Newsgroups: comp.lang.scheme
Path: cantaloupe.srv.cs.cmu.edu!das-news.harvard.edu!news2.near.net!MathWorks.Com!europa.eng.gtefsd.com!darwin.sura.net!ms!ky.cs.wm.edu!esharr
From: esharr@ky.cs.wm.edu (Earl S. Harris)
Subject: Re: definition of "Continuation"
Message-ID: <1994Sep26.172545.7749@cs.wm.edu>
Sender: news@cs.wm.edu (News System)
Nntp-Posting-Host: ky.cs.wm.edu
Organization: The College of William and Mary
References: <35rutm$a85@news.u.washington.edu> <BASILE.94Sep22174240@rosser.serma.cea.fr>
Date: Mon, 26 Sep 1994 17:25:45 GMT
Lines: 29

In article <BASILE.94Sep22174240@rosser.serma.cea.fr>, basile@rosser.serma.cea.fr (Basile STARYNKEVITCH) writes:
> 
> I'm not sure continuations are formally defined.
> 
> A.Appel in his book "Compiling with Continuations" Cambridge Press
> 1992 just says (p2) that `a continuation is a function that expresses
> "what do to next"'.
> 
> The Revised 4th Report on Scheme has a formal definition of the Scheme
> language, including the call-with-current-continuation primitive.

In the formal definition of Scheme, continuations are a special kind of
function.  Clearly, examples of continuations are presented.  What type
of "formal" definition are you looking for?

> 
> Tons of papers define or talk about "continuation passing style".
> 
> For me continuations = contexts = activation records = closures, but
> 

I don't think continuations are the same as activation records or closurues.
In the formal definition, an environment is more like an activation
record.

For me, continuations is involved in specifying control flow.  It is
like a structured branch.

					Earl Harris Jr.
