OLPC Should Be Three Related Organizations


Newman Inc is a great company with great products and great goals, but I do not think that OLPC should replicate the Paul Newman Strategy of a "for-profit non-profit" business model. Let's see what One Laptop Per Child wants to do.

  • Promote 1:1 computers as a learning and social development tool for the developing (and maybe developed) world. In order to achieve that it proposes
  • Hardware that is affordable, durable, power consumption-lean, interconnected (mesh) and outdoors usable (when necessary).
  • A UI/Application/content package appropriate for young kids.

Three ways to achieve this

So it should be 3 entities, related but not necessarily connected with organizational ties.

1) OLPC foundation.
Should be a mostly PR non-profit organization where Nicholas Negroponte and co will try to convince the world that this is the way to go in education, and come out of poverty, "and by the way these are your available options". Eg OLPC should try to generate and increase the interest and market for the "products"

2) XO Inc.
Thanks to OLPC the netbook era has arrived. The market already has machines of various capabilities costing from $100-$400+, improving with a fast pace every year. What XO Inc should do, is have the MIT wiz-kids push the envelope even further in the category and drive the market in the desirable direction according to aim 2 above. This should be a for profit organization Newman-type. If they offer to build the hardware locally, even better. If another company does it better, no big deal! Is the product that we are aiming for. Not the company.

3) Sugar Org
This is the most crucial part that needs the major attention and effort. It needs to define a minimum set of hardware requirements and then build the UI that can run on any hardware with these specs and over many if not any OS. Linux, Darwin and (yes!) Windows. In this way it can have the support of a big pool of companies and organizations and minimal enemies (OK... M$ will be always there...).

However, this would be the easy part of Sugar Org! The real challenge is to develop the educational package. Some of it, as math and sciences is universal, so it should provide 5-10 different activities for each of the math, programming, physics, logic etc subjects. Few of these are already there and some exist in other platforms (see a variety of "educational games"). However, many need to be developed. This should be done in co-operation with teachers from all over the world.

For some other disciplines as language, history, ethics/religion etc, the major part of the curriculum in many educational systems, one size does NOT fit all. In this case the EASY application development environment should be provided so the content can be customized locally to fit the needs.

This means that you provide a GUI application with boxes, action buttons, and menus where the local developer/teacher defines what material is placed and where (the more forms accepted the better), interconnects it in a suitable way (consecutive, reciprocal, indirect, opposite, random etc), defines the user responses (pick, move, navigate, introduce-remove element, type, etc), defines if they are going to be cooperative, student-to-student (competitive) or teacher-student (testing), and is done!

In short something like a GUI Python (or any other interpreter language)! Provide also some sample applications and scenarios as guides to trigger imagination and showcase possibilities. Finally provide truly-universal or region-appropriate content to populate applications.

So after all these, Mr Negreponte comes on his white horse and says: Computer education bla...bla... bla... will change your life for ever. We offer and support a FREE package that can do all these marvelous things (but for real this time). You can use any hardware you wish! Hardware from company A has these advantages and these limitation, from company B this and this, from C... etc.

We recommend the XO Inc hardware that is affordable, durable, requires minimal energy consumption, forms ad hoc networks, can be used outdoors and has minimal running cost. It also has these shortcomings but they way fewer than the shortcomings of other systems (if true). AND you can build it locally or negotiate which company is going to build it for you. Go for it!

Everybody is happy, for-profit and non-profit companies, the recipients/users feel that you offer them a fishing boat and not just a fishing pole that can only fish tuna, and the open-source community has its work cut-out for it... Importantly after all these you really increase the chances to achieve "one laptop per child" for the developing or the developed world.

As usually the devil is in the materialization details and the personalities that will undertake these endeavors but if the goal is common, the boundaries of each organization well defined and the approaches to be taken by each one defined internally but with consultation, it might have a hope

Mavrothal originally wrote this article for OLPC News Forum.

Related Entries


Dumb question - is XO based or very linked to python? Could python make a program by which students see spelling words listed and then must choose the one that is right or wrong and the program reinforces (good job!) or gives the answer and puts that wrong spelling word back in the queue for extra practice?

Short answer: yes.

Long answer: The XO isn't exactly locked into running only Python programs, but programming in Python is encouraged by both the system software architecture and by OLPC's policy. And Python is completely capable of being used for an application such as you describe.

Got any guesses as to how much it would cost to develop such an application whereby a database stores the words and recording of the words being said and is updateable by the teacher?

No, but it sounds like a worthy project. My own Python skills are not up to the task, but there's nothing in principle that would prevent this sort of activity from being created. There are existing activities that do parts of this job already, such as playing sound clips and accessing shared data.

Author wrote:

"What XO Inc should do, is have the MIT wiz-kids push the envelope even further in the category and drive the market in the desirable direction according to aim 2 above (Hardware that is affordable, durable, power consumption-lean, interconnected (mesh) and outdoors usable (when necessary)."

Nice "plan", but isn't that *exactly* what Negroponte promised but never delivered? (it sounds good on paper, but is not so easy in practice).

The rest of the suggestions are very reasonable-sounding, but equally empty of meaning, as they all lack a coherent implementation plan.

Sorry to be so abrupt, but that's the truth...


I do not really intent to comments on comments etc. This post is part of an ongoing discussion in the thread mentioned at the end of the post. It would be nice if you see the rest of it before you draw any conclusions or comments.
I would also love to hear your proposal for a plan B, C or D as part of that discussion.



A hardware consortium between gee whiz kids at MIT and specialist component manufacturers making open hardware prototypes. Even big manufacturers might be interested in participating just to get advance info so they are the first to market. Part of the open hardware spec to be a compliance test suite.

Sugar doing the UI and the software tools.

Education ministries releasing their local history, language and literature texts under CC licenses (In most small countries the primary school textbooks are published by the ministry).

Development charities doing the same.

A wiki for lesson plan development with local teams in each country sugarising the ministry texts(could be allied with Sugar but the Wikipedia foundation might be a better fit)

A variety of manufacturers making machines certified to be XO1 compliant. Certification provided by certification bodies testing against the test regime agreed.

Local companies in every country to do rollout to schools and logistics and political cover.

Commercial sale of the computer anywhere and everywhere. Every supermarket and souk from Cairo to Kamchatka. Commercial version to be a different colour from the schools version. I would make the commercial version with no keyboard. Make the keyboard a USB plug-in addition = no transformer hinge, no hardware localisition, no complaints about small keys.

Lessons available for download by everyone - lifelong learning. New business opportunity - terminals to rent by the hour in a storefront where you can study. Professional development - terminal available in every business and prison. Open architecture so other can come up with new uses we haven't thought of.


Would you care to add a USB Toilet, to maximize computer use by malaria-stricken students in remote villages?

Now Irv

That wasn't nice.

This post seems to focus on the dire need for educational software, and the ease of creation of specialized educational software.

My guess for this ease of use might be something like hypercard? I also noticed on the developer mailing list a suggestion that: given no-one knows python, but a lot of people know web technologies, harness those to open up the number of potential developers. Then a government could potentially hire a local web developer to create specialized content.

Looking at the available "activities" for sugar is disheartening. Half of them are programming related (either Python, LOGO, or something else), a quarter of them are games, some with dubious education value, and the remaining quarter are close to being educational (this includes things like musical tamtam). Many of the activities are alpha or beta quality at best. I also haven't seen a lot of the promised content like e-books (after all these things can replace a hundred textbooks. Right?)

The truth is, the vast majority of people don't care about programming. The computer is a tool. Unfortunately in the free software volunteer environment, developers scratch a personal itch. They care more about programming than teaching math or science.

Until I see good useful education content, and proof that it actually helps learning, I have a hard time supporting this project.

You may not have been following the news, but this is *exactly* what has happened over the last year.

One Laptop Per Child handles dealing with Ministries of Education and organizing deployments.

Pixel Qi, a for-profit, is working on future display and other technologies, and almost certainly has a huge role in designing the XO-2 hardware.

Sugar Labs develops Sugar and its Activities.