>>Hi all,
>>
>>What is the advantage when we allowing Null data type to each data field?
>>
>Aston,
>
>Null's are usefull when value in a field can be unknown. It mostly used with datetime data type to represent unknown or not entered yet value. I tend to use them only if I've to because they can be a pain to work with.

They're quite useful if you are working with data that will be used for any statistical purposes -- means, medians, standard deviations, and the like. In that case NULL can mean "not applicable" and these values are left out of the equations. When I worked on statistical apps in FP 1.* and 2.* I always had to write complicated procedures to duplicate what I can now accomplish with standard function calls.