I don't recommend drawing too much of a conclusion from these tests, about anything other than printing in <your language>. Once the algorithm is somewhat opt

I don't recommend drawing too much of a conclusion from these tests, about anything other than printing in <your language>. Once the algorithm is somewhat optimized, the benchmark is completely dominated by printing to stdout, not the algorithm in question. Pulling from the near the conclusion here, bugaevc finds: "And the measured time reflects how long it take to actually compute the solutions (≈1.6s), not how much time it takes to output/buffer them (≈59s)." source: https://github.com/renatoathaydes/prechelt-phone-number-encoding/pull/1 https://github.com/renatoathaydes/prechelt-phone-number-enco...