Data Types Used in Mimer SQL

---

The following sections explain how to compile applications using floating point data types, and what data types Mimer SQL uses internally and externally.

Compiling Applications Using Floating Point Data Types

OpenVMS supports various floating point types. Mimer SQL uses the types that the C compiler uses as default. Unfortunately, this differs between the Alpha and Integrity platforms.

F_FLOAT	4 bytes	Used by Mimer on Alpha
S_FLOAT	4 bytes	Used by Mimer on Integrity
G_FLOAT	8 bytes	Used by Mimer on Alpha
T_FLOAT	8 bytes	Used by Mimer on Integrity
D_FLOAT	8 bytes	Not supported by Mimer
H_FLOAT	16 bytes	Not supported by Mimer

Internal Mimer SQL Representation

Mimer SQL uses only two data representations internally: numeric and character. These two types are used to store all data in the database.

The numeric data type stores numbers in packed BCD format with a maximum precision of 45 digits. Every number stored has an exponent with a range of -999 to +999.

The character data type uses the ISO 8859-1 standard character set, also known as Latin 1. This standard specifies graphic characters and the representation of each character.

On most platforms, including OpenVMS, the ISO 8859-1 character set is used by Mimer SQL to transfer data to and from the application, i.e. the same character set as is used internally.

The ISO 8859-1 character set is presented in the Mimer SQL Engine Reference Manual.

External Data Types Supported by Mimer SQL

Any value stored in the database may be read into host language variables as described in the Mimer SQL Engine Programmer's Manual.

Mimer SQL will perform all the necessary conversions and will signal an error if the value to be converted is not compatible with the destination type.

Mimer Information Technology AB Voice: +46 18 780 92 00 Fax: +46 18 780 92 40 info@mimer.se