>>Beware, do not use TYPE(). It was added for speed but makes little difference and is
>>known for returning the wrong result at times. I have seen this myself.
>
>I know, it may return C or something else when the value is null
>
>As in
>

>x='qq'
>x=null
>?type('x')  && C
>

This is not a wrong result. X was previously defined to be character. When a NULL value is populated into X it is a NULL Character. The same holds true if you use x=10 and x=null. It now reports a NULL Numeric. It is reporting exactly the type of variable it is ... it's just that now instead of being populated with data, it is populated with empty data (a null).

This is often useful for testing for unpopulated data. Try this:

x = 10
x = null
@ 5,5 get x
READ

It initially shows ".NULL." but when you try to type input, it only allows numbers. If the user types anything (even the letter "k") the value changes to 0. If the user doesn't type anything in that field then it remains NULL.

This is a perfectly valid and legal concept in both VFP and computer science in general, and one which relates to the underlying programming languages of C/C++ down through assembly and machine code and the concept of a "null pointer".