Path: utzoo!attcan!uunet!samsung!uakari.primate.wisc.edu!uflorida!haven!wam!walrus
From: walrus@wam.umd.edu (Udo K Schuermann)
Newsgroups: comp.sys.amiga
Subject: Re: Distributed Amiga processing. Is it possible?
Message-ID: <1990Mar26.192906.18583@wam.umd.edu>
Date: 26 Mar 90 19:29:06 GMT
References: <595@ucsvc.ucs.unimelb.edu.au> <10346@cbmvax.commodore.com>
Sender: usenet@wam.umd.edu (USENET Posting)
Reply-To: walrus@wam.umd.edu (Udo K Schuermann)
Organization: University of Maryland at College Park
Lines: 24

In article <10346@cbmvax.commodore.com> valentin@cbmvax (Valentin Pepelea) writes:
>In article <595@ucsvc.ucs.unimelb.edu.au> U3364521@ucsvc.ucs.unimelb.edu.au
>(Lou Cavallo) writes:
>>Does/should the Amiga OS have the potential/capability in future for distri-
>>buted processing?
>
>There are several problems involving this. One is that we need device 
>independent graphics for distrubuted processing to work. How else could
>we run an GUI based application on a remote machine.
>
>Another problem is that the usual vm/protection/multi-processing MEMF_PUBLIC
>memory flag definition is unused. What if a program running on a remote machine
>wants to send a message to the local machine? How do you do that?

If you intercepted all calls from a particular process to the system
routines (fairly costly) you could redirect them to the other system.
     I don't grasp the full extend of this job, but couldn't screen,
window, and port addresses be used to determine ownership, transmitting
them as messages to the remote machine where they'll act as if locally
issued?
     Maybe we should continue this in .tech ...

Udo Schuermann
walrus@cscwam.umd.edu