Could someone confirm whether my understanding is correct (step by step)? From the mockup exam, http://joppa.appliedreasoning.com/javaCert/html/exam.html Ques No: 23 (byte) 0x81 >> 2; Steps: 1) a) Each hex digit must be represented in four binary digits (zeros and ones) and each octal digit must be represented in three binary diigts. So, when we represent 8 in four binary digits, we get 1000. Similarly 1 is equivalent to 0001. So, the binary representation of 0x81 is 0x 0000 0000 0000 0000 0000 0000 1000 0001 2) Since cast operator has higher precedene than shift operator, (byte) 0x81 is now 0x 1000 0001 (truncated to byte) 3) a) Now, when we apply shift operator, L.H.S should be promoted to int( if it is short, char or byte). b) Since sign bit is 1, we need to extend the byte to int with bit 1. So, promotion to int gives us 1111 1111 1111 1111 1111 1111 1000 0001 c) now shifting to two bits right, we get 1111 1111 1111 1111 1111 1111 1110 0000 (since the sign bit is 1, we use 1 bit to fill up the gap produced by shift) So, the result is 0xFFFFFFE0 ( I saw the answer is correct, but i want to know whether the logic what i am thinking is correct...) ********************************************************** Next question from the same mock up exam (Ques no: 24) 0x81 >> 2 As step 1 (above): we get 0000 0000 0000 0000 0000 0000 1000 0001 now, shifting 2 bits right gives us 0000 0000 0000 0000 0000 0000 0010 0000 (here the sign bit is 0,so we use 0 bit to fill the gap) So, the result is 0x20 Right? ****************************************************** Next question from the same mock up exam (Ques no: 25) 0x1 << 36 Here the L.H.S is int. So, we need to reduce the right hand operand to modulo 32, which gives us 0x1 << 4 Now, 0000 0000 0000 0000 0000 0000 0001 << 4 gives us 0000 0000 0000 0000 0000 0000 0001 0000 (for left shift, we need the gap with zeros) So, the result is 0x10 If i understand wrongly, then please do correct me?

Everything is right except that the two last answers should be 0x00000020 (not 0x20) respectively 0x00000010 (not 0x10). But the way you presented it is not wrong since you can make abstraction of the leading zeroes but it is more comprehensible that way since we are speaking of integers and not bytes... HIH ------------------ Valentin Crettaz Sun Certified Programmer for Java 2 Platform