Path: utzoo!attcan!uunet!samsung!usc!apple!genbank!agate!saturn!pcg@emerald.cs.aber.ac.uk
From: pcg@emerald.cs.aber.ac.uk (Piercarlo Grandi)
Newsgroups: comp.os.research
Subject: Re: message-passing vs shared memory
Message-ID: <9827@saturn.ucsc.edu>
Date: 18 Nov 89 14:10:59 GMT
Sender: usenet@saturn.ucsc.edu
Organization: Coleg Prifysgol Cymru
Lines: 29
Approved: comp-os-research@jupiter.ucsc.edu


In article <9745@saturn.ucsc.edu> mendozag@ee.ecn.purdue.edu (Victor M Mendoza-Grado) writes:


    I am trying to recall the reference where the set of primitives
    for message-passing and for shared memory mechanisms are shown
    to be equivalent in power.

Not just equivalent in power; actually different ways of doing
exactly the same thing, with exactly the same runtime behaviour
and efficiency. It is even demonstrated how a program written in
one style can be easily converted to the other...

    I that thought maybe someone in the net
    can lend a hand, which I'll appreciate very much.

It is Dr. Bijarne Stroustrup's doctoral dissertation, from
Cambridge.  It has been summarized in some article; I think it
was in SP&E, but maybe you are right in remembering it in a LNCS.

You have not asked my opinion on it, but I will venture to say
that he proves the equivalence of the two things having defined
them in a way such that the only differences are syntactic :->.
The dissertation is very interesting, especially because of the
discussion on performance.
--
Piercarlo "Peter" Grandi           | ARPA: pcg%cs.aber.ac.uk@nsfnet-relay.ac.uk
Dept of CS, UCW Aberystwyth        | UUCP: ...!mcvax!ukc!aber-cs!pcg
Penglais, Aberystwyth SY23 3BZ, UK | INET: pcg@cs.aber.ac.uk