Professional ADO.NET 3.5 with LINQ and the Entity Framework

Chapter 9. Raising the Level of Data Abstraction with the Entity Data Model

Microsoft's ADO.NET Entity Framework (EF) represents a seismic shift in the SQL Server Data Programmability group's approach to querying and updating relational data. Since the initial release of Visual Studio .NET in 2002, the DataSet has been central to Microsoft's data management toolset strategy. The DataSet represents an in-memory cache of relational DataTable(s) that programs can populate from and persist to XML documents in a file system or tables in a relational database management system (RDBMS).

DataSets extended the "disconnected client" architecture introduced by COM-based ADO Recordsets to an extensive ecosystem of graphic designers for generating C# or VB classes to define strongly typed DataSets, DataTables and DataViews, databinding components and data-bound controls. Wizards orchestrated combinations of these elements to autogenerate simple Windows form projects that enabled displaying and editing relational data in bound text boxes and data grids. The ability to create a working — but trivial — database application with 20,000 or more lines of code in a minute or two was a big hit among most attendees at Microsoft conferences but didn't impress the more jaded developers of data-intensive projects at the enterprise level.

While the Data Programmability group was investing heavily in developer tools for relational data, the Developer Division's C#, VB, and VS teams were emphasizing object-oriented ...