Replication Server version 15.0 adds support for the unitext datatype. unitext is a variable-width, nullable Unicode datatype. Although independent of the text datatype, unitext mirrors its behavior, and can be used wherever text is used.
Like the data in the other Unicode datatypes—unichar and univarchar—unitext data is encoded in UTF-16, which is essentially a 2-byte, fixed-width encoding of Unicode. unitext can hold as many as 1,073,741,823 Unicode characters, or the equivalent of 2,147,483,647 bytes. unitext has no connection to the default character set ID or the default sort order.
The main advantage of Unicode datatypes is efficiency. The UTF-16 character types are approximately 33% more space efficient than UTF-8 for Asian characters.
To fully support unitext, the primary and replicate Replication Server must have a site version of 15.0, the route version must be 15.0, and the LTL version must be 700. If the LTL version is less than 700 at connect-source time, RepAgent converts unitext columns to image.
rs_subcmp supports the new unitext datatype.
