Interview: Jim Gettys

Jim Gettys was a colleague of Keith Packard in the past, but he isn't coming to FOSDEM to talk about X11. He told us his keynote will be "a talk for us geeky developers, with a bit of motivation of why we're doing [the OLPC project] and why FOSDEM's developers may want to help".

What was your motivation to come to FOSDEM?

Several things: to increase awareness and appreciation of the problems and opportunities OLPC faces and presents to the European free and open source development community.

I hope to establish ties with people and projects who may be interested in furthering our mission.

And last, but far from least, to thank the European Free and Open Source developers for their efforts, without whom this project would not be possible. Without you, OLPC could not be done.

It seems the OLPC has quite a powerful wifi chip [ed's note: it has an integrated ARM microcontroller and some memory]. How will this flexibility be put to work?

One of the key abilities of the OLPC system is its ability to participate in the mesh on behalf of other systems at very low power. If the main CPU and motherboard had to be turned on, we'd be using 5-10x the power, and it is very unlikely that kids would/could leave their systems on to keep the mesh network alive for others.

So this autonomous mesh operation is absolutely essential to aid the "last kilometer" problem in parts of the world lacking reliable electric power, which is a majority of the Earth today.

Which OLPC innovations in software and hardware will make it to the mainstream in the next few years?

I think a number of our hardware/software innovations will be mainstream within five years, including the screen, which can be used in sunlight, uses a tiny fraction of the power of current screens, and kept on with the CPU entirely off, while costing 1/3 of a conventional display will make a serious change in systems everywhere.

Note that a large part of why these hardware innovations are possible is that we are able to simultaneously design hardware *and* modify the system software, something very difficult to do in the conventional PC ecology, and fundamentally enabled by the fact that our software is free and open source.

We also hope to show the value of presence in user interfaces; while Sugar, our UI, is intended for relatively young children (the bulk of the school children of the world only get 5-6 years of primary education), the concept that presence is always present in our user interface we hope will inspire a wave of collaborative applications.
And it is time to stop inflicting user interfaces designed for developed world office workers on children who are still learning to read, the world over.

Will the OLPC software be available as a separate distribution for people that want to see the software but don't have the hardware?

It already is; either as an image, or as software you build and run on a conventional Linux desktop.

How do you feel about the competitors of OLPC, like the efforts by Intel and Microsoft?

"It's an education project, not a laptop project."
â€” Nicholas Negroponte

I think to differing extents, both companies lose sight of this fundamental fact.

I don't see either company as competitors, and the two companies' behavior in public and private is very different. We build laptops right now since no one is building ones that will work for most kids, a majority of whom have no reliable electricity at home, to name a fundamental reality that neither company has internalized.

We are fundamentally about serving unserved children.

Hypothetically, say we succeed at getting 5-10 million machines into the hands of kids in our first year. Sounds like a large number, doesn't it? It would represent just about the fastest hardware product launch in history, if we do this our first year of shipment.

So this sounds huge, until you stop and do the math; for simplicity's sake, we'll keep the numbers very round, and say there are a billion kids to serve, and we hope to make a computer that lasts five years. The steady state need is therefore *200* million/year. So in reality, OLPC is a drop in a very large bucket indeed. If OLPC succeeds at shaking Intel and Microsoft up to realize the magnitude of need and begin to serve the needs of the world's children, we will have succeeded.

As to how I feel about them, I feel frustration and anger if/when they expend energy opposing our project, rather than expending that energy toward what *they* could do to improve the children's opportunities.

That's a very big scale indeed... What's the current status of OLPC deployment?

Small scale trials are about to start; large scale deployments starting during the second half of the year.

Which members of other projects do you plan to meet at FOSDEM 2007?

It is people I don't know and should that I plan to meet at FOSDEM. Please do look me up, and don't be shy, if you have anything you think would be useful in helping the world's children.

Additional links:

This interview is licensed under a Creative Commons Attribution 2.0 Belgium License.