Zimbabwe
A lot more information about Zimbabwe, some history, and also a look at their draconian proposed wiretap law (it's possible even worse than in the US!)
« August 2006 | Main | October 2006 »
September 23, 2006
digg this
del.icio.us
reddit
Slashdot
Stumble It!September 21, 2006
The limits of ICT4D
Someone has to pay the 'net bill:
Internet traffic in Zimbabwe has come close to a standstill after an international satellite firm slashed its bandwidth because the cash-starved government failed to pay the bill.Government-owned TelOne, which owns the country's main satellite Internet link, said satellite firm Intelsat had cut its international bandwidth because it failed to pay the $700,000 fee.
"The link is slow because they reduced the megabits on our satellite link until the payment is made," TelOne spokesman Phill Chingwaru told Reuters on Wednesday.
"We have approached the Reserve Bank of Zimbabwe for foreign currency and they are working on that, but meanwhile there would be delays in browsing because of the partial cut-off."
September 19, 2006
Nica Tech Writeup: The importance of local
Directions in Managua
But I digress. I feel that IT capacity building among these small non-profits is of huge value. It's a bit removed from front-line development work, but in the two short weeks I was there I was able to greatly increase their workflow efficiency and capacity, which will enable them to do more with their current setup and to grow, hopefully, in the near term.
Some of this was just training in some relatively universal best practices; backing up, using some tools more efficiently, and so on - nothing particularly unique about it. However, the most important parts are where traditional IT breaks down. The electricity in Managua goes out, unpredictably, for 2-8 hours at a time, often in the middle of the afternoon. Though they have a shared document folder, it is on the only "permanent" computer (i.e. non-personal laptop) in the office, which is a desktop plugged into an old UPS that doesn't last very long. Ideally, some version control system on a remote Internet machine would be a great solution, but then the Internet can fail for up to months at a time, leading to huge problems with binary files (text files could just be merged, but .docs are binary and resist such useful things). So we came up with a (human) process to save locally during power outages and replace on the server when the power returns. It's not perfect, but it's locally relevant. This is, I feel, the real key to ICT development strategies - realizing when the "best practice" is contextually invalid for social or environmental externalities, whether they be hungry goats chewing on phone lines or cultural differences.
Free and open source tools also provided me with a great opportunity. Not only are they reliably internationalized (all software has to be available in Spanish for obvious reasons), they are within the budget and customizable to local needs. There's a lot more to say about OSS and ICT Dev, and I'll save that, and how I punched down ethernet jacks with my CVS keychain card, for another post.
Morality of Development
Development, like politics, is a metaphorical room where you're amazed at just how many elephants can fit simultaneously, and yet be ignored. These elehpants are conjured through some central, unanswered questions. A former Bank employee friend of mine has a fine one, for example - ask any Bank defender how the incentive structure for determining loan recipient validity and reliability works when the loan agent is encouraged to issue loans, and the recipient doesn't have the option to default without serious global consequences.
Another question is the morality of development. Think of psychology or anthropology, and the extensive, exacting processes these professions have to approve studies involving humans, ensuring the safety of the subjects, measuring risk and making sure the subjects have been clearly educated on, accepted, and signed off on any risks involved. Now compare this to development, where entire nations are subject to the economic theory de jure, with, if they're lucky, pilot projects (which, if they're really lucky, are quantitatively measured and evaluated before being used as blueprints for a larger strategy). This new book review at the Boston Globe discusses this with an example of the response to the 2005 Pakistani earthquake, and problems in distributing aid to remote regions (or, indeed, effectively at all). An organization tried to implement a simple logging feature to get reports from aid agencies on where they were sending supplies to make sure that easily-accessible communities weren't over-compensated at the detriment to more remote areas. The response from the agencies was predictable - "people are dieing and you want us to fill out forms??"
This is a bit extreme on one hand, this was an unusual event, a disaster relief project, not your normal more organized development project, but at a certain level, the lesson still applies - yes, people are dieing, and guess what? they're also dieing because of your failure to organize. There are limits of bureaucratic inefficienies, but there are also inefficiencies in entirely horizontal structures with no coordinating effects.
This reminds me strongly of where software development goes awry in too many cases - the business addiction to urgency too often overrides the software development need for careful planning and forward-thinking organization.
In both software and international development, planning, testing, and strong, honest evaluation methods create better results. Similarly, the incentive structures are very hard to implement, as they take time and money while (profits, people) are being lost, so even though they "pay off" in the long run, it takes talented and willful management/policy makers to ensure that these structures are in place.
Making sure the code compiles and is bug-free is something that both the Microsofts and the World Banks of the world should keep an eye on.
Technology and the University
More tangentially related tech info, my former employers, The University of Texas' Office of Technology Commercialization are hosting their next big conference to feature commercializable UT research. Last years had tech ranging from backpack-totatble HIV/AIDS field testing units to creepily-good evolutionarily-learning AI . This year they're also including some fascinating talks on technology creation and diffusion in the University/government/private enterprise triangle, and a panel discussion on Open Source, including an appearance by Google's OSS maven, Chris DiBona. Check it out!
The Economics of Free
While not strictly dev/ICT related, this blog is tracking the economic implications of Open Source, (focusing on the university software development context). Interesting stuff.
September 01, 2006
Technology and personal networks
I'm reading Technology and Social Inclusion, which is a fascinating book on the role of ICT in development, looking internationally and within underserved/poor/rural US communities as well. Today on Slashdot, there's a story on the Internet improving, not replacing social ties/networks, which connects well both with the Human/Social resources chapters of this book on the importance of communties of practice and social networks, and of course the social capital theories of Granovetter and crowd.
