RDF: The next Operating System

Speaking of Danny, he’s pointed to two terrific RDF resources:

The first is ActiveRDF allowing access to RDF data from Ruby-on-Rails. This one I can’t wait to try.

The RDFRoom — using RDF to create a virtual 3D room. Appropriate after my last post.

I wonder how OPML could be used to create a 3D room? It would probably consist of one, looong, staircase.

This entry was posted in Technology. Bookmark the permalink.

6 Responses to RDF: The next Operating System

  1. I’m still reading over Practical RDF. I’ve learned a lot from it. Though so far I haven’t figured out exactly when I should prefer RDF to some other XML based format. I suppose the examples and the ontologies given at the end of the book might give me a better grasp of that when I get there.

  2. To answer my own question above, I’d recomend this article by Tim Berners Lee. It’s from 1998 and it clarified for me the advantage of RDF over XML in situations where you want to run queries against the data.

  3. Shelley says:

    Actually, Tim said it best recently when he said you can easily and instantly merge _any_ data modeled in RDF.

  4. I’m getting that. I think when I picked up your book “Practical RDF” I was expecting a different version of RSS. I knew just about nil about RDF. It’s taken me awhile to realize how different RDF is.

    I had initially assumed RDF was a different vocabulary (perhaps slightly better?) for solving the same problems that RSS/Atom address, but really RDF is trying to solve a completely different set of problems. I imagine there must be a large number of people who are under the same impression I was.

    I’ve also been reading “Beginning RSS and Atom Programming” by Danny Ayers and Andrew Watt. I think the biggest surprise for me these last few weeks was realizing the extent to which RDF goes beyond Atom.

    I do think RSS has some advantages for beginners. RSS is almost an English-language language compared to RDF. You look at the Google Base extenions to RSS and you see stuff like “expiration_date”, “immigration_status” and “sexual_orientation”. One feels that one can immediately understand what these words mean. Of course, a computer can’t know what these words mean, which is the advantage of RDF – it’s less human readable but more machine readable, it takes out the ambiguity for whatever query might be run on the data. But it takes awhile to understand that, or at least, it did for me.

  5. Shelley says:

    Lawrence, you’re comparing a single-purpose specification against an entire data model. The two are not alike.

    RDF is meant to be used with any domain data–syndication feed, medical, photo, even my little bits here in the weblog. RSS is a syndication format based in XML and designed for a single purpose supporting a limited set of data.

    RSS is meant to be consumed by computer; so is RDF. Some of the structure causes some XML purists to get all huffy, but it is no less ‘readable’ to a person who has never been exposed to XML than RSS.

    Now, RDF does require more upfront time to work with it from a developer perspctive. But then, it takes more time to work with Oracle 10g, then a does to work with a comma delimted text file.

    RDF is comparable to the relational data model–different mathematics, but the same concept.

  6. You’re comparing a single-purpose specification against an entire data model. The two are not alike.

    I do get that, though I only recently learned that. I think, before I started reading up on it, I thought RDF was a jazzier kind of Atom. I’ve since learned how wrong I was.

    but it is no less ‘readable’ to a person who has never been exposed to XML than RSS.

    Doesn’t the limited nature of RSS ensure it will be easier to comprehend? I suspect your sentence above would be more true if you changed “XML” to “HTML”. For someone who’s never seen HTML, then I’m sure RSS is opaque. The brackets and forward slashes must be confusing. But for someone who is used to seeing HTML, I don’t think RSS is much of a stretch.

    <name>Donald Duck</name>
    <employer>Disney Inc.</employer>

    If you’re used to HTML I suspect that is slightly easier to figure out than basic RDF. This is the basic example they give for RDF over at w3.org

    <contact:Person rdf:about=”http://www.w3.org/People/EM/contact#me”>
    <contact:fullName>Eric Miller</contact:fullName>
    <contact:mailbox rdf:resource=”mailto:em@w3.org”/>

    But, I’d also agree, it is not a huge difference.