Here's another "as it happens" for the archive, taking in a talk on Creating and Supporting Free Software in Africa by Prof Derek Keats, University of the Western Cape (UWC), South Africa -- in some ways a continuation of the previous edu-sig session. About 20 of us in the audience. He's a "CIO" and initiated African Virtual Open Initiatives and Resources (AVOIR). He'll probably podcast the audio, as he's recording to his iPod. Most of the world is not like Stanford. Let's start with "alarming perspectives". Making areal displays with numbers, e.g. numbers per capita with tertiary education, results in many distorted maps (worldmapper.org). Shanty towns around Cape Town is more his school's focus. There's a lack of critical mass in many areas, such as computer science. How to solve? Through collaboration (the only way). Building new universities and staffing them can't happen quickly enough. Free and open source software, applying lessons learned to building community, infrastructure, is the vision of AVOIR (pan-African FOSS). AVOIR is "established as a network for capacity building in FOSS engineering". Key links: http://en.wikipedia.org/wiki/Chisimba http://avoir.uwc.ac.za/ http://ics.uwc.ac.za Twitter: dkeats Facebook: Derek Keats email: dkeats@uwc.ac.za Phase 1 (2005-2008) has involved networking universities around Chisimba, a Web 2.0 AJAX application with attendant software (13 nodes so far). Kabul Polytechnic is part of the alliance (eQuality Alliance). Picture of training workshop in Kabul, re the portal, eLearning. Georgia Tech another partner... KEWL training in the Philippines. We haven't seen the interface yet (we're getting to it), am trying to visualize what Chisimba looks like (booth 818 here at OSCON). Word for "framework" in Malawi. Chisimba is highly modular, per Debian community (ala Synaptic), includes a package management system. Implements MVC. Yes, there's Python involved, hooked to gstreamer. Also: CURL, FFmpeg, Java... lots of toyz, went by pretty fast. Now Keats is demoing the remote package management component. You can start a new eLearning environment in just minutes, given enough bandwidth. Everything else is component based, e.g. adding a blog or some other resource is a matter of mouse clicks. Some components allow real time audio, shared white board annotation, filtering. With cut and paste, you can embed the live component in a blog, in a moodle or whatever. www.dkeats.com for more examples. USAID, Sun Microsystems, Geek Corps are part of the alliance somehow (USAID focused on animal health). University of Nairobi, NOLNET in Namibia are working on implementing AVOIR, as is National University of Rwanda. Electronic Thesis and Dissertation system is another of the components. Sun is providing hardware in six of the thirteen partner institutions so far. Social Content and Networking for Schools connects like 50 schools in poor areas to create an intranet (not connected to Internet) with social networking and eLearning. Social tools help curriculum content to grow. Chisimba reminds me of Plone, sort of, but it's built very specifically for customized eLearning environments. Sometimes Chisimba users become potential interns, helping with Chisimba's development. Those who've become interns have contributed usable code without exception. Realtime podcasting application is a 3-click process. Start, Stop, Publish. Bandwidth is the biggest challenge. Some institutions don't realize the value of networks. Culturally, there's often a tendency to keeping quiet, to be deferential, whereas on the Internet you're encouraged to be more egalitarian (out of necessity in a lot of ways). Salary structures and high turnover are other problems. People move on before they have time to become Chisimba developers. Questions: what about OLPC (one laptop per child). Getting computing resources to students is a good thing, not limited to the XO initiative. Providing labs on a massive scale is the only way to provide a lot of access in many cases, as the students have no access to laptops. This is where Sun comes in, in some cases. Government funding has been problematic in that the South African department with jurisdiction has had some turnover, some friends have left. Why start a new framework? In part to encourage local participation, hard to break in to ongoing open source projects in some cases. A goal here was to be self-managing, running all aspects, including the version control, servers. Sometimes it's just easier to start from scratch. AVOIR is about developing competence and confidence among young African software developers. It's an exercise in community building, as well as a project aimed at producing quality software (they go together). I asked about fonts, internationalization. A lot of the languages just use Latin-1 characters, but in Afghanistan they're working on the Farsi translation (of Chisimba). The Chisimba package itself is sophisticated about language issues, uses UTF-8. The animal health project has a mobile devices API although Dr. Keats isn't sure how far they've come along with that. Kirby