Snip of Jim's e-mail.

Jim wrote:

here is what I found for setting different float vals
into each primitive type:
             1.2345 1234.1 70000.1
bool exc exc exc
byte no exc exc exc
char no exc no exc exc
short no exc no exc exc
int exc exc exc
long exc exc exc
 
So:
- bool is handled ok
- byte, char, short are handled ok with your interpretation that
  just losing precision is ok
- int and long seem to be broken. There should be no exceptions
  in any of the above for int and long, shoult there?