Refurb or eWaste? You be the judge

 As mentioned elsewhere, Computer Aid International has released results of their study determining "the preferred low-cost, low-power computer solutions appropriate for rural settings in developing countries." Yes, it's a bit comprehensive, their claim. After all, some rural settings have limited electricity, some have none, some rural settings have the Internet, some have none, some rural settings are in schools, some are in post offices (and some go wee wee wee all the way home). 

Why does this make any difference? First, the test is skewed against the OLPC dedicated solution (which, while it might be great with first-graders, would make a college student feel like a clod) and perhaps against others. Second, the test doesn't address important considerations such as: maintenance, reliability and support; physical security; usability. Third, well, third is that the market for refurbished PCs is shrinking faster than the stock market these days, and rather than find a new way to be of service, CAI is opting somewhat disingenuously to plump the potential for refurbs in thin clients arrays. 

The ever-charismatic Wayan Vota of OLPC News offers a gentle take-down of the CAI report based on the test's concentration on computer-use in university settings:

Did you every wonder if the XO laptop was designed for use in African universities? I hope not. In even the name, One Laptop Per Child reminds us that they are designing technology for children....Pitting the Asus 701 & 900 laptops against the XO-1 may seem logical to a refurbished computer vendor - to them everything shiny and new must be confusing - but its illogical to expect university students to have the same use case as primary school students.

Wayan gets to one of the major flaws in the report: By focusing primarily on cost, power consumption, use of office-style software and video, the tests are biased against the OLPC laptop, which is designed to be a world-class dedicated solution for primary and lower-secondary learning. The test doesn't score for the battery of learning-related software that the machine offers, or for its collaboration-centric user interface, or for its mesh-networking feature. And goodness knows what stuff it overlooks on the other platforms... (NOTE that this is not an endorsement of the OLPC option!)

But the report is most disturbing in its concluding recommendations, which "imagine" three "ideal" scenarios for 20-user installations: one using the Asus eeepc (the preferred machine, based solely on "power, performance, and price"); one based on nComputing's x300 thin-client network with new PCs; and one combining the nComputing solutions with Computer Aid refurbished PCs.  

Excuse me? How do refurbs creep into the picture? Sure, they can be arrayed in thin-client networks, but they are at least as power hungry as new models. And the CAI scenario is based on, per the report, a "Pentium 4 PC with flat screen (when available)." How often have you seen old flat-screen monitors set out on the sidewalks of your street? (Here in lovely Oakland, Calif., we see a lot of stuff that could be called eWaste set out with a handwritten sign, "Free!" Might be there for a week or two, in fact, before someone gets fed up and junks it.) It would be helpful to know how often CAI ships container-loads of flat-screen monitors, or Pentium 4s for that matter. 

But it's a tough time to be in the refurbished-computer business: Prices of new PCs have dropped a lot and are still dropping; requirements for computing power have increased, at the same time, to the point that even 5-year-old PCs are challenged by some software; the "refurb reputation" has, as a result of bad-faith exports, fallen low enough that countries (Trinidad and Tobago for one) have adopted standards for the importation of refurbished computers (none of that eWaste on our sidewalks!). And now, there's all this hubbub about low-price, low-power solutions designed specifically for use in developing countries! What's an eWaster refurbisher to do?  

In all fairness, I should mention that CAI is very publicly advocating that developing-country governments crack down on eWaste. 

I should also, however, mention that I have a degree of personal experience with refurb-based projects. In 2005 in Rwanda, to take one example, I visited a number of schools that had received 20-computer labs courtesy of World Links. I found installations that were 50% Pentium I (one) computers (!), with a mixture of Pentium II's and III's. None of the Pentium I's had optical drives, so students couldn't use CDs, several were missing monitors, and a station in one school (at least one) was comprised of a monitor, keyboard and mouse, without any CPU. In 2005, Pentium I computers were well-over 10 years old. Needless to say, there was not much value in this project. I'll leave you to guess who the hardware supplier was...