Client datasets use the CloneCursor method to let you work with a second view of the data at runtime. CloneCursor lets a second client dataset share the original client dataset's data. This is less expensive than copying all the original data, but, because the data is shared, the second client dataset can't modify the data without affecting the original client dataset. 

CloneCursor takes three parameters: Source specifies the client dataset to clone. The last two parameters (Reset and KeepSettings) indicate whether to copy information other than the data. This information includes any filters, the current index, links to a master table (when the source dataset is a detail set), the ReadOnly property, and any links to a connection component or provider. 

When Reset and KeepSettings are False, a cloned client dataset is opened, and the settings of the source client dataset are used to set the properties of the destination. When Reset is True, the destination dataset's properties are given the default values (no index or filters, no master table, ReadOnly is False, and no connection component or provider is specified). When KeepSettings is True, the destination dataset's properties are not changed.