From: Christian Neukirchen <chneukirchen@...>
Date: 2006-09-03T23:39:39+09:00
Subject: Re: Integer to byte string - Speed improvements

"John Johnson" <johnatl@mac.com> writes:

> As far as it being faster, dunno, probably. Here's the benchmark:
> irb(main):041:0> Process.times  {
> s=Marshal.dump(0xCC8080808080808080808080808080808080808080808080808080808080AA)
> }
>
> => #<struct Struct::Tms utime=1.02, stime=0.54, cutime=0.0, cstime=0.0>

I at first thought to have missed something, but Process.times doesn't
do what you think it does:

irb(main):005:0> Process.times { p "foo" }
=> #<struct Struct::Tms utime=0.18, stime=0.05, cutime=0.0, cstime=0.0>

The block doesn't even get called. 

     Returns a +Tms+ structure (see +Struct::Tms+ on page 388) that
     contains user and system CPU times for this process.
                                        ^^^^^^^^^^^^^^^^
>   JJ
-- 
Christian Neukirchen  <chneukirchen@gmail.com>  http://chneukirchen.org