Metadata Wins Again!

Volume 6, Issue 72; 21 Aug 2003

Explicitly making use of metadata to drive the production of this website has been a total win several times.

All things are contingent. And there is chaos.

Spalding Gray

At bottom, this website is really a loose collection of essays. There are several views instantiated over the collection, and they give the appearance of order. That's a good thing. But the fact that there's very little intrinsic order imposed by the publication system is even better.

On more than one occasion, I've been saved by the fact that I can influence the publication through no more than the addition of metadata. For example, most of the images on this site are JPEGs and I rely on embedded metadata in those JPEGs to provide information about those images. But the animated image in one of my Paris essays is a GIF. Because I don't extract metadata from GIFs, it appeared momentarily that I'd have to do implement all sorts of special case logic to handle the GIF image.

Not so! The metadata doesn't actually have to come from the image. By simply adding appropriate metadata to the system, I was able to make the publication process handle the GIF image without any changes whatsoever.

I won again this morning!

Edd Dumbill asked me to write a piece for XML.com. Now, fair is fair, XML.com asked me to write the piece, they should get first dibs on publication of it. But I still want to put the essay in my RSS feed.

My first thought was that I could write some sort of placeholder that pointed to the XML.com article. But then I realized that the RSS feed, like everything else, just comes from the metadata! So I added a snippet of RDF to describe the essay and “Bob's your uncle!” it appears in the RSS. It even appears in the date, topic, and subject indexes which I hadn't even thought about.

This metadata stuff, it's got legs.