Xref: utzoo comp.edu:2905 comp.lang.ada:3169 comp.lang.misc:3882
Path: utzoo!attcan!uunet!clyde.concordia.ca!rutgers!uwm.edu!rpi!zaphod.mps.ohio-state.edu!usc!apple!fox!portal!cup.portal.com!pgl
From: pgl@cup.portal.com (Peter G Ludemann)
Newsgroups: comp.edu,comp.lang.ada,comp.lang.misc
Subject: Re: Teaching Concurrency
Message-ID: <25876@cup.portal.com>
Date: 13 Jan 90 07:34:59 GMT
References: <7588@hubcap.clemson.edu> <602@agcsun.UUCP>
  <10330@june.cs.washington.edu> <5598@bd.sei.cmu.edu> <6219@ubc-cs.UUCP>
Organization: The Portal System (TM)
Lines: 15

In my experience, most "multi-tasking" turns out to be nothing
but coroutining with message-passing for communication.

On an AS/400 (for example), a task switch is only slight more expensive
than a subroutine call (basically the cost of saving the 16 registers
plus a return address), so the additional clarity of writing with
multi-tasking is well worth it.

Furthermore, in some dialects of Prolog (e.g., NU-Prolog), a
corouting style can be used to vastly INCREASE the efficiency
of programs (the programs are written in a multi-tasking style but
are actually run as coroutining).

- peter ludemann  	pgl@cup.portal.com
	--- my opinions are my own responsibility, of course ---