SqlXml and Yet Another Blog Engine
In Don Box's post Yet Another Blog Engine (Don, if you read this, the ?key=XXXXT takes me to a blank page every time :) ) he asks about the "a-ha" moment when he may get SQLXML or see it as a preferred technique over others.
Here's one scenario for a blog engine. You develop a backend database and the basically acts as a backend to your RSS data (indexing and whatever...). You deploy a client that updates that database based on the Xml from the RSS file(s). You then want to extend, say with a new namespace set of elements (say DC). Easy, add the columns to a new table in the database and ??? There are no doubt a million number of ways this could be done, but what could possibly be easier than annotating the centralized Xml Schema for the RSS documents you have to map the new nodes to columns in the database (and vice versa when getting them back)?? Sure you could update your code that uses the IDataReader, but then there is re-compilation and re-deployment. In SqlXml you have no real overhead - and update to the database (which happens in either scenario) and an update to the schema, which may or may not be centralized, but certainly doesn't require a build/install.
I have done this (not for blogs) and after we got the Schema sorted it was easy. Imagine you wanted to import your Blog Archive to some other database (analysis/reporting) - BulkUpload and SqlXml make this a very easy thing to do. Anyway I can interact with Xml I love it, and SqlXml is not perfect, but there are some situations where it is hard to beat!
XML Query will make it even better in the future and i'm looking forward to it.
vtgo.net