A recent The Guardian interview with Canonical CEO Mark Shuttleworth reveals this gem:
TG: Will you be coming out with a tailored version of Ubuntu for the ultraportable sector?MS: We're announcing it in the first week of June. It's called the Netbook Remix. We're working with Intel, which produces chips custom-made for this sector.
Canonical's Mark Shuttleworth
Though they're working closely with Intel; with any luck a "lightweight" version of Ubuntu would also be a natural fit for the OLPC (and perhaps Intel's Classmate?). Naturally, the OLPC community already has Ubuntu and other Linux versions and/or window managers running on the XO-1, but further developer support on creating an ULPC/4PC desktop system that can compete feature-to-feature with Windows could be a great asset for the anti-XP/MS/Closed source crowd.
The interview reveals two insightful pieces of how the power of community has shaped both Canonical as a business and Ubuntu as a Linux desktop. First is Canonicals "rather unusual way of picking" their original employees:
I simply read a large amount of correspondence between the developers on one of the projects that is key to the way we do Ubuntu, the Debian project. It's amazing how much jumps out in terms of the way people think, the depth of their experience. So open source is not only a great way to develop your own talent and skills, but it's also a great way to get a job, and a great way to go looking for people.
An innovative way to find and pre-screen for the exact style and skillset you are looking for. The details of the deal with Dell are a fantastic vision of the strength of a community to focus on a feature and push for it:
We found out about it after it was a fait accompli. [Dell are] very much a numbers-driven company. They asked their users what they wanted to see. They had a lot of data and that data pointed to us. That was a little unsettling, because we didn't have a relationship. But it was a significant step up in our corporate profile. It will be very interesting to see what we're able to do with companies like Dell, which are aimed at a wider audience. That's my number one challenge: how to make the Linux desktop something that you want to keep on your computer.
Both good ideas to keep in mind for the OLPC community as we face changes and/or outright removal of Sugar, the move to a closed-source XP-based OS, and the general change of the guard at 1CC.
this is the why the idea of porting sugar as a wrapper over other OS is not so bad. Maybe canonical could handle the part where the olpc system actually works, and sugarlabs could focus on improving the collaboration experience and making a theme for ubuntu
"this is the why the idea of porting sugar as a wrapper over other OS is not so bad. Maybe canonical could handle the part where the olpc system actually works, and sugarlabs could focus on improving the collaboration experience and making a theme for ubuntu"
Sugar already works on Ubuntu -- it was packaged for it with minimal changes, no real "porting" was needed.
Porting to drastically different environments such as Windows would be a project of a completely different scale, and it would have to pollute the original code base with Windows-specific pieces that will have to be maintained as the code is being developed. All projects I have seen that were ported to Windows ended up with original developers maintaining Windows-specific code that their sources ended up laced with.
On the other hand, if someone will ask me, I can build a Ubuntu-based or even Fedora-based live CD with Sugar that can be booted on any otherwise-Windows-based desktop and provide full Sugar environment right now, and in complete isolation from the original Windows configuration, except for storing users' files on a hard drive, USB stick or school server.
"this is the why the idea of porting sugar as a wrapper over other OS is not so bad."
Sugar is, to some extend, a window manager with desktop environment. Any OS that allows a choice of window managers should be an easy target for porting. Any version of *nix should qualify, and most things that can handle Xwindows.
One of the prime problems with MS windows is that it does not distinguish between the window manager and the kernel. This is one of the prime sources of security breeches. It also makes it hideously complicated to graft a new window manager onto MS windows. You can run "sugar" inside Xwindows for MS windows, but you seem to be unable to REMOVE the underlying windows desktop.
Winter