NewsForge: What is your general impression of migration to and acceptance of FOSS in South Africa and on that continent?
Hall: In 2001 I was invited to Johannesburg to talk about FOSS software by LinuxAfrica, a local user group that was putting on a small but nice event. In 2003 Hewlett Packard invited me to both Johannesburg and Cape Town for a FOSS "road show." In 2005 I was invited to Linuxworld South Africa in Johannesburg, and this year I came back again. I have seen a steady increase in the interest and usage of FOSS in South Africa. I believe this has been boosted recently by the actions of Mark Shuttleworth and the Ubuntu project as well as the efforts of the Meraka Institute and its projects.
I think it is exciting that OpenOffice.org and other FOSS software has been able to be translated into African languages so it is easier for African people to use.
NF: What were the developer and user concerns you encountered while in South Africa?
Hall: I would encourage South Africa to help facilitate lowering the costs of books and computer magazines for the South African consumer. The price of books is anywhere from two to three times what we would pay for them in the United States. Technical books and magazines on computer science are necessary.
Also, the price of telecommunications and the Internet there are too high. The FOSS community works best when communications are of a "no concern" cost.
In both cases (high cost of books and telecommunication) the flow of useful information is hampered.
NF: At the conference, you suggested that refurbishing machines might hold more promise for Africans than only going with a program like Nick Negroponte's "One Laptop per Child" (OLPC) initiative. Could you please elaborate on your feelings on this?
Hall: The OLPC initiative is aimed at providing a laptop for every child who wishes to learn. A laptop computer is a wonderful thing, particularly when it is low-power and built to survive conditions that might be found in a place far from a reliable source of electric power. Charge the battery to full and you have at least a few hours of uninterrupted computing.
One way of getting the price of the OLPC down is through high-volume manufacturing. This is why Mr. Negroponte wants to have millions of these laptops committed. I would guess that most of these would be manufactured in Taiwan or China, not in South Africa. Therefore, millions of rand (dollars, dinar, yen) will flow into China, not stay in South Africa.
On the other hand, there are lots of computers being upgraded by banks and companies. They will be "throwing out" good system boxes that would run Linux perfectly fine, and which could be donated to a local charity. By gathering these boxes up, pulling their components apart, reconfiguring them, installing Linux on them, and selling them for $100 -- or even $50 -- you could give a person a good job. They could probably refurbish and reconfigure four or five computers a week (or more). The money would stay in South Africa.
If this task were coordinated by a trade school, you could even teach students how to make the choices of equipment and software for building systems as they earned money by selling the units. These skills would be applicable to new systems also.
A lot of people do not need a laptop. They just need a computer and an Internet connection. While I would not want to see someone pay $50 for an Intel 486 box, a 500MHz Pentium with a 4GB disk and 128MB of RAM would run Linux fine. This system could even be expanded to be a small server system, for holding data for a community, or for setting up a small Web site.
Another solution that is being touted these days is thin clients, which consists of a diskless box that sits on the Internet and depends on a remote server for all computation. There is no local systems administration, no viruses to worry about, etc. This might be a solution in some cases.
I am sympathetic with the fact that a student having a lightweight notebook in areas where dependable electricity is either non-existent or very expensive would allow the student to use the notebook in both school and at home. I use a notebook almost exclusively for everything I do.
Finally, all the hardware in the world will not help if information is not available that will be useful to the people, in the language that the people will understand. Systems will sit idle (as they do in some places in South Africa, locked in rooms to prevent theft) unless there are useful programs and information for people to use. Of course I believe that this is a "chicken and egg" problem, and I hope that local people will help to generate, or ask for, the information that they need when they have the ability to ask and receive it.
Right now I am sitting in a hotel room in Sao Paulo, Brazil, watching a program on BBC about Vodacom in the Congo. People thought that cell phones would have very little sales in the Congo, but in the first month they had eight times more sales than had been predicted for the entire year. All of the same types of criticisms of the OLPC project were made for Vodaphone, but most of the issues around Vodaphone have been overcome. I anticipate the same will be done with OLPC.
I support the OLPC program, but there are other alternatives and concepts that people need to consider as a total solution.
NF: What's your overview on the issue of training and how it's being addressed?
Hall: There are some universities that are now using FOSS to train their computer science students, but others are still using closed source proprietary software. I do not understand why any university is not at least starting to go down the path of FOSS software. FOSS software not only teaches you what the software does, but allows you to see how it does it. It also allows you to participate in creating the software.
I also have experienced high schools teaching with FOSS software in other countries. They assemble their own computer labs out of cast-off hardware and install their own software. This teaches the students pride in accomplishment as well as computer science.
Most FOSS people train themselves, but there are more and more companies that supply training. For example, there are at least 15 companies in South Africa that supply training.
NF: Let's talk about the developing world in general, vis-à-vis FOSS. What insights can you share with us?
Hall: I think the developing world is embracing FOSS more than the developed world, and the developed world outside the United States is embracing it more than the US, for several reasons.
Having resources inside of your country that can fix and enhance the software in times of need is essential for national security. A lot of software is developed or controlled in the United States, and therefore under control by the US government. FOSS code is developed all over the world by many peoples with many eyes. This makes it hard to limit access to needed software technologies by embargo or terrorist attack on one main development campus, or to have trojan horses and trapdoors put into the code by any one developer from any one country.
Balance of trade is another issue that affects developing world nations more than developed nations -- having a lot of money flow out of the country to some other country to pay for royalties on packaged software when the money could be used to employ local people in tailoring the software.
Often the developing world cannot afford to pay the license fees of software companies, so they "pirate" royalty-based software. Free software is given to them freely, which means they do not pirate it.
Language support is another issue. While a lot of people speak English, most of the world does not. Being able to translate the software users can make it easier to use for themselves.
Finally, in a lot of places Linux may be the only computer system they have ever seen. They had never been able to afford the high-priced software and hardware to use computers. Linux, running on refurbished, donated hardware may give these people their first access to computers.