In Brief

• The .NET Compact Framework DataSet supports a variety of methods to load and save the contents of a DataSet in XML format.

• The .NET Compact Framework DataSet supports the standard XML format extensively, but only supports DiffGrams well enough to enable passing DataSet in Web Services calls.

• XML data can be loaded into a DataSet in a variety of ways, such as from a file or a stream. There are multiple methods in the DataSet for loading XML data depending on the source of the data.

• A unique way to transmit a DataSet in XML format between two devices is by using the devices' infrared port. The IrDAClient class makes this possible.

• If XML describing a DataSet lacks schema information, the DataSet can infer the structure of the relational data, a process called inferring schema .

• The contents of a DataSet can be saved as XML in a variety of ways, targeting such outputs as files or streams. There are multiple methods in the DataSet for saving XML data depending on the destination for the data and the format of the XML written.

• When writing XML data from a DataSet , it is possible to write schema information with the data, to write only schema information but no data, or to write data only with no schema.

• XML Schema is used to describe the relational layout of the data in a DataSet . An XML Schema can be used to describe a variety of relational data constructs such as table relations, autoincrementing fields, and parent-child relationships.

• Many of the data types possible for a DataColumn can be also be described in an XML Schema.

• A DataSet can load the schema information from an XML document, ignoring the data values. Doing so causes the DataSet to set up its internal structure in such a way as to match the relational format specified in the schema.

• The Visual Studio XML editor is an easy way to manipulate and alter existing XML data that a DataSet can read.

• Although the DataSet can read and write DiffGrams, there are many restrictions on how it can do so. Specifically, the DataSet can only write all of its data at once to a DiffGram; it cannot write out only records which have been updated. Also, when a DataSet reads a DiffGram, the DataSet must initially be empty.