I imagine so! In my example, sorting/insertion was fast enough that q was still faster than Jd.
> On Feb 1, 2022, at 5:28 PM, Elijah Stone <elro...@elronnd.net> wrote: > > On Tue, 1 Feb 2022, Vanessa McHale wrote: > >> easy to imagine that selections and joins (on the ordered keys) are going to >> be faster since one can implement them differently. > > I thought k maintains insertion order? I can imagine fast joins for sorted > keys, but not for keys which happen to be in the order they were inserted in. > > -E > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
