Blogs, Semantics, Feeds and Databases. I've been following some of the stuff being written about structured blogging and semantics. Some of it makes a lot of sense to me, and some of it doesn't. (Here are some recent references: Structured Blogging and the filling-out-the-forms issue, Building the Recipe Web, and follow other links there.)
- Allowing Semantics makes sense. Sure, there are other kinds of posts that might be quite useful, like storm reports or traffic updates, which are timely and have a relatively short shelf-life. And if there was some kind of agreement (hard) about a schema we could do some really cool stuff. I mean I think we are still wrestling about the schema of the proto-posting, the news item.
- Feeds are not Databases. Where I get lost is when folks use examples like Recipes or Movie Database information. Why? Because it seems to me that the notion of a 'feed' that is refreshed regularly is pretty core. Architecturally and habitually we are oriented to a feed which reflects contents recently added. The notion of using these new found semantics to perform searches seems at odds with fact. Either a search would have to be constantly building a copy of the archives of the feeds it wanted to query, or each feed server would have to be include a query API to support federated searches. This latter idea isn't bad but I haven't seen it discussed.