Friday, January 04, 2008

XO the Hardware and Sugar the Software

I would be disappointed if OLPC scaled back the ambition of their software goals and switdhed to a more conventional desktop on the XO hardware platform. But I am also frustrated that people seem to have trouble evaluating the hardware and software separately. As hardware, I think the XO is an unalloyed success. To the extent there are hardware problems, they're primarily driver and firmware problems, so arguably they're really software problems (I'm confused about what's up with stylus input on the "wings" of the trackpad... whether it is a hardware or software problem).

Sugar, however, is a raw work in progress. Whether or not the ideas embodied are good ones is an open question -- I think they are -- but there is no question that this software is not done. They had to start shipping laptops, and they whipped together what they had pretty well, but if you read over the design documents on the wiki even if everything they've deployed worked perfectly, it would only be a fraction of what they imagined, and what they deployed is certainly less than perfect.

What happens next is anyone's guess: Sugar is partially or completely replaced? Sugar matures and fills out over the course of the year? Sugar drags the whole project down like an anchor? I'd say it is equally likely that any of those happen.

But it is frustrating to read Doug write:

I'm still wishing for a practical student laptop. The ASUS Eee and OLPC XO are exciting, but not there yet.

Specifically, he wants:

  • Weighs less than two pounds?
  • Runs at least eight hours on a battery charge?
  • Is 802.11x compliant?
  • Can be dropped without breaking?
  • Comes only with a full featured web browser for software?
  • Has a screen that can be read for a long time without eyestrain?
  • And sells at a price point most parents can afford – let’s say under $200?

I'm not sure that weight is realistic. My high-zoot ultra-portable ThinkPad is 3.5 pounds, and the XO is a touch over 3. Asking for 2 pounds seems pretty arbitrary. The XO's power consumption and management are state of the art; not getting 8 hours yet, but that's a software problem. I've got the advanced power management running on mine, and it is pretty impressive. The cpu goes to sleep with the display on and seems to consume very little power. XO handles wireless pretty well; it would be better if it wasn't also trying to do the crazy mesh stuff. XO handles drops better than anything short of a Toughbook. Lots of people are hacking their XO to do less in software; it is only a few commands to make it run a lightweight window manager and Firefox instead of Sugar. It has a great, innovative screen for reading without eyestrain. It could be sold in quantity for under $200.

So the XO is really very close to what Doug wants in hardware. He mostly just wants less software, which is really easy to provide. So yes, in a very practical, direct sense, the XO is "not there yet," but I think it is also short sighted to think that the next device will be the solution. The potential of the XO as a hardware platform is barely tapped. What's needed is some patience, curiousity, and ingenuity, not hope that the next solution will be the one to emerge perfectly formed.

No comments: