Noble goals and politics at the OLPC

The "One Laptop Per Child" project (OLPC), the brainchild of Nicholas Negroponte, was envisioned as a way to bring the cost of computing down to a level where more in the developing world could afford it, either directly, or more likely, through purchases by governments and NGOs aiming to encourage development. The goal is certainly a noble one.

The "One Laptop Per Child" project (OLPC), the brainchild of Nicholas Negroponte, was envisioned as a way to bring the cost of computing down to a level where more in the developing world could afford it, either directly, or more likely, through purchases by governments and NGOs aiming to encourage development. The goal is certainly a noble one. Unfortunately, noble goals don't automatically result in easy sailing.

Concerns range from hardware issues related to the innovative design of the OLPC, clashes with newfound competitors scrambling to prevent the OLPC from undermining their revenue model (originally, the OLPC device would only use AMD chips, which was bound to annoy mighty Intel), and the exodus of key staff related to OLPC software development efforts. Walter Bender, former OLPC president of software, explained his departure by noting his devotion to the cause of open source and free software.

My personal interest is in helping build a community of developers, educators, and learners dedicated to advancing the quality of free and open source software for learning and the sharing of pedagogical approaches in this community by adopting the spirit and methodology of the open-source movement.

As a contrast, consider what Negroponte said in an email posted to the same list only two days later:

Our mission has never changed. It has been to bring connected laptops for learning to children in the poorest and most remote locations of the world. Our mission has never been to advocate the perfect learning model or pure Open Source.

Granted, Negroponte is clearly pro-open source, following the previously quoted statement with this: I believe the best educational tool is constructionism and the best software development method is Open Source. However, Negroponte just wants to get as many laptop computers into as many childrens' hands in the developing world as possible, however that might be achieved...even, if necessary, through a deal with a large software company based in Redmond to ensure that his OLPC device runs Windows. Many of the people drawn to his project, however, had additional goals, even if they highly valued what Negroponte hoped to achieve through OLPC.

The OLPC project has long been the darling of the open source / free software world because its early decision to use Linux as its operating system, combined with its low price and the fact that it is targeted at the most numerous group of potential computer users on the planet (low-income users in high-growth nations), had the potential to upend the established computing world order. Though Windows (and increasingly the Mac) are dominant in the western world, there would be pressure to "go open source" if the overwhelming majority of the planet's population used it.

I don't deny that every person attracted to the OLPC saw strong merit in the goal of giving laptops to children. However, the "buzz" surrounding the OLPC project's early days attracted people interested in more than just putting laptop's in children's hands. This was bound to create trouble once Negroponte began to suggest that Linux shouldn't be given exclusive attention in OLPC-related software.

I was gratified, however, to hear confirmation that OLPC developers implicitly understand why one platform tends to rule triumphant in software markets. I use that reason as the basis to explain to antitrust regulators that the goal should not be to eradicate dominance, so much as mitigate market power. I somehow doubt that C. Scott Ananian would agree with me.

What are we to make of this? Are you serious about Sugar on Windows or not? If you are, then you need to immediately hire *at least* 10 windows developers to actually perform the port, and inform the deployment countries that we are placing a hold on new development for at least 6 months while the port is prepared. And the result, of course, will be a new version of Sugar which is guaranteed to run *no better* than the one on Linux. From an IT management perspective, this is madness.

Cross-platform is hard (though mitigated somewhat if you start with the goal of being cross-platform, which, apparently, the OLPC project didn't). That's why most organizations, particularly new ones, tend to avoid it.

Speaking of development goals, a word of warning. Though I am hardly a George Ou-level blogger, some may have noticed that I have been blogging less frequently of late. That is mostly due to the fact that I have been spending every night and weekend completing software which will serve as the basis of a project that will take me to Africa for extended periods of time (possibly a week a month, but starting May 3, for about two weeks).

I'd like to provide more details, but I think that would be unwise at this juncture. I will, however, be bringing my video camera along. So, hopefully, there may be some interesting video extras to add to this blog...

...that is, when I come back. I have no assurance that I will have regular Internet access from some of the locations I am visiting. If I do, well, expect some blog-based "postcards from Africa."

Anyway, this is the "odd career twist" I mentioned last week, one that will grow odder as the months roll along.