Printable View
I just finished the 2nd Chapter in Beginning Visual C++ 6 by Ivor Horton, and when doing the exercises found that I did not really understand how Bitwise Operators work, and why they are of any importance. So if anyone can give me a short, yet logical explanation or point me towards an online tutorial that covers this please reply. Thanks! :)
>and why they are of any importance
look up hash code generating using bit indexing. Bit wise operators become very usefull here.
heres a short and maybe logical explanation of how they work
you just perform the operations on the individual bits.Code:bit1 = 1001
bit2 = 0101
bit1 & bit2 = 1&0, 0&1, 0&0, 1&1
= 0 0 0 1
bit1 | bit2 = 1|0, 0|1, 0|0, 1|1
= 1 1 0 1
etc....
They are used often for 'flags' in programs, in which a variable in a single byte must store multiple on/off switches (one for each bit).Quote:
Originally posted by o0obruceleeo0o
...and why they are of any importance.
Also, if you wish to use modulus on an even computer number (2^n,n=integer), then it is the same as using a bitwise AND, if you use the number one less than the modulus. Example:
n % 8 is equal to n & 7
n % 32 is equal to n & 31
There are tons of other reasons.
Hi,
I'm studying that book too. It's not as thorough as Ivor Horton's Beginning C++, but in some ways I think it might be better for beginners because they won't get bogged down in too much detail--it took me a long time to finish Beginning C++. With Beginning Visual C++, I skipped the first five chapters, and I'm reviewing classes and working my way towards the windows programming and MFC part.
However, I remember the bitwise operator section from Beginning C++ because it almost caused me to abandon the book. It was so complicated, I was forced to skip the section, and since nothing built on it later, I was OK, and the book turned out to be great. In my opinion, bitwise operators should be in an appendix because it's a great way to turn a beginner away from the language. You may never need to use bitwise operators, so you shouldn't worry about them. If you understand the basic AND'ing and OR'ing of two numbers together, that's plenty.
PM me through the bulletin board if you have any specific questions about examples or problems in the book.
It should be explained in the book, since you will come across bitwise operators, and you must understand them. So, it cannot be just an appendix. But, I agree that not enough examples or explanation of their use is supplied in the book.Quote:
Originally posted by 7stud
In my opinion, bitwise operators should be in an appendix because it's a great way to turn a beginner away from the language.
Hmmm, I thought there was something in the tutorials or the FAQ... didn't see it just now.
You can do a lot of programming without worrying about bits, binary, or hex. But, people who program embedded processors, program in assembly language, or work in the hardware world (like me) use binary, hex, and bitwise operators all the time. One bit might be used to turn on an LED or a relay, or to monitor the status of a switch. I'm often using an oscilloscope to check the state (or activity) of individual bits. My test programs often set, clear, or read individual bits.
I don't have the Horton book, but if I was writing a "bitwise" chapter, I'd start with a single-bit boolean logic... 1 AND 1 = 1, 1 AND 0 = 0, 1 OR 0 = 1.. etc. Next I'd work on binary and hex number systems, then finaly bitwise operations on bytes, words, etc. So, you might want to study-up on boolian logic and binary and hex first. Hmmm... I might actually start my chapter with a tiny bit of digital circuitry, where 0 = 0 volts and 1 = 5 volts... to make it a bit less abstract (?). I learned about boolean logic and logic circuitry before I learned how to do it in software.
This must seem very abstract and confusing to someone with no knowledge of digital circuits who is learning a high level language, unless you've already learned this stuff in a math class. But, it's really NOT complex.
[EDIT]
Here's some boolean logic info: www.cplusplus.com/doc/papers/boolean.html
Thanks for all the great replies! I understand that 1&0=0 etc.... but I guess my knowledge of binary is too limited to understand how that helps me much. On a side note I just started reading "Code" by Petzoid, and im on the binary chapter, so hopefully that will help me with all the binary stuff :D.
The problem is that complex bitwise operations are being presented in chapter 2 of a beginning C++ book in which the basic variable types are introduced. I was able to figure out and understand 880 pages of the C++ language, but I never did figure out the complex section on bitwise operators. I don't think that's the way to order the material. However, if your goal is to slap down anyone trying to learn C++, that's the best way to do it. Anyone giving up at that point would have an impression that C++ is a complex bitwise manipulation language, which is ridiculous. In fact, in the rest of my book, I don't think bitwise operators were ever mentioned again.Quote:
I don't have the Horton book, but if I was writing a "bitwise" chapter, I'd start with a single-bit boolean logic... 1 AND 1 = 1, 1 AND 0 = 0, 1 OR 0 = 1.. etc. Next I'd work on binary and hex number systems, then finaly bitwise operations on bytes, words, etc. So, you might want to study-up on boolian logic and binary and hex first. Hmmm... I might actually start my chapter with a tiny bit of digital circuitry, where 0 = 0 volts and 1 = 5 volts... to make it a bit less abstract (?).
If you started your chapter with digital circuitry, you would cause 90% of the readers to give up. There are so many more important things to learn about C++, that bitwise operations should be reserved for the last chapter of a beginning text, if included at all. Personally, I think any mention of bitwise operators early on in a beginning text should start and end with basic AND'ing and OR'ing.
7stud:
You're absolutely right. I just mentioned the digital circuitry stuff because it helps me actually visualize the little ones and zeros in there. And, to give an example of what it's good for... because otherwise why in the heck would anyone want to use this stuff? :) For me, it makes it less abstract, but I suppose for someone who doesn't know what a volt is, it's even more abstact! This also would be very easy for someone who already understands the math / logic.
o0obruceleeo0o:
If you understand the AND, OR, NOT, and XOR stuff, then you're 90% there. I think what makes this difficult, is that C++ is not very "binary friendly." (You can't just stick binary numbers into you source code, etc.) You're probably getting bogged-down in the Binary to decimal (or hex) conversion. BTW - you shouldn't have to deal with number base conversion in the 2nd chapter either!
Let's try this... Easy in binary!
01010101
00001111
------------ AND
00000101
01010101
00001111
-------------OR
01011111
In hex...
55 h
0F h
--- AND
05 h
55 h
0F h
--- OR
5F
In decimal
085
015 *
---- AND
005
085
015 *
---- OR
095
Examples:
Testing / manipulating bit 1 (Starting with bit zero at the far left.)
if(X & 2) // If bit 1 is high.. Note- 0010 binary = 2 decimal
X = X & 2; // Force bit 1 high
Darn! I must be in a wordy mood today!
*[Corrected... oops!] :o
Here it isQuote:
Hmmm, I thought there was something in the tutorials or the FAQ... didn't see it just now.
In hex...
55 h
0F h
--- AND
05 h
55 h
0F h
--- OR
5F
In decimal
085
255
---- AND
005
085
255
---- OR
095
Ok, now you've got me interested. Can you explain the results from AND'ing and OR'ing with hex and decimal?
>Ok, now you've got me interested. Can you explain the results from AND'ing and OR'ing with hex and decimal?
try converting those hex and decimal numbers to binary and doing the and's and or's. then convert binary back to hex/decimal. you'll see bitwise operations work just the same :)
I knew that was in there somewhere. I just sent a PM to hammer suggesting that he change the title.
[EDIT]
To expand on what perspective said: Everything is stored in binary. The computer doesn't care what base you use to enter or display it. You could even do bitwise operations on ASCII characters!
The advantage of hex over decimal is that it's easier (for humans) to convert between hex and binary, than to convert between decimal and binary. You can learn to do hex-to-binary in your head, because each nybble (half-byte =4 bits) converts to a number between 0 and F.
1010 1111 = AF Hex (1010 Always converts to A, etc)
1111 1010 = FA Hex
1010 1111 = 175 Decimal (Had to use a calculator)
1111 1010 = 250 Decimal
The advantages of hex over binary are that C++ handles hex directly. (You can't cin or cout binary). And, when your bit patterns get longer than one byte (8 bits) they get very hard to read... you have to start counting bits.
[EDIT AGAIN]
I used the same numbers in different bases:
1010 1010 binary = 55 hex = 85 decimal
0000 1111 binary = 0F (or just F) hex = 15 decimal (correct now)
0000 0101 binary = 05 (or just 5) hex = 5 decimal
0101 1111 binary = 5F hex = 95 decimal
ok I think i got it now, thanks a ton everyone! Can anyone give me a non complicated example of when you would want to use this? It still doesnt make sense to me because I cant count past 10 in binary without writing it down :D. Once again, tyvm for the replies!
Hi,
I can think of an example. Suppose you had two sources in your code sending information on what the propeties your window should contain to properly display data and provide for certain required functionality for your window. Let's say the first digit is the color(blue or white), the second is the font(large or small), and the third digit is the ability to resize the window(resizable or not resizeable). You decide the appropriate action is to OR them together for the final look because if one source requires a certain background color, text size, and a resizable window, you have to have those propeties to display ALL the data properly and provide the proper functionality for the window.
1 0 1 (from source a)
0 0 1 (from souce b)
-----------------------OR
1 0 1
Based on that, you would create a window that had a blue background with small font and a resizeable window. The alternative would be to parse each source and with a bunch of compound if-statements choose the appropriate properties. (Note: I have no experience with this stuff, but it seems like a plausible example)
ahhhh!!! They used an example like that in the book "Code", i think I got it now! Thanks a ton!!! By the way, if anyone else is having problems with binary etc, may i suggest picking up the book, Code by Charles Petzoid. It starts out with explaining Morse code and then Brail, and then Binary. I new nothing of what binary was other than a bunch of 1's and 0's and now i understand how it works :D. It also covers Boolean Algebra!
Ok, I read all the post here and i found something i think is VERY usefull and was not mentioned.
By declaring a structure member to be a specific amout of bits,
e.g a char is 8 bit, and int is 16 bit, ect.....
Well, if you have to only save date as a yes or no why not declare a structure member with only one bit. 16 times smaller as an int but you can still either be a 1 bit or a 0 bit. Or what i like to call true or false.
If you used only 2 bits you can say "zero = 00 one = 01, two = 10, and three = 11. You get four posibles with only 2 bits!!!
3 bits can hold 7 values, 4 can hold 15 ect... That also made me realize why computers store info in binary =)
Then your only job is to define answers or values you want each combination to equal.
Example: using people and thier appearance....
I hope that helps you understand a great way to use well, i guess not exacty bit operators although im sure you could use them somehow with this. But using bits as a database alone with bit shift could make things pretty simple. Note: I forget if the colon operator is used the same in C++, i think i was using the c stdio.h when doing this. Thanks.Code:
struct appearace {
unsigned hair_color : 3; //defines # of bits
unsigned hieght :2; //lets say tall, ave, short, or NA
unsigned dead :1; //living or dead
};
//then define an array or whatever
struct appearance people[100];
//then assign you values as unsigned int, the equation for how
//many number for you bits size in 0 through 2(#of bits) - 1
people[1].hair_color = 3 //or bits 11
people[1].hieght = 0 // or bit 00, tall
people[1].dead = 1 //they are dead
-Thenrkst
>3 bits can hold 7 values, 4 can hold 15 ect
you mean 3 bits can hold 8 values ( 0 - 7), 4 can hold 16 (0- 15)...
Lets see....
3 bits, 000, 001, 010, 100, 011, 110, 101, 111...
yep 8 ...=)
That's OK. I have a couple of calculators that do decimal-hex-octal-binary conversion. You can also get a program to do it. (Or write one yourself!) Anybody who does lots of conversion uses a tool! Don't get bogged-down with this shtuff!!! It may be a long time 'till you really need to use it.Quote:
I cant count past 10 in binary without writing it down
Now, people who work with this stuff alot do learn to count to "F" (15)... Well, actually we memorize the 16 conversions between 0 and 15 for binary, hex, and decimal. Then, we can convert between hex and binary of any size! (I tend to forget some of 'em when I don't work with it for awhile.)
Converting to and from decimal is usually done with a calculator or program, although my boss is pretty good at doing it in his head! Again, this is why hex is used... hex-binary is much easier than decimal-binary.
Actually... this may sound weird... but when we work in binary and hex, we usually don't care what the decimal value is... its just a "bit pattern" and all we are doing is bitwise operations... no other math.
EVERYTHING in the computer's memory is stored in binary. The compiler / operating system normally convert to-and-from decimal (or ASCII) for you. Any conversion is done ONLY during input/output. So, you can cin a decimal number and cout the same variable in hex, and the actual variable in memory is NOT affected by the "conversion".
Greetings,
Elaborating on thenrkst's ideia, and if speed is not that important you can avoid bitwise operators alltogether:
Makes the code easier to read, but if you want speed stick to the bitwise operators.Code:#include <iostream>
typedef unsigned char byte;
typedef unsigned int uint;
typedef union {
struct {
uint bit0:1;
uint bit1:1;
uint bit2:1;
uint bit3:1;
uint bit4:1;
uint bit5:1;
uint bit6:1;
uint bit7:1;
} bits;
byte val;
void display_bits();
void input();
} TBYTE;
void TBYTE::display_bits()
{
std::cout << bits.bit7 << bits.bit6 << bits.bit5 << bits.bit4
<< bits.bit3 << bits.bit2 << bits.bit1 << bits.bit0
<< std::endl;
}
void TBYTE::input()
{
int tmp;
std::cin >> tmp;
val = tmp;
}
int main()
{
TBYTE tbyte;
std::cout << "Byte me: ";
tbyte.input();
std::cout << "Bits in byte: ";
tbyte.display_bits();
// You can check/set/unset the individual bits:
if (tbyte.bits.bit7 == 1) { // same as '(if byte & 128 == 128) byte^=128'
std::cout << "Bit 7 is set. Unsetting: ";
tbyte.bits.bit7 = 0;
tbyte.display_bits();
}
byte b = tbyte.val;
if ((b & 128) == 128) {
b ^= 128;
}
std::cout << int(b) << std::endl;
std::cout << int(tbyte.val) << std::endl;
}
The formula to calculate this is pretty simple: the number of possibilities is 2^n, where n = number of bits of storage.Quote:
Originally posted by thenrkst
Lets see....
3 bits, 000, 001, 010, 100, 011, 110, 101, 111...
yep 8 ...=)
So, 8 bits (1 byte) has 2^8 = 256 possibilities.
16 bits = 2^16 = 65,536 possibilities
32 bits = 2^32 = 4,294,967,296 possibilities.
24 bits = 2^24 = 16,777,216 possibilities (this is why you usually see true color mode [24-bit color, 8-bit for each R, G and B] as having 16,000,000 colors).
and so on...