I guess this depends on -. e. -. or _. = _. now being 1. IMO if they give _. or 0 instead of 1 will be more reasonable.
That said (as Miller would say), all bets are off (as Hui said). Raul Miller wrote:
In that case, you can ensure whatever result you want. For example: inde=:1 :'u`(_."_)@.(_. e. ,)"0' 1 2 3 <.ind 1 _. 3 1 _. 3 (You could also define inde tacitly, but explicit is a bit more concise.)
-- regards, bill ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
