Tuesday, January 22, 2008

One Laptop Per Child

Over the past week, I've had a chance to play with the XO laptop from the One Laptop Per Child (OLPC) project. Conceived in 2005 by MIT Media Lab co-founder Nicholas Negroponte as a "$100 laptop" for poor children in developing countries, the project reached production in November 2007 -- albeit with a shipping price of $200.

What follows are my first impressions, after several days of testing.


The case uses heavy-duty, sturdy materials. What's more, the laptop has no moving parts internally -- it uses solid state Flash RAM rather than a hard disk for storage. It can definitely take a beating.

The device uses very little power (less than 2 watts), is completely silent, runs cool to the touch, and has good battery life.

The screen has high resolution, and features both a transmissive and a reflective mode -- for use both indoors and out.

Open-source software.
The laptop leverages free, open-source software. That includes the OS, which is based on Red Hat's Fedora Linux distribution. There is also a suite of innovative educational software applications that are guided by the Media Lab's philosophy of enabling children to learn by creating (as opposed to just consuming) their own media, programs, and ideas.


User Interface. It can be a healthy thing to challenge existing conventions of user interface design, but the team behind the XO laptop has taken this approach to such an extreme that real-world usability suffers greatly. While elegant, the case design is so eccentric that I was literally unable to open the clamshell without consulting online documentation. Once the laptop is turned on, the user interacts with the Sugar UI, which runs on top of Linux. With its disregard for commonly accepted UI conventions, heavy use of cryptic icons, and lack of on-system help files, Sugar forces users to relearn the process of using a personal computer from scratch. It's a bit like trying to drive a car in which the steering wheel and pedals have been replaced by hidden new control surfaces. After consulting online documentation yet again, I was able to complete basic tasks such as web browsing -- but one wonders how teachers and students in developing nations will fare without extensive training materials. The built-in educational applications have no help files whatsoever.

Performance. The XO laptop is quite sluggish in use, but the biggest problem is its lack of real-world web application support -- particularly with respect to multimedia and Web 2.0. The web browser's support for Flash isn't just marginal, it's abysmal. Even the Yahoo home page isn't supported by the browser. And you can forget about watching YouTube or CNN videos. The upshot is that the XO's utility as an information device is severely constrained. A massive amount of Internet content is simply inaccessible to students using this laptop.

Bugs. A couple of strange bugs surfaced after a few days of use. For example, the keyboard stopped responding and had to be fixed by a reboot; no teacher wants that kind of distraction in the middle of a lesson. Also, the web browser displayed inconsistent behavior, refusing to let me navigate a simple site after I had accessed it just fine the day before. Other reviewers have noted similar issues.


The XO laptop is a great concept that embodies some very innovative ideas. But the execution just isn't where it needs to be. The eccentric user interface design, poor real-world performance, and system bugs are putting its mission at risk.

I hope Mr. Negroponte's team gets the execution right soon, because this laptop represents an idea that deserves a chance to succeed.