1 T.'' 6 T.&''&>i.0 1 T.'' 7
I think what's happening here is that there's an exploratory execution of T. as a part of getting the type and shape of the empty result. But that mechanism should not spawn new threads. (Conceptually, I think T. should generate an error in that context, which would inform that mechanism to use a fallback to the simplest case result prototype.) Thanks, -- Raul ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
