Newsgroups: comp.lang.apl
Path: watmath!watserv2.uwaterloo.ca!mach1!torn!howland.reston.ans.net!zaphod.mps.ohio-state.edu!cs.utexas.edu!uunet!s5!hui
From: hui@fid.morgan.com (Roger Hui)
Subject: Re: Seek examples of lev and dex
Message-ID: <1993Apr24.174008.11891@fid.morgan.com>
Organization: Morgan Stanley & Co., New York, NY
Date: Sat, 24 Apr 1993 17:40:08 GMT
Lines: 18

Eythan Weg wrote on 1993-04-12:

> Try
>    s1=.'%&x. @ (] - &y. -&x.)':22
>
> I see this translator also as a teacher.  I often wondered if its
> scope is exhaustive.  If it is, one does not need to learn the
> semantics of trains of conjuctions and/or adverbs.  A related question
> has to do with the quality of the translation, which may bear on the
> previous conclusion.  Is it optimal in some sense?

The :21 and :22 translators (from explicit adverbs and conjunctions 
to tacit ones) are not exhaustive.  For example, they assume that the 
arguments x. and y. are verbs, and giving noun arguments to the 
translations may yield incorrect answers.  The translations are 
not optimal, typically having an extra level of depth (through the 
use of [. and ].); as well, sometimes there are global simplifications 
that the translators would miss in their local approaches.
