All the results get the same speed if I run the test repeatedly, it is likely hot optimizing it, and because it is so simple it is quite efficient to optimize
— Jsperf can be sort of nondeterministic though
— Same argument for the function, monomorph
— That was buggy 0-1 index
— So.. chrome was faster in push long time ago, others did not.. still there is assignment thing and overwrite thing that push cant do and it will be slower in that situations
Message permanent page
— A = b[b.length] = ..
— A[++i] = ..
— All the same, except cached variant🤤
— Whenever diff is around 1% I don't give a gosh darn diddly