Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!uwm.edu!spool.mu.edu!uunet!mcsun!ukc!harrier.ukc.ac.uk!dat
From: dat@ukc.ac.uk (D.A.Turner)
Newsgroups: comp.lang.functional
Subject: Re: thunk's
Message-ID: <7418@harrier.ukc.ac.uk>
Date: 24 Apr 91 15:21:48 GMT
References: <1151@creatures.cs.vt.edu> <2444@m1.cs.man.ac.uk>
Reply-To: dat@ukc.ac.uk (D.A.Turner)
Organization: Computing Lab, University of Kent at Canterbury, UK.
Lines: 24

In article <2444@m1.cs.man.ac.uk> jk@cs.man.ac.uk (John Kewley ICL) writes:

>>I have a question that has been nagging me for quite some time: why do we
>>call closures "thunk"s? ...

>and, are they they the same as wigglies...

Essentially, yes.  The term "wigglies" or "wiggly values" was introduced
by  Christopher  Strachey,  circa  1970,  when  giving an account of the
denotational semantics of Algol-60 call by name.

In verbal presentations he would say something like "I call these values
'wiggly'  because  they can be different each time you look at them." In
the later published account [1] the notation "W"  is  retained  for  the
domain  of  wiggly  values,  although  the reason for the choice of this
letter in the paper was obvious only to those who had  heard  Strachey's
lecture.  His lectures were always very colourful!

In a functional language lazy evaluation does not make  values  "wiggly"
because  there  are  no side effects - but in a procedural language like
Algol-60, call-by-name produces this rather strange effect.

[1]  C.  Strachey  "The  Varieties  of  Programming   Language"   Oxford
University Compyting Laboratory Technical Monograph, PRG-10, March 1973.