the press release states that the CDB is a single-copy data repository, but can it cope with multiple copies that may be required for different security classifications? For example, UNCLASSIFIED and SECRET versions of the same geographic area depending upon the customer requirement.

You may create and maintain as many CDBs as you like, each containing different levels of classified data. In fact it is quite straightforward to 'de-classify' a CDB that contains sensitive data.

The single-copy data repository description in the press release refers to the fact that the CDB you create can be used as a 'single-copy' database for a wide variety of client systems, such as a visual system, SAF/CGF or sensor. If you have created multiple CDBs, your CDB-enabled runtime systems should work with each of your CDBs if you have followed the specification conventions.

The CDB specification contains a versioning scheme that can be used to manage classified and non-classified data. Section 3.2 of the specification provides an outline of the process. Here is an extract:

"[the CDB versioning scheme] can also be applied to the handling of classified secure data. In this case, an incremental version can be used to hold the portion of the CDB that contains the classified information. The incremental versioning mechanism would be used to segregate the classified portion of the CDB onto a separate storage medium."