I understand all the fallacies of this question, but...
Well, I was a little curious to see what would happen if I made some deductions
about the purpose of the question did try to benchmark the following:
j = 0
benchmark "function":
for i in countup(1, x):
inc j
vs
j = 0
benchmark "operator":
for i in 1..x:
j = j + 1
The following code is only for without compiler optimization and on platforms
that support [rdtsc](https://en.wikipedia.org/wiki/Time_Stamp_Counter):
[https://gist.github.com/lbmn/d424dd0c8c85fa195743356fc133235a](https://gist.github.com/lbmn/d424dd0c8c85fa195743356fc133235a)
So now there's a new way to flip a coin!
