Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!wuarchive!uwm.edu!psuvax1!news
From: melling@cs.psu.edu (Michael D Mellinger)
Newsgroups: comp.sys.next
Subject: Re: ML
Message-ID: <ut8H8vnp@cs.psu.edu>
Date: 7 Jun 91 18:09:22 GMT
References: <19722@csli.Stanford.EDU>
Sender: news@cs.psu.edu (Usenet)
Followup-To: comp.sys.next
Organization: Penn State Computer Science
Lines: 22
In-Reply-To: murray@csli.Stanford.EDU's message of 7 Jun 91 14: 59:53 GMT
Nntp-Posting-Host: sunws9.sys.cs.psu.edu


In article <19722@csli.Stanford.EDU> murray@csli.Stanford.EDU (Peter Murray) writes:


   Hello NeXTworld and functional programmers

   Has anyone found/made an ML interpreter/compiler that can be compiled 
   on NeXT?  The NJ-SML doesn't compile and would (it seems to me) need
   significant reconstruction to be made to work.  

   Any ideas?  

It's a problem with the 68040 cache, it used to run under 1.0 on an
030.  I guess they're using self modifying code or something like
that.  Anyway, anyone who is interested might want to send e-mail to
the authors to let them know there is some interest in a NeXT port.  I
asked them a few months back and they said that they would eventually
fix it.  MIT Scheme 7.0 suffered from the same sort of problem.
Wasn't it fixed?  Isn't it just a matter of forcing the cache to be
discarded and then reread at the proper time?

-Mike