Following my previous article on separating application layers and using an In-memory persistence to "fake" the real data access layer, I've published a new article on implementing the actual data access using the previously outlined architecture and ado.net.
The article is: Building an Agile SQL Data Access Layer.
