On Feb 3, 2009, at 5:29 AM, Leigh Dodds wrote:
And within the context of the web as a whole, URIs will break. That is a given. But within a specific community of practice we might reasonably ask whether we can do better?
Yep, we can and occasionally should. One way (only one!) is to create a Persistent URL (or Persistent URI) for ones that you care about.
In the latest updates to the Persistent URL server code (see http://purlz.org ), we added the ability to add type information to PURLs. Creating a PURL with type information allows the creator or others with authority to manage a URL's resolution process to maintain it over time.
A PURL can be associated with metadata that describes to the world whether it should be interpreted as an information resource or as a "real-world" resource (returning a 303 response code and optionally relating to a target URL containing RDF metadata).
Of course, that is still only good for URLs (not all URIs), but I think it is a step in the right direction. Hopefully, we will be able to address the same issue for general URIs at some point in the near future and extend the PURL concept to deal with further computational separation of URI types. Particularly, I would like to be able to computationally separate information resources (e.g. a URL to a Web page about a dog) from "real world" objects (a URL or URI used to identify my dog) from concepts (a URI identifying the idea of a dog). We can do that now, but only via commonly agreed practices that are not yet official standards.
I'd love to hear from the LOD community regarding these concepts. Flame away, please.
Regards, Dave -- David Wood Partner Zepheira - The Art of Data http://zepheira.com/team/dave/
