What we are forgetting is that critics are bringing their world view with them, which is always jaded by personal/corporate experience. Anyone who has used a modern computer has had trouble with it, and will naturally be prone to assume that the OLPC experience will be no different. Their view of computing is highly structured and institutional, with layers of control coming in to play at every level. At the top there is the software company which exists to provide both software and support to the institution who is purchasing it. Next, there is the institution, who has purchased the software to meet some institutional goal or need and is providing the staff to support the software, as well as to seek support from the software company. Finally, at some point, the users gain access to the software, which in the end is painfully difficult to use because it was designed to meet the needs of the institution, largely forgetting the needs of the people who are actually using it. The traditional design is hierarchical, top-down, cumbered about by layers of control.
I was thinking this morning that it would be a good idea to do some very short, concrete posts about what collaboration will be like in a classroom with XO laptops running Sugar, if their code meets their vision. For example, let's look at a peer editing session in a writing workshop using XO's:
The kids split up into groups, and decide whose work they'll look at first. That student opens their document and hits a button to share it. The other students look at that student's laptop, note the unique color combination on this laptop cover's XO logo. Hit a key on their keyboard to switch to a view that shows icons for all the XO's on their network, find the icon that matches the author's XO logo, and click on the document icon next to it. A copy of the document is transferred to their computers. As members of the group do this, their icons move into a group around the document, so it is easier for subsequent students to find the document, and the teacher can see that they're roughly on task.
They read the document, make a few marginal notes that are visible to the author, and then do whatever they'd normally do for peer editing.
Now, I guarantee that this is inaccurate in some particulars (although I don't know which ones) compared to how the deployed version will work, but I think it captures the zeitgeist. Isn't this approach a million times more practical than what you're doing now in your classroom?