I am delighted to be able to write about OAI-ORE – on whose advisory board I am. I also had the chance recently to have a long discussion with two of the people driving it, Carl Lagoze and Herbert Van de Sompel who writes:
We thought you might be interested in the presentation about OAI-ORE given at the recent OAI5 Workshop which took place at CERN, Geneva, Switzerland.The presentation gives an insight regarding the problem domain in which ORE operates, and in the evolving thinking regarding potential solutions.The presentation was recorded on video and is available for both streaming and download at:We would also like to mention that Michael Nelson did similar presentations on behalf of the ORE effort at the recent CNI Task Force meeting and at the DLF Forum:GreetingsCarl & Herbert—Herbert Van de SompelDigital Library Research & PrototypingLos Alamos National Laboratory, Research Librarytel. +1 505 667 1267
ORE will be an extremely important development. Essentially OAI-PMH described the metadata that accompanied objects in a repository and works fine as long as the object and the metadata are eseentially the same – such as depositing PDFs. But for compound objects (for example an HTML file with several images) OAI-PMH is not able to cope – it cannot describe the relationships of the components. Which is why, for example, the only feasible way of repositing HTML files is to zip them.
ORE can describe compound objects, and will also be able to describe compound objects that intersect. Although ORE is formally not dependent on the objects being in repositories the motivation – OAI – means that is where the drive comes from.
My advice so far has been – keep it simple (e.g. I have to be able to understand it) – get early implementations out – and make sure the whole community knows what is going on. After all, every repository will have to be able to implement it, and that includes filing systems and directories on servers.
Then no-one will have any excuses for not putting their stuff in repositories.