Conversion of Primitive Data Types

[Zahid, Butt]

Hi
I was studying "Complete Java 2 Certification Study Guide ", Chapter 4. It clearly states that Primitive data types can be
converted to a "larger data type" e.g from a byte to a short.
This is known as widening conversion. The system will do this for you, its Implicit.
It then later states that a "long" can be converted to a "float".
How is this possible ???
A long is 64 bit BUT a float is 32 bit. This is not a widening conversion, rather its a narrowing conversion and hence will give a compile-time error.
Can somone please shed light on this, maybe a couple of code examples ???

Thanks in Advance.

[Rana Thakur]

Hi,
The value for a long variable will be stored as 64 bits.
However the floating point value will be stored in 32 bits in IEEE floating point format.i.e left most bit for sign bit, next 7 bits from the left for exponent and the remaining 24 bits for mantissa. Thus the value that can be stored using this format(maximum value being 3.4E+038) is much higher than that of the value that can be stored in a simple long variable(max value (2^63)-1). Hence the conversion is widening and not narrowing.