Yes, it's an oddball. The (unavoidable) trouble with fit/variant is that from the behaviour of f!.a you can not predict the behaviour of g!.b.
----- Original Message ----- From: Don Guinn <[email protected]> Date: Monday, April 6, 2009 5:00 Subject: Re: [Jchat] Operators and Functions To: Chat forum <[email protected]> > Like the"variant" operator. I guess the "fit" conjunction came > from that. > But "fit" has always bothered me in that it will not take a > named left > argument. Nor can it be used with user definitions. It's kind of > an odd-ball > of the primitives. > > On Mon, Apr 6, 2009 at 12:01 AM, Roger Hui <[email protected]> wrote: > > > A digital version of Ken Iverson's seminal 1978 paper, > > "Operators and Functions" is now available on-line: > > http://www.jsoftware.com/papers/opfns.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
