Trackback URL: Trackback
Permalink URL: Permalink
Sharing is a joy for all
July 3rd, 2006
These days networks to share information on the Internet have grown enormously in number, size and popularity. There are several social networks like SmsAc, Hyves, Hi5 or Internet repositories to store and share information like pictures (Flickr) or links (Del.icio.us) to name just a few. The social networks are meant to share interests, news and to send emails and meet other people. They are especially popular among young people and really fullfil a need to share and be part of a community. Apart from those social networks, there are networks that are more aimed at sharing data. They are formed for people of an (educational) institution, a large company or a community that shares a specific interest. They all have in common that there is a repository (a kind of database) that can be used to upload and download information and/or files. As such it is not so much different from the Internet itself (for the user at least) apart from three issues: the interface is always the same (which is convenient), the information units are files rather than a set of pages to be presented (i.e. it is not as much directed to browsing, but relies on a retrieval mechanism) and the information contained is to some extent controlled. Depending on the community and the kind of information they provide, they offer specific functionality in the interface. Repositories come in shapes and sizes. What they have in common is that they make the information everywhere available and that information can be shared of course.This is in particular interesting for educational institutes since students and teachers can work (partly) at home and inventions, nice explanations, lessons, texts, pictures, etc. can be shared to prevent work being done many times. Moreover, and maybe yet even more important: accessibility of other peoples work might be inspiring or a nice complementation of the educational material that would otherwise not have been available for lack of knowledge or resources.
What I want to argue in this short paper is that from the currently developed repositories on the Internet, there is a lack of middlesized easy to install Open Source repositories. We (i.e. the company RayCom in the European project called OpenDock) have been looking at several repositories and found lots of developments (been) going on. What separates them from one another is the software and platform it is based on and the model of distribution as well as the safety model.
I will quickly review the first two issues in the next paragraphs and give some examples.Since I suspect that the reader is not interested in technical details about pros and cons of specific software for repositories, I will just stick to a few remarks. First, it is important whether the actual software to store and retrieve the data is residing in one spot (the centralised model) or that it is distributed. In the first case lots of data traffic to a single point, demands for a heavy server that can handle that. If the workload can be distributed over several sites (the distributed model), the capacity of the server is less demanding, but some synchronising protocols are necessary. The other extreme is a peer to peer system (the decentralised model) where all sites operate rather asynchronously. This model comes close to the model of the Internet itself, with the difference being added functionality in case the user has registered himself via a login procedure.Secondly, for a new site to set up, it is important to know the platform to run it on. Is it based on proprietary software and which capacity and type of server (typically Windows Server or Apache) is needed? Of course these issues are vital for the availability of a solution in building a new site.I will give some examples of repository projects here with a very short description. I’ll provide some links at the end of this blog.
- Ariadne: A European project for a repository, closely integrated with a LMS called Moodle and Ine. Works with Oracle (this is proprietary software) as the database for data storage.
- Lionshare: A large project in the US, building on Java that is a peer to peer solution with some layers on top to harvest the information that resides in the peers. Only educational institutions can participate.
- Edutella: A peer to peer solution build on the Java package jxta. It uses the meta data model rdf, but provides translations to other meta data models.
- EduCommons: a project that is comparable to the OpenCourseWare of MIT. It is built on Zope and Python and seems to use a centralised model. The units of information stored are in principle courses which is a disadvantage in my opinion for a broader application of this software.
- OpenCourseWare: is a large project of MIT to publish all their courses online. It is one very large server with replication at other sites in the world.
- DSpace: A project from the Open University in Holland. It uses a centralised model of a (larger) Unix/Linux server and is built on Java-jxta, Java server pages and Tomcat. It supports the IMS Learning Design meta data model.
- Connexions: with a product called Rhaptos written in Python/Zope. Just as with EduCommons, the units stored in the repository are courses.
Of course I have named just a few. What seems to be missing is a freely available install that’s easy enough to handle by novice users. Some of the solutions have indeed easy installable programs for the peers, but they need to participate in a larger network with intermediating servers which might not be available for an ordinary user (Lionshare demands you to be an educational institute, EduSource requires you to be in Canada) or just different from the requirements of the community. All have their (dis)advanteges and the choices are overwhelming.
Just like in the seventies and eighties the proliferation of new computer programming languages was enormous, since everyone thought to invent the ultimate programming solution, this seems to hold for repositories in the new millennium (sofar at least). I think we have to accept that as part of an evolutionary process. Good work has been done by EduSource that developed ECL (educational communication language) in Canada to connect these repositories. And the esteem for open standards and Open Source Software and platforms (Java, Apache) is commonly high to enable quick development and connecting various initiatives.
The main problem is that every community has demands for specific functionality and user interface features (if not only to give the community a face) depending on the type of data they want to share and the information about the data they commonly want to store. Various models for structuring this meta data have been proposed. (We have, among others: Open Archive Initiative, DublinCore, and specific learning models like IMS-Learning Design and SCORM.). As a result there is always programming involved to set up a new community and the maintenance of the network requires some computer skills that teachers might not have.
We of RayCom think therefore that what is needed, is an Open Source scalable solution that is easy to install on a standard web server. This takes away the burden of technical details of the protocols needed to communicate between peers of the network. This responsibility lies with the server provider who has the technical skills to set up a site. This site then should be able to connect easily to other sites and even greater networks. The idea is that this intermediate layer is highly independent of the layers above as well as the actual user interface for the end user.
This means that the types of data and the adhering functionality for the data to be stored and retrieved should be highly compositional to cope with other network meta data models. Furthermore, the user interface should be totally detached from the actual repository. As far as we know, repositories of these kind do not exist yet.
Another thing I should mention about sharing for educational material, is that a solution should be found to cope with the intellectual rights. People are more likely to share their work if somebody else can not just copy the work and claim it to be his work. At the moment a very good initiative to deal with this issue, is the Creative Commons license. Its purpose is to let people share content/ideas without losing control over their work. As such it is familiar with GNU GPL licenses and the like as seen for a while in the world of Open Source Software. For more information on that, see my other posting in this blog.
Currently we are working in the OpenDock project on a prototype to test these ideas with a specific meta data model IMS Learning Design and all the material under the Creative Commons license. For more information about this philosophy and this initiative, see (What do we want from a floss repository?). I invite people to comment to this blog or mail me with ideas about what a good repository for educational material would look like. After all, if you share your view with us, it is a joy for us too.
References:
Ariadne: http://www.ariadne-eu.org/
LionShare: http://lionshare.its.psu.edu/
OpenCourseWare: http://ocw.mit.edu.index.html/
EduCommons: http://educommons.sourceforge.net/
EduSource: http://www.edusource.ca/
Edutella: http://edutella.jxta.org/
DSpace: http://dspace.org/
Connexions: http://rhaptos.org/
Entry Filed under: Repositories, Culture of Sharing, Content Storage
2 Comments Add your own
1. The Bazaar - Bazaar proje&hellip | October 10th, 2006 at 7:17 am
[…] At OpenDocument.net that will be launched soon, there will be software available that is based on PHP and MySQL too that will offer you a way to group documents in a repository (rather than information in general as for opendock.net). The way the metadata is organised is so flexible that new modules for different metadata formats can be added to the repository. In the blogs Sharing is a joy for all and What we want from a FLOSS repository it is argued how a repository for educational material should look like in our opinion. It supports IMS Learning Design and Creative Commons and allows the units that are stored to have a structure of their own. They comprise all the (meta) information and files or references that belong to their context. We believe that we can combine the benefits of having a flexible metadata system and a lightweight Internet repository service as to both group and tag information in a flexible way to facilitate new communities of practice. […]
2. The Bazaar - Bazaar proje&hellip | April 17th, 2007 at 1:07 pm
[…] This way the data stays closely to the content provider offering more (psychologically) control and guarantee of continuity plus the advantages mentioned before. What makes it different from other repository solutions, is that there is a strict separation between the interface that can be accustomed to the community (see an earlier article of me: Sharing is a joy for all) and a kernel that can easily be extended and modified so that with some effort a personalised repository can be built that talks with other repositories to share the information. […]
Leave a Comment
Trackback this post |
Subscribe to the comments via RSS Feed