Now that some of the furor over the OLPC XO notebook becoming -- at least in part -- the OL-XP-PC has died down, both the instigator of the project and its former software czar have announced where they're going from here.  The hardware's intriguing, but the software's here now.

First, the software.  Walter Bender, who formerly oversaw the development of the XO's software, has kicked off an outfit named Sugar Labs to focus on the further development of "Sugar", the system developed for the XO that's designed to be used by children and not depend on a particular language for its utility (the interface is almost totally iconic).

Sugar was open source to begin with, but Bender wants to make Sugar into something that spans multiple hardware platforms and Linux distributions.  One of the other machines that's being eyed to become a Sugar target is Asus's line of Eee subnotebooks.  It would be great if such a machine could be securely dual-booted between Sugar and a full version of Linux -- one for younger users, the other for more mature users.

Now, the hardware.  Nicholas Negroponte, chief of the OLPC project, has dropped strong hints about what the next generation of his XO machines will be like.  After the first generation of the "$100 laptop" ballooned to almost twice that price, people have good reason to be skeptical of what he's promising now -- a $75 machine (!) that runs on approximately one watt of power, courtesy of Intel's Atom processors, and sports a dual-touch display which also can work as a keyboard.

In short, it's more of an e-book device (Kindle, anyone?) than a full notebook computer -- although given how fluid the definitions of these devices have become, about the only thing that becomes the real discriminating factor is the price tag.  There's also the question of what software the device will run -- and, again, I suspect if it's the kind of thing that Microsoft thinks it can horn in on, it's going to try.

What I'm seeing here are two fundamentally different approaches to what might ultimately be two different problems.  Negroponte's vision is to make devices that are affordable enough to bring a certain degree of computing power to audiences that didn't have it before.  Bender's idea, by contrast, is to put together the software and make it available on just about any platform you can throw at it.  The assumption with Bender's approach is that the hardware is out there, waiting for the software -- and given how many computers still get pitched into the trash every year, his approach makes a lot of sense.  Is it new devices we need to make computing more broadly available, or just better use of the machines we already have?