Hacker News new | past | comments | ask | show | jobs | submit
How can benchmarks validate the wrong opinion just because it is popular, when they are benchmarking time and memory use?