Dalede wrote:Thanks for the info on precision. The using function is good for seeing what the real precision is. I was basing on the ECMA but this basic is no longer very close to ECMA particularly as far as it relates to limitations. 8 bytes mantissa would be 64 bits (8 x8). Do you mean the entire floating point number is 64 bits with 52 bits related to the mantissa and the rest to the exponent?
Dale
smart BASIC uses standard 64-bit float numbers representation, which follows 1-11-52 bit pattern for sign-power-mantissa parts.