Bob: And now it is time for Alice to say that it is time to test our codes, to see whether they still all give the same output.

Alice: Let it be said.

Bob: Three particles should be enough to check:

 |gravity> ruby mkplummer.rb -n 3 -s 33 > plum3.in
 ruby: No such file or directory -- mkplummer.rb (LoadError)

 |gravity> time ruby nb1.rb -t1 -d0.01 -s0.1 < plum3.in
 /home/makino/papers/acs/lib/clop.rb:310: warning: already initialized constant HELP_DEFINITION_STRING
 ==> The simplest ACS N-body code <==
 Integration method             : rk4
 Integration time step          : dt = 0.01
 Diagnostics output interval    : dt_dia = 1.0
 Snapshot output interval       : dt_out = 1.0
 Duration of the integration    : dt_end = 1.0
 Softening length               : eps = 0.1
 ./nbody1.rb:167:in `/': divided by 0 (ZeroDivisionError)
        from ./nbody1.rb:167:in `write_diagnostics'
        from ./nbody1.rb:86:in `evolve'
        from nb1.rb:144
 0.025u 0.001s 0:00.02 100.0%   0+0k 0+0io 0pf+0w

 |gravity> time ruby nb2.rb -t1 -d0.01 -s0.1 < plum3.in
 /home/makino/papers/acs/lib/clop.rb:310: warning: already initialized constant HELP_DEFINITION_STRING
 ==> The simplest ACS N-body code <==
 Integration method             : rk4
 Integration time step          : dt = 0.01
 Diagnostics output interval    : dt_dia = 1.0
 Snapshot output interval       : dt_out = 1.0
 Duration of the integration    : dt_end = 1.0
 Softening length               : eps = 0.1
 ./nbody2.rb:180:in `/': divided by 0 (ZeroDivisionError)
        from ./nbody2.rb:180:in `write_diagnostics'
        from ./nbody2.rb:88:in `evolve'
        from nb2.rb:144
 0.025u 0.000s 0:00.02 100.0%   0+0k 0+0io 0pf+0w

Here is your prettified brevity-rules version:

 |gravity> time ruby nb3.rb -t1 -d0.01 -s0.1 < plum3.in
 /home/makino/papers/acs/lib/clop.rb:310: warning: already initialized constant HELP_DEFINITION_STRING
 ==> The simplest ACS N-body code <==
 Integration method             : rk4
 Integration time step          : dt = 0.01
 Diagnostics output interval    : dt_dia = 1.0
 Snapshot output interval       : dt_out = 1.0
 Duration of the integration    : dt_end = 1.0
 Softening length               : eps = 0.1
 ./nbody3.rb:180:in `/': divided by 0 (ZeroDivisionError)
        from ./nbody3.rb:180:in `write_diagnostics'
        from ./nbody3.rb:88:in `evolve'
        from nb3.rb:144
 0.032u 0.008s 0:00.05 60.0%    0+0k 0+0io 0pf+0w

Bob: But I'm not happy about the small speedup. Maybe having three particles is too small a number. Let's revisit the one time step that we did with 256 particles. Or better both one and two time steps, given that the first time step had some extra overhead.

With the earlier version we had:

 |gravity> (time ruby nb1.rb -t0.01 -d0.01 -e0.01 -s0.1 -o2 < plum256.in) |& tail -2
        from nb1.rb:144
 0.038u 0.001s 0:00.03 100.0%   0+0k 0+0io 0pf+0w

 |gravity> (time ruby nb1.rb -t0.02 -d0.01 -e0.01 -s0.1 -o2 < plum256.in) |& tail -2
        from nb1.rb:144
 0.023u 0.002s 0:00.02 100.0%   0+0k 0+0io 0pf+0w

 |gravity> (time ruby nb3.rb -t0.01 -d0.01 -e0.01 -s0.1 -o2 < plum256.in) |& tail -2
        from nb3.rb:144
 0.023u 0.003s 0:00.02 100.0%   0+0k 0+0io 0pf+0w

 |gravity> (time ruby nb3.rb -t0.02 -d0.01 -e0.01 -s0.1 -o2 < plum256.in) |& tail -2
        from nb3.rb:144
 0.024u 0.002s 0:00.02 100.0%   0+0k 0+0io 0pf+0w

 require "profile"

 |gravity> time ruby nb3p.rb -t1 -d0.01 -s0.1 < plum.in
 plum.in: そのようなファイルやディレクトリはありません.
 0.000u 0.000s 0:00.00 0.0%     0+0k 0+0io 0pf+0w

Now all three:

 |gravity> (ruby nb1p.rb -t1 -d0.01 -s0.1 < plum.in) | & head -25 | & tail -10
 plum.in: そのようなファイルやディレクトリはありません.

 (ruby nb2p.rb -t1 -d0.01 -s0.1 < plum.in) | & head -25 | & tail -10

 (ruby nb3p.rb -t1 -d0.01 -s0.1 < plum.in) | & head -25 | & tail -10