Data Access and Transactions

Client-Side Data Access

The components that make up MDAC are designed for distributed applications that take advantage of the processing power on client, middle-tier, and database server computers. These components are part of a simple yet rich programming model for manipulating data and building applications that are easy to configure and maintain.

ADO and RDS are suitable for applications that need a high degree of database accessibility. When you create an RDS application for an intranet, you normally would use IIS 5.0 and Internet Explorer to communicate between the client and business services tiers. However, if the client- and middle-tier components are on computers within a local area network (LAN), you can also use COM to marshal the interfaces and method arguments across the network. RDS can also be used by applications written in Visual Basic to manage remote information.

ADO, with RDS, enables the Web-application developer to:

Bind ADO Recordset objects to intrinsic Dynamic HTML (DHTML) controls (and other data-aware controls) hosted in the browser, by using the DHTML “databinding” model.
Create and manage remote and disconnected ADO Recordset objects. ADO normally maintains a persistent connection to the database, but RDS works with locally cached, or disconnected, data.
Asynchronously request information from the server, and respond to events that are triggered when operations are complete. Asynchronous fetching is a feature specific to Microsoft® Cursor Service for OLE DB. It returns the first rows from a query and then continues fetching in the background, while the user manipulates the rows that have already been sent.
Invoke Automation objects on the Web server over either HTTP or COM. You can use RDS for applications that browse records, or that connect to and modify data on the middle tier.
Work with hierarchical and multidimensional recordsets. With the newly available Microsoft® Data Shaping Service for OLE DB (Microsoft® Datashape), RDS can finally overcome its previous limitation of only one recordset per query.

The RDS client-side and server-side components are described in detail in the following sections: