My name is Tony Anderson. I'm a retired computer professional. As an independent volunteer, I have been in full-time support of the OLPC concept and XO deployments since 2007.
As a volunteer initially at OLE Nepal, my assignment was to support the school server and later to work with Interns preparing to make XO deployments to remote areas of Nepal.
In 2010, started work as a volunteer in Rwanda, which has a national OLPC program (the XO is shown on their currency). Working with that program increased my direct experience with the requirements to support a deployment.
Supporting Laptops to Lesotho introduced me first-hand to the problems of deployments in very remote locations, especially the need for electrical support (both schools there are solar powered because the electrical grid doesn't reach - so there is no electrical experience in the community).
Recently, have started support of two deployments on Palawan in the Philippines. One is a primary school with XO-1.5. The second is a secondary school with refurbished Dell laptops (Core Duo). The latter provides an opportunity to see how much of the available school server content can be useful for older students. My goal is to install Sugar on Ubuntu on these laptops (Not immediately possible because James Cameron's Sugar desktop only works in 64-bit architecture),
For me, the goal of one laptop per child is to use computers to provide a more meaningful educational experience for students at community schools which are on the wrong side of the 'digital divide'. One primary assumption is that the value of the computer grows with the amount of time the student has access to it - hence, one laptop per child. This strongly contrasts with the conventional computer lab concept in which students are scheduled to use the laptops during specific periods to learn about computers. The educational value of the computer is in its support for learning across the curriculum, learning about computers is the natural result of their use (no one goes to class to learn to use a mobile).
Experience has shown that such a deployment requires a school server to provide local access to information from the internet. The economics are very favorable since very large hard drives are inexpensive (1TB for $60)and use of a server to deliver web pages and files does not require a high-powered processor. Currently a schoolserver can be provided for about the cost of two XOs.
My primary day-day effort is to strengthen the integration of Sugar on the XO with the school server and to add additional services and content. The BERNIE project is an attempt to put all of the software and content needed to deploy XOs with a school server on a single 1TB hard drive. This has proven much more difficult than expected because of the rapid changes to the software and the wealth of new content becoming available. At a school, the software and content needs to be updated only once a year at the beginning of each school year.
Sugar must move away from a dependency on the XO and be readily available for a wide range of platforms. This is existential because SugarLabs will fail without a large and growing base of users. SOAS is a valuable step, but we to be able to deploy laptops or tablets which requires installation of Sugar on the computer. James Cameron's efforts to provide Sugar on Ubuntu is one example. Lionel Laske's effort to provide a web based Sugar experience is another.
At the same time, SugarLabs has a responsibility to provide Long Term_Support for the XOs in the field (most of which are XO-1). Most deployments depended on outside donors to acquire their laptops and do not have the funds to replace them with newer equipment. These deployments will continue to use them as long as they function. Most deployed XOs are still using the original installed software (e.g. 0.82), a consequence of the high logistical cost to service them.
SugarLabs needs to extend the OLPC concept to include deployments with only a classroom set of laptops which are moved from class to class during the week. As Walter Bender pointed out at the Malaysia Summit, one laptop per child means that there are enough so that in use there is one available for each child in the class. This implies some attention to enabling the multiple users to save and access their own work.