Client datasets are specialized datasets that hold all their data in memory. The support for manipulating the data they store in memory is provided by midaslib.dcu or midas.dll. The format client datasets use for storing data is self-contained and easily transported, which allows client datasets to

• Read from and write to dedicated files on disk, acting as a file-based dataset. Properties and methods supporting this mechanism are described in Using a client dataset with file-based data.
• Cache updates for data from a database server. Client dataset features that support cached updates are described in Using a client dataset to cache updates.
• Represent the data in the client portion of a multi-tiered application. To function in this way, the client dataset must work with an external provider, as described in Using a client dataset with a provider. For information about multi-tiered database applications, see Creating multi-tiered applications.
• Represent the data from a source other than a dataset. Because a client dataset can use the data from an external provider, specialized providers can adapt a variety of information sources to work with client datasets. For example, you can use an XML provider to enable a client dataset to represent the information in an XML document.

Whether you use client datasets for file-based data, caching updates, data from an external provider (such as working with an XML document or in a multi-tiered application), or a combination of these approaches such as a "briefcase model" application, you can take advantage of broad range of features client datasets support for working with data.