RDB DSC 2 schema. The schema has been updated to support the Publish Data feature via the Categories table and CategoriesLookup table.

The Relational MR Database CDSC now supports unexpanded loops for reading and writing data. The UNICOM Intelligence Data Model 7.0.1 hierarchical data features are only available for the RDB DSC 2 schema and the hierarchical view of the data is available only when an MDM document is attached. For more information, see Upgrading older databases to the RDB DSC 2 schema.

Because the RDB DSC2 does not support unexpanded level variables under expanded level variables, another version of the household.mdd file and associated, backup relational database have been added to the UNICOM Intelligence Developer Documentation Library. These new files are installed to: [INSTALL_FOLDER]\IBM\SPSS\DataCollection\7\DDL\Data\RDB

The loop variable person in these new files have been updated from mlExpand to mlLevel in MDM2 Explorer.

The Relational MR Database CDSC now supports some types of WHERE expression natively, which can improve performance when filtering data for tabulation. For more information, see WHERE expressions supported natively.

In the past, the Relational MR Database CDSC could represent hierarchical data only in a flattened form in the VDATA virtual table. Now RDB DSC 2 can represent hierarchical data in HDATA hierarchical virtual tables. However, unbounded loops (which are loops for which the maximum number of iterations is unknown) are currently unsupported and the hierarchical view of the data is available only when an MDM document is attached. For more information, see Understanding hierarchical data and Loops, grids, and levels.

UNICOM Intelligence Data Model 2.7 comes with an OLE DB implementation of this CDSC in addition to the older ODBC-based implementation. The new OLE DB implementation is called RDB DSC 2 (the internal name is mrRdbDsc2). The existing ODBC implementation of this CDSC is essentially unchanged from UNICOM Intelligence Data Model 2.6.

RDB DSC 2 reads and writes case data through OLE DB and has the following main advantages over the older ODBC implementation:

RDB DSC 2 can read and write data in existing RDB databases using the UNICOM Intelligence Data Model 2.6 schema. However, when writing data to a new database, RDB DSC 2 automatically uses a new, improved schema. For more information, see Relational MR Database CDSC schema. Note that the new schema is inaccessible to the older ODBC-based RDB DSC, which means it is inaccessible to UNICOM Intelligence Data Model 2.6 and earlier. You can optionally ugrade existing databases to the new schema. For more information, see Upgrading older databases to the RDB DSC 2 schema.

When writing data, RDB DSC 2 creates a second SQL Server connection, because this offers a significant performance improvement compared with using a single connection.

Connecting to RDB DSC 2. When you connect to RDB DSC 2, you need to specify the Location connection property as an OLE DB connection string. You can also set the MR Init Custom connection property to favor either memory or speed. If you do not set this property, the favor speed mode will be used. For more information, see Connecting to a relational MR database using RDB DSC 2.

Case-insensitive databases. When using RDB DSC2, the LIKE keyword always searches text case-sensitively, regardless of the collation option chosen when SQL Server was installed.

Transferring data from older databases without an .mdd file. Changes in the Metadata Model (MDM) in UNICOM Intelligence Data Model 2.7 mean that it is not possible to transfer data from a relational MR database that was created in UNICOM Intelligence Data Model 2.6 or earlier without using an .mdd file. For more information, see Known problems in Relational MR Database CDSC.

Schema change. The schema has increased the number of tables from five to six. The new table is called OtherID and holds the maximum value in the database of OtherID, the unique identifier used in the OtherData table. This improves performance when writing records. For more information, see OtherID table.

New stored procedures. There are a number of new stored procedures. For more information, see Stored procedures.

Increased size limit for text responses. In UNICOM Intelligence Data Model 2.3 and earlier the limit for text responses was 2000 bytes (999 characters). This limit has now been raised to 8000 bytes (4000 characters).

Schema changes. The schema has increased the number of tables from three to five. The new tables are:

Case-sensitive databases. Relational MR Database CDSC supports SQL Server that has been installed with the case-sensitive collation option. However, note that if you are using this option, you must enter column names with the correct case in queries and the LIKE keyword will search case-sensitively. However, the UNICOM Intelligence Find.

Case-insensitive databases. Generally the LIKE keyword searches text case-sensitively. However, when SQL Server has been installed with the case-insensitive collation option, the LIKE keyword generally searches case-insensitively. For example, the following query is evaluated case-insensitively:

However, when the LIKE keyword is included in a complex expression (for example, one that includes a function from the UNICOM Intelligence Function Library) it is evaluated case-sensitively:

This behavior has not changed in UNICOM Intelligence Data Model 2.4, but was previously undocumented. This anomaly does not exist in RDB DSC 2.

For further information on expression evaluation, supported SQL queries, and the UNICOM Intelligence Function Library, see the UNICOM Intelligence Developer Documentation Library.

Connecting using an ODBC connection string. When connecting to a database using Relational MR Database CDSC in UNICOM Intelligence Data Model 2.4 and later, you can specify the Location connection property using an ODBC connection string instead of a data source name (DSN). This makes it easier to export data from a relational MR database using SQL Server Data Transformation Services (DTS), because you generally no longer need to set up a DSN for the database first because the ODBC connection string is automatically written to the .mdd file when the database is synchronized. For more information, see Connecting to a relational MR database.

Supported features. The documentation has been updated to list details of which CDSC features are supported. For more information, see Relational MR Database CDSC: Supported CDSC features.