I didn’t get the “Strongly Typed DataSet-Data Adapter Factory” thing at first. Chris Hollander showed me the light, and now I’m a true believer.
One of the few things I haven’t liked about Strongly Typed DataSets is having objects named things like “UserRow” instead of just “User”. Well, I stumbled onto a cool article on MSDN about annotating strongly typed datasets. Basically, you can use custom attributes in your schema to control the names of the objects created by the XSD tool.
<xs:elementname=”Users” codegen:typedName=”User” codegen:typedPlural=”Users”>
and “UserRow” becomes “User”. It’s not a big deal, but it goes a little way towards plugging what I consider a leak in the abstraction.