New script for making copies of the sample database    The newdemo.bat and newdemo.sh files are located in the bin32 or bin64 directory of your SQL Anywhere installation directory and can be used to create a copy of the sample database that includes all the data from the sample database. This script can be used to re-create the sample database or to make new copies of it with a different name. See Recreate the sample database (demo.db).

New selectivity estimate source type for the query optimizer   A new selectivity estimate source type, JOIN, has been added. This new source type is used by the query optimizer for selectivity estimations of atomic predicates of the form T.X = R.X. See ESTIMATE_SOURCE function [Miscellaneous].

Better handling of overflow errors   Arithmetic operations ( +, -, *, /, SUM, AVG ) may overflow because the result of the operation can not be represented in the data type. Previously, for expressions of type INT, this overflow returned an error, while for all other data types, the overflow resulted in an undefined value. Now, all arithmetic operations on all types detect overflow and return an error if the result can not be expressed in the data type.

Use the ALTER EXTERNAL ENVIRONMENT statement to set the location of dbmlsync   If, during synchronization, the dbmlsync executable cannot be located using the PATH environment variable that the database server is using, you can now use the ALTER EXTERNAL ENVIRONMENT statement to tell the database server the location of the dbmlsync executable. See ALTER EXTERNAL ENVIRONMENT statement.

Back quotes are supported as delimiters   You can now use back quotes (`) to delimit identifiers in SQL Anywhere. See Identifiers.

Japanese Unicode Collation Algorithm (UCA) collation tailoring option   A new Japanese UCA collation tailoring option is now available. You can use it to define a primary-level difference between all Hiragana and Katakana letters. This new tailoring option provides correct equality comparisons of Hiragana and Katakana letters in case-insensitive collations. See Collation tailoring options.

Changes to the server messages window and Windows system tray icon   The title bar for the database server messages window now specifies whether you are running a personal server or a network server. The tooltip for the Windows system tray icon also specifies the type of database server. As well, the database server About window includes the edition of SQL Anywhere that is running.

Information utility (dbinfo) enhancement   The dbinfo utility now returns information about the CHAR collation specification, the CHAR encoding, the NCHAR collation specification, or the NCHAR encoding for a database. See Information utility (dbinfo).

Controlling the amount of address space reserved for non-cache use   The -ch option now leaves more address space for use outside the cache, and the maximum non-AWE cache size on 32-bit operating systems has been reduced. See -ch dbeng12/dbsrv12 server option and -chx dbeng12/dbsrv12 server option.

Enhance priority management for I/O bound compared to CPU bound requests   The database server now dynamically detects requests that are I/O bound and increases their priorities over CPU-bound tasks to increase disk throughput and the use of hardware resources.

Improved robustness across power failures on Windows   You can set Windows Registry entries to improve robustness across power failures on systems using certain Intel storage drivers when deploying SQL Anywhere. Failure to set this parameter can result in lost data and corrupted databases in the event of a power failure. To determine if these entries are required as part of your deployment, see Windows Registry entries.

SQL Flagger enhancement   The SQL Flagger now supports the SQL/2008 standard. See Testing SQL compliance using the SQL Flagger.

Progress messages   Some SQL statements now support sending progress messages are sent from the database server to the client. See .