Harvard Open Source Courseware
mpawlo writes "Gnuheter reports that the Berkman Center for Internet and Society releases the H20 courseware software as open source. Two years and 1 million USD are invested in the software so far... The software has been tested at Harvard Law School, but should be suitable for other disciplines than law."
It's great to learn the content, if you're curious. However, you need the degrees the colleges provide. As a student, I think I could handle learning on my own with these "Internet courses", but there are only a few classes that are offered strictly as Internet courses. Maybe this will give the field a boost.
-----
Make Love not [Browser] War!
I know that they are two different things designed for different jobs, but I would think that Slash would be a more useful tool for online course content. It's open source, free software, well tested and works with a greater number of browsers than does WebCT.
I use FreeBSD at home, and WebCT seldom works for me, and when it does, it is slow. WebCT is worse than Windows, but unfortunately almost as widespread among people doing this sort of thing.
MIT Intellectual Commons (collection of related e-learning initiatives including dotlrn): http://web.mit.edu/cet/strategy/commons.html
What is .LRN? (from www.dotlrn.org ):
-A fully open source eLearning platform.
-A portal framework and integrated application suite to support course management and online communities.
-A scalable, secure, and enterprise-ready eLearning platform that can be deployed readily by small and large organizations.
-A modular architecture to permit flexibility and to drive innovation.
-A set of best practices in online learning shared in the form of source code.
The dotlrn project page has documentation, news, forums... It is hosted on the www.openacs.org site, which is the parent web framework upon which dotlrn is based. Besides the above, the framework has a rich architecture for managing permissions, users, groups, content management, course management, forums, email, and more.
One you have not mentioned is Ilias. Not pretty (yet), but it works well.
The rotisserie that they describe does not seem to be a collaborative tool, but rather an asynchronous discussion tool.
The obvious comparison is not other [expensive] courseware type systems, but Slashdot. Slashdot's system of open, anonymous graded peer review is probably at least as good a way of refining knowledge in this way. A side by side comparison would be very interesting.
Just think, if they had dropped a million on slashcode - it might even be able to spot dupes!
Humorous signatures are over-rated.
I am currently working at Middlebury College developing a GPL course management system called Segue.
segue.middlebury.edu - our current version, in use by over 100 courses (double the total number of course websites from last year).
The Segue Project Page
Segue approches Course Management from the "Course Website" paradigm as oposed to "course folders" paradigm of BlackBoard and WebCT. We feel that websites as they are, are a superior way of displaying information than the idea of posting documents for download. Our goals were to make a system that is platform independent and will allow even the most technically timmid professors to quickly and easily get their course information and discussions online. On both fronts we've had much success; professors find the system easy to use (even the foreign language departments) and all functionality is availible from any platform with the exception of WYSIWYG text formatting. We are looking for a browser-independent XML WYSIWYG editor to replace our ActiveX one for PCIE. Any recomendations on this front would be welcome.
Segue is written in PHP and currently runs on a mySQL database. As of May however, we will be using ADODB to support virtually all databases. In April our development team will be heading to MIT to work out OKI interoperability.
Our code is all GPL so check it out!
"When ideology and theology couple, their offspring are not always bad but they are always blind." -- Bill Moyers
Everybody should look at Moodle, it have lot's of features (assigments, quiz, peergraded assigments, forums...), authentication works with ldap, smb etc...
More features to come with 2.0 release, then it will fix to large enviroments too.
What strikes me, then, is how stuff like WebCT and others end up duplicating admin effort. Download your classlists from the central admin system, convert them to CSV, upload them to WebCT...and do this repeatedly to catch any modifications (or, more likely, the lecturers add names manually so that they can produce class lists that are completely out of sync with the admin system). What is needed is to start standardising on how to do XML-based messaging between systems so that students registering automagically get put on the courseware system.
Even more suprising - campus admin systems seem to be developed by folk that focus on finance and HR. The stuff that students and lecturers need comes in as an afterthought. The more I think about what academics and students need out of a registration/academic admin system, the more it looks like a courseware system that's grown a couple of new legs, like:
I've done some work at the university where I am on a homegrown registration/timetabling system. It isn't fun -- especially dealing with the huge inertia of "we didn't have this stuff in the old days (can check timetable while registering), why do we need it now?" I wish I had the capacity to add a baby course-CMS component to it all, as I'm sure this would replace 90% of the WebCT use on this campus (people still think you have to spend megabucks so that you can "make webpages!").
For that matter, I'd like to see similar stuff for campus admin systems -- because the requirements for presenting registration info can be (and in our case, are) very localised, so we may need to write a front end for advisors to use for registration that ties in tightly with the monster campus HR/finance/student systems from PeepSoft/Oracle/ITS/whoever.
It's seemed to me for quite awhile like courseware would be a natural target for opensource development. Maybe it will. Certainly the amount of $ you have to drop on a repackaged apache/mod_perl/buncha scripts and DBM backend is, um a little wierd. I looked around at various things to see if they would be easily integrable into our home-grown registration system, but didn't get very far here.
(The political use of the term Open Source can also be irksome. A major university in <name-of-country-witheld-to-protect-the-innocent&g t;
got a big cash award for being an "open source centre of excellence" after developing their own courseware platform on IIS/ASP + MSSQL (stupid, but not absolutely evil) that doesn't work with Mozilla (evil), and then complain in the press about the perception that OSS is inseperable from Linux (legit complaint, but get your house in order first...you expect me to dump my plans to deploy Mozilla for our students so that I can run your *open* (?!?) CMS?)).
Year before last, I taught a module on "Web Computing" to the Comp Sci crowd here, which I did an overview of HTTP, PHP programming, enough SQL to survive. I wrote a very quick and very dirty bunch of scripts to let the students upload and edit PHP stuff on the student server for doing their assignments, then I added stuff so I could record grades for the work. It was quick -- and easier than any of my previous struggles with WebCT. (I note that a couple of other people have tried WebCT here, and then gone off to write their own mini-courseware stuff).
Of course, then I had to dump the marks from the DB, convert them to CSV, load them in Excel and mail them to the departmental mark capture slave to type into the mainframe. Go figure.