FLOAT data type migration
Sybase FLOAT is generally used to save non integer numbers like fractions where no number of precision can hold the exact value. Sybase FLOATs do not store an exact value "It stores slightly imprecise representations of real numbers as binary fractions at the hardware level" http://www.sybase.com/detail?id=20313 Oracle has two data types, FLOAT and BINARY_FLOAT. http://stackoverflow.com/questions/332492/oracle-floats-vs-number FLOAT is really a decimal data type with exact values (basically it is a NUMERIC) BINARY_FLOAT is a binary data type which better maps to Sybase FLOAT data type If you migrate from Sybase FLOAT to Oracle FLOAT any value inserted will be treated as a specific explicit numeric, which will not behave like Sybase FLOAT. If you migrate from Sybase FLOAT to Oracle BINARY_FLOAT then any value inserted will be treated as a binary imprecise number, just like Sybase. SYBASE drop table testfloat go create table testfloat (floatcol float) go ins...