unsigned int and unsigned long int

so even though I do:

typedef long int int_4

and I do sizeof(int_4) that will be the same wherever machine I am?? I guess not right

Originally Posted by -EquinoX-

so even though I do:

typedef long int int_4

and I do sizeof(int_4) that will be the same wherever machine I am?? I guess not right

That will not necessarily be the same on every machine.

As tabstop mentions, no. That is why these typedefs exist in the first place. To map appropriate type into these typedefs so you can be sure about size.
You have to use lots of #ifs and stuff to make it right for every machine.

oh okay by the way I have to slip in one unrelated question to this thread, say I have a code:

Code:

#include <stdio.h>
#include <stdlib.h>

int main()
{
  float number;
  scanf("&#37;f", &number);
  printf("%f\n", number);

  return 0;
}

why doesn't it read directly from the command line when I typed in ./tiesto 0.5

it doesn't print 0.5 directly but it goes as I haven't typed in anything but after I typed it in again underneath it works...

scanf reads from the keyboard input, not the command line.

well I thought I can do that.. in the previous assignment I have it works

Doesn't work that way.
You need to use a special main and get the arguments from there.
There's a tutorial on the site maybe?

Originally Posted by -EquinoX-

well I thought I can do that.. in the previous assignment I have it works

No it doesn't. You may have read arguments from argv[], but not using scanf.

Code:

#ifdef NO_WAI_64_BITS
typedef int_8 something;
#elif NO_WAI_32_BITS
typedef int_4 something;
#endif

oh nevermind it works if I did a redirect from a file.. okay back to our topic again.. the point to all of this is this.. I want main to create a message say that I run the program and defining that it's on 64 bit machine but however it's on a 32 bit machine.. then I want to be able to print this error:

gcc -ansi -Wall -DMACHINE32 extract.c main.c -o main


ERROR: sizeof(int_u8) is not 8, it's 4
ERROR: sizeof(int_8) is not 8, it's 4

I think you're on the wrong track here.
The point of those typedefs is that uint64_t will always be 64-bit, whether or not it's a 32-bit machine or a 64-bit machine.
Same goes for any other typedef.

okay so the type def should all be the same either for 32 bit machine or 64 bit machine??

The point is that if you use, for example, uint32_t, then
sizeof(uint32_t) on a 32-bit machine will give 4
and
sizeof(uint32_t) on a 64-bit machine will give 4.

The headers are full of #ifs and such to make sure this is the case.
And of course, each compiler has its own such header to make sure it uses the correct type to make it's the right size when compiled.

The whole point with declaring typedefs (in this discussion) is that "you know what size the integer is". This helps in making sure for example that data structures are always the same size, data types match between different functions, etc.

In many cases, a regular "int", whatever size it happens to be, is a fine thing to use, but sometimes you need a type that matches some specific size, such as "32 bits", and if you want to be SURE that the type is exactly 32 bits, then you need a typedef that depends on the compiler/processor architecture, etc.

For example, a 64-bit compiler will make "long" a 64-bit number, but in a 32-bit compiler you need "long long" to make a 64-bit number.

--
Mats