I find it odd there are no pictures of the aforementioned damaged XO's to accompany this write up. I would be curious to see the states of disrepair XO's are ending up in. The fact that there are no pictures makes one wonder if there is an intentional reason these pictures are not available?

Hi Ned

I looked through my video clips from my visit to the school. I've just uploaded a small clip that shows a laptop with a broken off "ear" + a laptop with a screen defect. This is now uploaded to youtube: http://www.youtube.com/watch?v=Fm9bTrpjTZ0
This is only to prove my honesty in reporting from the school. Challenges like this I am sure would face any school that gives laptops out to kids that never before have had such a piece of technology in their hands.

I've added the video from Anders to the post and am a little annoyed that Ned seems to think there was a specific reason for the lack of images.

Anyone who has traveled on a tight deadline has many, many photo-less observations, none of which are intentional and all of which the traveler wishes they did have on video or image.

Kudos for Anders resourcefulness and quick response.

I just thought I would point out to anyone that didn't know that machines mentioned in the article were most likely b2's which only had plastic antennae. The newer machines have rubberized ears which are much harder to break.

People forget that most of these children have never owned anything and have no experience looking after things. I would expect them to be damaged or lost as time goes on.

I've always thought that the laptops would be better suited as school property and locked away, certainly over a school holiday period when they could not even be recharged in homes without power. Maybe 'One Solar Panel Per Child' is the next step?

Thank you Anders for another wonderful insight into this experiment. We are all learning so much.

Remember that these are BTest-2 machines from back in March. These are not only more fragile than current prototypes, but are also about half as fast and don't have enough memory to run any software version after 406.

The ears were certainly plastic - and from what I could see this was one of the real weak points, as I saw many kids carrying them in one "ear".

The headmaster told me that a number of laptops from the school had been sent back to OLPC headquaters for them to study the things that happened under intense child usage. So the donation from OLPC (sending aprox 400 machines to Galadima) paid off also a field trial with valuable reporting going back to the US.

Thanks again for this insight Anders!

From your obervations I take it that the main activity that is being used by the children in Galadima school is browsing the internet, right? Did you see any signs of the collaboration features of "write" being used? Is TamTam to be seen anywhere?

With regards to those broken X0s, did you anything about them being repaired onsite? And what would you say were the most frequent failures?

Thanks again for your reports, it's a delight to finally get some unbiased hands-on reports from somebody who knows his stuff. :-)

Have a great weekend!

The point of my observation was not to detract from the reporting but to highlight that we have yet to see ANY damaged XO's before Anders added the youtube clip of the damaged ears. Given we are all readily aware of OLPC's stated belief that holding back OLPC technology from western children might have benefited the OLPC project by suppressing what might have been negative commentary it behooves whomever writes about about the XO with the intention of being unbiased to be aware of such things. If the OLPC is to survive market forces(which YES it will have to survive, regardless of what NN thinks) the last thing it should have to weather is unintentionally cast reasons for doubt. Do not think for one nano-second that Intel or Asus wouldn't take advantage of such omissions, as unlike those behind OLPC they WILL do whatever it takes to further their product(s) which could compete with the OLPC or can be threatened by it.

What I would HOPE to see is what is the functional resiliency of the XO technology given the various states of disrepair...
1- How well does the networking work on the damaged ear Xo's
2- Can the children with cracked screens still see their work at all? How bad is the cracking? How well is the XO screen handling the impacts of day to day child exposure?
3- Of the Stolen XO...I thought the XO had all that award winning security we heard about. Did OLPC brick the stolen unit as they said they would with machines that were stolen?

From an design/engineering perspective the damaged XO's are more valuable than the non-damaged ones...Non-damaged means performing within Spec...Damaged on the other hand indicates something that ACTUALLY needs to be paid attention to.

Ned is right and has said what I was thinking. The process of introducing a new product is complex and messy - a point which may not have been fully understood by OLPCs upper management. If the product is new and original enough it will have to grow its own infrastructure - and this takes money and effort to support.

I strongly suspect that OLPC has begged this infrastructure development question by telling themselves that all they have to do is deliver the shiploads of XOs and the rest will take care of itself.

Can we find out what, if any, provisions were made by OLPC for the support of these beta-release units in the field? Have they even asked the schools to collect the broken units so that they can be (a) replaced and (b) carefully examined to learn the modes of failure and the thresholds involved? Or will we be told again "we are only twelve people..." and "it's an education project, not a laptop project"?

Lee,

"so that they can be ... carefully examined to learn the modes of failure and the thresholds involved?"

The fact the 'ears' in the latests models are made of rubber rather than plastic, like the broken ones mentioned here, clearly shows the process to "learn the modes of failure and the thresholds involved" has, in fact, been applied...

Actually, delphi, it does nothing of the sort. Someone in OLPC realized that the fragility of plastic ears was a problem and changed the specs. This took place before the news from the units in the field could have come in (it takes time to implement these sorts of things). How were the LCDs cracked?

I will feel more comfortable when I hear that someone connected with OLPC is actively seeking out problems that have developed in the field and feeding back information on the failures. Anders' report does not indicate such an effort.

Lee,

"Someone in OLPC realized that the fragility of plastic ears was a problem and changed the specs."

Not sure what you mean by "realized" - XO goes thru some of the most stringent tests[1] of any mass produced laptops and done by the largest manufacturer of notebooks computers in the world.

"This took place before the news from the units in the field could have come in"

A speculation rather than a statement of fact on your part - the B1 machines were available for testing since last year...

[1] OLPC News 2007-07-28
( http://lists.laptop.org/pipermail/community-news/2007-July/000066.html )

"4. XO testing: Quanta has put an enormous effort into testing the
laptops in each of the five builds to date (A Test and B1–B4). The
approximately 7000 prototypes have undergone temperature, electrical,
mechanical, durability, and environmental testing:
..."

There's still no substitute for looking at what happens at the actual point of use. How "enormous" is the effort OLPC is making in that direction?

Lee,

Quanta Computer is, as I hope you know, the largest laptop manufacturer in the world and I think it's reasonable to expect they know more about laptop hardware testing than just about anyone.

Delphi is to be thanked for providing us an excellent example of the condescending attitude which could well result in the undoing of the OLPC project.

From the beginning the project was based upon a rejection of research (claiming that all the necessary research had already been done), reliance on what was supposed to be common knowledge (which, when applied in situations different from those from which it emerged can lead to nasty surprises), and a reduction of the end users, their families and societies to photo-ops (grinning children holding laptops in benighted classrooms to show that all will be well).

If not OLPC, then someone has to get out where the dust and mud are real and find out what actually goes on and what goes wrong. All the product testing in China won't prepare you for life as it actually occurs.

Lee,

"Delphi is to be thanked for providing us an excellent example of the condescending attitude which could well result in the undoing of the OLPC project. From the beginning the project was based upon a rejection of research..."

Rejection of research ? How on earth a revolutionary design as XO could have been created if it were not based on research par excellence - surely you must realise how baseless and, yes, "condescending" your statement is.

"All the product testing in China won't prepare you for life as it actually occurs."

Would you have been happier if the testing was done in USA instead :? But seriously Lee, it's fine to have opinions and I'm sure you think yours is a very important one. But at the end of the day what matters is that those opinions are based on facts - I gave you sources were you can find out about Quants's testing procedures and results and we also know that hundreds if not thousands of XOs were distributed for field testing.

I've agreed before that marketing by OLPC could have been much better and deserves to be criticised. However constant degrading XO's software and hardware by some seem to be based on ignorance at best or be agenda driven at worst...

Delphi & Lee,

Let's keep this civil and professional, eh? Please focus on OLPC commentary, and not the commenter personally. You both are better than this thread's current direction.

Thanks.

Wayan,
I hope I'm not transgressing into the personal. I doubt that it would sound much different if I changed to the passive voice and said "The previous post displays the condescending attitude...". If it is unprofessional to involve Delphi by name, then please consider the name deleted. It's the attitude I want to highlight, not the author.

Delphi,
I can testify from first hand experience that research is not a prerequisite for successful product design. I've done it several times. You seem to be confusing research with development, and these are two different things.

Research starts from a premise and attempts to verify or refute the premise. Many cases are studied - in theory there is no maximum number of cases. Often more questions are raised than are settled in the process - this makes research productive in that it opens the way to more research.

Development, on the other hand, starts from a specification and attempts to design something that will satisfy the specification. Only enough cases are studied to provide justification for a decision to take one or another path in the design effort. The number of questions is reduced to the smallest possible number as quickly as possible, leaving many paths unexplored.

A negative outcome, such as showing that a premise cannot be verified, is considered an acceptable research outcome. Development admits no negative outcomes. Scientists do research - engineers do development.

During the eight years in which I worked at Interval Research Corporation, doing prototype development, I got to see the differences close up. I was neither trained, experienced, nor temperamentally suited to do research.

OLPC rejected research from the outset. When Nicholas Negroponte spoke with a well-known product development company, he was told that the first step of the project would involve ethnographic research. He rejected this approach and ended the relationship. At every step, when questioned about research, he would state that 35 (or however many) years of academic research was enough and that the task at hand was development.

In no way am I degrading the engineering effort being made by OLPC - I have said before that it is necessary work for a number of applications. However, it proceeds from a specification (which, while it often changes in small ways during the process, remains unchanged in its larger dimensions), and if the specification is based upon major unverified assumptions, the development work is at risk of being wasted, which is the outcome I fear.

Lee,

"You seem to be confusing research with development..."

You're mistaken. In my 30-odd years as a Software Engineer (almost 20 working on OO systems) I took part in a number of research projects - some of them were followed by a related development project, many were not. Thanks for the lecture anyway.

"OLPC rejected research from the outset."

This is easy to refute. The world class research done by Seymour Papert (AI, learning theories - constructivism ), Alan Kay (OO systems, user interface, early childhood learning), Mary Lou Jepsen (display technology) and others, and their ongoing principal involvement forms the scientific base for the OLPC project. The fact that you are not aware of the extend of that scientific research or, more likely, don't agree with it doesn't mean it didn't take place - please don't confuse the two.

Just because some people have done research in the past 35 years doesn't necessarily mean that it is directly pertinent to the proposition that OLPC has advanced, about constructionism as embodied in a laptop and the supposed benefits that would flow from widespread laptop distribution.

In no case has OLPC put forward its thesis in unambiguous terms, conmpiled data (from any source), analyzed the data in a fashion that can be repeated by others, and explained how the results validate the thesis.

Whenever I ask where this analysis can be found I am pointed to Papert and Kay among others. Yes, they did a lot of research, but they were not addressing the specific proposition that OLPC has advanced. It's all background.

I am surprised to see Mary Lou Jepsen added to the list of eminent researchers. She is a world-class engineer in the display area, but when I spoke with her in Tunis in 2005 and asked her about the research basis of the project she gave me the same answer I have have cited above, and made no mention of her own research.

I am most willing to be edified if you could direct me to the specific research upon which the OLPC project is supposedly based. Thesis, data, analysis, conclusion.

Lee and Delphi,

Have you considered the costs in terms of workforce, time, money, and overhead of doing educational research?

Developing something like the OLPC costs way more in research than parts construction. Going for the minimum on research is an economic and financial decision. This introduces risks, but there are also risks involved with research, eg, research can fail too.

Winter

Winter, that approach would certainly go nowhere in the commercial arena, where people get very protective of their money that you propose to put at risk.

In the political arena it may be another situation, since the money belongs to an institution (government or a foundation) and politicians are more likely to act based upon common knowledge. Accountability for bad decisions is much more diffused in such cases, also. All of which doesn't mean that the likelihood of success is any better - in fact, it will be worse due to the blurring of the lines of responsibility.

The big risk, as I see it, is that failure will be multiplied and propagated. Even when they get the recipe right, proposals for low-cost computer-based projects in the developing world will be turned away, because a new piece of common knowledge will have emerged - that all such projects will not work.

Lee,

I'm surprised that you dismiss, in context of OLPC project, work done by Seymour Papert and Alan Kay as merely a "background" (and when I mentioned Mary Lou Jepsen I meant her research into laptop display technology, of course - crucial, I believe, to XO usability in the target locations ) especially as they both are directly involved in the project. Alan Kay's , in particular, spent last four decades researching use of computers in education and many of his numerous papers from his work at PARC, Apple and more recent ones are accessible online - I think a good starting point, going back to 1972 (35 years ago!), would be:

A Personal Computer for Children of All Ages
( http://www.mprove.de/diplom/gui/Kay72a.pdf )

I see OLPC project's educational value (of course, it's introduction will have other benefits as well) in very simple terms: No one questions usage of books in education although their purchase and maintenance can involve a very substantial cost. XO machine, as opposed to any other portable computer currently available, not only can replace those books but provide unlimited access to information (even if it means, initially, only access to static Wikipedia on the local server). Add to it writing, educational software, internet browser, communication and collaboration tools and we come very close to Kay's Dynabook:

The Dynabook Revisited
( http://www.squeakland.org/school/HTML/essays/dynabook_revisited.htm )

"With regards to those broken X0s, did you anything about them being repaired onsite? And what would you say were the most frequent failures?"

Repairing the machines was on the agenda of the OLPNC initiative. However I did not see any examples of this happen while I was in Nigeria,

I did see several XO's with broken off ears + several with screen problems, and also a laptop which for one reason or the other would not boot up at all.

Delphi need not worry that I am unfamiliar with Alan Kay's work - it was a major influence on my work in early personal computers. I proposed that Osborne Computer build a Dynabook when the founders first met (in 1980).

In case Delphi has not read it, here is a paragraphi I wrote in these pages six months ago:

"If OLPC were indeed an education project then it would proceed from the basis of an analysis as to what is wrong with education in the developing world and how it could be fixed. There would be copious and detailed references to research results, there would be pilot studies under way and a coherent argument would be advanced as to how the laptop or some other system - not just a device - would function to attain the desired results. There would even be discussion and argument as to what the desired results are and how they would be measured."

This work has not been done (so far as anyone outside the project can see), the statements and actions on the topic from Negroponte display an antipathy to this process, and now that prototype units are in the field we look in vain for evidence of any organized effort to observe and evaluate functional performance.

FWIW here's the sad July 2008 update => http://www.olpcnews.com/countries/nigeria/olpc_nigeria_one_year_later.html
.