11.1 FLOSS and Ethiopia

In this section I will first make some reflection on the effective use of the Internet and then of FLOSS in Ethiopia. I will use the term effective use as described in section 2.3.1. That is, I will look at the physical, digital, human and social resources available for the effective use of the Internet and FLOSS. This discussions have to be seen in the light of section 5.7. Last I will discuss the HISP effort in Ethiopia in the light of my experience from Tigray, as well as in the light of my discussions about the effective use of the Internet and FLOSS.

11.1.1 Effective use of the Internet

Currently the physical access to the Internet is limited. The WAN inside Ethiopia is being actively developed to remedy this, and by connecting to the EASSy cable Ethiopia will get a much better connection to the international Internet. This is just the physical resource needed for the effective use of the Internet. There is also a need for relevant content accessible through the Internet. Ethiopia have quite a few government web pages, but those I have seen have not been very interesting. More interesting are Ethiopian newspapers, and weblogs and discussion forums where Ethiopians and expatriate Ethiopians discuss current political issues. Most of these digital resources are only available if you know English. There are some web pages available in Amharic, but for other local languages in Ethiopia there are barely any content.

The government of Ethiopia have high aspirations and there are quite a few students enrolling into ICT education. But the PC penetration rate is low and enrollment into first level education are still low. I don’t think it is a daring prediction to say that even if the higher strata of the Ethiopian society will catch up with the developed world there will be a significant digital divide within Ethiopia. This is typical for developing countries. Developing countries constantly have to catch up, leaving the lower strata of society behind. A question that is still open is how open the public’s access to the Internet will be. There are some disturbing signs of government censorship in the form of filtering out weblogs with government critical opinions.

Internet is the blood vein of the large majority of FLOSS projects, it is through this medium cooperation and contributions to FLOSS project are being made, and it is through the Internet you can get access to FLOSS software. Ethiopia’s high aspirations for building infrastructure for Internet connectivity can help in this regard. During my stay in Ethiopia the network was severely congested and slow. The time required to download a Linux distribution would be prohibitively high. The current international Internet connection have very limited bandwidth. To be able to effectively participate in an international FLOSS community, adequate bandwidth is important. The connection with the EASSy cable will improve this. The stability of the improved network and whether increased use will congest the network remains to bee seen.

The lack of Internet access can prove a major obstacle to the acceptance of FLOSS among the computer literate people of Ethiopia, as piracy copies of proprietary software is much more widely available.

11.1.2 Effective use of FLOSS

Even if work is being done to increase the capacity of the Ethiopian network this is an improvement only accessible to the elite in Ethiopia. This was also the case in the beginning in USA and Europe until the middle of the nineties. One big difference, however, is the higher general education level and wealth in USA and Europe. This gave a ground for a hobbyist culture around computers. It is little reason to believe that this will be a widespread trend in Ethiopia. It is much more difficult to get a PC for private use and people generally have more pressing needs. To be able to participate in the FLOSS virtual community you have to have three things, provided you are interested in participating; access to the Internet, adequate computer knowledge and time to spend. It is safe to guess that there is not many people in Ethiopia who have this three things. There is limited ground for a hobbyist community in Ethiopia, computers are still prohibitively expensive relative to the average income. It is only in universities, colleges, government bureaus and in the small computer business sector it is likely for anybody to use and contribute to FLOSS. Ethiopia is likely to benefit most from FLOSS in the education sector, and the broader government sector.

Among the educated elites in the universities and colleges there should indeed be grounds for usage of and participation in FLOSS, if the students are made aware of FLOSS and given the opportunity to experiment with software. Using thin-client networks like those promoted by Ndiyo and remote administration, it seem feasible to use FLOSS in schools as part of the e-school program and SchoolNet. This require the education of a group of expert network administrators who are competent in FLOSS software, with special emphasis on Linux and the Unix class of operating systems. At the federal level there is a need for a small group of experts responsible for support and for making helpful documentation, howtos and tutorials. In each region groups responsible for going to the physical location of each school to install and maintain the networks can get help from the core group. In each school a person with basic computer knowledge can be responsible, as this person can phone a regional expert or get help through the Internet. Most maintenance can be done remotely by the regional groups. If there is a need for making changes to the source code of a particular software, in order to make it useful in Ethiopia, the core group with help from students, faculty and professional developers can make the necessary changes. This changes can be contributed back to the FLOSS project maintaining the software.

The limited participation in the FLOSS community by Ethiopia is predominantly focused on translation of software packages. This is a natural place to start. In Ethiopia with its federal system where every region can decide on the working language of the government and schools, the possibility to translate software into many different languages is a great advantage. If the government is willing to sponsor such projects it will give ICT educated people in Ethiopia jobs and save foreign capital, and giving Ethiopia long term advantages in the form of giving people computer access in local languages, in the form of making Ethiopia independent of ICT vendors and in the form of building local ICT capacity.

The building of institutional support for FLOSS have just recently started in Ethiopia, through the formation of efossnet.org and an Ethiopian LUG, in additions to the efforts by Dawit Bekele and Daniel Yacob. GFF also come into this picture, which is mostly an effort by Yacob to localise FLOSS software. Relevantive had an idea for an information center for FLOSS. By having such centers where it is possible to get FLOSS software and support it will be much easier to effectively use FLOSS. You cannot get FLOSS software from the regular computer marked in Ethiopia. This efforts are limited to the small minority of computer literate people in Ethiopia, but I think it is a step in the right direction.

Prime Minister Meles Zenawi have expressed the opinion that there is a need for a minimal level of training to implement FLOSS solutions, and that Ethiopia have to wait five to ten years before they can choose. I do agree that FLOSS solutions, like all solutions, need a minimal level of education, but I think it is better to start right now to build the necessary capacity. Ethiopia have a chance to choose now when it comes to university and college education. The university curriculum should be reviewed and a narrow focus on proprietary technologies should be removed. My colleague in Tigray, Kalkedan, was only familiar with Microsoft technology after education in Mekelle University. It is also possible to start an education program to build the capacity to implement and maintain LANs in Ethiopian schools, it is not necessary to rely on foreign capacity to do this. Maintaining LANs in schools is an ongoing effort which should not be done in one huge effort and later laid to rust.

For Ethiopia to benefit from and contribute to FLOSS there needs to be a critical mass of people with computer knowledge and familiarity with FLOSS. This is best addressed through the higher education sector. FLOSS has its background from academic institutions because it was there people with the necessary interest, knowledge, equipment and time were found. This is even truer for Ethiopia since the general population is much poorer. The Ethiopian government should also play a major role by sponsoring translation efforts and building capacity to support the effective use of the planned SchoolNet and WeredaNet programs. If this capacity is not relevant for a job in the developed world it is only an advantage that will limit “brain-drain” from Ethiopia.

MS Office and MS Windows was available on all computers I saw in Ethiopia, with or without a license. Software piracy is common in Ethiopia and you can get expensive proprietary software for just a few dollars. The lower cost of FLOSS is currently not an argument for the regular private computer user. The government and business sectors usually need to be more concerned about having licenses for their software, especially if they are to comply with the TRIPS agreement.

11.1.3 HISP, Tigray and Ethiopia

The planned improvements of the Ethiopian WAN will make it easier for the Ethiopian HISP node to collaborate in the HISP network. It will also make it easier for the HISP people in the field to tap into the overall HISP network. The awareness of FLOSS is on the rising in Ethiopia. HISP-Ethiopia have the opportunity to show that FLOSS can be feasible in Ethiopia.

I did not manage to make a usable plug-in framework for DHIS 2 within my time constrains, but DHIS 2 still is much more modularised than DHIS 1.x. This will avoid the problem of making an incompatible fork to support ICD codes. Now the necessary adaptions done to the DHIS core can be sent upstream to the central HISP development team and incorporated in the main DHIS releases, with less likelihood of the changes making problems for the other users of DHIS. DHIS 2 is neither dependent on the client having specific versions of MS Office and MS Windows, all the necessary software can be bundled on a CD. Neither is it necessary with a costly proprietary program to make an installer. This solves the problems we had in Tigray with installation.

The HISP effort in Ethiopia can be a model for how it is possible for Ethiopia to benefit from FLOSS. HISP had, during my stay there, efforts going on in five regions. We had a team in each region implementing DHIS in the region and in some pilot districts. If this is possible for HISP, it is also possible for the education sector in Ethiopia. Communication between the different HISP teams was limited, a better computer network and a web page with documentation and an e-mail list would improve this. We could not improve the network, but we could have made a web page and e-mail list. Everyone in Ethiopia was busy with what they where doing in the regions and had no time to build a central node in Ethiopia. At the time of writing HISP Ethiopia has built a web site ( http://www.aau.edu.et/faculties/dis/site/hisp/).

In Tigray we had to “compete” against an installed base. There already existed a computerised system used for data capture and analysis, namely EPI-info. As noted in the comparison between EPI-info and DHIS in section 8.3 there are important design differences between this two systems. EPI-info is designed for surveying possible decease outbreaks and is good for non-routine data collection where the data that need to be collected varies between each case. DHIS on the other hand is designed to gather routine data from the primary health system.

I will say that DHIS in Tigray is best suited for routine data. DHIS will make reporting and data analysis easier as it offers an unified system for routine data collection. Data can easily be exported, sent and imported at the unit above. Data analysis can easily be done for any number of month and aggregated at any level in the hierarchy. If you want data for a district the data from the primary health units in that districts are aggregated together. Epi-info on the other hand can be better for non-routine data collection which the DPC department might need to make. Epi-info is designed with decease prevention and control in mind. DHIS 1.3 do not support weekly reporting intervals used by the DPC department, but later versions do. We only adapted DHIS 1.3 in Tigray, so EPI-info is better for such reports at the current time. DHIS and EPI-info fulfill different needs and are not mutually exclusive.

The DHIS core implements structure and processes common among routine health information systems. DHIS is designed to be relatively easy to adapt to different health systems by facilitating the process of technology translation. By isolating many complex implementation issues in the stable core a lot of the development and maintenance burden is shared with HISP and other countries using DHIS. The development of DHIS is a community process which makes it more likely that the software will stay in sync with changing demands, as opposed to an in-house developed IS only relevant for a specific country. By using the FLOSS licensed DHIS each country do not have to reinvent the “wheel”, get a “wheel” as a donation or pay somebody to reinvent the “wheel”.

The Ethiopian HISP node have been relative active in DHIS 2 development. The ICD extension have been incorporated into DHIS 1.4, and work to implement an ICD extension for DHIS 2 is being done. Tigray have not started to use DHIS 1.3 for routing data collection in the region as we hoped for, but the Tigray health bureau is interested in testing DHIS 2.