long bigvalue; bigvalue = 400*400;
You’d expect bigvalue to hold 16000 right? No. Apparently arithmetic operation are done using the data type of the operand. So if you do any arithmetic operation (+-*/) that exceeds what the operand data type can hold, for example integer {+32767,-32768} you’d get some random overflown garbage result, even if the variable to hold the result is capable of containing it.
A workaround is to put the constant into a variable compatible with the result. Like
long bigvalue; long bye = 400; bigvalue = 400*bye;
The compiler will use the larger data type if mixed data types are used in an arithmetic operation, and in the above case produce the correct result.
