My friend Andrew Heavens has been engaged in a great adventure – installing Ubuntu on one of his laptops in Addis Ababa. He’s found it harder than he expected – in an earlier post, he observed that he was only able to do the install because he has the luxury of two machines to play with – one running Windows, which he uses to access the internet and get help, and the other to run Ubuntu – most Ethiopian internet computer users don’t have this setup.
(Predictably, Andrew got caught up with the whole Winmodem thing – most Windows computers emulate a modem in software, rather than actually having a hardware modem… which Linux strongly prefers. Poor guy – I remember trying to debug this years ago under Debian and finally going out and buying a real moden – one that modulates and demodulates on a chip, not using the CPU. Probably not such a realistic option in the fine technical emporiums of Ethiopia…)
His most recent post centers on a subject I’ve always found interesting – does it make fiscal sense for people in the developing world to move to open source? Andrew observes that many pirated, cracked software packages are for sale in Addis at very attractive prices – Adobe’s Creative Suite for 80 birr (about $9) rather than the $1100 it would cost on Amazon. But it’s not just the cost that matters:
So the Ethiopian computer user can choose between paying nothing for an open source graphics package like The Gimp, or saving up 80 Birr to buy Photoshop CS2.
For 80 birr s/he gets to learn how to use the most popular graphics package on earth – a skill that could land him/her a job with one of the high-paying NGOs or UN organisations scattered across Addis. (Most of these international organisations have IT departments that limit their branch offices to a list of authorised software, almost all of it from Microsoft). Or for nothing, s/he gets to use a package that is very cool and opensource, but totally obscure.
I know which way I would go.
The obstacles to adoption of open source software in the developing world are myriad. Yes, there are countries like Brazil and China that are making governmental investments in open source. And there are organizations like the Shuttleworth Foundation – main sponsors of Ubuntu – who work hard to make open source software available and useful.
But there are lots of other countries where the reason to get a computer is so that you can develop marketable skills so you can get a good job with an NGO or a branch of government… which means you need to learn the software your potential employer uses… which ain’t Gimp.
This is just one of several difficulties open source has in gaining traction in less developed nations. Another is the lack of information and documentation. If Photoshop is $9 and Gimp is free, but you know how to use Photoshop and don’t know how to use Gimp… and manuals are impossible to get and hugely expensive if you do get them, Photoshop just might be a better deal.
I wrote an essay about this years back for Linux Journal: “Free beer doesn’t sell” Since I was writing it for Linux enthusiasts, it was a bit less critical and more solution-oriented than I tended to be in my practice with Geekcorps. When we worked on projects in developing nations, we tried very hard to push Unix when people were working on server-side applications… which meant pushing Linux or FreeBSD, because none of our clients were lining up to buy Sun workstations. But we almost never pushed Linux on the desktop – generally speaking, it would have been a disservice for our clients, who had an easier time supporting their pirated copies of Windows, an easier time hiring employees who knew the software, and very few worries that the BSA would shut them down any time soon.
Linux has made some real strides in the last few years. Ubuntu is a pretty, pretty thing and may be my next operating system if Apple can’t build a MacBook Pro that doesn’t catch fire. But Andrew’s experiences reveal that it still has a long way to go before it’s an OS for everyone in Africa.
Do you think that Linux advocacy should then focus on enlarging market share in the first world, as a way of marking it more attractive to people in the third world?
One could argue that the above is also true of many organizations in the first world as well.
I think the biggest obstacle for open source is bandwidth, which is painfully lacking in many parts of the developing world, especially in Africa. A fresh Ubuntu install is fine, but to make it day to day usable, you need to add some more packages to it. To make it great you need to be able to google regularly, and participate in the open source communities. It is not the software that makes open source better than closed source, it is the communities, mailing lists, irc, bulletin boards, without these and without manuals figuring out how to do stuff is hard.
Here in Malawi we develop software to help poorly trained health care staff provide top notch HIV treatment. We are in the process of migrating from a Visual Basic on Windows approach to a Linux + Ruby on Rails in the browser approach to make us 100% open source. In a country where Microsoft tools are used almost exclusively this is a big switch, but there is a lot of excitement, and I think this is mostly due to the developers sensing that with open source they will be part of something big and exciting and cutting edge. Next week all of their machines are dumping Windows for Ubuntu. Lets hope we can pull it off!
Interesting. I had never thought about teaching Microsoft or Adobe products as a possible resume booster for smart people in the global south. It’s definitely worth considering.
What I notice here at Interplast is the need for integration with lots of different computing programs all around the world. The people in developing countries with whom I work do not generally have high budgets for technology, and are generally running older machines with older software. Despite all the benefits of open source software, it’s hard to convince them that everything will be ok if they leave the big boys.
Based on my experience, I concur, using Linux/BSD in developing countries is a real challenge and partly this has to do with the ease of, and acceptance of, the use of pirated software; a practice not tolerated by most developed country businesses or NGOs, but considered fair game by companies and NGOs in the developed world.
Andrew’s article is a good introduction to one of the reasons why Open Source is having such a hard time in Africa.
Bandwidth, in my opinion is not as big a problem as one might think. People in developing countries have no problem duplicating CDs.
Recently, I learnt that a site where we had installed a Linux machine was recently re-imaged with Windows. This was done because the local techie hadn’t learnt Linux, nor did he have much interest in it. I have heard other developing world techies state that they perceive Linux to be for the poor and that they want to use the rich man’s operating system. Though we spent considerable time teaching Linux to this techie and customizing that machine, that experiment failed. Partly, it was because we had customized the machine so that it was limited to its intended purpose, a studio machine. Now the machine is a slow, but ready platform for any software and to be used for any purpose. Partly it was because he would have to re-learn how to use a new system. Partly, I suspect that the Volkswagen just doesn’t shine next to the Mercedes. And frankly, some of the proprietary tools are better for audio editing, but, the question is did the Volkswagen not get the job done?
Image and ease are critical in most countries that I have visited, and sadly open source is just too plain.
As well, Microsoft has been a strategic and apt donor. It funds many projects and on many of my projects, Windows is installed to apease Microsoft or the American government who promotes, with its funds all things American (naturally). I have been involved in at least 5 projects in the past two years where Linux would have been considered had Microsoft not contributed its little sum.
Thanks Andrew, Ethan for noting this issue. I will be at a conference next week in Ottawa, the Linux Symposium, where this very topic is one of the topics to be discussed. I’ll be sure to forward your posts.
I will continue to promote Open Source and to train people to use it. Those who learn it well are very employable, as there are few who chose to learn it and there remains many applications of open source that are invaluable off of the desktop.
Ian
Ian I am afraid you are missing my point. It is not the software itself that makes open source so great. It is the communities: the ability to google and discover other people with the same problems, to login to freenode irc looking for help and end up with friends. You can’t replicate this through burning a CD. Local communities can help this to some extent, but you miss out on the long tail of knowledge that the internet brings. Or as one of my favorite sayings goes:
chmod 777 world
Pingback: The Cool Macbook Pro blog / Ubuntu, and the ongoing difficulty of marketing free beer
Pingback: One Laptop Per Child News
Comments are closed.