Xref: utzoo comp.lang.fortran:5746 comp.lang.c:40330 comp.lang.c++:14293
Path: utzoo!utgpu!watserv1!watmath!mks.com!tslwat!louk
From: louk@tslwat.UUCP (Lou Kates)
Newsgroups: comp.lang.fortran,comp.lang.c,comp.lang.c++
Subject: Re: Running long memory intensive programs!
Message-ID: <654@tslwat.UUCP>
Date: 22 Jun 91 00:26:56 GMT
References: <1991Jun12.203754.9356@noose.ecn.purdue.edu> <1991Jun13.160602.23726@afit.af.mil> <4117@ssc-bee.ssc-vax.UUCP>
Reply-To: louk@tslwat.UUCP (Lou Kates)
Followup-To: comp.lang.fortran
Organization: Teleride Sage, Ltd., Waterloo
Lines: 23

>  Well, we used MS C 5.0 along with PharLap's DOS Extender and
>  things have turned out well.  We support many customers internal
>  to Boeing, and it was not a feasible thing to have to tell our
>  customers that they had to install Unix on their PCs to run our
>  product.
>
>  Check out PharLap's DOS Extender.  It works.
>-- 
>|~~~~~~~~~~       David Geary, Boeing Aerospace, Seattle, WA.       ~~~~~~~~~~|

You could also look into Ergo's DOS  extender. We use MS C and MS
FORTRAN together  with Ergo's DOS extender. With it you write two
processes: a real  mode process that contains DOS dependent stuff
and a  protected mode process that  has everything else. You  use
remote procedure calls between the two which seems a rather clean
way to   do  things.  It uses  the  same API on  both 286 and 386
processors. We have  some large  optimization programs  that  run
under it. We did notice that by switching  to a compiler that can
generate 386 code (the MS compilers can only   generate 286 code)
that we could get quite a speed up.

Lou Kates, Teleride Sage Ltd., louk%tslwat@watmath.waterloo.edu
519-725-0646