Insert motto here

October 3, 2002

This just in


Forgot to mention…

One of the problems with posting when you’re falling asleep (a mistake I’m this close to repeating as I write this) is that you forget stuff. Like, the major reason I wanted to mention Metropolis was that I rented it from the local video store, which all of a sudden has an anime section. It’s no Allen Street Video (an establishment near Penn State which nourished many a budding anime fan), but it’s a heck of a lot more than their previous selection (none).

Of course, Suncoast also has a big anime section, and Cartoon Network has whole anime blocks now—which fans with long memories still find ironic.

I should explain that: Back when Pokémon had not reached American shores, there was a controversy over a particular episode of the anime which apparently caused seizures in a number of kids. At the time, some Cartoon Network exec swore that their network would never show this “anime” stuff, because it was too complicated and violent. (I don’t have a reference handy, so I can’t promise complete accuracy.)

To go even further on a tangent, I’ve seen the seizure episode (untranslated, not that it mattered much). No ill effects, but I also wasn’t sitting right in front of a big-screen TV, and no one was saying I could stay home from school if I had had seizures. (An insinuation I make entirely without evidence, by the way. Heck, just disregard this paragraph entirely.) #

Dumb blond-related reporting

Apparently, a number of news organizations reported on an allegedly recent study asserting that blond hair would die out in 200 years or so. (I encountered the BBC report via Doc Searls.) Supposedly, too few people now carry the “blond gene” for it to last much longer. Hilariously, it also suggests that “so-called bottle blondes may be to blame for the demise of their natural rivals.”

Naturally, this spread widely through the blogosphere, as can be seen in the Blogdex listings for the BBC article (1, 2; for some reason, the BBC provided multiple addresses for the article). The problem, as you may have guessed, is two-fold:

  1. The study is actually several years old
  2. The proposition makes no sense

As the New York Times (registration required) puts it, “Apparently it fell into the category ‘too good to check.’”

First off, blondness is not caused by a gene. Hair color is much too complex to be the work of a single gene. My hair, for example, is brown, which is somewhere between my parents’ hair (black and blonde). Second, genes don’t just disappear unless there’s something bad about them. Non-blondes may still have recessive blond genes that could result in blond children. Hair colors mix, yes, but it isn’t like mixing inks. The component parts still exist and can show up in unexpected combinations in later generations.

Honestly, didn’t anyone pay attention in biology? #

RDF Channel

Want to know what I’ve been spending time on lately? Well, one of Shelley Powers’s query to the RSS community got me thinking about the essential nature of RSS: What is it, and what does it need to do what it does? I concluded that, in terms of its data model, the thing RSS does which makes it special is associate a resource (called a channel) with a list of resources (called items) which, at any given time, are considered current. The most common uses for that are things like headline feeds, such as a newspaper (a channel) listing its most recent articles (the items).

In terms of RDF, all that is needed is a single property (call it “current”), which associates a resource with a collection of resources. Everything else, titles, addresses, descriptions, saying what the contents of the collection are, can be handled by the core RDF vocabulary or common vocabularies like the Dublin Core.

I quickly threw together a description of this idea which I linked to from a comment on Ms Powers’s post. To my surprise, this actually lead to an outside comment from Timothy Appnel, who called it “quite an intriguing and interesting concept.”

That, apparently, was all the encouragement I needed. So, I have now posted a much more detailed description of RDF Channel. The basic vocabulary is slightly different (Item is gone, but there is now a distinction between Channels and Feeds), but the big addition is the lengthy guidelines on how to describe a channel and the examples. It’s not quite complete, but it should be enough to get a rough sense of what it’s about.

Naturally, this means I disagree with Ms Powers’s suggestion that RDF is not suited to describing transient information such as syndication feeds. In fact, to prove that it could be done, and that RDF Channel is sufficiently detailed for real-world use (and to give me some experience using Python), I’ve put together a working aggregator that reads RDF Channel feeds (which include RSS 1.0 and TDL feeds). Given some more time to complete its primitive UI and add support for some new ideas, I’ll have something I’ll can post here as an example. For now, I’ll just say: It can be done. #