Jeremy Hylton : weblog : 2003-11-04


Tuesday, November 04, 2003, 12:44 a.m.

I learned a bit about Technorati tonight. I started with a Cosmos report on the Daily Python URL.

Cosmos tells you about links among weblogs. It reports when a weblog was last updated. The link search reports on links from one weblog to another. You can sort it be date -- the most recently made link first -- or you can sort it be authority. The most authoritative weblog is the one with the most links to it.

How does it decide whether something is a weblog? My first guess would by RSS feed, but the Daily Python URL doesn't have one. It can't be registration, because there are 1,179,981 weblogs watched.

David Sifry, who created Technorati, explains how it works. Among the more interesting notes: It counts links from the weblog home page, not from the archive. So it captures thinks of recent interest. It spiders weblogs, but you can also submit updates. It doesn't appear to have my weblog, which may not be popular enough to be found automatically.

What's the difference between Technorati and Google? There are obvious differences, but the basic idea is similar.

David's weblog entry on the subject is filled with spam posts. I don't see the point of comments posted directly to a weblog. People can post their own comments on their own weblogs -- and software like Technorati can find them.

There is a Technorati API for programs to interact with Technorati. I haven't found an API definition yet. If you sign up as a member, you get a long Technorati key that is good for 500 queries a day.

After I registered as a Technorati user, I tried to "claim" my weblog. I got an email asking me to put a link to Technorati on my weblog home page or my blogroll. (There's a lot of jargon to learn; a blogroll is the customary set of links to other weblogs that you read.) I had to update my weblog software to add the link. I really need to add a template system, but that's a pretty big project.