October 2024 - version 2.7.4
-- version 2.7.4 jar requires JRE 11 or later - can be used as a java module  - tested up to Java 21
-- version 2.7.4 alternative jar requires JRE 8 or later
-- added support for SQL:2023 LISTAGG aggregate function
-- added support for update and delete visiblility in JDBC scrollable and updatable result sets
-- fixed a number of reported regressions and bugs (see tracker)

June 2024 - version 2.7.3 
-- version 2.7.3 jar requires JRE 11 or later - tested up to Java 21
-- version 2.7.3 alternative jar requires JRE 8 or later
-- SQL:2023 updates 
-- source code refactoring to use Java Generics - no longer compiles with pre 1.8 JDK
-- added support for CAST with datetime template argument
-- enhancements to TO_CHAR, TO_DATE, TO_TIMESTAMP functions; now using java.time.format.DateTimeFormatter
-- added TO_CHAR for numeric values
-- added JDBC 4.3 support
-- added alternative Java Trigger method to the interface, maintaining backward compatibility
-- fixed a number of reported regressions and bugs (see tracker)

March 2023 - version 2.7.2
-- version 2.7.2 is a release with minor enhancements and corrections

20 Oct 2022 - version 2.7.1
-- version 2.7.1 jar requires JRE 11 or later - tested up to Java 17
-- version 2.7.1 alternative jar requires JRE 8 or later
-- CVE-2022-41853 disallow calling java methods - setting the hsqldb.method_class_names is required for accessing java methods
-- enhancements to RECURSIVE CTE's

30 May 2022 - version 2.7.0
-- version 2.7.0 jar requires JRE 11 or later - tested up to Java 17
-- version 2.7.0 alternative jar requires JRE 8 or later
-- added support for JSON constructor functions
-- added support for microsecond precision CURRENT_TIMESTAMP and similar functions
-- added support for GRANT and REVOKE on all objects in a schema
-- added support for CSV file data import and export
-- added support for regional time zones such as 'America/New_York' in conversions
-- added SET DATABASE SQL MAX RECURSIVE (sql.max_recursive) setting
-- changed the default for SET DATABASE SQL SYS INDEX NAMES (sql.sys_index_names) to true
-- changed the properties reported by JDBC DatabaseMetaDat::getClientInfoProperties() method
-- improved  JDBC DatabaseMetaDat::getTablePrivileges() method to cover fine-grained privileges
-- improved error messages for SQL settings statements
-- improved documentation
-- added some minor enhancements
-- fixed a number of reported regressions and bugs (see tracker)

21 October 2021 - version 2.6.1
-- version 2.6.1 is a bug fix release
-- added sql.lowercase_ident property for extra compatibility with PostgreSQL and MySQL unquoted identifier names
-- fixed an issue with performance of large updates in mvcc transaction mode
-- fixed an issue with query timeout settings
-- fixed a number of minor issues and regressions

21 March 2021 - version 2.6.0
-- Version 2.6.0 marks the 20th anniversary of HyperSQL.
-- The project maintainer would like to thank core developers Blaine Simpson, Campbell Burnet and Robert Preston for their contribution over many years, 
--    and all developers and users who contributed code, made useful suggestions, or reported issues.
-- version 2.6.0 main jar contains the module org.hsqldb and can be used as a Java module
-- version 2.6.0 jar requires JRE 11 or later - tested up to Java 16 RC
-- version 2.6.0 alternative jar requires JRE 8 or later
-- version 2.3.9 jar can be compiled with Java 6 (or 7) for JRE 6 and 7 and fully compatible with 2.6.0 except for the new JDBC features
-- The DatabaseManager class is no longer included in the Jars. Use DatabaseManagerSwing instead. 
-- added support for additional REGEXP_ functions.
-- added support for adding extra columns to FOREIGN KEY indexes; see ALTER CONSTRAINT in the Guide
-- added support for RECURSIVE_TABLE in WITH RECURSIVE queries
-- added system role SCRIPT_OPS for export / import of database structure and data
-- added extensive features to the collection classes in org.hsqldb.lib (maps, lists, queues etc.) which can also be used in user apps
-- improved MySQL syntax support for ENUM
-- fixed a number of regression in 2.5.1, e.g. use of MODULE for LOCAL TEMPORARY tables, truncated microsecond precision of java.sql.Timestamp values
-- fixed several minor issues
-- removed all use of Java (up to JDK 16) deprecated methods
-- revamped Ant and Gradle builds, now updated to latest Gradle

24 June 2020 - version 2.5.1
-- version 2.5.1 jar requires JRE 8 or later - tested up to Java 14
-- version 2.5.1 can be compiled with Java 9 or later as org.hsqldb Java module jar
-- version 2.3.8 jar can be compiled with Java 6 (or 7) for JRE 6 and 7 and fully compatible with 2.5.1 except for the new JDBC features
-- added support for GROUPING SETS, ROLLUP and CUBE grouping operations
-- added support for PERFORM CHECK ALL TABLE INDEX to check all CACHED table data and indexes
-- added support for use of PostgreSQL ODBC driver to access a HyperSQL Server
-- added support for calling stored procedures with INTERVAL parameters via JDBC CallableStatement
-- added support for TRANSACTION_UTC, ASCIISTR UNISTR functions
-- added support for reuse of memory space for lobs in mem: databases
-- changed behaviour of WITH RECURSIVE with UNION ALL for compatibility with other systems
-- improved role-based, row-level data access control with more complex filter expressions that can be applied to separate actions
-- improved MySQL syntax support in the ON DUPLICATE KEY UPDATE clause
-- improved loading of text files with null or empty fields into TEXT tables with columns defined as NOT NULL with DEFAULT 
-- many minor enhancements
-- fixed issue with ALTER COLUMN data type modification between SMALLINT and INTEGER in CACHED tables
-- fixed an issue (regression) in some server shutdown cases
-- fixed an issue (regression) in some cases when JDBC connection timeout is used
-- fixed several minor issues
-- changed and improved how file spaces are set up for existing cached tables 
-- removed legacy support for opening version 1.8 databases
-- removed legacy support for non-incremental internal backup of the .data file

- 1 June 2019 - version 2.5.0
-- version 2.5.0 jar requires JRE 8 or later - tested up to Java 12
-- version 2.5.0 can be compiled with Java 9 or later as org.hsqldb Java module jar
-- version 2.3.7 is released as a separate jar for JRE 6 and 7 and fully compatible except for the new JDBC features
-- added Temporal System-Versioned tables feature 
-- added offline distributed database feature based on System-Versioned tables
-- added role-based, row-level data access control
-- added new properties, memory_lobs and hsqldb_tx_interrupt_rollback
-- changed effect of Thread interrupts by user application in in-process database, transaction is rolled back if hsqldb_tx_interrupt_rollback is true
-- changed return type of the UUID() function to UUID; the SYS_GUID() function will continue to return a BINARY(16) value.
-- improved reporting of line number in compile error messages for routines
-- improved syntax compatibility modes
-- fixed issue with formatting of some TIMESTAMP nanoseconds returned by the TO_TIMESTAMP function
-- fixed issue with TIMESTAMPDIFF with TIMESTAMP WITH TIME ZONE
-- fixed issue with IN (UNNEST(NUMERIC ARRAY))
-- fixed issue with transaction size reported in INFORMATION_SCHEMA.SYSTEM_SESSIONS 
-- fixed issue with storing plus or minus infinity DOUBLE values
-- fixed issue with CREATE DOMAIN ... COLLATE not persisting
-- fixed issue with CREATE TYPE with custom COLLATION
-- fixed an issue with multiple FOR loops in SQL routines
-- fixed several minor issues
-- changed calls to Java methods deprecated in JDK 9 - 12
-- removed Java 5 code switch and build options

20 May 2018 - version 2.4.1
-- version 2.4.1 jar requires JRE 8 or later
-- version 2.3.6 is released as a separate jar for JRE 1.6 and fully compatible except for the new JDBC features
-- added support for EXPLAIN REFERENCES FROM | TO object statement
-- added support for HEX, UNHEX, TO_BASE64 and FROM_BASE64 functions
-- added support for deployment in docker containers that do not support file rename
-- added support for generated keys in update and merge
-- improved CHECKPOINT DEFRAG to allow very large data files
-- fixed issue with some temp table queries
-- fixed issue with some SELECT ... INTO in routines
-- fixed issue with parsing some TEXT TABLE source files
-- fixed issue with auto-generated identity values in MERGE statements
-- fixed issue for WITH named non-recursive subqueries in routines
-- fixed issue with some WITH recursive queries
-- fixed issue with duplicate values in UNNEST used with IN predicates
-- fixed issue with WHEN clause in Java TRIGGER declarations
-- fixed several issues with implementation of JAVA 8 JDBC method
-- fixed some issues with transaction management
-- fixed several minor bugs

9 April 2017 - version 2.4.0
-- version 2.4.0 jar requires JRE 8
-- version 2.3.5 is released as a separate jar for JRE 1.6 and fully compatible except for the new JDBC features
-- added support for new JAVA 8 JDBC methods to ResultSet and Statement classes
-- added support for several functions SINH COSH TANH NEXT_DAY DATEPART DATENAME EOMONTH NEWID SYS_GUID
-- added support for execution of DDL statements concurrently with other transactions
-- added support for table spaces for CACHED tables, which helps file space reuse after dropping tables or mass deletes
-- improved support for WITH named subqueries in routines, allowing parameters to be referenced
-- improved support for WITH RECURSIVE named subqueries to be used in other subqueries
-- improved support for UUID in ResultSet to return a Java UUID with getObject()
-- improved support for UUID in PreparedStatement setObject() with VARCHAR parameters
-- improved use of allow_empty_batch connection property
-- improved field param setting of DATEADD DATEDIFF TIMESTAMPADD TIMESTAMPDIFF functions
-- improved ORA, MSS and MYS syntax compatibility modes for type names and functions
-- fixed issue with cancel() method of java.sql.Statement with in-process connections
-- fixed issue with the timeout setting of java.sql.Statement
-- fixed issue with duplicate row with INSERT followed by UPDATE and ROLLBACK in MEMORY tables
-- fixed issue with dropping indexes and constraints in MEMORY tables
-- fixed issue with queries on DECIMAL or NUMERIC columns containing very large values
-- fixed issue with ResultSet when calling next() multiple times was not followed by a get(...) call
-- fixed issue with temp table indexes
-- core code refactor of iterator for rows of tables and results
-- fixed several minor bugs

16 May 2016 - version 2.3.4
-- added support for storing live Java Objects in mem: databases with sql.live_object connection property
-- added support for UUID data type
-- added support for CREATE SYSNONYM FOR, for tables, views and sequences
-- added support for ON UPDATE CURRENT_TIMESTAMP for auto updated TIMESTAMP columns
-- added REGEXP_REPLACE function
-- added support for PERIOD and CONTAINS, EQUALS, PRECEDES and SUCCEEDS predicates in SQL
-- added support for UTF-16 text source files for TEXT tables
-- added support for the cancel() method of JDBC Statement and PreparedStatement
-- added support for END STATEMENT to end long-running statements
-- added support for CallableStatement parameter reference by name
-- added date range checks to setDate and setTimestamp methods in PreparedStatement
-- added support for lowercase patterns for TO_DATE etc.
-- added support for VARCHAR string literals with sql.char_literal property
-- improved MYS syntax mode for INSERT, REPLACE and ON DUPLICATE KEY UPDATE
-- improved DB2 syntax mode for queries
-- extensive review and improvement of transaction management in muti-threaded access
-- fixed issue with user application Thread.interrupt disabling in-process database session
-- fixed regression with interaction of FK constraints with ON DELETE CASCADE and NO ACTION
-- fixed regression with opening database with CACHED table with files_readonly 
-- fixed issue with SEQUENCE increment in multi-row inserts
-- fixed issue with incorrect results in multiple LEFT JOIN with NATURAL or USING
-- fixed issue with ClobInputStream 
-- fixed issue with missing references to views used in routines
-- fixed issue with .lobs file space reuse at CHECKPOINT
-- fixed issue with readonly session blocking in MVLOCKS mode
-- fixed several minor bugs

28 June 2015 - version 2.3.3
-- extensive review and improvement of concurrent updates
-- fixed an issue with CHECKPOINT DEFRAG that could cause minor inconsistencies
-- fixed a rarely occuring error immediately after CHECKPOINT
-- fixed regression with comparison of TIMESTAMP values when one value is WITH TIME ZONE and the other WITHOUT
-- fixed reporting of DISTINCT and DOMAIN types in JDBC DatabaseMetaData getColumns results
-- fixed issue with new FK constraints, existing rows are now checked
-- fixed restrictions with FK constraints with NO ACTION, all valid outcomes are now allowed
-- fixed regression in 2.3.2, multiple rows with partial null values could not be inserted in UNIQUE indexes
-- fixed issue with removal of IDENTITY when using ALTER COLUMN <name> <type>
-- fixed issue with getExportedKeys reporting wrong DEFERABILITY
-- fixed issue with CURRENT VALUE FOR usage
-- fixed issue with LIKE with parametric escape character
-- fixed an issue with CACHED table persistence
-- fixed issue with TEMP table with ON COMMIT PRESERVE ROWS and RESULT MEMORY ROWS setting
-- fixed several minor bugs
-- improved usage of DECLARE CURSOR and DECLARE TABLE in SQL routines
-- improved ORA, PGS, MSS and MYS syntax compatibility modes for CREATE TABLE and other statements
-- added support for MYS syntax INSERT IGNORE, REPLACE and ON DUPLICATE KEY UPDATE
-- added support for ALTER CONSTRAINT <name> RENAME TO <new name>
-- added support for LIMIT in DELETE and UPDATE statements
-- added support for DELETE in MERGE statement

12 February 2014 - version 2.3.2
-- fixed several minor issues
-- fixed bug with incomplete rollback of MVCC transaction that inserts and updates the same row
-- fixed issue with parsing of long BIT strings in SQL
-- fixed issue with SQL triggers when columns were added or dropped from a table with SQL triggers
-- fixed issue with an index added to a TEMP table with ON COMMIT PRESERVE ROWS and containing data
-- added URL property hsqldb.digest which can be used with alternative secure hash algorithms for passwords
-- changed the class used for offline backup and restore to org.hsqldb.lib.tar.DbBackupMain
-- extended the scope of SET DATABASE SQL REFERENCES TRUE to catch ambiguity with identical table aliases
-- extended support for the DEFAULT keyword used in INSERT and UPDATE to apply to columns with no default
-- improved support for recursive queries
-- improved ORA and MYS syntax compatibility modes

08 October 2013 - version 2.3.1
-- fixed several minor issues
-- fixed bug with some queries using ORDER BY DESC on indexed columns used in conditions
-- fixed issue with NPE is some complex subqueries
-- improved JDBCPool to be created with DataSourceFactory
-- added support for LIKE with parametric escape character 
-- added support for routine SIGNAL messages using variables

08 July 2013 - version 2.3.0
-- improved data exception error messages
05 July svn 5259 snapshot 52
-- added connection property, close_result
30 June - svn 5256 snapshot 51
-- added INSTR function
15 May - svn 5241 snapshot
-- added support for backup of database as set of files BACKUP DATABASE TO <path> AS FILES
12 April - svn 5228 snapshot
-- added TIMESTAMP_WITH_ZONE function
-- fixed issue with AT <time zone> use with TIMESTAMP WITHOUT TIME ZONE
-- fixed issue with TRIGGER with non-zero QUEUE setting
4 April - svn 5222
-- improved data space reuse
-- improved optimisation of SELECT DISTINCT and GROUP BY
-- improved MYS syntax compatibility mode to support backtick quoted identifiers
25 March 2013 - svn 5214
-- added support for statement timeout via java.sql.Statement and PreparedStatement
18 March 2013 - svn 5208
-- improved support for setting collation strength and case-sensitivity
-- VARCHAR_IGNORECASE is no longer a type and is translated into the collation name
10 March 2013 - svn 5203
-- increased default max size of BLOB and CLOB to BLOB(1G) and CLOB(1G)
-- added support for compressed and encrypted lobs
-- improved DatabaseMetaData reporting of actual settings for null ordering
-- improved DB2 and ORA syntax compatibility modes
06 February 2013 - svn 5168 snapshot
-- improved backward compatibility for use with embedded OpenOffice / LibreOffice databases
-- fixed regression with some subquery tables used in OUTER joins
-- fixed regression with some in MERGE statements
04 January 2013 - svn 5139 - snapshot
-- improved reuse of empty lob spaces
-- fixed various minor issues 
09 December 2012 - svn 5127
-- added support for CONCAT_WS and UNIX_MILLIS functions
-- added support for table spaces
-- added INFORMATION_SCHEMA.SYSTEM_TABLESTATS
-- added SET DATABASE SQL NULLS ORDER and the property sql.nulls_order for the default sort order of nulls
-- improved name resolution in queries within SQL routines
-- fixed various minor issues
01 November 2012 - svn 5103
-- fixed issue with some correlated subqueries in where clause
-- fixed issue with ROWNUM in inner selects
-- fixed various minor issues
-- improved support for DATE and TIMESTAMP subtraction without explicit INTERVAL type
-- added support for ADD_MONTHS, LAST_DAY, MONTHS_BETWEEN and other Oracle style functions
24 October 2012 - svn 5096 snapshot
-- fixed various minor issues
-- fixed issues with some disk-based TEMP tables used in multiversion transaction modes
12 October 2012 - svn 5089 snapshot
-- fixed issue with access to LOBs in res: databases
-- fixed regression with some nested views
-- added global IGNORECASE property
10 September 2012 - svn 5071 snapshot
-- fixed regession with the result returned after a SHUTDOWN 
-- fixed regression with non-correlated scalar subqueries causing NPE
-- fixed regression with FUNCTION and PROCEDURE local tables data change logging
-- fixed regression with script file with CREATE TYPE with character types
-- fixed issue with DECLARE CURSOR in routines using write locks and requiring MODIFIES SQL DATA
-- fixed issue with TRUNCATE function causing NPE when used in column GENERATED ALWAYS expression
-- fixed issue with escape in Unicode string which caused only the part before the escape to return from Scanner
-- fixed issue with upgraded version 1.8.0 databases which did not have lob space
-- fixed issue with logging changed password after ALTER USER ... SET PASSWORD ...
-- improved text table source textdb.allow_full_path=true to allow absolute full paths
-- improved sql.enforce_references=true to catch ambiguous references in ORDER BY columns
-- improved ORA syntax compatibility mode
-- added support for terminal DROP TABLE ... IF EXISTS for local temporary tables

05 August 2012 - version 2.2.9
-- code review and minor fixes
-- improved support for use in OO and LO version 3.2 and later
-- new simple connection pool class org.hsqldb.jdbc.JDBCPool
2 July 2012 SVN 5017 snapshot
-- added integrated support for large data
-- enhanecments to Servlet and WebServer
28 June 2012 SVN 5009 snapshot
-- improved optimisation of queries with views and subqueries
-- improved support for complex, nested queries
-- improved name resolution in SQL statements in functions and triggers
4 June 2012 SVN 4997
-- fixed issue with server sometimes not exiting at shutdown
-- fixed regression with recursive queries
-- END is no longer a reserved keywords
-- property for SQL logging is now persisted
14 May 2012 SVN 4991
-- fixed issues with hot backup
-- improved name resolution with nested subqueries
-- improved support for collations with NO PAD
-- added support for MySQL style DATE_ADD, DATE_SUB and interval expressions
22 April 2012 - SVN 4977
-- fixed issue with datetime TRUNC and ROUND functions
-- fixed issues with a number of functions
-- fixed issue causing reduced speed and memory leak with disk based result sets
-- fixed issues with comparison of lobs and lobs file exceeding 2GB
-- extended batch execution to procedure calls
26 February 2012 - SVN 4944
-- fixed regression in DECODE function
-- fixed issue with ROWNUM when correlated supqueries were used
-- added support for COUNT(DISTINCT ...) with multiple arguments
-- added support for SQL log level 1-3
-- extended ORA syntax compatibility mode
-- improved garbage collection of dropped MEMORY and TEXT tables
-- improved name resolution with joined tables
28 January 2012 - SVN 4925
-- fixed issue with min(), max() optimisation for indexed columns
-- fixed issue with some usages of UNNEST
-- fixed issue with some NOT IN usage with NULL in list

22 January 2012 - version 2.2.8
-- fixed regression since 2.2.6 that affected database crash recovery
-- added support for hot backup

14 January 2012 - version 2.2.7
-- fixed issue with NPE thrown with nio data files
-- added SET DATABASE SQL NULLS FIRST and the property sql.nulls_first for the default sort order of nulls
-- added SET DATABASE SQL SYNTAX DB2 and the property sql.syntax for DB2 syntax compatibility
-- extended SET DATABASE SQL AVG SCALE and the property sql.avg_scale to the results of division
-- changed the JDBC object returned for INTERVAL values to String for the default jdbc.translate_tti_types=true setting
-- improved syntax compatibility modes
-- several minor improvements and bug fixes

19 November 2011 - version 2.2.6
-- fixed issue with memory use of large strings in cache
-- fixed issue with FOR loops with ORDER BY in routines
-- fixed issue with TIME ZONE types JDBC type reporting in server
-- fixed issue with type resolution of arrays
-- fixed issue with identical prepared statements in session
-- fixed issue with cached tables in jar databases
-- fixed issue with certain server properties
-- fixed issues with MVCC transactions
-- fixed issue with some row comparison predicates
-- added code branch supporting huge multi terabyte databases with tens of billions of rows
-- added property sql.avg_scale for decimal results from AVG and MEDIAN aggregates
-- added support for AVG with datetime types
-- added two-arg REPLACE function
-- added new datetime and bit functions 
-- added support for ROWNUM() in UPDATE and DELETE statements
-- added backward compatibility for IDENTITY options with commas
-- added support for new SQL properties
-- added Gradle build
-- improved SYSDATE, NOW and similar functions to accept optional empty arg list
-- improved JDBC support for LOBs
-- improved concurrency of CREATE and DROP DDL operations
-- improved data type determination for dynamic variables
-- improved routine variable scope in MERGE statements
-- improved compatibility modes
-- improved persistence reliability
-- many minor improvements and bug fixes
-- major improvements to SqlTool

5 July 2011 - version 2.2.5
-- fixed UNNEST in subquery
-- fixed issue with INFORMATION_SCHEMA.ELEMENT_TYPES view
-- fixed missing exception when table functions were used as SELECT items
-- fixed regression with server ACL file load
-- fixed issue with query not compiling with certain types of IN predicate 
-- added support for UNNEST with multiple ARRAY args
-- added SYSTIMESTAMP function
-- dba privilege required for LOAD_FILE function
-- improved handling of misconfigured external logging platforms
-- improved query speed for some joins
-- improved parameter type allocation for IN lists
-- minor improvements

12 Jun 2011 - version 2.2.4
-- fixed reported issue 
-- improved INFORMATION_SCHEMA concurrency
6 Jun 2011 - version 2.2.3
-- fixed issue with limit offset with indexed IN predicate
-- added INFORMATION_SCHEMA.ELEMENT_TYPES view and related enhancements
-- added hsqldb.full_log_replay property
-- added REGEXPR_SUBSTRING and REGEXPR_SUBSTRING_ARRAY functions
-- added support for multiple unnamed savepoints
-- improved IFNULL etc. functions that are based on CASE WHEN expressions
-- improved Transfer Tool in HSQLDB transfers
-- minor improvements

22 May 2011 - version 2.2.2
-- increased .data file limit to 2TB
-- improved EXPLAIN PLAN
-- improved query speed for some joins
-- improved TRUNCATE statement
-- added support for text tables in mem: databases
-- minor improvements

16 May 2011 - version 2.2.1
-- fixed regression in 2.2.0 with server TLS
-- fixed regression in 2.2.0 with ResultSetMetaData schema names
-- fixed type limitations of Connection#createArrayOf
-- fixed DatabaseManager running as Applet
-- added contents to INFORMATION_SCHEMA.SQL_IMPLEMENTATION_INFO
-- added auto checkpoints execution in separate session
-- improved query speed for some LEFT OUTER joins
-- added deletion of the temp directory at shutdown

10 May 2011 - version 2.2.0
-- fixed regression in 2.1.0 with database log created after a backup is completed not used for recovery
-- fixed regression in 2.1.0 with alter table operation failures causing loss of table data 
-- fixed issues with disk based result sets
-- fixed issue with multiple ALTER TABLE statements used as a single statement
-- reserved keywords now include DEFAULT and exclude ADMIN
-- added support for ROWNUM() function
-- added support for GET DIAGNOSTICS statement and DIAGNOSTICS function
-- added support for POSITION_ARRAY and enhancements to SORT_ARRAY (also ARRAY_SORT)
-- added support for WIDTH_BUCKET function
-- added support for SQL execution logging SET DATABASE EVENT LOG SQL LEVEL 3
-- improved column nullability reporting in ResultSetMetaData
-- improved BOOLEAN type support in compatibility modes
-- improvements to Server, including new shutdown methods and properties file name argument
-- improvements to MVCC transaction management
-- improvements to text table properties support
-- improved table and database locking for DDL statements
-- minor improvements and bug fixes

14 Mar 2011 - Version 2.1.0
-- added support for WITH RECURSIVE
-- added support for ALTER INDEX <name> AS (<new column name list>) to change the indexed column list of an index
-- extension of MySQL and PostgreSQL compatibility modes. Includes type conversion and sequence functions
-- extension to IDENTITY column support allows the use of a named SEQUENCE
22 Feb 2011 - SVN 4096 - 2.1.0 rc4
-- improvements to avoid creation of new .data files at checkpoint and allow incremental enlargement of this file
-- additions and improvements to INFORMATION_SCHEMA
-- added MS SqlServer function compatibility
-- added more extensive support for ALTER SPECIFIC ROUTINE
-- added initial support for JDBC 4.1 - Java 1.7
-- added support for opening most 1.8.x databases that are not explicitly SHUTDOWN
-- event logging format has changed slightly, and new level 3 allows more detail in the log
-- added MEDIAN aggregate function
-- added ARRAY_SORT function
-- routine variables and parameters can be used in LIMIT / FETCH / OFFSET clauses
-- improvements to lob use with PreparedStatement 
-- improvements to calling procedures with JDBC
10 Dec 2010 - SVN 3940 - 2.0.1 RC3 
-- added support for cached tables for lob catalog
-- added support for CURRENT VALUE FOR <identity sequence>, same as CURRVAL
-- added TO_TIMESTAMP, LPAD, RPAD functions
31 Oct 2010 - SVN 3836 - 2.0.1 RC2
-- added support for CREATE TABLE ... IF NOT EXISTS ...
-- added support for ?user=xxxx&password=yyyy connection in URL string
-- improved query optimisation for LIKE
-- added support for password complexity check functions
-- added support for external user authentication
-- added UUID functions
19 Oct 2010 - SVN 3824 - 2.0.1 RC1
-- added support for compressed scripts
14 Oct 2010 - SVN 3814
-- added support for recursive SQL functions and local table variables
04 Oct 2010 - SVN 3808
-- added support for translation of Oracle types VARCHAR2, NUMBER, RAW, etc. with SET DATABASE SQL SYNTAX ORA TRUE
-- added support for more procedural statements in SQL triggers
-- added support for variables in database URL's
24 Sept 2010 - SVN 3794
-- fixed issue with new sql compatibility properties applied to older databases
-- fixed bug with IN predicates in arguments of aggregate functions
-- fixed bug with table functions used in subqueries
-- added support for Oracle syntax elements including DUAL, ROWNUM, NEXTVAL, CURRVAL with sql.syntax_ora=true property
-- added support for clustered indexes
-- added enhancement for column visibility in nested joins
-- added support for datetime ROUND and TRUNC functions
-- added new connection properties
-- added support for FOR loops in routines
11 Aug 2010  - SVN 3750
-- extended getGeneratedKeys to UPDATE and MERGE operations
-- extended ALTER SESSION statement
-- added SEQUENCE_ARRAY and SESSION_ID function
28 July 2010 - SVN 3725
-- improved changing type of columns
-- added support for adding GENERATED ... AS IDENTITY to an existing column
23 July 2010 - SVN 3721
-- added FILTER clause support to aggregate functions
-- added functions ACTION_ID, TRANSACTION_ID, TRANSACTION_SIZE, UNIX_TIMESTAMP
22 July 2010 - SVN 3716
-- improved default reporting of WITH TIME ZONE and INTERVAL types in metadata and system tables
-- improved getGeneratedKeys with user defined column sets
-- fixed bug #3032488 - multi-column IN condition without a full index on columns did not work
19 July 2010 - SVN 3709
-- added support for procedures returning result sets
-- additional relaxation of BIT used as BOOLEAN
13 July 2010 - SVN 3693
-- added aggregate functions ARRAY_AGG and GROUP_CONCAT
-- added TO_DATE function
09 July 2010 - SVN 3688
-- added function LOB_ID
-- added BLOB, CLOB, VARBINARY support to Java functions and procedures
-- fixed bug with OCTET_LENGTH function
07 July 2010 - SVN 3686
-- added property sql.longvar_is_lob
-- added missing CONCAT function
-- added ARRAY support to Java functions and procedures
-- added query optimisation for DISTINCT and GROUP BY on indexed columns
-- speed improvement for some large results
-- fixed bug related to persistence of lobs after row updates that change the lob
-- fixed bug #3023253 - multiple count(*) with duplicate tables in outer join
-- fixed bug #3022106 - mix of IS NULL and other conditions in outer join
26 June 2010 - SVN 3674
-- added property sql.enforce_types and related SQL statement
-- added ALTER SESSION statement
-- fixed bug #3021175 with setFetchSize()
-- relaxation of type comparison and null types - controlled by sql.enforce_type
22 June 2010 - SVN 3667
-- changed names of three system functions ISDATABASEREADONLY() to IS_DATABASE_READONLY(), etc.
-- fixed bug with Clob.getInputStream()
-- improvement to TO_CHAR function
-- some relaxation of DECIMAL dynamic params
16 June 2010 - SVN 3658
-- fixed bug #3016375 Server remote open with no initial database
-- fixed bugs #3016397 - multiple UNION CORRESPONDING did not work
-- fixed bugs #3016481 - TABLE T did not work with views
-- minor improvement to MVCC transactions
13 June 2010 - SVN 3654
-- fixed issue with third parameter of LOCATE function
-- added TRUNC and changed return type of TRUNCATE function to reduce decimal scale
-- added synchronized blocks to allow mass creation of mem: databases
-- added support for jdbc:default:connection for Java functions
07 June 2010 - SVN 3646
-- fixed bug - SET FILES LOB SCALE did not change the store without a shutdown
-- fixed bug #3012585 - multi-column FOREIGN KEY with mismatched order of PK columns did not work correctly

06 June 2010 - version 2.0.0