Wierd round on a float

[Boda]

I was practicing/learning C++ while came up with a problem that requierd me to derange large nunbers to units one at a time.
Like 157 get 1 then get 5 and then get 7 separatly.
While preforming this I noticed that multiply by 10 was rounding low my nunbers.
This got me 3 hours back and forth coding difrently and googling with no results.
Here it goes a demonstration of my problem.
I'm using wxDev-C++ with default settings.

#include <stdio.h>

main()
{
  float x,count=0;	
	
  scanf("%f",&x);   //Im talking about something larger than 100//
  
  while(x>10){
    x=x/10;         //getting base
    count++;
  }
  
  while(count>=0){
    x=x-(int)x;             //clearing unit
    x=x*10;                 //source of problem (I think) 
    printf("\nX=%f\n",x);   //geting output
    count--;               
  }
  
  fflush(stdin);  
  getchar();     
    
};

The output of this stuff to 157 is:


X=5.699999 //the problem starts right here
//instead of 5.700000 is that...
X=6.999993 //which keeps going enlarging the mistake

X=9.999933

I dont know how comon this is nor how to solve it.
I've been for something like 3 hours around this, can't find solution on google, or compiler settings, kind of going insane now.
Thnaks in advance =)

ps:Can this be a local problem? Caused by the compiler or machine?

[PiMaster]

I think it has to do with how numbers are stored in a float variable. Floating-point variables approximate a number, which is why the numbers are almost what you want, but not quite. Maybe using double would give you the precision you need?

[rocketboy9000]

Don't use floats when you need absolute exactness. For this use ints and the % operator, then reverse the string.

[Boda]

Thanks ppl.
The % and int is really well thought.
But without major changes on the code using double instead of float completely solves it.

Happy holidays. =)