Thread: varaible help...?

>Though I believe technically you can leave it in.
Yes...if you discard everything past the decimal place. Integers and floating-point are different beasts fundamentally, even though you can use floating-point types to represent integral values with a little effort.

>>The largest possible standard built-in type in C++ is long double.
Ah. I was just about to post that "unsigned long long x;" gave me a compile error.

Apparently, on my C++ implementation/platform/whatever, long doubles are the same as doubles. Both are 8 bytes.

>>Though I believe technically you can leave it in.
Well, it won't compile on MSVC so I'm assuming that you can't (since MSVC tends to allow more than it disallows).

>So I can get that many numbers crambed into the int?
Only if by "int" you mean "several int manhandled to work together so that they look like one int".

>Well, it won't compile on MSVC so I'm assuming that you can't
Whoops, I was thinking in abstract terms when the question was about type declarations. Silly me.

>>But I also heard they can be off when adding 2+2.
I certainly would hope that computers are smarter than that. Otherwise the world will be doomed when they take over.

>>a long double float
float and double are both floating-point values. double is twice the size of a float. You should realize that double is a built-in datatype just like int

And if a byte = 2 letters or numbers (or 1 of each) how can a long be 8 bytes and hold so much. Or are you talking about 8 bytes * the int. What the heck how does that work?

A byte is 8 bits. A bit is either 1 or 0. So with all possible bit combinations, that makes up to 256. A long has 4 bytes, therefore 4 * 8 = 32 bits -> 2^32 possible values. 8 bytes -> 8 * 8 = 64 bits -> 2^64 possible values.

[edit]The 2 letters or numbers you're referring to is the hexadecimal representation. Hex is a base 16 number system, so you go 1,2,3,4,5,6,7,8,9,A,B,C,D,E,F,10,11,... It actually has 256 possibilities.

>double is a whole diferent type of varaible, right?
There are three floating-point types: float, double, and long double. You can think of them in similar terms as short, int, and long int, except for floating-point.

>So then can it handle + and - and decimals?
Yes, it's a numeric type, so naturally it supports the numeric operations.

>and can I still put a long before it?
Yes, but keep in mind that long double is technically a different type. C++ is picky about types matching up, and double is the default floating-point type.

>>double is a whole diferent type of varaible, right?
Yes.

>>So then can it handle + and - and decimals?
Yes.

>>and can I still put a long before it?
Yes. Except the additional 'long' isn't guaranteed to make it any bigger (as my quick test showed, with both double and long double being 8 bytes).

>>So basicly I have to pick 1 type before I start making my program?
Yes.
If you're going to use built-in math functions, use double; if you're not, use long double if you're going to need all the storage you can possibly have.

>A byte is 8 bits.
Not necessarily, though octets (as they're called) are the most common by far.

>A long has 4 bytes
At least.

>alright so a long double + double = messed up?
No, because of type conversions, long double + double = long double.