Re: Shootout (sum-file)

"Richard Heathfield" <rjh@[EMAIL PROTECTED]
> wrote in message 
news:cpKdne1e4ZCqrr3VnZ2dnUVZ8vmdnZ2d@[EMAIL PROTECTED]
> Bartc said:
>
>> "Richard Heathfield" <rjh@[EMAIL PROTECTED]
> wrote in message
>> news:Z76dnZQYG-54sb3VnZ2dnUVZ8q2dnZ2d@[EMAIL PROTECTED]
>
> <snip>
>
>>> If it's insignificant, why bother to raise it in the first place?
>>
>> Because you were testing a benchmark with such a small runtime that it
>> became significant!
>
> Words fail me.

You'll have to explain. My point is merely that with timings near zero,
and 
with normally tiny overheads becoming significant, performance cannot be 
compared reliably. Such as the C and C++ timings.

>
> Observe:
>
> #include <stdio.h>
>
> int main(void)
> {
>  double u = 0;  /* init velocity */
>  double a = 2;  /* accel due to thrust */
>  double g = -1; /* decel due to earth's gravity */
>  double v = 0;  /* current velocity */
>  double t = 0;  /* time of flight */
>  double h = 0;  /* current height above sea level */
>
>  puts("Launching small coin NOW");
>  do
>  {
>    ++t;
>    v = u + (a + g) * t;
>    h += v;
>  } while(h > 0);
>  puts("Contact re-established with ground.");
>  return 0;
> }

And again. (This coin will accelerate towards outerspace (despite gravity 
staying constant) because it seems to have built-in thrusters.)

-- 
Bartc