help me:what 's the difference among these 3 mothods?

[Mike Lin]

byte m1() {
final short s=2 ;
return s; // 1
}
byte m2() {
final short s ;
return s;
}
byte m3(final short s) {
return s;
}
comile m2()or m3() will complain of:
" possible loss of precision"
but m1() is ok.
why???
I really get sick !

[Alfred Kemety]

The short S in m1() is what we call a "Compile Time Constant", its value is known at compile time and is unchangiable during run time since the variable is declared final.
2 is in the range of byte, so there will NOT be any loss of precesion at run time, the compiler knows that.
In m1(), m2() although the variables are declared final, they are not assigned a value yet, so maybe they will have the value 140 which when converted to byte from short will be chopped and loses preceision, so the compiler warns.
HTH

[Mike Lin]

thanks!:-)