Subtype | Internal Size | Value Range |
Text | 2*(len+1) + overhead in Windows 4*(len+1) + overhead in Unix | Up to 2 billion characters in Windows (32- and 64-bit) Up to 2^64 characters in 64-bit Unix |
Short | 2 bytes | 16-bit signed or unsigned binary integer, therefore range is: Unsigned: 0 to 65535 Signed: -32768 to 32767 |
Integer | 4 bytes | 32-bit signed or unsigned binary integer, therefore range is: Unsigned: 0 to 4294967295 Signed: -2147483648 to 2147483647 |
Float | 8 bytes | Precision of about 16 digits. Standard IEEE 64-bit floating point. |
Decimal | Varies | Precision of 76; scale of 38. Actual size depends on output format. |
Date | 12 bytes | For value range, see comments for DateTime. Based upon DateTime with a time of 0. |
Time | 12 bytes | For value range, see comments for DateTime. Based on DateTime with a date of 0. |
DateTime | 12 bytes | DateTime can represent date and time to nanosecond precision, starting with January 1, 1 AD. Julian to Gregorian transition uses England-Scotland transition, where September 14, 1752, immediately follows September 2. |
Boolean | 1 byte | True or False |
Name | Varies | A string parsed into these name parts: • honorific • first • middle • last • title |
Quadword | 8 bytes | 32-bit signed or unsigned binary integer, therefore range is: Unsigned: 0 to 18446744073709551615 Signed: -9223372036854775808 to 9223372036854775807 |
Interval | 32 bytes | Range depends on output format and interval type. |