Last night, Herbert Van de Sompel announced the availability of the draft specifications and user guide for Object Reuse and Exchange (ORE). This effort, under the auspices of the Open Archives Initiative (OAI), seeks to define a standard for the description and exchange of aggregations of web resources.
If you stop to think about it, the world of the web is at once a simple and a complex place. It is simple in that it is made up of resources (this HTML page, the jester's cap graphic in the upper right corner, the, etc.) that are uniquely addressed by URLs. ((Technically called "URIs", but let's not go down that path for the sake of trying to speak the vernacular of the public web and not that of the technicalities of the underlying standards.)) It is complex in that there is a relationship between this HTML page, the jester's cap graphic, and the audio version that is not explicitly stated -- at least not stated in a way understandable by machines processing content on the web. Which is to say, how does a web crawler know that the linked audio file is an alternate version of the text of this posting rather than a link to some holiday music file that I am critiquing?
This is where OAI-ORE comes in. It offers a machine-parsable way to describe relationships between various web resources. In a manner of speaking, an "ORE file" is a new kind of web resource that encapsulates and describes the complexity of real-world compound documents that already exist on the web. The neat thing now, though, is that there is a formal specification for those compound documents that allows their nature to be understood by web crawlers, indexers, and content repositories.
There is lots more about OAI-ORE in the user's guide and specifications. Keep in mind, though, that these are alpha specifications that represent the best thinking and compromises of a couple dozen people. ((I'm honored to be among those serving on the ORE Technical Committee.)) There are undoubtedly issues and ideas that haven't been considered yet, so don't go too far in coding up an implementation that may change (in subtle or dramatic ways) as the public conversation gets going. An OAI-ORE Google Group has been set up to discuss these documents. Those working on ORE to date welcome your comments there, and to reinforce that, I've taken the until now unprecedented step of closing comments here on DLTJ -- put your comments into the Google Groups area.
Also remember that there are two open meetings scheduled to focus on the specifications:
We're looking forward to your questions and comments on the OAI-ORE draft documents.