OLPC software task list
Introduction
OLPC will need to establish software work groups as general framework to enable the successful deployment and use of the One Laptop per Child (OLPC) initiative. Most of these work groups will have national counterparts that may vary in how they choose to prepare for and manage the program in order to take advantage of local strengths, skills, resources, culture, and strategic directions.
There is some natural overlap of workgroup missions so naturally they will have to remain coordinated. A coordinating body that includes representatives from each workgroup should meet on a regular basis. OLPC is unique. Adoption of universal access to connected computers for education presents opportunities previously virtually impossible. Serious thinking and creativity can bring widespread social benefits. By being open to innovative ways of thinking combined with an understanding of local conditions, the workgroups can contribute mightily to OLPC success.
The critical success factors include creating:
- High quality educational examples of OLPC use
- Conditions for the development of people and ideas
- Mechanisms for the spread of good ideas and tools
Fostering the development of the early adopters and helping their ideas and examples to spread are key. As OLPC is a multi-national effort, each nation does not have to invent everything for itself. They can and should learn from each other. The technology itself facilitates the spread of the ideas and examples. There also will be the workgroup meeting at MIT in March, 2006 to help facilitate this process.
A computer is only as useful as the software that runs on it. The OLPC will come with certain general purpose software available (e.g. browser, word processing, utilities, etc.). There also will be several types of software for learning that take specific advantage of computers (e.g. LOGO, Squeak, Scratch, etc.). A major decision was to base the OLPC on open source software. This provides countries with the freedom to decide for themselves what to place on the machines.
The task of the software workgroups is to determine what principles to follow and what mechanisms to put in place to provide widespread access to high quality educational and general software for children and their families. The taskforce also needs to analyze and make plans for any needed localization. Rather than re-invent the wheel, the workgroups can take a strategic view and adopt high quality software developed elsewhere, localize at the level required (whether for language, local content, or local culture), and develop new software to fill in the gaps. In this way overall costs can be diminished.
Moreover, the software workgroups can take a strategic view towards industrial development by fostering a positive climate for local developers to create new software, tools, and content for digital media. Software that truly facilitates learning and that utilizes the best of digital, computational environments (that is, that they are expressive, dynamic, multi-modal, collaborative, visual). This workgroup will need to coordinate extensively with all the other workgroups as the software is a key component towards the success of every group. The software group also can study how to best use the emerging connected computational infrastructure to further other social goals.
OLPC Software Task List
At its most basic, we need: an operating system; a web browser; an eBook reader; and some other base-level utilties, such as audio recording and playback, video playback, word processing, tools for drawing, etc.
With 1GB of flash memory, and 256 MB of RAM, the $100 laptop will not be tied to a specific set of software or content. It will have the flexibility to change with the requirements of the student. Individuals will inevitably find individualized uses; certainly local, regional, and national requirements will drive much of the base configuration.
Indeed, the faster OLPC moves from designing the OS to adding applications and localization, the better for everyone.
OLPC Fantasy Software Teams
Kernel
The responsibilities include base device support, power management, memory and OOM management, flash file system, Athleros driver, X EXA device driver, distributed file system (e.g. Coda).
Who: Jim Gettys, David Woodhouse (JFFSn), Marcello Tosatti (general help), Mauricio Lin & Sampsa Fabritius (Nokia, OOM), X driver (AMD & Tungsten Graphics (?)), Satya (?) & people to be hired.
Base System Environment
The responsibilities include derivative distribution repackaging, base libraries, "standard" out of the box applications and desktop environment (e..g. window management, application launcher, file browser, system configuration). Internationalization and localization technology, accessibility technology all belong here. Mesh networking power aware networking belongs here.
Who: Jim Gettys, Mako Hill, Dan Williams, David Zeuthen, Marco Gritti, Phil Blundel (?), Owen Taylor, Mathew Allum(?)
Suggestion: The Base System should be the same for all laptops and do not change very often. This would be the OLPC API. The benefit is that students can just download and use an OLPC software (doesn't matter if it's a brazilian student getting a chinese program). Also, developers can adjust their software to use the standard OLPC libraries avoiding library duplication, reducing application size and memory usage.
Python development
The target for many application developers will be the OLPC Python Environment. This is composed of the Python language, GTK GUI and Gecko HTML engine.
Network Environment
The responsibilities include software services such as backup, software distribution, email, web services, VOIP, IM, printing, etc. Additionally, services for schools themselves belong in this group (e.g. administrative software). There are also non-software-related issues this team will confront, such as regarding regulatory regimes, spectrum availability, use of "universal access" funds, carrier participation, etc.
Who: Michail Bletsas, Dave Reed, Demi Getschko
Localization and Universal Access
The responsibilities include fonts that will provide an acceptable coverage for the required languages, translation of the software (user-interface and help files), writing support (keyboard layouts and input methods), creation of locales, etc. The system must fully use Unicode (UTF-8) instead of legacy 8-bit encodings.
Educational Software
The responsibilities of this group include selecting suitable educational software, (with an understanding of memory consumption of the software). Ebook and book authoring technology belongs here. Other issues include teacher preparation, migration of exisiting content, a framework for new content development, school-server technologies, a "Kids Corps" for maintenance.
Who: Walter Bender, David Cavallo, Seymour Papert, Brian Smith, Alan Kay, K12LTSP(?), Mitchel Resnick, Roseli de Deus Lopes
There is also a need for basic tools for programming and authoring. LOGO, Dia, TuxPaint, etc. all should be considered.
However, let's keep in mind that there is a difference between Edutainment and and Playful Learning. And we want Playful Learning instead of Edutainment. Also, we don't want kids to be punished by rewards. Some understanding of constructionist educational philosophy will be useful for developers.
Deployment
Unless the country will immediately supply laptops for every child, decisions must be made about how to roll out the computers until the goal of one laptop per child is reached. The immersive nature of one-to-one computing is essential to making dramatic improvement in education. Therefore, we strongly recommend distributing computers on a one-to-one basis throughout an entire region, and not to sprinkle them in low density throughout the country.
Choosing the order of districts for receiving the laptops is not an easy decision. It is our experience that the greatest gains can be made when the initial sites set strong exemplars for use. This is possible through selection for high likelihood of good educational practice with computers as well as by addressing critical social needs. One tactic that has worked well is to invite local communities to propose what they will contribute in order to be chosen. This does not imply funds; rather it often is support by parents, or the offer to give teachers time and support to learn how to successfully use the computers, or plans for innovation in education enabled by the computers.
Digital Content
It is strongly advisable to begin to develop new content that takes advantage of the widespread presence of computers. Rather than just re-hosting existing content on the computers, new content and new approaches are possible. It has been our experience that certain limitations in the learning environment that seemed intractable can be ameliorated not just through new methodological approaches, but also through new content enabled by the technology.
Copyright Problems
The OLPC community, and the country educational bureacracies, will need to "cut out the middleman" in procuring textbooks for OLPC. That is, we or they should contract directly with textbook authors to write textbooks for a one-time fee, publishing the textbooks under a free (CC-ShareAlike or GNU FDL) license. Older, existing textbooks will not be usable, even if scanned in, unless their copyrights can be purchased. The author and editor can be paid what they would usually be paid by an educational publisher.
- We should ask publishers for CC-style licenses (no-cost, allowing adaptation and republishing into all currently unprofitable languages) for their textbooks and all other educational materials, to be translated into the languages of the developing countries. Or all languages, with distribution only in target countries.--Mokurai 22:07, 17 October 2006 (EDT)
While new interactive teaching programs should also be developed in parallel, the process of buying static textbooks will be the fastest path to books that are usable on the OLPC by the very first kids to get them. This process should start immediately. There's no need to wait for hardware. The bulk of the work can be delivered in Unicode plain text. Markup (e.g. OpenDocument or HTML) will only be needed to incorporate graphics -- and that markup can be inserted later, once the book-viewer software's specs have settled.
- There are considerable advantages to using XML markup from the start. With a suitable DTD template, writers can edit visually in Quanta Plus (Free Software). Then we can use standard tools such as Cascading Style Sheets to reformat the material as books, HTML, PDF, or other formats.--Mokurai 22:07, 17 October 2006 (EDT)
It will take qualified authors and editors a year to write a good textbook, so it's time to get started today. Most such textbooks can be translated into a variety of languages, speeding up the process for subsequent languages (though these will still require final editing by an expert in the subject and in the language). But we don't even know yet what grade the first kids will be attending (high school? elementary?), or what subjects will be included in their course of study. As soon as these things are pinned down (or guessed well), the process of finding qualified and available authors and editors can begin.
- The process can begin for all grades immediately. We know that first graders will get Laptops within a few years in one country or another, barely enough time to get a reasonable set of textbooks together. We know that all subjects will be needed. Some Free textbooks are already available--Mokurai 22:07, 17 October 2006 (EDT)
If we don't watch out, textbooks will end up on the critical path, slowing down hardware distribution. (And if hardware isn't on the critical path, then we should wait for better hardware before producing a million of these!)
I am worried at the fact that in Mexico, the content of the "Free Grammar Textbooks", for example is not available over Internet because of "Copyright reasons". Also an recent development in Educational Material called the "ENCICLOMEDIA" is not available for Private Schools, for the "same reason"; if the greedy friends of governments that develop Educational Material, behave like that and try to extract money from the PUBLIC FUNDED developed material, I think is unfair... Country governments should warranty that will NOT HAPPEN --Dagoflores 03:13, 19 March 2006 (EST)
Could material be adapted from existing open sourced documents and then of course edited for accuracy? Obviously for in-depth course studies this would not be an option,
- Actually, we can get high-quality, in-depth materials. There are a number of free textbooks, and we can make the pitch to academic authors that getting to hundreds of millions of children should be worth more than money.--Mokurai 22:07, 17 October 2006 (EDT)
but on certain more limited topics, would this be a possibility? This would of course eliminate all costs but editing/verification. What makes me consider this is certain articles/items from the various Wiki* that are well written, in-depth, and well cited. --Daniel Patterson (dbpatterson at exeter dot edu) 16:41, 2 September 2006 (EST)
- Or we could just package Wiki* mirrors on village servers in places without full-time Internet.
Using living books (eg. Ambleside Online)
Ableside Online http://amblesideonline.org/01sch.shtml has a complete curriculum from K-12 using living books as the foundation. Most of these books are already past copyright, and a majority of them are already in electronic format. This is part of that project. Charlotte Mason's eductional philosophy is the foundation of the Ambleside Online book list. CM believed in using living books instead of textbooks, so that would eliminate the need for textbooks completly. Math would be the exception, but there are old math textbooks that are already scanned in, eg. http://www.donpotter.net/math.htm. Also, current science and history from the last 100 years would need to be addressed, but sites like wikipedia could provide the foundation to start with.
Collaboration
The OLPC enables new, more customized environments for learning. New communities of learners can be enabled because every child will have a laptop and the laptops will have connectivity. The borders of the classroom walls are removed and learners can get information, work with others, pursue their interests and passions, and work in ways previously not possible. How to support this, what tools are needed, how to re-arrange the curriculum, are all areas for serious study.
Strategic Research
OLPC is new. Very little experience exists regarding how to use one laptop per child effectively. There is virtually no experience with this in developing countries. Thus, it is highly recommended that an ongoing panel of thinkers study, reflect, discuss, and propose based upon what is working, what needs improvement, and what new possibilities are enabled. It is also highly recommended that this group be broad-based, not limited only to those in the education field. We propose including other intellectuals, artists, civic leaders in order to provide a diversity of experience and expertise.