RDF in a Nutshell

Okay, so the past few days have seen some major flamefests break out over the new RDF specs that were just published. Everyone is wading into the topic, it seems and all I really want to add is my two cents from the perspective of someone who's been trying to get RDF and the Semantic Web into his small little pea brain.

First, thanks to Danny for The Resource Description Framework in 500 Words which is as concise a description of the concepts as you can get. (Please add the next 500 words Danny!) Here's a snippet:

Descriptions are made in RDF using statements. A statement has three parts: the thing being described, the characteristic of interest and the value of that characteristic. For example, the thing being described might be a book, say "A Christmas Carol" the characteristic of interest (property) the author, and the value would be the name of the author, "Charles Dickens". In RDF jargon these three parts are the subject, predicate and object, and together they form a triple. The subject is a resource, the predicate is a special kind of resource and the object can either be another resource or literal text.

It took a while, but I finally got something very important into my head: The concepts of RDF and its implementation are quite separate. I was going into the whole thing from RDF/XML on up - trying to figure out what the heck all that fugly markup was trying to do. It turns out that starting from the top down is actually a nicer way to go about the whole process. The concepts are very clear and immediately obvious in their usefulness and power. However, where the rubber meets the road is where the problem lies.

Bitworking has several recent posts which encapsulate all the opinions out there by Shelley Powers, Mark Pilgrim, Tim Bray (the originator of XML), and more. And it includes this very brutal summary (in his opinion) of RDF:

Frankly I think it's about time the tide turned on RDF. For a long time it has been beyond reproach. The attitude has been that because RDF was thought up by a bunch of really smart people and that it's a W3C standard that it was above criticism. I think a healthy dose of skepticism and a critical eye turned on it by people outside of the usual circle would be helpful to RDF, and it certainly couldn't hurt the XML serialization.

So to summarize, my sources of animosity towards RDF:

  1. The XML serialization.
  2. The push by RDF proponents to turn every format into RDF/XML.

And Tim Bray really has a lot of really good points along these lines to make in this email to the xml-dev mailing list.

I have to say that I'm on this side of the fence. RDF/XML is really, really obtuse. Personally, I don't think we need to throw out all the work so far and start from scratch, but there needs to be a serious effort on the part of all the RDF people to make the technology more accessible. Otherwise it's going to just get replaced, and quickly. RDF seems really cool - it's a shame that its rate of adoption is so slow because smart people forget that dumb people have to use their technology in the real world.

By the way, I've been linking to Danny for a while now because he's both a Java developer (and author of a book I actually own) and an expat blogger living in Italy. Now it turns out that he's quite the heavy-hitter in the Semantic Web space. Why this surprises me I'm not sure, noting the name of his weblog...


< Previous         Next >