---Raul Miller wrote: > The expression old + loc * new-old isn't that complicated, > but perhaps Roger will modify the interpreter's treatment > of loc}old,:new to avoid having to buld old,:new?
My suspicion is that the expression old + loc * new-old will be more susceptible to "errors" due to floating point arithmetic than the other numeric method? ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
