# http://blog.managedopensource.com
# http://blog.managedopensource.com/2007/10/31/reversal-of-code-performance-optimizations-from-ruby-1-8-to-1-9

# Ruby benchmark output
First run...
      user     system      total        real
for   4.560000   1.400000   5.960000 (  5.962244)
each  4.800000   1.360000   6.160000 (  6.158466)
      user     system      total        real
<<       5.410000   1.450000   6.860000 (  6.855109)
collect  4.980000   1.430000   6.410000 (  6.423387)

Second run...
      user     system      total        real
for   4.570000   1.350000   5.920000 (  5.916540)
each  4.800000   1.370000   6.170000 (  6.165264)
      user     system      total        real
<<       5.550000   1.360000   6.910000 (  6.918924)
collect  4.970000   1.440000   6.410000 (  6.398015)

Third run...
      user     system      total        real
for   4.500000   1.480000   5.980000 (  5.986721)
each  4.810000   1.320000   6.130000 (  6.135532)
      user     system      total        real
<<       5.460000   1.420000   6.880000 (  6.871511)
collect  4.980000   1.420000   6.400000 (  6.392726)

Fourth run...
      user     system      total        real
for   4.490000   1.410000   5.900000 (  5.900563)
each  4.730000   1.400000   6.130000 (  6.131811)
      user     system      total        real
<<       5.560000   1.360000   6.920000 (  6.912522)
collect  4.970000   1.440000   6.410000 (  6.407939)