different alternatives for lispOS

William A. Barnett-Lewis wlewis@mailbag.com
Sat, 26 Apr 1997 22:27:32 -0500

>Date: Thu, 24 Apr 1997 19:21:04 -0500
>To: Richard Coleman <coleman@math.gatech.edu>
>From: "William A. Barnett-Lewis" <wlewis@mailbag.com>
>Subject: Re: different alternatives for lispOS
>At 07:40 PM 4/24/97 -0400, you wrote:
>>Ok.. I've been looking into various alternatives for
>>a lispOS.  Here are some various possibilities, starting
>>with the most ambitious.
>If were going to do it at all, I think either 1 or 2 are the best bets.
>The OS Toolkit provides most of the gnarly elements already done.
Essentially all that remains is to graft the VM of choice on top. As they
point out on their page, they have done this recently with the kaffe JVM.
This approach would allow for easier integration of custom VM/GC or threaded
schedualer sections.
>Going with a stripped down Linux kernel, would give a similar result in the
end. But first you would have to get down to the point where the folks in
Utah already are. Figuring out just what the minimal bootable Linux kernel
is, doesn't seem like fun to me! 
>Whichever is used, might I suggest the Boehm-Weiser collector
<URL:http://reality.sgi.com/employees/boehm_mti/gc.html>? It looks to be the
easiest to integrate into a c/c++ based VM system.
>One other tidbit- I vote for lisp over scheme. 
>>Richard Coleman
William A. Barnett-Lewis
"We are artists.  Poets paint motion  and light.  Historians paint stills.
It can be dangerous to get history from a poet.  It can also be the greatest
						Larry Miller Murdock