What I saw at the revolution: .LRN/ OpenACS meetings in Heidelberg, Germany
This post originally appeared on my first blog, hosted by Harvard Law School's Berkman Center.
Last week I went to Germany for the OpenACS and .LRN meetings in Heidelberg, hosted by .LRN Executive Board members Michael Hebgen and Carl Robert Blesius of Heidelberg University.
(For the un-initiated, OpenACS -- Open Architecture Community System -- is a free, open-source software application toolkit (including many modules that work "out-of-the-box", as well as a sophisticated development framework) for online communities. Originally developed at ArsDigita (acquired by Red Hat in 2002), OpenACS is maintained and has been significantly extended by the OpenACS project, a community of more than 7,000 people around the world who use it in their organizations and as the foundations of their software and IT services businesses. The OpenACS architecture is the foundation for Siemens' award-winning global knowledge management application, Sharenet, which has been the subject of a Harvard Business School case study. But, it is also used by many other smaller organizations, like Greenpeace, and by countless individuals around the world, since it supports an entirely free-and open-source software infrastructure stack as an option for those who don't/ can't use tools like Oracle's RDBMS.
.LRN, also free and open-source, is a related project that uses OpenACS as its foundation and extends the toolkit for use in a variety of educational settings. MIT's Sloan School and nearly a million users at more than 25 other major universities and other organizations around the world are now using .LRN extensively in ten languages with another 20 on the way, in many cases as the main e-learning application platform for thousands, and tens of thousands of users at each school. It was also selected as the foundation application technology for the E-Lane project, which was organized by Spain's Telefonica and recently received a 3-million Euro grant from the European Union. Along with the better-known OpenCourseWare Project, .LRN is part of MIT's Intellectual Commons.
The .LRN pitch is simple. The educational world is still figuring out how best to deliver e-learning, especially to external audiences. While commercial vendors are adding zeros to their prices, .LRN is providing a suite of applications and a development framework free and open-source so you can save your dollars and have the flexibility for the innovations that will work at your school. The more people that adopt and contribute to the toolkit, the more powerful the starting point.)
Based on what I saw in Heidelberg, the projects are thriving. The number of registered users at openacs.org has doubled in the last two years. More than 60 people came from various parts of Europe and the Americas to get to know each other, present their work, and coordinate their efforts. Examples of the work presented:
>An advanced Learning Management System (LMS) developed by Vienna University and now used by tens of thousands of users there for online self-testing. Despite a fixed budget, by law Vienna U. has to accept all comers, and so needs to use e-learning extensively to leverage the school's physical and human resources. Vienna evaluated and rejected several commercial options as inadequate to their particular needs for features and support, and eventually selected .LRN as the basis for their platform. Aware that most instructors at Vienna are already familiar with Microsoft Word, Vienna developed a set of custom styles which test authors can use to designate certain parts of their documents as explanations, questions, instructions, answer options, etc. When saved, the document translates these styles into a set of XML tags, based on SCORM, that transform the Word-based test into a structured data file that can be imported into .LRN. In Vienna's implementation, this structured data file is then administered as a test using the .LRN "Complex Survey" module. Vienna has pushed the envelope on scalability: this highly dynamic application, using AOLServer as the web server and PostgreSQL as the RDBMS, gets 3.5 million page views per day and serves users with page load times that average 0.5 seconds. Vienna has agreed to contribute all of this to .LRN shortly under GPL.
>A multi-player (as in hundreds) simulation for Leiden University's law school. This application uses the workflow and notifications capabilities of the OpenACS core to coordinate interactions among the players in the simulation. While today all of the "work" is manual (students write and instructors and others respond), the system has been architected for increasing levels of automated response using rules-based AI that will be implemented over the next several years. This also will be contributed to .LRN under GPL.
>a highly sophisticated knowledge management system being developed for the UK's Camden City Council (the "county"-equivalent counterpart to the municipality of London). This too will go into OpenACS/ .LRN under GPL after it launches in August 2004 to over 1,000 initial users (government staff workers).
There were several other ambitious and useful innovations using OpenACS and .LRN shown during the meetings that will also be released into the public domain under GPL shortly. (Professionally-shot and -edited videos of these presentations will be streamed shortly from a server at Heidelberg University, courtesy of Michael Hebgen and his staff, and I'll post the the URL when it's available.)
The meetings offered useful insights into the dynamics of successful open-source projects. This was not the stereotypical gathering of hacker volunteers with outsized anarchic visions. 80% or more of the work being done on OpenACS / .LRN is done in-house by, or funded under contracts with major organizations and institutions. Several CIO's from these organizations attended, as did CEOs from several new IT services firms looking for alternatives to commercial software partnerships that have not met their clients' needs. .LRN/ OpenACS development is being done by extremely talented and experienced engineers, who have now worked together long enough to take advantage of each others' strengths and work around their shortcomings. Usability is a major focus area, and the community includes users and administrators to help guide that. The toolkit itself is now nearly ten years old, and the community has survived multiple process-related upheavals to prove its viability. Finally, the community is now organizing itself in a more formal way to fund promotion and development.
The perfect weather and setting helped assure a great time -- the meetings were held in the Marstall (originally the royal stables) with views of Heidelberg's lovely schloss and the Neckar out the windows. And our hosts were extraordinarily gracious and well-organized (thank you again!). With the momentum I saw here, the next year should be very interesting to watch.