem the amount of bytes per minute would be: bytes per minute of 28.8 kbps modem (1.5 x 285357= 387535)

Using the above figures, we can calculate the cost per kilobyte of different means of communication.

TABLE 3 Raw data used in computation

Date and Time

Number of bytes transferred

Time taken


10 July at 6:00




11 July at 6:00




11 July at 1:00




11 July at 5:24




Source: Log files of the PADISnet host July 1995.

In addition to the reduction of communications costs, email made the following contributions to the research community in Ethiopia [Adam, 1994]:

· It facilitated the organization of international seminars, joint authorship, and the execution of joint projects that otherwise would have been expensive and time consuming. According to the Director of the Institute of Ethiopian Studies, Dr. Bahru Zewde, "Global conferences in Ethiopia and those on Ethiopia in 1994 were cost effective and successful because of email connections."· It inspired the introduction of standard computing operating systems such as UNIX. Recently the School of Information Science for Africa (SISA) purchased a UNIX server as a networking backbone. The Mathematics Department of Addis Ababa University received a UNIX computer in connection with its networking effort from Ethiopians residing in North America. Deans and department chairs were forced to plan to "wire" their offices and create local area networks (LANs) to cope with an increasing demand from researchers to have independent accounts on work stations.· It facilitated a switch from the hierarchical model of organizational procedures to the informal interaction among researchers in the country and those worldwide. User feedback indicates that delays in signing regular correspondence were eliminated upon arrival of email. (See Table 4.)

Email also brought about the reunification of the research communities in Ethiopia with those elsewhere. It improved family and informal ties between Ethiopians residing overseas and those at home."

Improved Training Options

Electronic networking facilitated the availability of more national and institutional trainers. It created the expansion of knowledge on the use and benefits of networking and TCP/IP connectivity. The methodology used in training contributed to improved skills in the country. The national host uses the following techniques to train users.

TABLE 4 Cost Comparison Between Fax Message and Electronic Mail Between Addis Ababa and London


Average kilobytes/per minute


Cost ratio to a fax





Email using 14.4 modem




Email using 28.8 modem





Users are introduced to key features of networking. This covers basics of local networking? what it involves, cost, and types of services.

Site configuration and training

Site training is meant to connect researchers to the network. It involves configuring the hardware, setting up telephone lines, providing hands-on training on system usage, and training in basics of troubleshooting.

Ongoing assistance

The host provides continuous online assistance to improve users skills in dealing with different utilities and to work on improved and cost efficient techniques.


This is used both as marketing tool for the national node and for advanced training. Experience indicated that initial introductory workshops overload users with new concepts. Successful workshops could only be held after usage of the networks. Advanced workshops were found to be useful in promoting connectivity.

Improved Skills

Another considerable impact of the network was an increase and improvement of the skills at the main national node. We developed skills in managing networks, dealing with administrative problems, developing techniques to respond to users' problems, managing system malfunctions, developing tools and guidelines, and gathering data. The experience we gained was immense.

Improved Access from Rural Areas

The spread of the network to the rural area is one of the most significant achievements of the project. Despite technical problems, such as unreliable connections and more focus on the capital city, a number of users were able to link to the PADISnet from outside of Addis Ababa. NGOs with field offices outside of the capital found networking a convenient tool for sending logistical information and data on field situations.

Doctors with no telephone connections were able to use networking. Medical doctors from Medicins Sans Frontiers, headquartered in Belgium, who were working 800 kilometers from Addis Ababa were able to send messages and data to their correspondents all over the world. Lacking telephones at their field site, they send messages via small airplane to Addis Ababa for uploading to their host to resend to PADIS. All international messages received are sent back by the same plane the following day. Other mobile field researchers from the same location were able to link their modems to the nearest city where phones are available.

A number of colleges outside of Addis Ababa are discovering the value of networking. They face double isolation - from both the developed world and the country's capital where most of the activities and decisions take place. Everything in the country is hierarchical and centralized. Decisions at higher levels (at the capital) take a long time to reach institutions in remote areas. Networking can reduce the isolation and make communication more efficient.

Connectivity to the rural areas facilitated the coordination of resources and the diffusion of technology. Connectivity at the Mekelle College (710 kilometers from Addis) recently stimulated the establishment of a LAN for connecting individual professors to the global network. A project run by the Ethiopian Science and Technology Commission to link over thirty-six colleges outside of the capital stimulated further international cooperation between the colleges and other universities worldwide. The Commission intends to "wire" all colleges and high schools in the country while working on a full TCP/IP connection at the capital.


Lesson 1

The resources required for achieving access to knowledge and wider national coverage are a fraction of the benefits gained by building networks. Coordinating activities at a national level assists in realizing cost effectiveness and efficiency. Networks can subsidize themselves. Lack of an infrastructure should not be seen as a major bottleneck to networking.

There are two major goals of electronic networking of the research community:

· wider coverage of users by reaching many scientists, researchers, students etc.; and· improving access to knowledge by bringing global information resources closer to the scientific community.

Attaining these two goals needed institutional and infrastructure readiness from the outset. When these two goals were first proposed to managers and policy makers their initial reaction was that "there are no resources" and "we can not do it under our telecommunications situation"!

Lesson 2

Training is the main element in infrastructure development. Do not start training on DOS based networking at universities, start with UNIX' which I believe to be the multi-user system of choice for networking universities.

Wider national coverage requires local infrastructure development. Local infrastructure means not only the physical set-up of machines and software, but also building a sense of awareness and increasing the participation of users. Training should cover a wide range of groups at different levels and should introduce all standard technologies such as UNIX to the academic community from the very beginning. Starting training in the DOS environment limits universities' capacity to move quickly on to sophisticated techniques. Using DOS for networking universities instead of UNIX is like using typewriters instead of word processing. Good documentation saves time.

Lesson 3

Build as many local area networks as possible. A LAN is a building block for Internet connectivity.

Capacity building in universities requires setting up of a university-wide network for inter-connecting many LANs, serving individual departments, libraries, and administrative offices, and offering the connections to the Internet through gateways. LANs are building blocks for good research networks.

Lesson 4

Good network management practices should begin when the number of users is very small. Good network administration can prevent user and donor frustration.

The professional management and administration of a network should begin as soon as you have two users. A well-run network needs a good billing structure, well developed methods for gathering data on the type and nature of users, and mechanisms to backup log files, document problems, and track progress. Put these systems in place before the number of users grows to an unmanageable size.

Lesson 5

Networks should be interactive on both an inbound and outbound basis.

Interactive sysops are the best assets of good networks. Inbound interaction with local users and outbound links to other networkers worldwide assist in sharing knowledge, resources, and time. Documenting tips and passing them on to users helps networks to expand.

Lesson 6

Using software with simpler user interface is important.

One of the major components in infrastructure building is the provision of simple tools to users. Users should get uncomplicated tools. Using software with good user interfaces reduces frustration and improves users' interests in networks. The community of Macintosh users should not be neglected.

Lesson 7

Do not focus only on global information sources. Assist in building local capacity in STI exchange.

Improved knowledge access can be achieved by creating a collaborative atmosphere within a country and by connecting to global networks. National STI sources should be strengthened to enable local exchange of information and collaboration. Networking should be built with maximum participation of local STI providers and users.

Lesson 8

Do not promote the concept of subsidy. Promote the concept of "pay a small amount to keep your network healthy."

Networks should sustain themselves. Income-generating should start at the beginning. Users should be told to pay for services. Subsidizing researchers is important; researchers should, however, know the costs and participate in paying for services from the beginning. The payment could come from university budgets or projects.


Scientific and technological information in Africa needs to be strengthened. African input to global information resources emerges from the local STI institutions. Databases and low cost networks will ultimately serve African information needs. This is especially true for local loops where telecommunications and other infrastructure continue to be problematic. The international link itself is also inadequate. Except in South Africa, the other eleven African countries with Internet connectivity have a maximum speed of (usually) 19.2 Kbps. Reaching the global community and providing African STI to global users need efforts in strengthening the local capabilities both in networking and in information generation, management and dissemination, while upgrading the bandwidth for international connectivity.

Quality, reliability, and sustainability of African STI and networks is becoming another critical focus. Surveys and experience indicate that databases on STI lack good quality and are mostly unreliable. Database and information systems development in the region need to build quality tools and strategies for sustainability of African STI databases. The quality and sustainability of low cost electronic networks serving STI in Africa should also be improved.

Electronic communication will become a major tool for distributing global information to Africa. This does not mean African researchers need and can afford full TCP/IP connectivity. The key questions become: Is the African STI environment ready for full Internet connectivity? Will communication be two-way (from Africa to developed world and vice versa)? What will be the overall implications of Internet for a few in the capital cities versus those who are isolated due to a poor infrastructure and who need to break the isolation?

The answers to these questions remain a challenge to STI providers in Africa. African STI institutions need Internet connectivity. Internet connectivity means full participation of local providers and STI institutions in information generation and usage on the global networks. The current trend in most African countries focuses on Internet "proximity": bringing Internet closer to the users with no expansion plans for wider connectivity using low bandwidth links. Encouraging STI institutions to participate in information exchange will become the next challenge after bringing an Internet link to a country.

The situation in Ethiopia, which is often similar to that in other countries in Africa, indicates that careful analysis of the national STI environment should be made to design the most appropriate solution to improve overall research capacity in a country.

Efforts should begin by building the capacity of national science and technologv institutions, such as research institutions, colleges, and small STI service providers. They should be encouraged and trained to manage information and use global information sources. Interim innovative services, such as connections to major databases or Internet services via local Bulletin Board Systems, should be available. Users should be made fully aware of the services available to them even through existing low-cost, store-and-forward technologies.

One of the major tasks facing African STI networks is a lack of knowledge of what is happening in the same region, the same country, or even the same institution. We should develop directories, dissemination tools, and guides on STI networking efforts. We should disseminate via our networks STI meta-information (information on scientific and technology information). Without a solid regional STI base, connectivity alone cannot serve its purpose.


1. Ethiopia's longest crisis and war years were between 1974 and 1991. During the transition period (1991-1995) the government attempted to eliminate centralized and rigid policies to make favorable conditions for economic growth.

2. International telephone linkage to major cities in the world is relatively good. Users can connect with up to 28 Kbps to London without difficulty. A 300 to 900 CPS connection can be achieved using error-correcting modems. It is often easier to telephone internationally than across regional cities. The call from Addis to Mekelle is often compared to that of Addis to London. Cost of international communication is also high. The tariff was raised by 75% in May 1994.

3. The Ethiopian socialist government which had fallen apart in 1991 promoted strict policies on communications. Modems were not allowed due to reasons related to national security. Any move towards opening up communication was looked upon as a "war against socialism."

4. The national host was set up at PADIS and the network is called PADISnet. It is open to all categories of users on a fee-for-use basis.

5. Since April 1995, the number of installed sites has increased by an average of 5 per week, thereby nearly doubling since that time.

6. WCP is a UNIX based store-and-forward technology. Developments were fast and based on international cooperation between programmers. It is a suitable start up technology for the academic community.

7. Fidonet is a store-and-forward networking technology by which computers poll each other to exchange email and conference messages. Ability to work over poor telephone lines and software that enables off-line working make Fidonet attractive to NGOs and African institutions with small budgets.

8. Leased line costs to South Africa are almost 1.5 times those of links to the United States. There is double hop to reach South Africa.

9. In June 1995, PADIS used an intern - an Ethiopian PhD student from Columbia University - to look at university-wide networking and to improve the reliability of local systems.

10. The PADIS information system consists of bibliographic databases on African socioeconomic development, databases on African experts, on-going research projects in Africa, research institutions in Africa, and other numerous complementary and specific databases on Africa.

11. The author received numerous requests from Ethiopians studying abroad to link to their families at home. Families separated during the war and those who fled economic problems are beginning to use networking to reunite with their families. Ethiopian researchers abroad who discovered the existence of the linkage are connecting back with their former teachers and institutions.

12. Fax costs vary based on text type. Graphics can be 100K or more. Regular fax message can be 10 to 70K in size. An average size was taken for this calculation.

13. A 28.8 Kbps was taken to be 1.5 times faster than a 14.4 Kbps modem.


Adam, Lishan. ( 1994) Sustainable Academic Networking in Africa: System Operator's Perspective paper published for Advancement of Science,Accra, Ghana.
Beyond Windows: Globalization of Windows. Byte June 1994, pp. 177-183.
International Association of Universities ( 1991) University Based Critical Mass System for Information Technology.
National Research Council ( 1991) Status of Scientific and Technological Information Systems and Services in Selected African Countries National Academy Press, Washington, D.C.
Parker, Ben. (1994) The Hornet: The Horn of Africa's Electronic Information Exchange. Addis Ababa, Ethiopia

Networking in West Africa

by Moussa Fall

Moussa Fall is network manager for ENDA-Dakar node in Dakar, Senegal. He also works for the CABECA project and through that project has installed electronic mail nodes in Morocco, Chad, and Mali. He has plans for working in Mauritania, the Gambia, Guinea, Guinea-Bissau, Niger, Togo, and Benin. He was an administrative assistant at ENDA before getting into the networking field in 1991. He asks that his case study be dedicated to "Pape, gone so early, and to Touti and the whole tribe for their help and support."

"Networking means connecting people to people and people to information; it does not mean connecting computers to computers." Wendy D. White - Growing the Internet in Africa, /nternet Society News, 1994, Vol. 3, No. 2, p. 28.


I work for an organization called Environment and Development Action in the Third World (ENDA). It was founded in 1972 in order to:

· work with grassroots groups on the basis of their needs and objectives;· contribute to the search for alternative development possibilities at all levels as well as to the various kinds of training programs that will make this development possible; and· contribute to intellectuals' and trained personnel's involvement in the setting up and implementation of development programs in the service of the largest number of people possible.

ENDA has also had a leading role in various global networks dedicated to habitat, energy, street youth, pesticides, and other issues that have an environmental component. ENDA works in all parts of the Third World but the headquarters, where I work, is in Dakar, Senegal, in West Africa. Senegal is a francophone country.

Since 1975, my role at ENDA, among other administrative tasks, has consisted of organizing seminars, conferences, and training sessions. These outreach tasks have helped to put me in contact with many people and I believe that I started networking from there.

I first met Doug Rigby1 in 1991 , at an Interdoc meeting in Epe (Netherlands), at which there was an introduction to electronic networking. He spent considerable time talking to me about Fidonet technology and the NGOnet project. (See Box 1.) Back in Dakar, I pushed my boss to get an email account with a European private email provider. We accessed this account through the Senegalese PTT X.25 and we extensively used it for email and for the fax service.

BOX 1 The NGOnet Project

The NGOnet project was instrumental in bringing networking to Africa. It started as a project of the Environment Liaison Centre International (ELCI) in Nairobi where a Fidonet bulletin board system had been set up to provide a conduit for electronic mail traffic in the region and to NGOs worldwide. This was done using a highspeed modem to make daily calls to the GreenNet Fidonet (GnFido) gateway in London.

The project was based on a survey that found that there were significant numbers of non-governmental organizations that had computers but were not using them for electronic mail. To provide NGOs with the cheapest access, emphasis was placed on establishing a series of hosts with high speed modems distributed throughout Africa. These then provided NGOs with local support and a local call to connect to the global electronic network. Four prototype hosts were set up, one for each region of Africa - ELCI in Nairobi, MANGO in Harare, ENDA in Dakar, and ENDA-Arabe in Tunis.


In June 1992, the NGOnet project organized a one week workshop in Dakar to introduce Fidonet to the non-governmental organization (NGO) community. Doug Rigby was the organizer and, of course, I was among the participants. To encourage NGO networking, the project donated one PC 386 for the node and four modems. ENDA was chosen to host the node because of its commitment to immediately acquire a dedicated phone line and to allow me to give part of my time to the networking project.

Here, I have to confess that I had no DOS skills - ENDA only used Macs. But, because I was really interested in this new technology, I bought books on DOS and computing in general and started learning. When you are very motivated you can "move mountains". . .

For three months, the four NGOs that had received training and modems started communicating just among themselves. There was no international traffic! Then, in August 1992, the network received another donation from the NGOnet project - a high speed modem (Telebit Trailblazer 2500) - which made us ready to connect to the rest of the world.

Fidonet - A Grassroots Network

My introduction to Fidonet was so important that I should take time here to describe this form of networking in more detail. Fidonet is a grassroots electronic communications community that has been hard at work for over a decade devising ever cheaper and increasingly sophisticated tools to serve its needs through dialup, store-and-forward, and modem-based connections. Fidonet technology has proven to be a powerful do-it-yourself tool for establishing initial footholds into the world of electronic communications. It offers users an affordable option even if they lack institutional affiliations, financial resources, or are located in a country where the nearest electronic communications link requires an expensive dialup call over international phone lines.

Because Fidonet technology emerged in an environment where individuals operated each system independently and covered their own costs for phone calls and equipment, it had to be very flexible, decentralized, and designed to operate inexpensively with standard modems and microcomputers connected over ordinary phone lines. The "handshaking" and file transfer protocols built into all Fidonet-compatible software incorporate compression, error correction, and error recovery capabilities that squeeze as much data as possible into the shortest transmission time that the hardware will allow. Instead of using packet-switching, these independent systems establish gateways with larger, international electronic mail systems using high speed modems. At regular intervals, the independent systems dial into the larger systems to swap incoming and outgoing messages. This approach keeps down the cost of international calls without requiring sophisticated computer equipment.

Fidonet messages are sent along a hierarchy. At the top are five very broad geographic zones. Africa is Zone number 5. The zones are divided into regions, then into hosts, then into hubs, and then into nodes. Just like all the levels above them in the chain, nodes offer local email, pass new messages to and from the hub, and collect messages to and from the point - the lowest level of the Fidonet hierarchy. Point operators have systems that are configured with all the software necessary to call (or poll) the node and upload and download messages and files whenever it is convenient. The computer can do all the work automatically, making the calls into the central system at a time when the lines may have less traffic or the costs may be lower.

Fidonet is a communications technology that many consider to be less advanced and, therefore, less useful than other technologies. It does not offer all of the sophistication that other, more costly systems do. However, as you have electronic mail capabilities, you can access important Internet tools like file transfer protocol (FTP), Gophermail, World Wide Web (WWW), Veronica, Wide-Area Information Servers (WAIS), and listservs - even if you do not get the results immediately as in a direct Internet connection. (See Box 2.) Fidonet technology has limited expansion capabilities, insofar as it will always remain a store-and-forward, modem-based network. It lacks the capability for online information retrieval, database searches, remote-login, and remote-execution that other systems offer. However' while the expansion of more advanced computer networking technologies iS often constrained by prohibitively high costs and inadequate telecommunications infrastructure, Fidonet technology is not.

The Association for Progressive Communications (APC), a global network dedicated to NGOs, was involved in the NGOnet project.2 The APC was our natural partner. Through theAPC London-based member, GreenNet, we connect through their Fidonet gateway (GnFido) to the rest of the world. GnFido serves as a gateway for small Fidonet hosts not only from Africa, but also from Asia and Latin America.

Through this system we have been able to use email, download conferences, and use the GreenNet fax service, which costs very little compared to the national PTT charges. The computer conferences received were on various topics: environment, development, health, AIDS, and so on. We also followed the United Nations Environment Programme's preparatory conferences and the NGO discussions regarding these conferences. ENDA now runs a private conference for its own use, linking its various offices around the world.

BOX 2 Internet Navigational Tools

TELNET- Telnet, the internet standard protocol for remote terminal connections, is used for logging into and searching other computers connected to the Internet.
FTP - FTP stands for File Transfer Protocol and allows users to exchange files between their workstations and remote computers connected to the Internet.
GOPHER - Gopher, a software program for browsing and information retrieval, provides a menu-driven interface that shows what is available on a server. The user burrows through a set of "nested'' menus to get closer and closer to a specific topic. Gopher collects information scattered across many computers and presents it on the same menu.
VERONICA - Veronica is an indexer that can query every gopher on the gopher system to search for a keyword or phrase in a menu title. It then gives the address of all the menus with those key words.
WAIS - WAIS, Wide Area Information Server, searches the full text of a document to look for specified key words. WAIS accepts commands in plain English, processes them at the user level, and relays the processed information from the user to the selected databases.
WORLD WIDE WEB (WWW) - WWW is a tool for working with collections of data, or databases. It is a hypertext based system that provides access to a variety of files and information. Hypertext allows a user doing research on one document to jump to a related item in another document through hypertext links. In WWW, each document contains highlighted items for which additional information is available. The additional information is contained in another document that is displayed when the user selects the highlighted item. With appropriate software, such as Netscape, the user can view not just text but pictures, sound files, and video.


How Users Are Trained

Users are trained either on an individual or group basis. After only one hour, the trainees are able to send messages and check their mailboxes. Usually training is done gradually. It is better to have three sessions of one hour each than just one session of three hours. During the second session, users are shown how to encode files to be sent and how to decode the received ones. The last session teaches them how to subscribe to conferences, how to contribute to a conference, and how to quit a conference.

Because the point software we use is user-friendly, after the training sessions, users need very little help. This they can request by sending a message to the postmaster who will respond to them as simply as possible.

For organizations using desktop computers, training is organized on site. For individuals or organizations using laptops, we ask them to come to the ENDA offices where things will be easier in case there is a problem.

Development of the Network and Perspectives

Accessing our network is relatively cheap3. Since the host is a non-profit, non-governmental organization, it does not intend to make a profit from providing networking services.

Our active users are our best advertisers! They help us "sell" the benefits of networking. Presently the ENDA-DAK node has over 110 users - coming from national and international organizations, individuals, government agencies, universities, and even the private sector. Our users come not only from Senegal, but also from other neighboring countries, including the Gambia, Mali, and Burkina Faso. The demand is increasing locally and also in all the neighboring countries.

Connecting Problems

During the period of April-May-June 1995, we experienced problems in our connections with GnFido in London. The Senegalese phone system was not working properly and users complained that their mail had not been delivered in time. We tried several modems and, paradoxically, the modem with the worst reputation behaved better than all those of a supposed higher quality. Now the situation is better and we are presently using a Zyxel modem.

Regional Networking

In 1994, I was a consultant to the United Nations Environment Programme, Global Environment Facility Project, which consisted of linking national teams doing research on greenhouse gases in Senegal, the Gambia, Morocco, Uganda, Tanzania, Nigeria, and Kenya. My part of the work consisted of developing teaching tools for national teams in Senegal, Gambia, and Morocco. I also did system installation, provided training, and helped link these teams to the conference set up for them. They were thus able to exchange data on their research, exchange views on methodology and software used, and share experiences. It should be noted that this conference was bilingual and I did most of the translation into [French.

Since 1994, I have also been involved in the PADIS/IDRC CABECA project (see Box 3), which tries to give access to countries where there is no connectivity at all or where the existing infrastructure is too expensive for the NGO community to afford. An agreement between the ENDA Executive Secretary and Nancy Hafkin, who is the head of PADIS and the CABECA project, allowed me to work full time in electronic networking.

This agreement was easier to reach than might be expected. I should point out here that first our two institutions shared a common goal: to give electronic access to communities at the grassroots level. This facilitated the collaboration. I should also note that Nancy Hafkin's devotion to African networking and her confidence in this continent and its human resources made this institutional collaboration possible.

Through the CABECA project, I installed nodes in Rabat, in Chad, and in Mali. Another CABECA goal is to train node operators who will themselves become trainers.

In the CABECA project, we are trying to implement a regional network that is badly needed in francophone West Africa. Among these countries the telephone exchange is relatively good and the costs are relatively low. The idea is to have a regional hub - ENDA-DAK - which will be the focal point for the region, through which all regional mail shall be routed. With this infrastructure, there is no need for mail addressed to the neighboring countries to be routed through European gateways. Putting our resources together will make regional networking more efficient and will reduce our costs for international traffic. We believe that once this infrastructure is set up, the region will be a zone for intensive information exchange. Organizations in the region have a great deal of information to share' the countries have the same language, the same currency, participate in the same economic organizations, and often are in the same ecological zone. Common interests and experiences should encourage the growth of networking in West Africa.

Some international organizations understand this very well and are discussing with us how to connect their projects or field offices to this low cost infrastructure. We are presently discussing with HealthNet the best way to for Senegal and other countries in the region to collaborate with this project.

Methodology for Sensitizing

In each country visited under the CABECA project, we organized a half-day workshop to sensitize NGOs and individuals to electronic networking. Some workshops, such as the one in Chad, brought together as many as 35 people. After a preliminary introduction to the technology, using a phone simulator I always carry with me, we split participants into three groups, each one sitting around a computer. That way, each group can prepare one or two messages addressed to the other two groups and then, through modems and the phone simulator, exchange mail. They can see how fast the message can go and also the other possibilities of the system.

After this hands-on training, all groups get together for a final discussion on how this technology can serve the national community, the advantages compared to technologies such as fax or even conventional mail. We always focus first on how this technology can serve information sharing and exchange within the country before opening up to regional and international networks.

Here we must not forget the context in which we work in many African countries: there are poor phone lines and electrical shortages and outages. In some of these countries, you can get a dial-tone only after working hours. In some others, it is even worse, as not only do you have to wait a long time for a dial tone but, once you get it, you are not sure your international call will succeed. In one of these countries you can dial directly only to France!

Difficulties in Sensitizing

There are people who are allergic to new technologies. In one of the countries I visited for sensitizing purposes, the Government computing department is now attached to a national service. During our meetings, the head of that service had been completely against the introduction of this new technology, despite the fact that minister and all his colleagues agreed that their country should be part of the process. They all wanted to join the growing networks and reap benefits from them. Yet' the head of the service was able to block progress. Every month, his office sends data concerning the country to their subscribers who pay for the information. The documents are sent by regular mail. After many demonstrations and taking into consideration the volume of information sent monthly to North America and Europe, that person still retains his original position - which makes his colleagues quite unhappy.

BOX 3 Capacity Building for Electronic Communication in Africa

Funded by a grant from the Canadian International Development Research Centre (IDRC), the project on Capacity Building for Electronic Communication in Africa (CABECA) builds on the experience gained over the last four years from a number of IDRC-funded electronic networking projects in Africa. A public corporation created by the Parliament of Canada to support research designed to adapt science and technology to the needs of developing countries, IDRC has taken the lead in financing African electronic networking initiatives with the aim of demonstrating that the technology for deploying electronic connections throughout Africa is readily available and can be implemented at relatively low cost, taking into account the economic, social and political difficulties the region faces. The CABECA project design attempts to address the problems of the African region that have isolated it from the international networking phenomenon.

CABECA's overall objective is to provide technical assistance to bring about sustainable computer-based networking in Africa, at an affordable cost, accessible to a wide variety of users from both the private and public sectors. To build African capacity for computer networking, it will train a corps of systems operators who can train others in their area and offer continuing support to fledgling users to ensure the sustainability of national nodes with connections to international networks. The project's aim is to offer inexpensive and easy access to local and international information services on systems run by local operators and sustained by revenue from users. They will be able to exchange electronic mail worldwide at a fraction of the cost of fax or telex; they will also have access to conference mail, file transfer and databases. Efforts will be made to facilitate African connectivity to the expanding range of Internet information services.


Some of our users told us that they were able to get rid of the international line they used to send faxes after they discovered our fax service. Now their fax machine is mainly receiving faxes from their correspondents who do not have email and this has dramatically reduced their communications costs.

NGOs have been changing their way of communicating - especially with their northern partners. They are not only sending faxes but they send and receive email messages and files. Some of them have been subscribing to listservs while the others, because of language problems, wait for the opportunity to have access to conferences in French.

NGOs have also benefited from the technology by being better able to participate in the preparations for such events as the United Nations environment conference. Through the NGO forum, even small and isolated organizations can be heard. ENDA itself, during the UN Environment Conference, issued a daily newspaper in French giving accounts and comments on what was happening. The newspaper was uploaded to Dakar and issued in both places on the same day. The newspaper was broadly diffused among NGOs in Africa via networking technology. (For another example of the benefits of email, see Box 4.)

BOX 4 Using Email to Solve Problems

In Burkina-Faso, there is a daily newspaper that is issued at noon every day except Sunday. The editor told us that to send an issue to a town located 400 hundred kilometers from the capital city they have to transport it via "bush taxi." Due to lack of road infrastructure, the newspaper can be sold only the following morning. Using email technology, the newspaper can be transferred very easily and very quickly through the phone system, printed on the spot, and issued at the same time as the capital city.


· Sacrifices are necessary to begin a network. You must be prepared to work long hours, spend your own money, visit users who have problems that have nothing to do with the electronic communications system, and fix hardware and software problems.
· If you give people your home phone number, then you will be disturbed for anything.· You also need to be patient with people. You may find people with laptops who will take minutes to write a single word.· For strategic purposes, you will sometimes have to give free accounts.· Fidonet is a simple technology to master for NGOs which do not have the resources and the time to invest in too sophisticated technologies. Presently this is an appropriate technology for some African countries taking into account the situation described above.

Fidonet has been criticized for its limitations, but for people who are interested in email only and fax services, this is no doubt often the best choice. In spite of all what is said, IT WORKS. . .


Here is probably the best recollection I have from networking. Two years ago, I was surprised to find a message in my mail box saying:

"I have been searching the Internet for an electronic address in Africa and found yours for Senegal. My name is..., I am living in .... of the USA and I am 12 years old." For a while this young boy and myself have corresponded.

This would surprise any African - as we have not yet reached this level of computer use in Africa. Taxes continue to make computers unaffordable for individual use. And because of this we have a very low level of computer literacy.

One should note also that almost all NGOs use computers only for word processing, which really shows lack of computer skills.

When you try to convince NGOs or individuals who have access to computers to join the network, they are afraid of investing a lot of time learning something new. Those who do not have access to computers do not understand the necessity to invest money in a new technology of which they are afraid.

At this level, those who are connected are only interested in email and sending faxes. This is understandable as communication costs are very expensive in our countries.


1. Doug Rigby came to Nairobi in 1989 to work at the Environment Liaison Center International (ELCI), which served as the node for several networking projects. While there, Mr. Rigby did much to extend networking technologies, principles, and training throughout Africa. Many local systems operators credit their current enthusiasm and success to Mr. Rigby.

2. The APC has contributed much to African networking. For more information about this organization see page 189 in this volume.

3. The following rates are applied: installation US$ 20.00; monthly fee US$ 5.00; charge per kilobyte sent US$ 0. 10, charge per kilobyte received US$ 0. 10.


I am deeply grateful to Jacques Bugnicourt, ENDA's Executive Secretary for his generosity and his open mind. He understood very quickly how crucial this technology was for his own organization and for the NGO community. And without his personal commitment and his encouragement, I would not have been able to work in this field.

To Abou Thiam, from ENDA, who first mentioned email to me.

Thanks to Doug Rigby for the role he played in electronic communications in Africa. He is the one who introduced the technology to me and to many African system operators.

Thanks also to wonderful friends and colleagues who played and continue playing a crucial role in African networking: Bob Barad, Karen Banks, Mike Jensen, Cesare Dieni, Youba Sokona and Ann Heidenreich.

Thank you also to the very good friends in the APC networks.

Background Summary: SatelLife and HealthNet

SatelLife is an international not-for-profit organization whose mission is to improve communications and the exchange of information in the fields of public health, medicine, and the environment. It is an initiative of the International Physicians for the Prevention of Nuclear War, recipient of the 1985 Nobel Peace Prize. Not wanting to see space become a battleground, Dr. Bernard Lown, IPPNW cofounder and co-president, suggested the creation of a satellite-based global health communications system as a means of demonstrating that space can unite rather than further divide humankind.

SatelLife works in developing countries where libraries, hospitals and other organizations often lack adequate funds or foreign currency to subscribe to medical journals or to maintain ongoing subscriptions. Where communications facilities are intermittent, of poor quality, or expensive, the use of direct "real-time" voice or fax connections can be inefficient and frustrating. Health workers can become isolated from one another because of the high cost of telephone calls, fax services, and travel. Even with today's improved networks, communication in remote and rural areas is still difficult or impossible, and always extremely expensive. SatelLife has thus put much of its efforts into providing and improving access to "store-and-forward" message systems.

HealthNet, administered by SatelLife, is an information service that connects health care workers around the world. Using the most affordable and appropriate technology, HealthNet offers electronic mail and conferences as well as access to several electronic journals and publications. It also provides access to databases and experts. SatelLife initiated the Library Partnership Program to facilitate access to medical literature for libraries in the developing world. Finally, for its Internet users, HealthNet offers pointers to useful health mailing lists, Worldwide Web homepages, Gopher and FTP sites on the Internet. HealthNet is currently operational in the following African countries: Botswana, Burkina Faso, Cameroon, Eritrea, Ethiopia, Gambia, Ghana, Kenya, Malawi, Mali, Mozambique, South Africa, Sudan, Tanzania, Uganda, and Zimbabwe.

SatelLife began by linking medical education centers in Africa with medical libraries and other centers in the United States, Canada, and Europe via a lowearth orbiting (LEO) satellite. The satellite was expected to be an essential part of the solution to the communications problems in developing countries but, since its launch, modern telephone switch equipment has been installed in many cities in Africa, providing good international connections. Because of rapidly evolving technologies, dialed telephone circuits and error-correcting modems are often a more economical and efficient solution for international transfer of electronic data in developing countries than the LEO satellite.

Fidonet-based networks were in widespread use in Africa in the late 1980s and SatelLife joined the many NGOs supporting Fidonet use there. Fidonet provides an economical alternative to Internet email systems. The evolution of HealthNet has produced a large network of Fidonet nodes. The network is still growing rapidly, although much of its growth is in the "points" connected to network nodes rather than nodes themselves. SatelLife is committed to supporting this low-cost access system along with advanced Internet services so that services can be brought as close as possible to the end-user.

Since 1993 SatelLife has operated the Internet domain "" to support access by its member networks. While the cost barriers to providing international permanent leased lines are still prohibitive for most public health and medical education programs in the developing world' there are now several cases in which full Internet access is a realistic alternative to store-and-forward networks. SatelLife is helping its member networks take advantage of these circumstances as they arise.

The HealthNet satellite system offers Internet message services to any remote area where the local telecommunication infrastructure is poor or inefficient. The ground equipment needed to contact the satellite consists of an IBM-PC compatible computer, a Terminal Node Controller (TNC), a satellite radio, and antennas. At this stage, SatelLife is field testing a new radio design that merges the satellite radio and the TNC in one single box that can be portable.

Advances in satellite technology, opportunities in radio technology, and the growth of Internet connectivity to Africa, all mean that SatelLife will become increasingly diverse and complex. The only practical approach to handling this complexity is to base the system on a powerful software environment called UNIX. In the past two years, Linux, a public domain software for UNIX on personal computers, has become a stable and respected (and very low cost) alternative to commercial UNIX system software. SatelLife is working to bring together many different networking technologies (including full Internet, satellite packet radio, and Fidonet) into an automated system based on Linux. Its own engineering task will be to provide software to simplify the control and monitoring of this package so that the amount of training required is kept to a minimum.

This new network node technology will offer many new options in network services and will make HealthNet more useful and attractive to users of standard Internet mail systems. Using whatever technologies are available SatelLife and its HealthNet service network will continue to address the twin problems of an acute shortage of current health information and of the severe isolation of African health care workers from their colleagues.

MUKLA: Evolution of a Homegrown Network in Uganda

by Charles Musisi

Charles Musisi is Network Manager of MUKLA (Makerere University in Kampala) and pioneer of electronic networking in He has an electrical engineering background. He has shared his networking expertise with colleagues in Kenya, Tanzania, Nigeria, and Ghana.


This paper tells the story of the establishment of the MUKLA Electronic Network in Uganda. There is much in the establishment and growth of an electronic network that can be learned from our experiences. MUKLA is an example of a sustainable network bred from a homegrown desire to network. I say this because:
· Our primary emphasis is on self-sustainability;· We are now in our fourth year of operation, with an increasing base of users, expanding facilities, and a financially viable operation;· Starting from a university and grassroots NGO base, we now have established strong working relationships with various levels of government and other sections of the community;· We are internationally connected to the Internet, the Association for Progressive Communications networks, United Nations organizations, major commercial networks, and facilities around the world; and· All of this was accomplished with very little donor funding and no government grants whatsoever to commence operations. A small team of dedicated individuals in a unique government institution is running a sound business-based electronic mail service.

The need to communicate across distances on vital issues is far from new. An ever-increasing range of technologies has been applied to this need, from rudimentary tools before the age of transport to the constantly emerging suite of sophisticated services offered by the information age.

The ESANET Project

The MUKLA node began with the Eastern and Southern Africa Networking (ESANET) Project. Inspired by the communication needs mentioned above, the IDRC-funded ESANET research project was aimed at investigating various microcomputer-based methodologies for communications. The countries of the five participating institutions are all members of the Preferential Trade Agreement (PTA) region. The participating institutions themselves were the Institutes of Computer Science at the University of Nairobi and Makerere University, and the Computing Centres at the University of Dar es Salaam, the University of Zambia, and the University of Zimbabwe. The long-term goals of the ESANET project are given in Box I below; the specific objectives of ESANET were to:

· experiment with alternative techniques for data communications between the five nodes in five countries;· evaluate the technical, economic, sociological, and management aspects of the communication network experiments,· disseminate information to the research community within the region about the development and the results of the project with a view to increasing the awareness of possibilities, stimulating new and wider applications, and inviting feedback on related topics; and· make recommendations to the research community (users and telecommunications authorities in the region) on cost-effective data communication modalities, and appropriate network models and policies for specific environments and applications.

The introduction of electronic networking to the wider community in Uganda began in May 1991 as a natural spin-off from the ESANET project. MUKLA was mandated to provide email services to non-governmental organizations (NGOs) within the NGOnet-Africa project, which is described more fully in Fall's case study on page 142. We also sought collaboration with the HealthNet project whose aim was to facilitate communication among health professionals within the African region and with their peers elsewhere. (See information on page 153.) While the mode of communication chosen by HealthNet was a store-and-forward, lowearth orbiting satellite (LEO) with tracking ground stations, the regional interconnection of the ESANET and NGOnet nodes was to be across regular, dial-up telephones lines using Fidonet technology.

Activities of ESANET

A start-up meeting for ESANET in November 1990 brought together representatives from all the participating institutions, as well as those from IDRC, SatelLife, and the Nirv Centre/Web of Canada. The start-up workshop focused on identifying suitable methods of communication and on drawing up preliminary lists of hardware and software requirements. We decided that the project would support a series of workshops in the different participating countries.

At a design workshop held in May 1991, in Harare, Zimbabwe we formally adopted Fidonet technology as the technology of implementation. Our review of the status of licensing of the HealthNet ground station revealed only Zambia at the time had obtained a license. Other highlights of that meeting included confirmation of lists of requirements by different nodes and the topology of regional mail traffic exchange. The nodes in Uganda and Tanzania were to route their regional traffic and international traffic via Nairobi, while Zambia would poll (or generate a computer call to) Harare, which would in turn poll Nairobi for regional mail.

The University of Zambia (UNZA) hosted an experimentation and review workshop in November 1992. Our aim then was to enhance internode telephone calls (polling) and review the progress of the project to date. At the time of the review workshop, we observed that some interesting trends had clearly emerged and we realized that the assumption upon which the topology had been drawn was evidently unattainable. For instance, the interregional polling in East Africa was not possible due to several factors: delayed arrival of equipment purchased from overseas suppliers; wrong power specification supplied; or poor telephone lines at the node on an old analog exchange. Hence, while activity was well under way in Uganda by February 1991 and a little later in Tanzania, the MUKLA node only became operational in August 1992.

MANGO, the NGO network in Zimbabwe, was an established Fidonet bulletin board by the time the ESANET project began and so only needed to integrate the ESANET activity early in 1991. UNZA was also fully operational early in 1991 but routed its regional and international traffic through Rhodes University in South Africa. Interregional polling between these two was never reliably successful owing to poor interconnecting telephone infrastructure between Zambia and Zimbabwe. All of the ESANET participating institutions and bodies met in September 1993 at an evaluation and closeup workshop held in Uganda.

BOX 1 Long-term Goals of ESANET

The long term goals of the ESANET project were to:

· achieve more cost-effective and relevant data collection and assessment by better identification of users, in both the public and private sectors, and of their information needs at the local, provincial, national and international levels:· strengthen local, provincial, national and international capacity to collect and use multi-sectoral information in decision-making processes and to enhance capacities to collect and analyze data and information for decision-making;· develop or strengthen local, provincial, national and international means of ensuring that planning for sustainable development in all sectors is based on timely, reliable and usable information; and· make relevant information accessible in the form and at the time required to facilitate its use.


As it happened, Doug Rigby, a networking consultant at the Environmental Liaison Center in Nairobi but also working for the NGOnet-Africa Project, visited Kampala in February 1991. I was then a final year student of Engineering at Makerere University but on forced vacation as a result of student riots that caused the closure of the campus.

My meeting with Rigby was helped by the student riots as much as by my volunteer work at an environmental NGO, called JEEP, a grassroots network organization. JEEP staff asked me to meet Rigby at the Entebbe airport. This was perhaps my longest wait at an airport for any visitor coming from abroad - as the scheduled flight, I later learned, had been a phantom one. All the same, I persevered until he arrived on a late evening plane and took the 35 kilometer journey back with him to Kampala. After hearing his stories enroute, I never regretted the long wait and I slept that night dreaming about this intriguing new concept of electronic mail! The following week was to usher in many more exciting ideas that have led, as it turns out, to a new and unforeseen career.

With Rigby, we went through the drills of Fidonet: installing a modem; communication software for end users; and even basic DOS commands. The team of trainees had now grown to five: the then three folks from the Institute of Computer Sciences: one representative from the Developmental Network of Voluntary Associations (DENIVA), an umbrella network body for local and indigenous NGOs in Uganda; and myself. A windup workshop of Rigby's mission brought together the first batch of potential users from different organizations. Among them were people from CARE (the relief agency), the Centre for Basic Research, the Makerere Medical School, a handful of DENIVA members, JEEP itself, the Makerere University's Vice Chancellor, and other independent participants.

While I may not have been the most conspicuous participant, I had apparently caught the eye of some people and I was chosen as the person to carry the mantle of making it work.


The Critical Years

I quickly realized that my new role would have to be accommodated within my academic life when campus reopened. That I did by working at the node after classes, not an easy thing to do for an engineering student. The next thing was that I had to do to this work as a volunteer. There would be no pay at first. I still remember those first meetings with prospective users - environment groups, missionaries, and other NGOs. This was new to all of us. Even as I crawled under desks to connect modems to telephones, and even as I made test polls to Nairobi, not one of us was sure that we were not going to get enormous telephone bills. I wasn't all that sure that I wouldn't be able to pay the actual telephone bills even if they were not enormous! (See Box 2.)

The harder electronic networking seemed the more determined I was to prove that it worked. It was the best thing I could have done. The process taught me a great deal, refined my thoughts considerably, and led me into contact with many people whose expertise I would need if this was going to succeed.

Originally, I had planned to set up a non-profit network for NGOs to run parallel to MUKLA. That might have been ideal but, with my unfinished engineering degree in the way and a lack of clear institutional support from the NGO community, this seemed a formidable task. I labored to convince the director at the ICS that there was more to be gained in merging ESANET with NGOnet activities than there was in operating them separately.

Shortly after the university reopened, however, we reached agreement and installed the ESANET equipment and procured a dedicated telephone line. The MUKLA node got under way and the real fun began! At the end of our first year of operation, we had nearly 50 users, quite a remarkable achievement and a good omen for an exciting future. We were beginning to fulfill many people's dreams for use of this media. At this point, I stood as the systems operator (sysop) and manager of an ever-growing network. I was confident that I had established a sound management structure to introduce cost recovery. The director's secretary would handle billing and accounting while I concentrated on software and hardware developments as well as other technical matters.

Project Synergy

The ESANET project, with its idea of electronic networking at the Institute of Computer Science, took second place to the higher profile and better funded UNESCO Intergovernmental Informatics Project (IIP). The IIP was aimed at sensitizing decision makers in government and other sectors of society on the use of informatics tools in decision making and management. The attention paid to the IIP reduced high-level interference in the day-to-day running of MUKLA. Electronic networking nevertheless benefited from the many IIP workshops organized where email was always on display.

As the rush to sign on to MUKLA grew, so too did the pressure on the hardware - then a 386 SX, 2 RAM 80 megabyte hard drive personal computer. So, by July 1992, with the user base standing at well over 150, the node personal computer was upgraded to a 486DX 33 MHz, 8 RAM and 170 megabyte hard disk from the IIP project consignment.


At this level of expansion, MUKLA posed a real management challenge. With technical capacity to handle new installations and user support overstretched we made a decision to get more members of the teaching staff involved. One person was put in charge of documentation and promotions, while a technical assistant was assigned to help me. (See Box 3.)

Billing and accounting were firmly put in the hands of an accounts clerk who also doubled as the director's secretary. I believe that the stability of our network caught the attention of other regional networking projects, based mainly in Nairobi, and that they borrowed ideas from our experience. We then had over 10 groups or users dialing-in on long distance trunk phone calls from Nairobi and other parts of Kenya.

Around this time I was drafted into a number of regional initiatives. 1, for instance' participated in the design and implementation of the UNEP/Global Environment Facility (GEF) Greenhouse Gases project to link up researchers in six African countries, namely: Uganda, Tanzania, Nigeria, Senegal, Morocco and the Gambia. I was also hired as an independent consultant for the CABECA project to setup and train systems operators in Uganda, Nigeria, and Ghana. (See CABECA box on page 147)

Between the UNEP/GEF and the CABECA project I have established a total of nine nodes in the above mentioned countries. I have also trained a score of incountry technical support people and have installed several end-user sites for access to email and the Internet. I have also set up email access for the UN's World Food Programme (WFP) offices in Dar es Salaam, Mombasa, and Kampala. Only recently I was on mission to war torn Burundi on an International Alert of London mission. There, l set up email so that the UN's special Representative to Burundi could access information and participate in discussions aimed at better informing the world and also the warring factions in conflict resolution.

Present Situation

From around mid-1993 to the present, we began routing Ml)KLA's international traffic through the GreenNet's GnFido node in London via four daily polls. This link works satisfactorily well, though disruptions in the past often occurred due to factors such as adverse weather conditions. Lately, the unreliability of the GnFido node can be attributed to over-stressed hardware facilities coupled with lack of support staff.

As of January 1995, MUKLA had an installed user base of over 300 sites. The majority of these are around Kampala but there are about 15 sites in Entebbe (35 Km from Kampala), five in Jinja (80 Km east), three in Mbale (150 Km east, close to the Kenya border), three in Mbarara (220 Km southwest), and three in Kabale (400 Km southwest, close to the Rwanda, Tanzania, and Zaire borders). The upcountry installations all have to make long distance calls to Kampala.

It therefore became imperative to set up local access points in these towns. In February 1995, a node was established in Entebbe at the Uganda Virus Research Institute with equipment supplied by the HealthNet project. Similar nodes are planned for Mbarara, Kabale, Mbale and Jinja during the first half of 1996.

Even now a few users call in from Nairobi, though the majority of regional traffic flows to MUKLA via Sasa Communications System. This company, based in Nairobi, is an initiative of the East Africa Internet Association. (See Box 4.)

Recent Developments

Throughout 1995, I consulted for a commercial enterprise (StarCom) to help establish a full Internet (JP) link for the Uganda market. In August 1995, another company (InfoMail) opened full IP access from Uganda - thus becoming the first site in East Africa and the Horn of Africa region to give full access to the Internet, including facilities such as World Wide Web. Starcom became operational in November 1995. It is noteworthy that both these companies implemented international access using very small aperture terminal (VSAT) satellite technology in preference to the overpriced and unreliable digital leased lines from the Uganda Posts and Telecommunication Corporation (UP&TC).

As seen from the information in Appendix A, the electronic network market in Uganda is now fully liberalized with not less than five providers with services ranging from email access to full Internet. However, the usage charges remain high. The charges given are accurate as of January 1996.

BOX 2 Seeing Is Believing

I really thought everyone would be so excited that they would just put in some money and we would be up and going in no time. Not so, I learned. There was a lot more to do - and the hardest lesson of all was that, until I could prove that this thing saved money, there was no chance of it getting off the ground.

BOX 3 Help from Some Friends

There are many funny stories from those first few months. l grappled with the establishment of a node facility for which there were no textbooks - only a concept and some guiding hands from friends abroad. Invaluable help and support handily came in from Doug Rigby at ELCI, Mike Jensen at WorkNet in South Africa, Karen Banks at GreenNet, and later Bob Barad in Washington. l had considerably underestimated the technical complexity of what we were doing.

BOX 4 The East African Internet Association

The EastAfrican Internet Association (EAIA) is a not-for-profit group formed in April 1995, seeking to promote and expand cooperative electronic communications and inter-networking in the East African region. Its members include the majority of electronic service providers, serving at least 3,000 users, in the following countries: Eritrea, Ethiopia, Kenya, Tanzania and Uganda. Partner networks abroad are also members, as are a number of interested individuals. Membership is open to all interested parties for a nominal fee. The Association is in the process of being officially registered in each country.

EAIA's aims of cooperative internetworking in Africa are also supported by: Capacity Building in Electronic Communication in Africa (CABECA, UN Economic Commission for Africa), GreenNet (London), UN Environment Programme (Kenya), UN World Food Programme (Kenya), RIO/ORSTOM (France), Web (Canada), Wolfnet Communications (London) and SangoNet (South Africa).

Discussions take place on a private Internet mailing list.


Beyond Email

While the bulk of our traffic on MUKLA remains electronic mail, there is a new trend towards user participation in electronic conferences. MUKLA at the moment carries over 50 different conferences mostly from the Internet and APC networks, with a few specifically regional or local ones. Perhaps the most notable among these is a discussion list on Uganda-related issues and news, appropriately called Ugandanet. Through this conference, over 600 Ugandans from all over the world link with each other on a daily basis to interact with lively discussions on various topics ranging from the constitutional process, to entertainment, to sports and news.

Perhaps responsible for the bulk of growth in the beginning was the email/fax service that MUKLA provided. Users are able to send' at the price of email, faxes to destinations all over the world. Its popularity was mainly due to the exorbitant international charges levied by Uganda Post & Telecommunication.

Sectoral Involvement


Electronic mail has had a particular attraction for research activities to many researchers. Students are also increasingly getting to use email. About 25 percent of MUKLA traffic is for university related activities.

Non-Governmental Organizations

These form the single largest group of MUKLA users accounting for over 40 percent.


The involvement of government in any major way has not come yet, although from the early days government departments were interested in what we were doing and, in 1992, we were approached by many government bodies for connectivity. Presently, however, only isolated projects within government departments are connected, usually for very specific reasons, such as easy access to a donor office abroad. We have been approached by the Ministry of Foreign Affairs to explore possibilities for linking Ugandan embassies, especially in particularly hard-to-reach places like Zaire and Rwanda.

Business Sector

This perhaps remains a weak and under-represented section of the network. Nevertheless, from the early days, business users concerned with sustainable development issues began to join us.


There are many other individuals and independent groups that MUKLA knows are interested in electronic networking. We have yet to tap this group.


Though MUKLA benefited from the ESANET research fund as seed capital, cost recovery was instituted at an early stage to supplement this. As of now, and at completion of the ESANET project, MUKLA is fully self-financing through fees levied on users. A check is put on users via shadow billing whereby an itemized bill for all users is prepared at the end of the billing quarter to check on excessive use and possible abuse of our lenient terms. There is no additional charge for the fax facility though strict monitoring is done for each of categories above. When excessive use is noticed, the user is upgraded to the higher billing group.


The ESANET project introduced the idea of electronic mail communication to researchers and other users at Makerere University campus - though the focus of the project was then on experimenting with various computer-based communication technologies and working out the technical bugs involved with poor telephone lines, erratic management of the telephone long-distance dialing system, and hardware and software equipment.

As of September 1993, most of these problems had been ironed out. The electronic mail system using the GnFido system has reached a level of reliability that surpasses that of fax machines. A cost-effectiveness analysis was carried out at the end of the ESANET project and this provided the evidence needed to prove that this venture could be sustainable, given the proper setting and management.

The ESANET project established that microcomputer-based electronic communication was a viable, sustainable technology and appropriate to the context of the region in which it has operated. There is a proven demand for electronic communication, both regionally and internationally. Following are the recommendations from the ESANET project:

To the research community, we recommended that:

· Relevant institutions be encouraged to consolidate and expand their user base.· We should undertake more research into improving and expanding technologies used.· Manpower requirements for this activity be consolidated into the establishment of the institutions concerned.· Appropriate mechanisms be established for operational cost recovery and institutional funding to cover system operation and expansion.· Cognizance be taken of the ever-improving nature of electronic communication worldwide and relevant upgrading be considered.

To the telecommunication authorities, we requested:

· Further investment be made in improving regional telecommunication links to enhance electronic viability of using national telecommunication infrastructure for modem related activities · Special tariffs be applied to the academic and research communities for packet switching and leased line facilities thus enabling greater access to computer based communications regionally and internationally.· Liberalization of user terminal equipment connected to national telecommunication networks be encouraged.· Pragmatic policies be established with regard to licensing of alternative communication methodologies such as packet radio and low earth orbit satellite.

To the donor community, we recommended that:

· In view of the fact that national communication networks are an indispensable component of national development, providing as they do for an efficient and effective information delivery system in diverse sectors, continued funding be made available for establishing infrastructure related to computer-based communications for capacity building in order to facilitate transfer of technology.

To regional governments, we recommended that:

· They recognize the importance of national and regional networks for all aspects of development and for human resource and manpower interactions. The final meeting of the ESANET project paid tribute to the invaluable contribution that IDRC had made through this project in furthering the case of electronic communication within academic, health, and other related communities. We accordingly asked IDRC to:

- Ensure that ESANET nodes be covered in any forthcoming regional electronic network support programs.
- Share the findings of this project with other donor and development agencies.
- Be receptive to future requests from ESANET community to consolidate and expand what has been achieved so far.


The ESANET project provided the participating institutions the opportunity to experiment with regional and international microcomputer-based communication. The aim of the experimentation was to establish the viability of the regional computer networking for data communication from the technical and management perspectives. We acquired data and experience from the experimentation phase of the project using Fidonet and packet satellite technologies. Based on this experience and data, we made the following specific observation on hardware, software, connectivity, network management and research activities.


For modems, we have observed that:

· 9600 baud or higher modems (e.g., Telebit Worldblazer) perform better for international connections, while 2400 baud modems (e.g., GVC 2400) are sufficient for local connections.· External modems perform better and are easier to handle than internals.· Modem power supply should be 240V AC.· Modems should be Hayes compatible.

For the node computer, we recommend a machine with the following specifications:

· 386 or higher,· a minimum of 4 megabytes RAM,· 120 megabyte hard drive minimum (max 12 ms cached),· dual floppy,· fast serial ports,· 220 V AC supply, and· UPS with stabilizer.

There is always a need to have a backup computer.


We used the following software packages in most of the our node operations

· FrontDoor 2.1+ - mailer· Gecho - conference mail processor· Msgtrack - mail tracker, ReDir - mail redirection GUS, Echovol - for conference tracking, AC - for accounting and billing


We observed that GnFido, Rhodes University, or WorkNet polling into ESANET nodes was cheaper and easier than direct polling between ESANET nodes This has been the case for regional as well as international traffic. We thus disproved the earlier assumption on which the interconnectivity topology had beer based.

User Base

We also observed that:

· A massive effort was required to develop and support a user base. We needed advertising, workshops, and maintenance visits.· There was a need to clarify to the users what installation and support imply and discourage frivolous requests.

Network Management

To sustain the network it is important to:

· train system operators and establish official positions for system operators at various universities and· put into place efficient and cost effective cost recovery mechanisms.

Some Fidonet management tools are available in basic node software. Additional software was developed during the project period to supplement these basic tools. Examples are: AC (accounting) from the Zimbabwe node and MTMON and HISMON (monitoring) from the Zambia node. However there is still need to develop more tools. Whereas there was sufficient documentation for point operation (the lowest level of the Fidonet hierarchy), there was need to develop more documentation for node operation. Nodes are the individual systems that belong to Fidonet. They are responsible for passing mail between the points and the next hierarchical level of Fidonet - the hub.

Research and Development

We need to encourage technical innovations to improve the system and we can use student resources for such projects. Technical innovations are required in the following areas:

· Terminal Node Controller (TNC)· The packet satellite upload program (PG)· Modems and modem testing/evaluation· Mailer interfaces

Beyond the ESANET Project: Future Developments at MUKLA

While the interest in using electronic mail among researchers is strong, the technical capacity to meet this demand remains severely constrained due to lack of adequate funding. Additional funding is needed both for the personnel involved in its spread and for the purchase of hardware upgrades for the present installation. Although a number of groups and individuals have expressed interest in using and learning more about electronic mail, there are no adequate pedagogical materials nor the funding to have technical experts available to advise on both hardware and software problems as well as to do the installation and training.

Given the lessons learned from the experimentation phase, with the outlined constraints' it is imperative that MUKLA seek to improve its institutional capacity to effectively spearhead a development plan. We will seek support for recruiting trainers whose job will be to build on the achievements to date and enhance the existing network amongst all these groups. We will also seek support to update the hardware at the existing nodes and to create a pool of modems to serve as seed investments in new areas where electronic communication is to be introduced.

Specific goals should be towards:

· upgrading the hardware and software at the MUKLA node to full Internet status via leased line;· developing appropriate educational materials and formats for workshops and short-courses,· providing technical assistance and training; and· facilitating communication via email amongst users within the country and region.

There are two sets of activities to look at in the process. The first would involve the acquisition of the necessary funding and identification and purchase of the necessary hardware for a full IP connection. Simultaneously, we would prepare training materials for the use of electronic mail, instruction on both hardware specification and troubleshooting, and hands-on training with the software. The second set of activities will include the provision of regular and ongoing technical support to users. This will be necessary both to iron out technical problems encountered in the daily operations of the electronic mail, as well as to sensitize and train new users.

New Nodes

We expect that the user base will have grown to an estimated 2000 installed sites by the end of 1996. It is time now to consider new nodes to improve services in areas outside of Kampala. Some possible sites for expansion of MUKLA are given below.

Entebbe is a major seat of government ministries and headquarters to many international United Nations bodies and NGOs. It is ripe to have a node established that could serve as hub to all users in and around Entebbe. This node would periodically poll MUKLA to transfer national, regional and international traffic. There is potential for up to 150 installed sites within one year; this would also relieve the pressure on MUKLA.

Kabale located about 400 Km from Kampala has good potential to play a node role for Rwanda, Burundi, and some parts of Western Tanzania that fall within the Kagera Basin. Kabale is the gateway to these countries that have traditionally used the northern corridor trans-Africa highway to link them to the Kenyan seaport of Mombasa. There is an automatic exchange that is part of a large telecommunications project linking the countries that comprise the Kagera Basin Organization (KBO): Burundi, Rwanda, Tanzania and Uganda. With 3,000 lines, the exchange is the most modern to be installed in Uganda. The KBO regional telecommunications project consists of a microwave transmission system interconnecting the four countries. It is aimed at the promotion and development of agriculture, forestry resources, and telecommunications links between member states that share the river Kagera.

Goals of Expansion

The nodes in Entebbe and Kabale, together with a series of training workshops for users in Jinja and Mbale, would constitute the first phase of the expansion program.

Existing national and international mechanisms of information processing and exchange, and of related technical assistance, would be strengthened to ensure effective and equitable availability of information generated at the local, provincial, national, and international levels, subject to national sovereignty and relevant intellectual property rights.

National capacities would be strengthened, as would capacities within governments, non-governmental organizations and the private sector, in information handling and communication, particularly within the East African region.

Improved Services

MUKLA would offer its clientele the full benefit of a wide range of Internet services. Electronic mail and conferencing, public and private access to Internet mail and other worldwide networks might be provided via partnership with the Association for Progressive Communications. MUKLA can also provide consultancy services on network establishment for organizations and government departments.

Future Needs and Possibilities

One of the difficulties we face as we prepare for future developments is to clarify the role of what is essentially a well meaning private enterprise with the responsibilities in this area of various tiers of government and public in Uganda. The question of appropriate ownership structures for networks like ours is quite complex. Our decision to operate as a private company was indicated at the time of our birth. The people are sufficiently interested to devote time and financial energy to the project and do not have any particular philosophical attachment to a mode of operation.

However the structure has served us well amidst the politics of governments and NGOs. We have always regarded ourselves as the carriers of information related to these issues, rather than as a policy body or an arbiter of truth and best practice. In the early days, this structure and philosophy allowed us to engender cooperation between NGOs who did not see eye-to-eye on all issues. In later days, the same stance has allowed us to assist cooperation between tiers of government and government departments, which have been known to jealously guard their own interests rather than cooperate. And, more importantly, we have managed to remain credible with both government and its more radical opponents. It is probably our strongest point, that we can operate independent of the political will of any tier of government or its funding priorities.

The future holds many specific problems to address - questions of cooperation, questions of standards, and questions of access to information will have to be addressed. Much work will have to be done towards these ends by both ourselves and government bodies.

Regional Interconnectivity

There will be immense savings by sharing costs and collaborating with other regional providers. The East African Internet Association (EAIA) effort could be one way to foster further regional interconnectivity. (See Box 4 on page 164 for more information about EAIA.)


My experience suggests that there is far more to establishing a successful network than purchasing and learning the technology. Indeed, technology expenses have been a minor part of our budget. Outlined below are some basic principles I believe are essential in establishing viable networks.

User Friendliness

Experience indicates that the issue of user friendliness is one where there should be no compromise. Simplicity of interface is crucial; however, simpler interfaces, those that use graphics, for example, also require more computing power. It is important to have machines with adequate power, memory, and speed to perform the necessary tasks. This is where a full needs analysis is necessary. The simplest looking interface may not do the job; nor will old or obsolete equipment provide enough sophistication to run programs such as Windows.

Adherence to Standards

The only trade off in user friendliness that might occur would be in the area of adherence to standards. Here, important standards have to be considered if contact with global networks and global relevance of data collected are concerns.

Promotion and Education

The most common mistake we can make in large scale electronic messaging installations is believing that somehow the system will "introduce itself." Not so! You can only successfully train the converted. The network has to be actively promoted. Its benefits have to be known before people will use it. Its applications to work areas and advantages have to be received with enthusiasm. Without this, the basic aims will not be achieved.


Equally, training is an absolute must for a successful implementation. Links should be made to existing training organizations. Training materials must be available for any software used. (See Box 5.)

Product Champions

The concept of a "product champion" is often mentioned in sales literature. It essentially refers to enthusiasts who promote concepts and products willingly because they believe in them. All networks need them. Product champions sometimes create problems for organizations with their over-enthusiasm and are rarely popular with administrators. But they are totally necessary and need to be identified and supported.

Adequate Funding

An under-funded initiative that fails can delay a concept such as sustainable development networks in a country by a decade or more, and a few failures can ruin the concept altogether. It is far too easy to attempt to stretch available funds too far and to leave behind a string of underdeveloped projects with little chance of success. It is also far too easy to get carried away by enthusiasm and to start a project without sensible financial plans to ensure viability.

Appropriate Ownership Patterns

This is a difficult question for which there is no immediate formula. An appropriate ownership pattern has to be one which will not restrict the participation of any governmental or non-governmental body whose cooperation is needed.

Plans for Financial Sustainability

Unless the venture is to be a continual financial burden to funding organizations, a realistic business plan has to be adopted to ensure that the network is self sustaining within a given period (perhaps 2 - years).

Managerial, Sales and Technical Expertise - in that order!

The need for managerial expertise must be obvious, as is the need for technical expertise. What is less obvious is the need for sales expertise. Even if a facility is not expected to raise revenue, it surely is expected to engender use and that's a sales job.

In conclusion, I would state that the world is not a series of isolated ecosystems bearing no relationship to one another and capable of resolving their own problems. National sovereignty does not rule the atmosphere, nor the oceans, nor indeed the rivers that meander happily across borders with no care to the politics of the government of the day. Global cooperation is vital, and access to the experiences and knowledge of others is essential if we are to solve the problems facing us.

Indeed' this paper is written in the belief that what we have achieved and learned in Uganda is valuable, and may be of assistance to people in other countries wishing to establish similar facilities. We would be happy to assist and advise based on our experiences here.

Box 5 Training Needs

A typical mistake made in electronic messaging implementations is to spend all available funds on hardware, bandwidth, and software development and find that no funds are available for training. This has been noted in many universities where typically less than 13 percent of academics actually use systems. It's not surprising - they are presented with very basic interfaces, no help desk, no manuals, and only the very brave and technically-inclined actually make it.


1. MUKLA Institute of Computer School
Makerere University P.O. Box 7062, Kampala
Contact: Charles Musisi, Network Manager
Services offered: Electronic Mail, Listservs, APC conferences/Usenet, Faxing

Usage charges


$ 10

A monthly individual/small NGO

$ 30

A monthly corporate rate

$ 50

Big corporate/International

$ 100

2. StarLight Communications (U) Ltd (STARCOM)
Sheraton Complex, 14th Floor, Ternan Ave., P.O. Box 10524, Kampala
Contact: Kiggundu Mukasa, Internet Manager
Services offered: Email and email fax, World Wide Web, Telnet, FTP

Usage charges:

E-mail only (unlimited usage)

$ 30

Shell account

$ 50

SLIP/PPP account

$ 100

Full IP


3. Infomail (U) Ltd
Plot 2 Clement Hill Rd. P.O. Box 11465, Kampala
Contact: M.M. Otyek, Manager
Services offered: Email, World Wide Web, Telnet, FTP

Usage charges:






$ 50

3 hours





8 hours



Big corporate


20 hours



Very big corporate


50 hours



4. Transmail Ltd
Blacklines House, Suite 2B4, P.O. Box 7482, Kampala
Contact: Patrick Mawanda, Manager

Services offered: Email, Electronic Fax, Mail Broadcasting, Conferencing

Usage charges:

Traffic not exceeding 420KB


Traffic not exceeding 910KB


5. InfomaNet

4th Floor Impala House' P.O. Box 8945, Kampala
Contact: Christine Nantongo
Services offered: Email, Fax

Usage charges:

Quick start plan:



Email sending

$1.20 per page

Email receiving

$0.70 per page

Fax sending

$2.95 per page

Power user plan:



Security deposit




Email sending


Email receiving


Fax sending


The UNIMA Fidonet Network: Computer Networking for Communications in Malawi

bv Paulos Nyirenda

Dr: Paulos Nyirenda is Head of the Departnent of Physics and Electronics at the University of Malawi. He has a PhD in electrical engineering from the University of New South Wales. He has been working on networking at the University of Malawi since 1992. He asks that this chapter be dedicated to his late wife, Gemma.


Malawi is a landlocked country located on the southeast side of Africa. It shares boundaries with Mozambique, Tanzania, and Zambia. The country has a population of about 9 million, of which about 90 percent live off subsistence farming. Malawi's economy is agricultural-based, with few manufacturing or mining industries.

The University of Malawi is the only university in the country. It is made up of five colleges spread out over a wide geographical area in the central and southern sections of the country. There are many research institutions in the country, most of which are linked to the agricultural sector and, as such, are located in remote locations across the country. Most of these research centers have access to direct exchange telephones that work most of the time but are of relatively poor quality, just as in some other African countries. Also, as in other African countries, the telephone penetration into the population in Malawi is very low.

Communications in Malawi is difficult and expensive. Research and data communications have been achieved primarily by physical travel to a site - often over seasonal roads in poor condition - or by fax where available. This is also the case for most governmental and private sector communications. Most needed research and other data and information do not reach the people and decision makers who require this data to make important national and international decisions. However, even as early as 1992, most research, governmental, and major commercial offices had computers. Most of these were IBM-compatible desktop computers or Apple Macintosh computers.


Realizing the difficulty of communicating among widely separated university colleges and research institutions, I proposed a project titled Study of Computer and Telephone Network Based Communications in Malawi. The proposal was submitted for funding to the University of Malawi (UNIMA) Research and Publications Committee (RPC) in February, 1992. The original budget was for Malawi Kwacha (MK) 5123' or roughly $1,000. The main objectives of the project were to:

· investigate the feasibility of establishing computer and telephone-based communications in Malawi; and· demonstrate more efficient and effective communications among researchers and academics within Malawi as well as among those in Malawi and outside.

With these objectives and minimal base funding, the Malawi Fidonet network was started. The network, now called UNIMA, has grown to be a public, nationwide network serving all sectors - government, non-governmental, and commercial - of the Malawi economy. It thus serves a much wider population than the academic and research communities originally envisioned and for which I had budgeted. It is a public network in the sense that anyone and any organization in Malawi can be linked to the network - provided they have the basic resources required for such linkage. The project has since attracted funding from various national and international organizations.


The proposal I made to the University led to the establishment of the UNIMA Fidonet network. The specific objectives of the proposal were to:

· experiment with modems, microcomputers, and the telephone network for data communications within the University of Malawi;· design and build electronic interface devices for connecting direct telephone lines and switchboard (PABX) extensions to one computer and modem node to handle both local college and direct line connections;· evaluate the quality of the Malawi telephone network for data communications, and · evaluate the cost of computer communications using modems and the telephone network and compare it to other electronic communications modalities.

At the time the UNIMA project was started, there was a regional project called the East and Southern Africa Network (ESANET) funded by the International Research Development Centre (IDRC) of Canada. (For a complete description of the ESANET project, see Charles Musisi's case study on page 158.) The UNIMA project had similar general objectives to ESANET, which linked university computer centers in Zimbabwe, Zambia, Tanzania, Uganda, and Kenya. Because I was involved in these activities, the UNIMA project benefited considerably from the ESANET project in the beginning, even though Malawi was not a participating country in ESANET. The startup modems and non-commercial Fidonet software that enabled UNIMA to take off were obtained under ESANET.

My proposal to study computer and telephone-based communications in Malawi was accepted by the RPC in January 1993. I next needed to request security clearance from the Malawi Government, as well as from the Malawi Posts and Telecommunications to start the project. The government issued the security clearance on 15 June 1993, one day after citizens passed the referendum that introduced multiparty politics into Malawi. The Malawi Posts and Telecommunications Department then cleared the project in July 1993. The direct telephone line required for the node or hub of the telephone-based computer communications network was installed at Chancellor College in Zomba on 24 September 1993.

We next installed the equipment for the node: an IBM-compatible 386 computer running at 20 MHz with a disk space of 40 megabytes (borrowed from the Physics Department at Chancellor College); a VIVA 2400 baud modem donated by IDRC under the ESANET project; and non-commercial Fidonet communications software (FrontDoor 2.02) obtained under ESANET and tested earlier. This startup setup was then used to negotiate the network address, initial mail routing, and initial polling procedures required for the network. In the startup phase, the network address was negotiated under ESANET to be a Fidonet point address off the University of Zambia electronic mail network. Malawi now has its own network number and the node established has the Fidonet address 5:7231/1.(2)

Connecting to the University of Zambia Fidonet node was very difficult and, when the connection was finally made, the line quality was poor. We decided in November 1993, to switch connections to the Africa Zone Gate directly at Rhodes University in South Africa. This improved the connection success rate as well as the quality of the connections. In May 1994, the UNIMA network started receiving funding from the Capacity Building for Electronic Communications in Africa (CABECA) project also funded by IDRC but managed under the Pan African Development Information System (PADIS) in Addis Ababa, Ethiopia. (For more information about the CABECA project, see Fall's case study on page 147.)

Under CABECA, the UNIMA network received more modems and assistance from the Southern Africa Non-Governmental Network (SANGONET) in Johannesburg. SANGONET polled the UNIMA system twice a day at first but this was upgraded to three times a day to cater to the increased volume. Also under CABECA, we upgraded the node with the installation of a Telebit WorldBlazer high speed modem. This modem was eventually replaced by a US Robotics modem.

Figure 1 shows the general Malawi Fidonet network and its linkage to other networks in the world. The number of points running off the node has increased rapidly even though the single most important constraint to the growth in the user base has been the availability of modems at an affordable price within Malawi.

Figure 1: Fidonet structure in Malawi

We replaced the computer borrowed from the Physics Department with a dedicated computer borrowed from UNIMA RPC. This was also an IBM compatible 386 computer with the same speed but with a 104 megabyte hard disk and more memory. This computer made it possible to perform multitasking using DesqView and QEMM. This allowed us to operate two telephone lines, using intermail as the communications program on the node. With this setup the node is now able to handle two data connections simultaneously, one on a direct exchange telephone line and another on a PABX telephone extension that caters to the local Chancellor College campus computers. Putting in the line from the local PABX helped to reduce congestion considerably on the direct line.

The use of a multiline mailer that enabled the node to handle more than one telephone connection at the same time made it unnecessary to develop hardware devices to achieve the same results. However, there were problems with rescanning the mail every time it was collected from the node. As the message base and user base grew, this mail rescanning took more and more time and eventually led to lengthy in-between call processing, leading to delays in answering calls at the node.

After observing the significant contribution that the Fidonet UNIMA network was making towards improving data communications in Malawi, the World University Service of Canada (WUSC) provided funds to the UNIMA network in February 1995 to purchase a Gateway 2000 IBM compatible 486 DX2 computer with 730 megabyte hard disk and running at 66 MHz. This has considerably improved the performance of the node, which can now process mail much faster and has removed the waiting that users experienced between calls while the node computer was processing received mail and extracting billing information. In addition the network has also received support from the Canada Fund of the Canadian High Commissioner to Malawi to purchase modems to assist the public sector in getting connected to the network so as to improve their communications.

UNIMA Network Operation

Under Fidonet, electronic messages and files are prepared offline to reduce telephone connection costs. The messages and computer files are then transferred onto the network in compressed files or packets to reduce the time of the telephone connections and to improve efficiency. When a point has a message to send, the message is normally sent to the node to be routed to its destination. If the message is destined for another point on the UNIMA network, the message waits at the node to be picked up by the destination point. Before April 1994, the node originated international calls to the Africa Zone Gate at Rhodes University for delivery and collection of international mail. All international mail is now routed via SANGONET, which polls the UNIMA node three times a day. International mail from Malawi is made to wait at the Malawi node in Zomba for international delivery during one of the polls from SANGONET.

When mail is received at the UNIMA node, it is processed by the non-commercial Fidonet mailer and other mail processors. Gecho.exe (version 1.01) is used as a mail and conference processor and also for mail compression and decompression. Netmanager, Netmgr.exe (version 0.99), is used as a general message processor for message-by-message identification, distribution, and redirection. Message tracking is done by msgtrack.exe, which produces a message-by-message log indicating such data as message origin, destination, dates, and volume.

I have written additional software to produce bills sent to users for cost recovery and network sustainability. Most of the billing information is obtained from the data produced by msgtrack.exe. Other data used to evaluate the performance of the network and the telephone system is collected from the various log files produced by the mailer and mail processors and analyzed using tools developed at the UNIMA node.


Network Performance

The network performance reported here was monitored at the UNIMA Fidonet node. As outlined above, the node software produces log-files (logs) for the day-to-day, telephone call-by-call, as well as the message-by-message activities that the node performs in sending and receiving electronic mail and files. The collection of the data presented here started in October 1993 and ended in March 1994. Much more data has been collected at the node but this has yet to be analyzed. By processing the log files, the monthly international mail volume can be determined. This is shown in Figure 2.

Figure 2: International e-mail volume and telephone bill using a 2400 baud modem.

Figure 3: Cost per kilobyte of international telephone calls for e-mail transfer using a 2400 baud modem.

Cost of the Service

Shown in Figure 3 are the costs of sending or receiving one kilobyte volume of mail and the monthly telephone bill in Malawi kwacha, as charged by Malawi Posts and Telecommunications Department. This calculation was done on a month-to-month basis as well as on a cumulative basis where the total bill and the total volume, up to the dates shown, were used to calculate the cost per kilobyte of mail. As shown in Figure 3, the cost of operating the network per kilobyte of mail has dropped from MK4.68 (four Malawi kwacha and sixty-eight Malawi tambala) in October 1993 to MK0.88 in March 1994 on a month-to-month basis and from MK4.68 to MK1.10 on the cumulative basis. Using these data, the node has recommended that users be charged at the rate of MK1.00 per kilobyte of mail sent or received internationally. This charge rate was still in place at the writing of this chapter.

Although the modem used for collecting the data changed from a 2400 baud modem to a 14400 baud one (with most good connections at 9600), the Malawi telephone charge per minute also changed upwards considerably from about $1.00 to $3.00 per minute in January 1995. We are currently collecting and analyzing data to check the effect of the modem speed change versus the telephone charge rate change.

A comparison between the cost of using the email system and using a fax to transfer text messages can be made based on the above data. One full page of text (as on this page) uses nearly 3 kilobytes and would therefore take MK3.00 to send by email. After examining several fax transmissions at Chancellor College, we found that such a page would take more than one minute to send but often less than two minutes. (The Malawi Posts and Telecommunications would therefore charge two minutes.) Most of the international mail transferred during the observation period has been for contacts outside Africa. The cost of sending such text messages internationally by email is therefore about 20 times cheaper than faxing similar messages at the current telephone rates and email charges at the UNIMA node.

Billing Users

Using the data collected from message tracking, a bill is sent from the Bursar of Chancellor College in the University of Malawi to each user once every three months. In addition to the volume fee outlined above, which now stands at roughly seven cents per kilobyte, each user account held outside the university is charged at 70 cents per month for account maintenance. Experience has shown that although these rates are very low, most Malawian users and organizations still have difficulties paying. I am currently proposing to conduct a study to find out whether this is due to genuine financial difficulties. I hope that user perception of "user pays" electronic communications will be better understood following the survey and analysis of the results.

I have shown in Figures 2 and 3 that even though the monthly bill and the monthly volume of mail sent and received internationally are increasing rapidly from month to month, the cost of sending mail per kilobyte has steadily approached a constant value. I expect that the speed of the modem used at the node will play a crucial role in determining the cost per kilobyte of international mail transferred at the node. At the moment this has been complicated by the international charge rate change made in January 1995. I hope that the further data analysis now being carried out will assist in clarifying the situation.

All the money raised from the bills is paid into a project account held at the Chancellor College Bursar office. Money in project accounts does not get absorbed into the University's pool account; although the money does not generate interest in such accounts, it is available when required to pay the bills incurred by the project. It also helps cover equipment maintenance and staff costs.

Failure Rates

Figure 4 shows the failure rate of telephone calls for data transfer as monitored at the node. All the calls at the node involve the Malawi Telephone system. Data transmitted during a call that fails in some cases have to be re-transmitted because the decompression of the data at the receiving side fails due to incomplete packets or files. Thus a failed call is a major concern to a user on the network. As can be seen from Figure 4, the proportion of calls that fail has come down considerably from nearly 100 percent in October 1993 to about 18 percent in March 1994. This improvement in performance can be attributed to various factors. The international mail route changed in November 1993 from via Zambia to via South Africa. We observed that fewer calls fail on the new route. We have also learned when to make the international calls in order to achieve higher success rates. This information was not available during the startup phase of the network.

Figure 4: Failure rate of telephone connections using a 2400 baud modem.

The number of calls to the node from Fidonet points within Malawi has grown considerably as a proportion of the total number of calls recorded by fdstats.exe at the node. This means that while the performance reported in Figure 4 applied more to international calls in the startup phase of the network, it applies both to national and international calls in March 1994. We can estimate from Figure 4 that on the Malawi telephone system about 18 percent of the calls for data transfer using modems will fail. More data needs to be collected and analyzed to determine the separate failure rates of national and international calls on the Malawi telephone system and to determine the rate which the graph in Figure 4 settles down to.


Network Use

The main services on the Malawi Fidonet network are electronic mail and file transfer. No remote login of any kind is currently supported. We need to conduct a survey to determine and categorize the purposes of sending mail by users within the University of Malawi and outside. During the experimental phase of the network, we have sought such information from users in the University and outside. The message tracking software at the node collects information on the sources and destinations of mail. We have observed that University of Malawi users use the Fidonet network mostly for the following reasons:

· contacts with academic colleagues and supervisors;· negotiating links with other universities at department level;· submission of research papers, proposals and reports;· applying for further studies abroad;· coordinating research for staff on study leave abroad'· library contacts;· seeking funding for research and other projects;· discussion of issues pertinent to Malawi;· ordering equipment from suppliers outside Malawi'· personal mail; and· file transfer using archie and ftp by email.

In using the network, University staff and other users are finding it easier and more efficient than before to request and access information from colleagues and institutions outside Malawi. I hope that as the network grows inside Malawi it will also become easier and more efficient to access colleagues and information sources within the country.

A number of users on UNIMA were already on email in Malawi but they used to make international calls to access email facilities abroad, such as CompuServe or CGNET. Since the UNIMA Fidonet network was established, many of these users have switched to using the UNIMA network and are finding it cheaper and more convenient since it only involves a local telephone call. (See Box 1.)

Toward the end of 1994 and the beginning of 1995, users in non-governmental organizations and commercial companies began making up a fast-growing proportion of the user base. These are high volume users who request and send large amounts of data over the network - mostly involving business related information. I expect that in the near future these users will be the main revenue generators for the running of the network.

Growing the Internet

With more than one hundred points on the UNIMA Fidonet network and over six hundred users, it is time for the establishment of a full Internet link in Malawi. The user base is there to support the network and our experience has shown that the users would also be able to pay the fees that such an installation would demand. At the point of writing, a leased line to South Africa seems too expensive and so we are suggesting that a VSAT connection be tried instead. VSATs or very small aperture terminals are satellites that provide two-way, high data rate services. VSATs appear to offer a low-cost telecommunication solution in developing countries.

Users on the UNIMA network indicate that they would not like flat billing. They would prefer message-by-message billing so that they pay for what they use. We will, therefore, have to carefully examine the current billing practices on UNIMA before full Internet connectivity is installed.

Equipment Shortages

The main constraint on the expansion of networks in Malawi will remain equipment based, with modems topping the list. I recommend to donors that, when projects are funded in a country like Malawi, each project should have an information technology section providing for the supply of computers and modems, when and where these are not available. Donors should also consider the expense of the installation and training of users of the information technology within the project. The minimum here should include electronic mail in any project proposal and implementation.

BOX 1 Reviewing the New Constitution

The UNIMA network made a significant contribution in the development of the new Malawi constitution. It made it possible for international lawyers in Malawi and outside to exchange information in a fast and more efficient way. It promoted discussion and editing as the constitution was sent worldwide for comments and review. This was reported in the local press in Malawi ( The Nation). It was said that, without this mode of communication, the review of the constitution would have been lengthy and it would not have been possible to meet the deadlines for the review.


A project to study computer and telephone network based communications in Malawi has been in progress in the University of Malawi since 1993. Under this project a nationwide electronic network using Fidonet technology has been installed and is being expanded. The network supports electronic mail and file transfer. This is a public network in the sense that anyone or organization with the required resources can link into the network. Most of the objectives of the project have been accomplished.

The project has shown that using the installed network for electronic communications is cheaper and more convenient than existing methods, such as fax and voice telephone. It has also shown that the Fidonet software can be used to generate data on a local telephone system that can be used to monitor the performance of the local telephone network. We expect that such data will be useful to the telecommunications operating corporation.

We have had many positive and congratulatory comments on the network performance and about how it has improved the communication efficiency of our users. So while the project began as a scholarly undertaking, mostly to determine the feasibility of establishing computer and telephone-based communications at the University of Malawi, it has resulted in the provision of improved communication services throughout the country. The small amount of seed money that we received from the University has been effectively multiplied and the impact of the project has been felt far beyond what we had originally intended.


1. The value of the Malawi kwacha has changed considerably from roughly US$ 1 =MK5 at the time when this date was collected to US$1=MK15 roughly et the time of writing of this report.

2. Fidonet uses a hierarchy to pass messages through the system. A Fidonet address identifies the zone. the hub, the node, and the point (if there is one) number of the user. This address, 5:7231/1, means Zone 5 (Africa), hub 7231, and node number 1. See Fall's case study for a more complete discussion of Fidonet.

Background Summary: A Global Computer Network for Change

The Association for Progressive Communications (APC) is an extensive global computer network, dedicated specifically to serving non-governmental organizations (NGOs) and activists working for social change. Composed of a consortium of international member networks, APC provides effective and efficient communications and information-sharing tools to NGOs and individuals. Member networks exchange email and participate in selected electronic conferences with over 50 partner networks worldwide. Many of these partner networks provide the only email access for NGOs in their countries. Through this global partnership, APC offers vital communication links to tens of thousands of NGOs, activists, educators, policy-makers, and community leaders in 133 countries.

APC member networks develop and maintain an informational system that allows for geographically dispersed groups to coordinate activities online at a much cheaper rate than can be done by fax, telephone, or for-profit computer networks. APC wants to assure that electronic communication tools are available to people from all regions in the world. By collaborating with over fifty partner systems in Southern countries, APC helps to increase the information flow between South and South. To further assure equitable access to computer technology, APC initiated a Women's Networking Support Program, which focuses on increasing access to networking technologies and information sources for women and women's NGOs.

The APC Networks are distinguished by the depth and quality of their information resources, by the global reach of their communication services, and by the diversity of their membership. APC developed from a dedicated grass-roots constituency and has become a true global community. The Network provides information in several languages, including Spanish, Portuguese, Estonian, Serbian, Macedonian, Bosnian, Croatian, Slovenian, Russian, English, Swedish, German, Polish, French, and Esperanto. In addition, APC provides extensive user support and training to help people access the system easily and learn to use it efficiently.

The APC Networks are designed for use by NGOs, United Nations agencies, research organizations, government departments, educational institutions, multistakeholder groups, and individuals who communicate with NGOs. People use APC as an information and organizing tool for a wide range of issues, including peace, human rights, public health, environmental issues, population, social and economic justice, development, education, poverty, and demilitarization.

Dial Locally, Act Globally

All APC member networks provide users with an Internet email address. APC users can communicate with any of the over 40 million people who have an Internet email address; as well as with most commercial, academic and other networks. Many APC networks provide navigational tools to help their users access information that is otherwise difficult to locate on the Internet. For example, users can access any public computer on the Internet, including universities, libraries, research institutions, and public databases. Several APC networks make information publicly available on the Internet via the Worldwide Web (WWW) and via Gopher P - a user-friendly, text-based, menu-driven information search tool. Several APC networks offer SLIP and PPP access which allows the use of a graphic interface to Internet services.

APC provides the following networking tools:

· Electronic Mail (email)· Electronic Conferences, both private and public· Databases· Fax and Telex· Internet Navigation Tools: Gopher, Telnet, Worldwide Web, FTP, WAIS· News and Information Services· International User Directory

The APC is a community without borders. NGOs and activists worldwide use APC Networks for both their internal organizational communications as well as their public organizing efforts. In addition, many progressive news services distribute their publications electronically on the APC Networks.

Bringing the Internet to Zambia

by Neil Robinson

Neil Robinson is a Senior Communication Software Engineer with ZAMNET Communication Systems Ltd. He has been working in Zambia for six years. This case study describes the work that led to installing a full Internet link in Zambia and the attempts to make the provision of Internet services a sustainable enterprise, independent of external donor funding


On 22 November 1994, Zambia became the fifth country in Africa (and the very first sub-Saharan country outside of South Africa) to have full access to the Internet, the world's biggest computer network. This achievement was made despite Zambia's official status as one of Africa's poorest nations and followed three years of development by the University of Zambia of an electronic mail network serving non-commercial interests within the country. It was also achieved in a country that lacks a Computer Science degree course and where advanced computer skills are scarce. However, the technological achievement is perhaps less important than the ability to meet the considerable costs of Internet connectivity from within and thus to establish a sustainable service. ZAMNET Communication Systems Limited is the company that has been established to do this.

Computer Skills in Zambia

Despite a steady increase in the numbers of computers in use in both government institutions and private companies, the number of skilled computer specialists working in Zambia is still desperately small. Top quality staff command high salaries and other benefits (housing and transport) that are beyond the reach of government-funded institutions and the majority of Zambian companies. Most of the top computer specialists work in the financial sector or for Zambia Consolidated Copper Mines in the (up until now) well-funded mining sector. There is a growing number of small computer sales and support companies in the capital, Lusaka, but the quality of service they provide is variable. Against such a background' the number of expatriate computer specialists working within Zambia is relatively high.

Zambia's highest national qualification in computing is a three-year computer studies diploma taught by Evelyn Hone College. The University of Zambia has no computer science department so people who wish to advance to degree level have to go outside of the country. The Mathematics Department in the School of Natural Sciences does provide a Mathematics and Computing degree but their department is desperately short of computers and is unable to provide the practical experience necessary in such a course. The locally produced graduates with the strongest computing background come from the Electronic Engineering Department of the School of Engineering. Many who succeed in acquiring the necessary skills and qualifications during training overseas fail to return to Zambia for very long, if at all.

The University of Zambia

The University of Zambia (UNZA) is the larger of Zambia's two universities (the other being the Copperbelt University in Kitwe). It has a student body of some 5,000 and approximately 500 academic staff. UNZA is wholly dependent on the Zambian government for its funding, although several schools within the University benefit from international donor support for their equipment requirements.

The University uses an old IBM 4361 mainframe computer for administrative computing requirements. The systems that run on this computer (personnel, payroll, admissions, examination systems, and so on) have all been written and are maintained by the University Computer Centre, a department of about 50 staff. Some 15 of these are technical programming and analysis staff. The rest are involved in the considerable data entry and computer operations requirements of such a mainframe-based system.

The Computer Centre is also responsible for providing technical support to computer users throughout the University. Up until about four years ago, this was largely confined to support for the mainframe computer to which students and staff were given access through a small terminal room at the Computer Centre. However, as elsewhere in Zambia, the number of microcomputers in use around campus has been growing steadily and the support requirement for the repair of hardware components and for the maintenance of microcomputer applications has also grown.

The vast majority of microcomputers within the University are donated and as such there is a diverse range of models and types of computers. However, in general, there are still far too few microcomputers to distribute access to the student level, and even in the mathematics department most student computer project work is carried out on the mainframe.

UNZA has recently embarked on a major project to downsize its administrative systems to a microcomputer-based network. A campus-wide fibre optic network is being installed to distribute the administrative tasks to the schools themselves. Once again donor supported (through the Dutch government), this should provide access to the first of the systems (student records) by the end of this year. It will also provide access for members of staff through departmental networks to the Internet, although the extent to which this happens will largely depend on the ability of each school to extend its own section of the campus network through its own buildings.


Since 1991, the Computer Centre at the University of Zambia has been the focal point for academic, non-governmental organization (NGO), and health-based email networks in Zambia. In that year, as a result of the International Development Research Centre (IDRC)-funded ESANET1 project, Zambia was given a microcomputer and modem to provide the hub or "host" of the first University email system. The very first recorded message through this system was sent to the Baobab, an African interests network based in Washington, D.C., on 30 September 1991. This and other international messages were initially routed through the Association for Progressive Communication (APC) network in London. The first recorded user of the system was the School of Medicine's Medical Library.

The rapid growth and development of what was then known as UNZANET was assisted by an arrangement with Rhodes University of Grahamstown, South Africa which, in November 1991, began providing a link with Internet via thrice-daily, computer-to-computer telephone calls. These were paid for by UNINET, the South African Universities Network funded by the Foundation for Research and Development (FRD). This arrangement remained until December 1994, when full Internet access was achieved and by which time some 270 email points were linked to the network.

The Early Technology

From the beginning UNZANET was a Fidonet system using dial-up telephone lines to transmit messages via dedicated host computers. When UNZANET was first established in 1991, Fidonet had been successfully tried and tested in Zimbabwe and in Kenya for NGO networks.2 Thus UNZANET was able to learn from the experiences of similar networks in the region and benefit from cooperation with them.

The UNZANET host system was configured to handle simultaneously three telephone lines attached to three separate modems. Two of these lines were direct external lines, while the third was an internal switchboard line for users on campus and with access to an internal telephone extension. The configuration of the Fidonet system remains in place today, primarily to serve rural users and those whose telephone lines are inadequate for interactive communication. It also serves users in advance of the completion of the campus network.

UNZANET'S Traditional Users

The original users of UNZANET were the schools and departments of the University. From the very beginning, a single site address or point number 3 was allocated to each school and administrative department. In the majority of cases, computers already existed in those departments and the email facility was installed using modems supplied as part of the ESANET project. Some of the computers used were old IBM personal computers, which were still perfectly adequate for email.

From very early on, it appeared to be inevitable and desirable that initiatives in academic networking would lead to collaborations with other sectors. Without expansion of the service into other, better-funded areas, the UNZANET system could never be sustainable. The development of a non-academic user base was encouraged especially for those users who had an immediate need for communication and for those who would have a growing need for communication in the future. These users would want to be part of a longer-term, more sustainable solution. However, with UNZANET's dependence upon donor support in mind, and particularly that of UNINET, the emphasis was always on non-commercial applications.

In Zambia, following the use at the University itself, there was widespread growth in the use of email among NGOs, health institutions, and aid or development agencies. Encouragement for the first two came about through external support, while the majority of aid and development agencies were able to fund the necessary infrastructure (modems, computers and telephone lines) themselves.

UNZANET'S Free Service

One of the reasons for the rapid growth of UNZANET was the fact that it was provided as a free service. Users only had to cover the cost of local calls into the UNZANET host or access the system through the University's internal switchboard. This allowed users to "taste" email and to experiment with it before they had to commit themselves to paying for the service. We feel that the growth of UNZANET would have been considerably slower had it been necessary to charge users from the beginning, although a culture of non-payment can have its disadvantages when a subscriber-funded service is introduced.

The Need for ZAMNET

Through the technical success of UNZANET, it was clear that email was a viable technology within Zambia. This was true not just for a rich urban elite, but also for smaller, less-wealthy organizations within Lusaka, the Copperbelt, and further afield. We had shown that email had the capacity to remove some of the communications barriers that otherwise engender a feeling of international isolation in professional communities. At the same time it was attractive to a large community of users outside of the academic sector. Email offered a fast, easy, and highly cost effective method of communicating compared to any other available technology. (See Box 1.)

However, a free service - provided through the generosity of others - could not be sustainable. The Computer Centre of the University, with its responsibilities to the University administration, was not in a position to sell and maintain a commercial service to customers from outside of the University, nor to pay the salaries of the high quality staff that such a service would require.

By the end of 1993, users were transmitting large volumes of international messages to and from the Internet, and there was a strong desire among users of UNZANET to expand the horizons of the system. The Internet was already expanding very rapidly: more and more data, journals, and information became available in electronic form; the nature of the information being offered broadened beyond the bounds of academic interest into business and recreation, and publicity about it was reaching the popular public media. The ability to transfer large files to and from other users; to directly access remote databases and electronic libraries; and to participate in electronic conferences and other peer group discussions had become of considerable interest both within and outside of the University.

To accommodate this growth and change in direction, the basic infrastructure of the UNZANET system and the way it was funded needed to be reviewed. The Computer Centre had already perceived the need for collaboration between the University and its fellow research institutions, as well as with the commercial and international organizations that had the resources to pay for the high quality, high volume access to information. Some of these other organizations were using UNZANET but, because they needed to transmit larger volumes of information more quickly, many had made their own direct links to their international head offices using a variety of electronic mail systems. Pooled access for these organizations to the Internet through a single service provider would clearly bring considerable cost savings, while improvements in speed and efficiency of communication, as well as the extra services available, would allow international organizations to better service their target communities.

BOX 1 Pooling Resources

At times it was suggested that the users in the NGO and health sectors should form their own networks (ZANGONET and HealthNet respectively) - independently of UNZANET. However, the need for a full-time system operator and the considerable advantages of pooling equipment and staff resources with UNZANET, dictated that both groups be maintained and administered from the university Computer centre. The advantages of such an arrangement are even more apparent under the current self-financing service, which enables non-commercial NGO's and health institutions to be subsidized to some degree by ZAMNET's 'commercial' customers.

The ZAMNET Proposal

By the middle of 1993, we made a decision to submit a proposal for donor funding to enable the University of Zambia to establish a direct link to the Internet. This was circulated among a number of major international donors but failed to attract any direct support. At the beginning of 1994, the University decided to establish a private campus company (ZAMNET Communication Systems Ltd.) to put in place the connection to the Internet and to sell access to the services that the Internet provides. Then the World Bank expressed interest in funding the ZAMNET project and agreed to fund 80 percent of the first year operating costs of the new company.

As a young campus company with no working capital aside from the grant from the World Bank, ZAMNET issued a nominal share holding, the major and controlling shareholder being the University itself. The board of directors is comprised of senior administrators from the University and the Deputy Minister of Health, with the Director of the Computer Centre as Managing Director.

The Budget

The project budget was estimated at $150,000 and covered the cost of:

· the leased data circuit to the chosen (cheapest) point of Internet access;· subscription paid to the service provider at that point;· extra hardware required to provide the Internet link and to provide a service to ZAMNET's customers;· hardware required to provide a training facility for ZAMNET's customers;· consultancy fees for technical Internet experts to assist in the configuration of the service; and· salaries for one administrator and two technical staff (to be met from customer fees from the beginning).

The Proposed Service

We intended to distribute Internet services to users throughout Zambia using a variety of methods including:

· local leased lines providing full Internet access to regular high-volume users with their own local area networks (LANs);· dial-up interactive access using SLIP or PPP for urban users on good quality telephone lines; and· Fidonet and perhaps UUCP access for users in rural areas where poor telephone quality prohibited interactive access.

We anticipated that Fidonet would continue to play an important role within Zambia providing as it did a quick, easy-to-use entry to electronic mail. It had already proved to be relatively cheap on remote trunk lines and generally reliable in a rural setting.

The Project Objectives

ZAMNET Communication Systems Ltd. came into being officially in February 1994, although it was unable to employ staff or provide a service to its customers until the following year. ZAMNET has the following objectives:

1) Put in place the hardware and technical support necessary to provide a reliable full Internet service. To do this, we would need to:

· establish a reliable link between the UNIX network at the University Computer Centre and an Internet service provider;· procure, install, and configure all necessary hardware to provide an Internet service;· provide an access point for customers to dial into the Internet with sufficient capacity to avoid congestion;· provide a software package for customers to use to access the Internet through ZAMNET;· connect the existing Fidonet service to the new Internet service, thus removing the dependence on the Fidonet-Internet gateway at Rhodes University;· develop locally based information services for local and international access;· provide a training facility and develop training courses for customers; and· produce or obtain user documentation for the system.

2) Put in place the administrative staff and procedures, and market the Internet service so that it would be fully self-sustaining after one year of operation, and after the World Bank funding had run out. To do this, we would need to:

· set up a computerized accounts system;· determine a pricing system; and· produce marketing materials (pamphlets, brochures, price lists, advertisements).


Although ZAMNET has been operating commercially for less than five months (at the time this report was written), the financial support from the World Bank is soon to come to an end. Thus, while in many ways the project is still ongoing, this is a good time to review progress on the ZAMNET objectives during the establishment of the Internet service to date.

Establishing the Link to the Internet

The establishment of the data communications link between Lusaka and (Cape Town, critical to our service, proved very slow and time consuming. The Internet Company of South Africa or TICSA (now Internet Africa) of Cape Town had already generously agreed to carry our Internet traffic for free (for the time being) when, in April 1994, we submitted the application to the Zambian PTT for a leased, four-wire, 9600 baud, designated data circuit between our own offices in Lusaka and the TICSA offices. After the necessary surveying work, we were notified that it would be commissioned in July 1994. However technical problems over the satellite link between the Zambian earth station at Mwembeshi and Johannesburg meant that the line was not available to us to test until October of that year. Unfortunately our problems were not over.

Our original choice of modem to serve both ends of our Internet link was the Telebit Worldblazer. We had good experience with Worldblazers through our work with the Fidonet system. However when we installed the leased four-wire data circuit between Lusaka and Cape Town, we did not know that the Worldblazer is purely a two-wire modem and, therefore, unsuitable for the type of circuit that we were using. Unfortunately our email messages (our cheapest and what should have been the easiest route to assistance) to Telebit often failed to solicit a response and it required several long distance telephone and fax calls to establish the facts and to find out what options were available to overcome this problem.

Our first solution was to install hybrids to convert the four-wire circuit to two-wire over the local segment of the circuit in Cape Town and Lusaka. Both the Zambian and South African PTTs were extremely helpful in immediately inserting hybrids into their circuits. However, the resulting signal contained too much echo for the Worldblazers to cope with and, despite the further insertion of echo suppressers into the circuit by the PTTs, the resulting signal was inadequate for the Worldblazers at either end of the circuit to communicate with each other.

Eventually, after seeking advice from various sources, we decided to purchase two (quite expensive at $1,100 each) Telebit Fastblazer modems through their South African agents. Despite the availability of email, communication with them proved as frustrating as trying to communicate with Telebit in the United States. However, eventually, two modems were purchased and installed. Alas our problems did not end there.

The line between Cape Town and Lusaka uses microwave technology as far as Johannesburg, satellite to the Lusaka earth station, microwave once again as far as Lusaka's main exchange, but then uses copper wire for the last six or seven kilometers to the University. The signal loss over this section is high and the signal coming from South Africa was already quite low. Over such a long distance, the quality of the line allocated between central Lusaka and the ZAMNET office was highly variable and it required some intensive work from the Zambian PTT to identify the very best connections to provide for our circuit. However on 22 November 1994 the line was finally strong enough to carry a signal from our mail server to the Internet and vice versa.

The line has since been quite reliable (above 95 percent), the modems training up to 14,400 baud. Although capable of connecting at 19,200 baud, it appears that the signal loss over the circuit as a whole will prevent us from achieving such speeds. When the data circuit does fail, the technical staff at the two PTTs (ZAMTEL in Zambia and TELKOM in South Africa) have been quick to respond and restore our circuit.

Procuring, Installing, and Configuring the Internet Hardware

Most problems encountered during the installation and configuration of the hardware were due to our own inexperience with UNIX and our ignorance of much of the technology with which we were dealing. The staff involved were subjected to a very steep learning curve and inevitably much of that learning was through our own mistakes. (See Box 2.)

The mail server

Through the RINAF (Regional Informatics Network for Africa) project, the University Computer Centre had received two identical Olidata (Olivetti) 486 computers. One of these was pressed into service as a new Fidonet host to cope with the rapidly growing number of users accessing that host. The other was set up on the Computer Centre network as a prospective mail server (

A copy of SCO UNIX was installed on puku and this machine was configured to run SENDMAIL (for mail delivery), to run a POP (post office protocol) server (for mail collection by ZAMNET's customers), and to exchange mail with the Fidonet host using UUCP. We experienced some difficulty finding free Internet applications software to run under SCO UNIX (a commercial product) and' when we employed an expert to assist in the configuration of our mail server, we were advised to switch to FreeBSD-2.0. This was done in December l994 - after the Internet connection had been commissioned.

FreeBSD is a relatively easy operating system to configure and to manage, but finding precompiled binary versions of Internet applications has proved difficult. Compilation of ports and patches is cumbersome for those inexperienced in UNIX and keeping abreast of the constant updates is time consuming. The only alternative free UNIX system that runs on a 486 personal computer is Linux. Having already started with FreeBSD, there are few advantages in switching at this stage.

Puku is now running as the ZAMNET name server, as well as the server for mail/POP, Gopher, and WWW. All of these services appear to be running quite well. However from a hardware point of view we have discovered that a system with eight megabytes of RAM and a 400 megabyte hard disk is insufficient to tackle the range of tasks and to serve the number of users that we are asking of puku. A gigabyte hard drive and 32 megabytes of RAM is essential on such pivotal system within an Internet service.

The router

The router required to connect ZAMNET to the rest of the Internet - a Cisco 3000 - was provided as part of the funding of the Zambian regional node by the RINAF project. ZAMNET itself purchased a transceiver to connect this router to the ethernet network. The Cisco gave no problems in terms of configuration; how ever since its installation two problems have arisen.

First, one of the primary requirements of the router is to provide statistics the can be used to bill ZAMNET customers who link their networks with ZAMNET' using locally leased data circuits. Such customers are billed for all traffic carried. The data to enable this kind of billing needs to be analyzed on another computer of the network that is running UNIX. We have yet to find any software to analyze this data that will run under the FreeBSD operating system on puku, our mail serve. The only such software that we have found is written for the SUN operating system (SUNOS) and we have since been advised that nobody would attempt to do Cisco accounting on a FreeBSD system!

The second problem is in trying to increase the speed of our link to the Internet. That link currently operates at 14,400 baud. It seems that by switching to an asynchronous connection using data compression it should be possible to increase the throughput of data on our link. To do this, however, the Cisco needs to have its auxiliary port configured to operate in asynchronous mode, which in turn requires a memory upgrade and a new version of the operating system to be loaded into the increased memory. With our whole system dependent on that one router we are somewhat reluctant to attempt this upgrade.

Providing Customer Access

To provide access to the Internet service to as many of ZAMNET's customers as possible simultaneously, 20 dial-up telephone lines configured as a hunting group were installed by the Zambian PTT, and a Livingston Portmaster 2e-30 with 30 configurable ports was purchased along with 20 Zyxel U-1496 modems to connect these lines to ZAMNET.

The telephone lines were installed when a new cable was being laid into the University and there were no problems in finding 20 new lines for ZAMNET. The hunting group is functioning well, although up until now (with 193 active, paying users as at the end of June 1995), we have rarely seen as many as seven of the lines in use at one time.

The modems that the ZAMNET customers dial into were purchased by mail order directly from Zyxel in California. Dealings with that company by email were very easy. The modems were purchased before the V34 (28,800 baud) standard had been ratified and they are all set to a maximum speed of 14,400 baud. However, given that this is also the maximum speed on our data circuit to the Internet, and that as yet there is no congestion on the dial-up lines, there is no disadvantage in restricting our customers to connections at this speed.

The Portmaster was chosen over a Telebit Netblazer on the grounds of its cheaper price and that its range of features was closer to our requirements. However we were unaware when purchasing the Portmaster that some of the software that is provided with it and that enables easy configuration and backup of the system would not run under the FreeBSD operating system that we are using on our mail server. The inability to back up the tables used to store user names and passwords has proved a serious weakness on two occasions when part of the user table was lost and had to be manually re-entered.

Provision of End User Software

For ZAMNET users continuing to use the existing Fidonet service or leasing local data circuits between their own LANs and ZAMNET, we did not need to provide any new software. The Fidonet installation had already been developed and proven over a number of years, while those linking through their own networks were responsible for their own software.

However' for our dial-up interactive customers we needed to find an easy-to-install, easy-to-use package for Windows, DOS, and Macs. When we found a set of books for each of these three categories, complete with software diskettes, we thought we had found the solution. However, we discovered that the DOS software contained a hard-coded login script that did not match the prompts provided by our Portmaster; the Mac software would not run on the systems on which we attempted to install it; and the Windows software contained frustrating bugs in its Mail program.

Software for DOS

The package distributed to DOS users is a slightly modified version of the SLIP/MINUET package developed by the University of Minnesota. While not as attractive as the Windows software, it does contain an easy-to-use mail program and a Gopher client. It lacks a WWW client and we are investigating how DOSLynx (a WWW client for DOS) might be integrated into the installation.

Software for Windows

After the frustrations encountered above, we borrowed the package distributed by GreenNet in London for their own PPP customers. Their package is based on shareware versions of the Trumpet Winsocket Manager, Eudora as a mail client, Netscape as a WWW client, plus FTP and Telnet clients. We modified this package for our own system and developed our own installation program. This package is stable, easy-to-install and has been found to be easy to use by our customers.

Software for Macs

The vast majority of computer users in Zambia use DOS-based computers. Consequently, providing support for the few Mac customers that ZAMNET has, has proved very difficult. The package distributed to Mac users is a combination of the diskette provided with the Internet Tour Guide and a second diskette obtained from GreenNet that they distribute to their own Mac users. The resulting package includes the Eudora mail client and Netscape and is very similar in its two main components to the Windows package. ZAMNET still does not possess its own Mac and so streamlining the installation is still some way off. Even copying the two distribution diskettes requires a visit to one of the two private individuals on campus who use Macs!

Integration of Fidonet and the Internet

When the Internet link was established in November 1994, we had just three weeks in which to commission the gateway between the Fidonet host and the mail server before UNINET started charging us for the use of their own gateway. On the recommendation of Bob Barad of the Baobab, we installed the new GIGO package and were pleased to find that it was very easy to install and configure, and what's more, it worked. Rhodes University had given us an experimental UUCP account and, for the three weeks prior to cutover, we polled (generated a computer call to) the Rhodes WCP host on a regular basis. We encountered no problems with this link and the modification of the software to enable connections between the Fidonet host and our own mail server went equally well.

The only disadvantage of the link at this stage is that it requires a modem and telephone line to be dedicated to the mail server for the UUCP connections (none of our users use UUCP directly). Even using local phone calls at 14,400 baud the telephone bill for this exchange of mail between adjacent personal computers is quite high. It would be better to move this mail over the LAN if possible.

Development of Local Information Services

The ZAMNET Gopher

In December 1994, a Gopher server was set up ( with a menu structure covering Agriculture, Engineering, Health, Communications and Networking, and ZAMNET News and Information. This proved to be a very useful way of publicizing our work and making contact and service information available to potential customers and other interested parties around the world. We also found the Gopher useful for advising our customers of the latest developments (or problems) on ZAMNET.

As customers became aware of the potential of the ZAMNET Gopher for the distribution of local information, they expressed an interest in storing their own information within its menus. To date a number of customers are paying for disk space on our Gopher server at the nominal rate of $1.00 per megabyte per month. For example:

· the National Farmers Union is providing weekly Agricultural Commodity Exchange prices;· the Ministry of Agriculture posts regular food security and marketing bulletins, and· the Engineering Institute of Zambia is publishing a range of information about the institute and its activities.

The Worldwide Web server

Our knowledge of the Internet has grown end we are aware of the rapid growth of the Worldwide Web (WWW) and the trend towards storing more information in hypertext form. Around April 1995, we installed our own WWW server. The ZAMNET homepage is accessible as and contains pointers to information about Zambia, about ZAMNET, to the ZAMNET Gopher, and to a small but growing number of pages developed for ZAMNET's customers. For example:

· articles from the Post Newspaper are currently published twice weekly before the paper actually reaches the streets;· the Zambia National Tourist Board has published information about travel and tourism, and· the University of Zambia has published the text of a short leaflet giving background information about the University.

We plan to develop all of these areas. In particular several commercial customers are interested in posting information about their companies and services within our WWW pages.

Provision of Training Facilities and Courses

When ZAMNET was formed in 1994, it occupied one small office within the Computer Centre. In April 1995 we got our own offices that include a training room with capacity for nine networked personal computers. The room was equipped in May with seven multimedia Compaq personal computers and we have set aside three mornings per week to provide training to ZAMNET customers.

In addition, this facility has been used to provide sensitization seminars to the staff, deans, and administration from each of the schools at the University. When not in use for training, it is open for these members of University staff to book for an hour at a time, and is also open for members of the public to use at a rate of $7.00 per hour.

User Documentation

The Internet is a new concept to many people in Zambia and there are few books or magazines about it in the bookshops in the country. Since it is important for customers to have access to information that helps them make the best use of the service they are buying, we decided to provide a book, The Internet Tour Guide, with the ZAMNET subscription. In practice, as explained above, the software included with the books has not been very useful and the books are very American in style and language. However, there are few alternatives until we can produce our own software specific guides. We have spent some time producing detailed installation instructions for the software that we distribute and these certainly appear to have reduced the number of queries that we field during customer installation.

Computerized Accounting

The initial chart of accounts was put together by an accountant who has subsequently kept an eye on our progress. We are using Mind Your Own Business (MYOB), a small business accounting package developed in the United States. While quite easy to use, it has been not been adaptable to the multicurrency system we operate (we accept payment in Zambian Kwacha and in U.S. Dollars). We are considering replacing it with a package better suited to our own particular requirements.

Determining a Pricing System

No two Internet service providers use the same method to charge for their services. Our own method of arriving at our fees was to draw up the budget over the next few years; decide how much emphasis we wanted to put on the basic subscription, the cost of international electronic mail, and the cost of connect time, estimate the number of customers and projected growth rate in each category; and to fiddle with the parameters until we could be sure of breaking even within one year.

Since we need to support our traditional users, we added a two-tier pricing structure for commercial and one for non-commercial customers. The resulting fee structure - which entails a signing-on fee, a basic monthly subscription, a per kilobyte charge for international email, and a per hour charge for connect time - has been well received within the country, although we have received some criticism from without. We are always reviewing these fees and plan to increase the free connect time to two hours per month. As yet ZAMNET is not sufficiently financially sound for any radical discounting of the prices.

Marketing of the Product

To promote ZAMNET to a largely unaware population, we employed a graphic artist to produce a leaflet and eight page brochure. This determined the "corporate image" of ZAMNET and its style has been copied on price lists, business cards, and advertisements. Over 1,000 brochures have been distributed so far. Advertisements have been placed in the Times of Zambia, Productive Farming, and Profit Magazine. However such has been the interest in the Internet that ZAMNET has benefited from free publicity in articles in all three of Zambia's leading newspapers, plus a lead article in Profit Magazine.

Staff Recruitment

The recruitment of the administrator/bookkeeper with a full Association of Accounting Technicians qualification and solid accounting experience proved a lot easier than the recruitment of the technical staff, although an initial plan to recruit a part time administrator was revised when it was fully appreciated just how much would be involved in administering the ZAMNET service. (See Box 3.)

Of only 30 applicants for the post of communications technician, six were suitable to interview. Few of these had UNIX experience, and none had any practical experience with internetworking technology. While one member of staff was recruited from the Computer Centre and had been involved in the development of the system from the beginning, the only external recruit is very much learning as he goes along.

BOX 2 Danger from Lightning Strikes

A real problem during Zambia's rainy season is the threat of high intensity lightning strikes. All telephone lines and the leased data circuit to cape Town are protected from lightning strikes using a surge protection box that diverts current surges down a thick copper wire into a copper stake driven into the ground outside of the equipment room. So far this system has thankfully not been tested by such a lightning strike.

BOX 3 ZAMNET Salaries

Even at the relatively (for Zambia) high salaries being offered by ZAMNET, recruitment of the right quality staff was difficult. The package on offer of a two year contract with a salary but no fringe benefits was perhaps not so attractive in a country where housing and transport costs are so high and many employers still provide housing (or housing allowances) and transport to senior information technology professionals.


As part of the billing process it has been necessary to gather comprehensive statistics about the use of the ZAMNET system from the very beginning. A very simple analysis of the connection time and email statistics reveals very rapid growth of the service. ZAMNET is growing at roughly the rate of one new account each day and at the current rate this will lead to ZAMNET more than doubling in size over the next twelve months.

Table 1 shows the growth in the number of interactive accounts and the connect time that those accounts have generated. Growth has been constant - with accounts connecting to ZAMNET for an average of just over four hours per month, although the average number of connections per month has increased from 38 in March to 46 in June. This increase perhaps reflects the increased regular use of the system for electronic mail rather than Internet browsing.

Table 2 shows the steady growth in the volume of email, although it should be noted that the mail volumes include mail from the Fidonet system which, unlike the connect time data, would have been present before March. Nonetheless the volume of international mail being sent has nearly doubled in two months. This does not include incoming mail. Interestingly, while the volume of mail to international destinations has increased steadily the volume of mail to local users has remained nearly constant, perhaps reflecting the international requirements of the newer users, and the benefits of ZAMNET in economizing over traditional communication methods (fax, telephone and courier).

TABLE 1 Growth of ZAMNET by Month, January 1995 to June 1995


Number of Accounts

Number of Connections

Total Connect Time





















TABLE 2 Growth of ZAMNET by Month, March 1995 to May 1995
































A Breakdown of the Users

The majority of urban users of the non-commercial Fidonet service prior to ZAMNET's formation have now subscribed to the Interactive service. Those remaining with Fidonet include
· Users in rural areas (notably the health community) from where telephone calls to Lusaka are expensive, and telephone line quality is often too weak to support interactive communication.· University users who will remain on the Fidonet system until the installation of the campus network has been completed.· Long term users of the Fidonet system who are due to leave Zambia shortly and therefore do not wish to upgrade to the new interactive service.· Some United Nations agencies that are currently planning to connect their own network directly to ZAMNET with a local leased data circuit but, in the interim, would prefer to continue to use the technology with which their users are familiar.

Looking at the geographical and categorical breakdown of the Fidonet and interactive subscribers, it is clear that the vast majority of users of both systems are Lusaka based. (See Table 3.) While this is not surprising, the significantly smaller concentration of users on the Copperbelt, Zambia's other major urban area, is more unusual and perhaps can be explained by the greater difficulty in fully supporting a service to users who are based outside of Lusaka. It will be necessary to consider a point of presence on the Copperbelt to rectify this and also to consider ways in which greater support can be provided to rural users.

TABLE 3 Users by Province





Lusaka Province








Southern Province




Western Province




Eastern Province




Northern Province




Northwestern Province




Luapula Province




Central Province








Analysis of the new interactive subscribers by category is particularly difficult. (See Table 4.) Beyond learning that companies are commercial, no attempt has been made to determine the line of work in which a subscribing company is involved. Many of the private individuals joining ZAMNET as non-commercial subscribers are attached to development organizations or international NGOs and use their points professionally. Again we have made no attempt to survey the uses made of email and so further analysis is impossible.

TABLE 4 Users by Category









































Tech Training




















Apart from the 55 private individual accounts and 46 commercial accounts, the largest category among the rest of the accounts is in health. As indicated by the number of accounts still using Fidonet, many health and agriculture accounts were users of the previous non-commercial service. Since the new ZAMNET service was put in place there has been a growing interest from a number of religious organizations active in Zambia (most of whom have head offices in the United States or Europe), and encouragingly recent interest from the Ministries of Finance and of Foreign Affairs.


ZAMNET has received very positive feedback from its customers within and outside of Zambia. Zambian expatriates write to us saying how proud they feel that Zambia is only the fifth African country to establish a full Internet service. Starved of information about their home country, they are eager to see expanded news and information services through ZAMNET.

ZAMNET customers within Zambia are particularly pleased with the cost savings that communicating by email has brought. With international telephone calls to North America and Europe billed at $7.00 per minute, the ability to make a cheap local telephone call and send an email message for about 20 cents per page is very attractive. To find that this message is delivered reliably and within minutes is an added bonus. (See Box 4.)

Media Coverage

As the Internet has received more and more coverage in the international media, ZAMNET has correspondingly come under the spotlight within Zambia. The Zambia Daily Mail, the Times of Zambia, and the Weekly Post have all published articles about the Internet and its arrival in Zambia (the Post subsequently subscribing to the ZAMNET service). Profit Magazine, Zambia's leading business magazine published a full leading article about ZAMNET complete with a front page image of some example WWW pages captured from the Internet.

The Zambia National Broadcasting Corporation has regular computing and business programs that make reference to information technology and the Internet. We hope that a full interview with the Managing Director of ZAMNET will be broadcast soon, while plans exist to connect ZNBC to ZAMNET in the near future.

Further afield, the efforts that Zambia has made to provide an email service and to establish a full Internet connection have been mentioned several times on the African service of the BBC World Service. In fact when the BBC began using email and started accepting questions to their Pop Science program, the very first email question that they received came from an email user in Ndola, Zambia.

BOX 4 Email to Fax Service

ZAMNET has recently reached an agreement with GreenNet in London that allows ZAMNET customers to send email messages to fax machines. These messages are delivered to London via the Internet and are then delivered from there at the cost of the delivering phone call. For faxes to Europe and to North America, this results in a total fax cost of less than 50 cents per page. An informative acknowledgment email message confirms to the sender whether or not the fax has been delivered. Delivery times are usually less than 20 minutes after the email has been sent, which is perfectly satisfactory for ZAMNET's customers.


Just four months after the ZAMNET product was made available to the public, its existing infrastructure is already creaking. Its communications link to the rest of the Internet is becoming congested at certain times of the day. Its mail server is overloaded and does not have the memory or disk capacity to cope with a significant increase in either the number of ZAMNET customers, or in the number of Internet users from outside of Zambia accessing its information services. Without a doubt the main area where ZAMNET might have been launched differently was in its technical capacity.

Capital Funding

ZAMNET's initial budgets should have been significantly higher and included items vital to the provision of a high quality, high volume' large customer base service. Namely, in hindsight, we should have begun with a SUN workstation, VSAT communications, a router capable of handling asynchronous communication, and a full set of equipment to provide backup in the event of system failures. While these items would have increased the initial cost of ZAMNET significantly, repayment of that cost could have been spread over a number of years.

Staff Training

Providing technical support from a position of considerable inexperience has been difficult. While the staff currently in place are learning fast, sustainability of the ZAMNET service is not only dependent upon them but also on the availability of suitable staff to supplement and replace the existing staff in the future. With this in mind there needs to be greater opportunities for Zambians to learn about Internet technology without having to travel outside of the region.


ZAMNET is providing an important and much needed service to all sectors of Zambian society:

· businesses and industries need a fast, reliable and economical communications system in order to compete within the newly liberalized Zambian economy;· international development organizations need to keep in regular touch with their projects in the field and their head offices overseas;· the academic sector and researchers need access to the latest research and need to keep abreast of international developments in their field;· government ministries need an efficient means of communication with district and provincial offices in the fields of agriculture, health and education; and· private individuals wish to keep in touch with their friends abroad or simply wish to use the Internet as a vast encyclopedia.

Both commercial and non-commercial customers have shown a willingness to pay the fees that ZAMNET has set in order to cover the considerable costs of its service.

Because its customers now rely upon its service, ZAMNET needs to ensure its future both technically and economically. At the current, very fast growth rate, the system will be overstretched within twelve months, by which time the customer base should have more than doubled to over 600 accounts. The resulting reduced performance could seriously damage the positive image of ZAMNET. Meanwhile the lack of technical backup within the system leaves the whole service vulnerable in the event of an equipment failure. The downside of this is the capital expenditure involved in safeguarding against possible disasters. ZAMNET urgently needs to look at ways it can improve and upgrade its services.

Increasing the Bandwidth

The capacity of the Internet link between Lusaka and Cape Town will not be able to support 20 simultaneously connected users without a significant and noticeable deterioration in performance. The most promising option for increasing the capacity of this link appears to be a direct VSAT link from the ZAMNET offices in Lusaka to our Internet service provider in Cape Town, or failing that, an alternative link direct to the United States or United Kingdom. Although continued cooperation with counterparts in the region would be desirable, progress in negotiations with Telkom (the South African PTT) have not been productive so far. Based on the cost estimates provided so far, we know that the hardware costs of a VSAT link, plus annual rental and license fees should prove cheaper than the line we are leasing from ZAMNET and TELKOM at a cost of over $60,000 per year. The disadvantages of this strategy will be the expertise expected of the technical support staff and the cost of paying for both the existing and the VSAT link during the transition. Indeed it may be desirable to maintain the existing link to provide redundancy in the event of failure of the VSAT connection.

Increasing Dial-up Access

With the current customer base, we have rarely observed more than eight ports on the Portmaster modem server active simultaneously. Further expansion to serve 600 customers, while also accommodating increased activity among the existing users, is likely to put severe pressure on our dial-up. This expansion can only be accommodated through the purchase of a second modem server and a further batch of dial-up telephone lines. Over the next year a number of leased line customers will also subscribe to ZAMNET thus putting further pressure on the number of free ports available on the existing system.

Covering for Equipment Failures

The current system is highly vulnerable in the event of any kind of equipment failure. ZAMNET service would be lost in the event of:

· the loss of the leased line modem at either end of the data circuit that links ZAMNET to the Internet;· the loss of the router linking the ZAMNET LAN to the Internet; or· the loss of the Portmaster providing local access to the Internet to ZAMNET customers.

None of this equipment is readily available within Zambia and, even if finances were available, the resulting loss of service while replacements were being shipped from abroad could last for several days and have a serious impact on the image of ZAMNET among its customers. Spare items to replace those listed above would cost about $8,000.

Increase Capacity on Mail and Information Servers

The current mail server is running on a 40 MHz 486 PC with just 8 megabytes of RAM and a 400 megabyte hard disk and is already overloaded. As the number of customers and the volume of local information being provided to users outside of Zambia increases, the burden on this machine will also increase. A second machine is currently being configured as a News Server to accommodate the growing number of useful log files that the system generates. However this machine is itself only a 25 MHz personal computer, albeit with a one gigabyte hard drive.

ZAMNET needs to invest in a much more powerful personal computer with at least 32 megabytes of RAM and another high capacity hard drive. Ideally ZAMNET would seek to obtain a SUN workstation or equivalent, which is capable of running some of the software needed for maintaining the Portmaster and for monitoring traffic passing through the Cisco router. However the cost of such a SUN machine would be $10,000 or more!

Expanding the Range of Provided Services

There is an immediate need to provide a News Server to enable ZAMNET's customers to participate in the many discussion groups available over the Internet. Work to configure such a Server is currently under way, although the impact of a full news feed on the limited bandwidth of the Internet connection has yet to be determined. It is likely that such a feed would be dependent on the implementation of the VSAT link.

As the Internet develops further, other applications will be developed and ZAMNET needs to be in a position to make these applications available to its customers (within the limitations of the bandwidth of its Internet connection). Current examples include "Real Audio" and the ability to communicate by voice over the Internet.

Increasing Technical and Administrative Support

ZAMNET currently employs just three staff, one administrator and two technical support staff. This team is severely stretched in its efforts to provide support and to develop the ZAMNET staff. We have decided to employ an assistant administrator to ease the workload.

With the anticipated growth, it will be important for ZAMNET to employ at least one additional skilled and experienced technical member of staff within the next twelve months. With both the contracts of the ZAMNET technical staff due for renewal at the same time, the possible impact of these employees departing at the end of their contracts also needs to be anticipated.

Looking Forward

This chapter was written in June 1995 and I want to provide a quick update. By January 1996, ZAMNET had grown to accommodate 417 interactive accounts. These accounts generated 9,558 connections totaling 1,600 hours of connect time and 12,862 international messages. This is a growth rate of 100 percent in just eight months. We have taken several steps to ease the growing congestion on both the Internet link and the server computers and to provide protection in the event of system failures:

We ordered the VSAT terminal equipment and it has been shipped from the United States. We need to complete some administrative procedures before it can be used but we hope that a 64 kilobaud VSAT connection direct from ZAMNET's offices to Johannesburg will be operational before the end of March 1996.

We maintain the Portmaster users table on the WWW server machine using Livingston's Radius software.

We moved the WWW server from the mail server to a second computer, which has been configured as a News Server, although we still do not receive a full news feed from the Internet. We just ordered a new mail server (a Pentium 120 with 32 megabytes of RAM and a 2 gigabyte hard drive).

We have recruited a new assistant administrator.

Within the next year ZAMNET plans to increase the number of access telephone lines to 40 and purchase an extra 20 V34 modems and a second Portmaster to serve these lines; employ two additional technical staff, a marketing manager, and another junior administrator; and provide local telephone access to customers based in the Copperbelt.

We have proven that the provision of an Internet service is viable and can pay for itself. The costs of expansion require capital expenditure and' ultimately' an injection of capital from investors either in the form of loans or in the broadening of shareholding is required. Most importantly, however, several months before the end of the World Bank funding, ZAMNET is self-sufficient and is able to buy new equipment from its own funds and to guarantee repayment on any loans it requires. This healthy position should enable ZAMNET to significantly reduce its fees within the next three months and to continue to expand services to meet it customers' demands.


1. The East and Southern African Academic Network (ESANET) is described more fully in Musisi's case study in this volume.
2. Moussa Fall gives a more detailed description of Fidonet in his case study in this volume.
3. Each member of the Fidonet system has a unique address. This address identifies the zone, the host, and the node. The point number will ensure that a message gets to a specific user in the node's subsystem.