What I would expect to happen with u/ Y where 1=#Y is: if the neutral element (E) is known then E u Y otherwise a domain error, with the ability to specify E through (!.). I think my mindset is a bit more pure where as the implementation is a bit more practical, albeit, as Pepe said, arbitrary.
On Thu, Sep 11, 2014 at 1:26 PM, Dan Bron <j...@bron.us> wrote: > Pepe wrote: > > My only question is: Does the Dictionary support > > [the behavior that y ? u/y when 1=#y ]? > > No, I don't believe it does. At the very least, it's not supported by the > vocabulary entry for / . That said, there are other areas where such > reasoning applies (e.g. ;. ), so maybe it's governed by higher-level > specifications. > > -Dan > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
