Transcript of MP3 Audio File Presentation by Ray Elferink, Raycom BV


 * Back to Learning Design Update Section.
 * Audio File of Presentation - Ray Elferink, Raycom BV (MP3 Format - 7.1Mb, Length: 8 minutes).
 * Notes from Presentation and Discussion on the OpenDock Project - Ray Elferink, Raycom BV.
 * Transcript of MP3 Audio File Presentation by Ray Elferink, Raycom BV.

My name is Raymond Elpherink. I'm from a Dutch software company called Raycom. Within the OpenDock Project, we have been working on a distributed repository called OpenDocument.net and I will give a very short overview of what we’ve been doing. I'll try to keep it very brief, so that we can go ahead and you know do the...

OK. So. Our aims... When we started this, what we wanted to do was build a system. All systems that are out there... There are a lot of repository-type systems out there today and they're all like huge central repository architectures, which are all built in Java and if you want to run them, you need a complete server and a whole crew of people that understand how to set that up and keep it running, and what we wanted was something else - something that can be run by a very small institute or bunch of people or whoever. So, we wanted low technical requirements and low maintenance, but we wanted to have high accessibility, so that means that we wanted to be able to get to the data which is in the repository from virtually anywhere - and we wanted to stimulate sharing and reuse because, well, actually, I believe that the content providers of the future are normal people or teachers or how do you call them, Mark? Practitioners. And those people need... In European law, if you create something it's yours, you have copyright. There's a lot of people that create stuff that don't really think that they will earn money from it. Nobody wants to buy it from them and they would be more than happy to share it with others. They could use it if they want but they forget to tell that to people. So we want to stimulate the notion of sharing and re-using, which people are doing anyway, and we wanted to support standards (which started off with supporting Learning Design but I'll get to that later).

So, our approach. We decided to build a LAMP-compatible web service. LAMP is Linux, Apache, MySQL and PHP. We decided to... we went for a distributed architecture. That means that it's not a central system, but it's like a peer system of very small installs that talk together. We built in support for Creative Commons, which sort of tackles the whole re-use and sharing issue from the start. So, if you put something in the repository, you can also say you know which Creative Commons licence it's on. Items are stored in containers, so just if you don't think about Learning Design if you put in a picture... Then it's not putting in a picture in a repository, but actually you get like a container in which you can put a picture so you can also put multiple pictures in, and in that sense, content packages (like the learning design UoLs) are actually containers with stuff in it. And one of the things that's in there is, of course, the manifest file and we can parse that to get to the internal metadata of the Unit of Learning. And next to that, we have built or are building in the possibility to leave comments on items which are in the repository and tag them.

OK. Distributed architecture means that every node of the network of small installs is actually the central system. So, if you set up one of these repository nodes then that's yours. The stuff that's in there is yours and your system is the central system of the whole network. So your system talks to all the other systems in the world, gets the data, and presents it to you or your interfaces. The same goes for all the other nodes in the network.

Metadata is distributed across the network and there are sequels between nodes for communications. So basically, if you have a bunch of nodes like that, you are master of your stuff. So, if somebody here wants to leave a comment on something you're doing, then the other nodes should be able to make a call into your nodes to leave that comment because it's yours. If somebody leaves a comment with your stuff, it's yours. And all the nodes allow RPC calls from interface applications - and that could be web applications, but that could be anything. So, we’re currently looking into the possibility to use RELOAD as an interface to the repository, so that you could from within RELOAD get to data, which is in the repository.

[Points to slide] This is the web interface that comes with it. Basically, I put this in because I didn't want to give a demonstration, but I put this in just to show that. Because it shows, you know, some basic information about the use of - in this case - the issues, the possibilities of the licence that leads to the Creative Commons website. You need to get the information that comes out of the Learning Design manifest file. As we put it in now, it's basic information but that can be filled out because we can parse the whole file, so basically at some point this can be a nice structured interface of activities or whatever. [Points to slide] And here we have the contacts, which is in the file. This is also searchable, so if you would look for a picture, then you could find, you know, the picture that you just put in the one container here, but we could also find a picture that is inside a unit of learning somewhere. In that way, we sort of up the possibilities for re-use. And if you download a container, you get it as a zip file. So basically, if you download a UoL, you get your UoL, and if you download one of these, you just get one of these.

Now - and this is my last slide - the road ahead. The support for Learning Design is built in as a plug-in. In a similar way, we can, you know, support and parse anything that has internal metadata. So basically, on a container level we could just as easily parse SCORM objects, or QTI stuff, or Open Office documents. On an item level, we could look into parsing internal metadata inside pictures for instance or inside, you know, other stuff which is in there. It's not that we are going to do all of that but it's something. It's an open source project, so at the moment it's all in our hands. But we are moving towards a first release into the world.