alexandre.ferrieux@xxxxxxxxxx <alexandre.ferrieux@xxxxxxxxxx> wrote: > > Because when this was implemented "highly asynchronous" was not on the > > radar. All users of this (that I know of) do in-order verdicts. > > So, O(N) instead of O(1) just because "I currently can't imagine N>5" ? Seems so. THis code was written 21 years ago. > Would a patch to that effect be rejected ? Probably not, depends on the implementation.