>
>If, as in the case with Fox, the most significant bit represents only the sign. If, as with SQL server, it represents the sign and is part of the value, 4,294,967,296.
>
>
>Where found you this sentence. This is wrong.
>
>VFP and SQL uses the same 2-complement representation.
>VFP reserve -2^31 = 0x80000000 for the Overflow status.
>On SQL you cannot store a Overflow status into a Integer field
>
>End.

Fabio,

Bit 31 is the sign and bits 0 to 30 are the value. This is why the maximum negative value is one less than SQL Server. You get the numeric overflow because the negative value you're trying to insert is beyond the range of the data type. In SQL Server bit 31 is the sign and part of the value.