ODBC Enhancements

The ODBC driver has been updated to Version 3.5.

Functions Supported by ODBC Driver

The Ingres ODBC driver supports all level one functions, as well as the following level two functions:

SQLExtendedFetch (through Microsoft Cursor Library only)
SQLForeignKeys
SQLMoreResults
SQLNumParam
SQLPrimaryKeys
SQLProcedureColumns
SQLProcedures
SQLSetPos (through Microsoft Cursor Library only)

Unavailable Features in the ODBC Driver

The Ingres ODBC driver does not provide the following features as of Release 2.6:

Executing functions asynchronously
Translation DLL
Support for Ingres SQL COPY TABLE command
Support for Ingres SQL SAVEPOINT command

JDBC Enhancements

The following JDBC 2.0 extensions have been added to Ingres:

Compatibility with GA release (protocol levels)
Execution in JDK 1.1 environment due to a new driver
Batch processing
javax—two-phase commit
javax—client connection pooling
Updateable result sets

The following are new features:

Support for Ingres intervals
Coalescing statement IDs
Utilization of VNODE passwords
Local connections without passwords
Support for procedure table parameters
Support for row producing procedure

Support for Unicode

This release contains the first phase of Ingres support for Unicode; further Unicode support will be added in future releases.

In this release, the DBMS supports three new data types:

nchar
nvarchar
long nvarchar

These data types store character data using two bytes for each character. Collation of these data types uses the standard collation algorithm as defined by the Unicode organization, and the data types can be used in indexes and database statistics.

The native two-byte (UCS2) format is supported and maintained through the entire process, from the front-end application, through the DBMS, to the data representation on disk.

The embedded SQL preprocessor for C and C++ supports declaration of wchar_t variables, which are assumed to contain multi-byte Unicode character strings.

OpenAPI version 3 was added to indicate support for these new data types.

VDBA also supports the new data types.

Support for the ODBC and JDBC drivers is present through their normal Unicode interfaces.

These new data types are not supported in any of the character-based tools or any of the terminal monitors.

This release does not support coercion between Unicode data types and non-Unicode data types such as char and varchar.

New Character Sets to Support Euro Currency Symbol

Two new character sets that contain the Euro currency sign (€, Unicode U+20AC) are added: IS885915 and WIN1252.

To set the money format to the Euro currency symbol you must issue the following command:

ingsetenv II_MONEY_FORMAT L:€

Alternatively, you can set this value in the Ingres Visual Manager (IVM).

Windows:

WIN1252 corresponds to Windows code page 1252 Latin 1. This is the common character set of most American and Western European Windows PCs, and includes the Euro sign. Users wishing to use the Euro symbol in a Windows GUI environment need to select the WIN1252 character set at installation time. To set this code page in a Windows command prompt environment, you must issue the following Windows command:

chcp 1252

The default font in a Windows command prompt does not provide support for the Euro currency symbol. For a workaround, set the font to Lucida Console. The Lucida Console font has moved the line drawing characters, used in Ingres forms, into an area not accessible to Ingres binaries, so we have provided rudimentary line drawing in the IBMPCD terminal entry. To set this terminal type, you must either issue the following command:

ingsetenv TERM_INGRES IBMPCD

or set TERM_INGRES through IVM or specify this terminal type at install time.

UNIX:

IS885915 corresponds to the ISO 8859-15 Latin 9-character set that is almost identical to the ISO 8859-1 Latin 1 set, except for eight characters; chief among them is the Euro currency sign (€, Unicode U+20AC).

If you have an existing installation and would like to change the character set, be aware that this is not typically supported because the new character set could display existing characters in your databases incorrectly. However, since the ISO 8859-15 only has eight characters that are different from ISO 8859-1, if you can verify that none of the eight characters are already present in your databases, you could safely change the set (by changing II_CHARSETxx).

The following table details these differences and provides the corresponding Unicode character names:

Hex		ISO 8859-1		ISO 8859-15
A4	?	Currency symbol	€	Euro sign
A6	¦	Broken bar	Š	Latin capital letter with caron
A8	¨	Diaeresis	Š	Latin small letter with caron
B4	'	Acute accent	Ž	Latin capital letter Z with caron
B8	?	Cedilla	Ž	Latin small letter Z with caron
BC	¼	Vulgar fraction one quarter	Œ	Latin capital ligature OE
BD	½	Vulgar fraction one half	Œ	Latin small ligature oe
BE	¾	Vulgar fraction three quarters	?	Latin capital letter Y with diaeresis

Differences Between ISO 8859-1 and ISO 8859-15 Character Sets