Repository as a Trusted Intermediary

The idea of a trusted intermediary that makes content both durable and usable with a "chinese menu" of added-value services is my new favourite definition of repository. These words come from the DuraSpace project's midterm report, and although they were not penned with repositories per se in mind, I believe that they provide an excellent description of their rationale ie to increase trust in material created in

• a random place on the Web.
• my rented niche in the Cloud
• my departmental filestore
• my own desktop.

So I am particularly pleased to congratulate the JISC EdSpace team on their recent upgrade to the EdShare learning resource repository at Southampton, because they have helped deliver on the first bullet point - adding trust to web resources.

I have been using EdShare to distribute material from the modules that I teach. Much of this material consists of PowerPoint lecture slides that I have created, but a significant proportion of it is material available on the open Web - perhaps other people's slides, papers or reports from their own web sites.

In the past I have had two choices: either deposit a link to the web page or deposit a copy of the web page. The former is a lightweight solution and obviously the right "Web thing" to do when you just want to provide a URL pointer to someone else's resource. But the latter is the right "repository thing" to do in terms of making a safe and durable copy. Except that I don't automatically have the right to clutter up Google space with ad hoc copies of the same material and reducing their Pagerank. So most of the time I have settled for "just linking", at the price of accepting that some of this material will move or disappear before I teach the topic again. In the words of Humphrey Bogart, I know that I'll regret it -- maybe not today, maybe not tomorrow, but soon, and for the rest of my course.

Now EdShare lets me have my cake and eat it. I can deposit and disseminate a link to the external material (as before) but the repository will make a dark copy and start serving that if the original disappears. Essentially they treat important material that I find on the Web in the same way that they treat important material that I move into the repository. Both get managed, indexed, thumbnailed and subjected to the normal range of repository services.

So I'm delighted that I can now do the right Web thing and the right repository thing at the same time.