decimal
is a .NET (also other frameworks, rdbms and languages) value type that stores floating point numbers with decimal precision, instead of a binary one.
Is there any way of saving these values to a SQL CE database without using binary precision in the process?