December 8th, 2008
When Linux fails
[The opinions expressed here are mine alone, and not those of Google, Inc. my current employer.]
Recently I was able to visit the Ontario Linux Fest. I love shows like Ontario, as they’re run by amateurs, not by professional show companies. Don’t get me wrong; the professional shows have their place too, but I don’t tend to listen to the other speakers at those shows as I’ve heard most of them before. I’m sure they’ve all heard my talks as well, so instead we tend to hang out in the speaker rooms trying to get the wireless network to work, and swapping airline travel horror stories.
I do listen to the talks at smaller shows. Amateur shows are a labor of love, and it shows. The line up of speakers at the Ontario show was fascinating. One talk in particular caught my eye, by Ian Howard, called “Free and Open Source Software in Africa - Emerging Opportunities for Linux”. The slides from his talk are available here. The talk itself didn’t disappoint, and it was quite shocking, at least to me. It was a positive and upbeat talk about his experiences in Africa promoting Linux and Open Source software. But what he was really teaching us was an understanding of how Linux can fail in places like Africa.
Ian had gone to Africa as part of an organization called Geekcorps, dedicated to promoting IT use in developing nations. He worked on a range of projects in Mali, a West African nation which, like most African nations, is very undeveloped as far as IT is concerned.
At first glance, Free and Open Source software should be perfect for places like Mali. The local economy is poor, and average salaries make proprietary software an unimaginable expense for most people there. Yet the place is overrun with copies of Microsoft Windows. This is the toxic effect of what is called software “piracy” of course, although it bears little resemblance to what occurs off the coast of Somalia and is better called copyright infringement, as that’s what it really is. Copyright infringement doesn’t sound as threatening or scary as “piracy” though, does it ? The outcome of this rampant illegal software copying is that Windows is seen as “the first world standard” and any attempt to push a cheaper alternative is strongly resisted. They consider it trying to cheat local people out of getting the same quality of software that is used in the developed world, even though it’s a legal way of getting quality software for free.
Ian’s group first worked on a custom Linux distribution called “Kunnafonix”, designed for local radio stations. Local radio stations are an incredibly important communication tool in Africa, and most of them run on proprietary systems imported from the West, ill-suited for the temperature extremes and power requirements found in rural Africa. Kunnafonix was designed to be easy to use, install and repair, could be run as a live CD, and contained copies of the Wikipedia encyclopedia and the Audacity audio processing software. Many problems in operation could be fixed by doing a simple one-click reinstall to reset the system into a known good state.
However, installation of Kunnafonix was resisted by many of the local organizations they had to work with. The local “computer support person” resented a solution that was so easy to use that it undermined the power and prestige they received by being the person to consult when a Windows computer had problems. It’s amazing to see the myth that Linux is hard to use, install, and support still being propagated in much of the media here in the U.S., when in reality it is resented by Windows administrators due to its ease of use and lesser requirements for professional support.
More successful was their project to extend the Internet into communities by wireless networking, creating innovative ways to extend the range of wireless networks. In the wonderful talk “A New Way to look at Networking” Van Jacobson, one of the creators of the modern TCP protocol, said “the Internet reaches everywhere in the world, it’s just that sometimes the latencies are really, really high”. Everyone in the world has access to the Internet, it’s just sometimes they get to it by bicycling to the next village to view content that someone has delivered to them by burning it onto a CD-ROM. Extending wireless networking to remote villages can cut that latency to the point where Internet access can make a positive difference to people’s everyday lives. They can check the local commodity prices to discover if it’s worth it before undergoing a day long trip to the market for example. This time they were able to teach local technicians to make and repair the wireless network infrastructure they were creating, which made a great difference.
In a region with no software development experience, even Open Source software isn’t going to help bootstrap a computer support or software industry where there is nothing to start with. Even though you can see all of the source code inside a Linux distribution, without local expertise and knowledge to support and maintain it, it might as well be a closed source Windows installation. In fact, as knowledge of Windows is already widespread in such areas, even though it’s due to illicit software copying, Windows may be a better choice until you can break through the network effects keeping it dominant.
What shone though clearly in Ian’s talk was that unless you can partner with the local people, and most importantly help them make money with the new systems you’re trying to get them to use, then you’re just another well-meaning interloper, trying to sell them something that probably won’t work. Making money is the key. Without the opportunity of economic benefits, people in developing countries simply don’t have the time to learn about Free and Open Source software, no matter how much it seems to fit their needs from an external point of view.
Jon “Maddog” Hall’s keynote talk at the Ontario Linux Fest also made this point in a very powerful way. Jon is a wonderfully entertaining speaker, and not afraid of controversy. Showing a picture of a child in the African bush holding a “One Laptop per Child” laptop he said, “I don’t care about this kid.” The audience drew a shocked breath. “He’s screwed,” continued Jon. “Five hundred miles of bush behind him, five hundred miles of bush in front of him. There’s nothing I can do to help here”. Jon flipped the slide to show a Brazilian “favela”, or slum city, with an incredibly dense population, seeming to cling to the side of a nearby hill. He said, “This is where I can help. These kids have electricity. They can get a network connection. I can do something with Open Source and Free Software here”.
Jon isn’t a callous person. He’s just decided to focus his resources on somewhere he knows he can help today. It’s hard to find fault with him for that.
Ian aimed higher, and when you aim higher you have further to fall when you miss. He’s recently completed an MBA, and is concentrating more on the business side of things than the technical. I learned a very valuable lesson from his talk though. Something I, like many Free and Open Source software geeks, often forget whilst concentrating on the technical side of the software we love. Sometimes, technical excellence isn’t enough. Linux and Open Source software can fail badly in the real world not because of technical issues, but because of economic issues. We have to remember the lesson learned in the US election of 1992, and again in 2008. Sometimes, “It’s the economy, stupid !”
Jeremy Allison is one of the lead developers on the Samba Team, a group of programmers developing an Open Source Windows compatible file and print server product for UNIX systems. Developed over the Internet in a distributed manner similar to the Linux system, Samba is used by all Linux distributions as well as many thousands of corporations worldwide. Jeremy handles the co-ordination of Samba development efforts and acts as a corporate liason to companies using the Samba code commercially. He works for Google, Inc. who fund him to work full-time on improving Samba and solving the problems of Windows and Linux interoperability.
