I am currently allotting data types to attributes. I would like know the usage of Float visa vis Decimal. Say I have got a field whose maximum holding capacity is Decimal(11,3) which the data may ocassionaly reach. Is it advisable to have float istead of Decimal. Will float have any overhead/benifits in terms of performance and storage. Float, I find is not commonly used in commercial applications.
Is there any overhead on Varchar as an Primary Key.
According to the SQl Reference manual, FLOAT and DOUBLE are approximations of real numbers. This basically means that the values stored as one of these types is not exactually accurate. And all math done on these types is not accurate for the same reason-they are approximations. If you need accuracy, stick to DECIMAL where the only problem is loss of precision in some math operations.