This appendix describes the new features of Ingres 2.6, including the following:
Enhancements have been made to the internal performance that concern row producing procedures.
This enhancement to the Ingres database procedure language addresses the ability of Ingres to read and return to the caller multiple rows from a select statement.
With server-executed database procedures, the program logic of the procedure is executed entirely in the server address space. Multiple SQL DML requests are executed in a single invocation of the procedure, with only one interaction with the client application. The ability to process and return multiple "rows" of some composite data types with a single call to a server-resident database procedure adds to the potential for improved performance of an application.
In a typical computing environment with applications executing on a variety of computers throughout a network, this approach can significantly reduce the footprint of the client application and the traffic across the network.
The ANSI compliant SUBSTRING function has been added to the SQL syntax. The SUBSTRING function is often easier to use than combinations of LEFT and SHIFT functions that Ingres traditionally supported. The syntax is:
substring(string-expr from start-column for length)
and the for length clause is optional.
Additional aggregate functions for statistical analysis have been added:
For details, see the SQL Reference Guide.
Prior to Ingres 2.6, character data types were limited to a maximum size of 2000 bytes; this restriction was imposed when the maximum size of a row was limited to 2 KB. This limit is increased to 32000 bytes, the maximum row size supported in Ingres 2.6.