Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!elroy.jpl.nasa.gov!sdd.hp.com!mips!pacbell.com!ucsd!sdcc6!crl!hartung
From: hartung@crl.ucsd.edu (Jeff Hartung)
Newsgroups: comp.binaries.ibm.pc.d
Subject: Re: 4DOS 3.02 and MS Windows 3.0
Keywords: 4DOS Windows problems
Message-ID: <18117@sdcc6.ucsd.edu>
Date: 9 Apr 91 16:19:27 GMT
References: <18059@sdcc6.ucsd.edu> <1991Apr8.201757.7530@menudo.uh.edu>
Sender: news@sdcc6.ucsd.edu
Organization: University of California, San Diego
Lines: 54

In article <1991Apr8.201757.7530@menudo.uh.edu> st12a@menudo.uh.edu (richard henderson~) writes:
>In article <18059@sdcc6.ucsd.edu> hartung@crl.ucsd.edu (Jeff Hartung) writes:
>>                                      One problem which I had been aware of is
>>that 4DOS does not like the version of HIMEM.SYS required to run Windows 3.0,
>>and will cause your system to crash HARD if you allow it to try to swap to
>>extended memory...
>
>That is odd, as I am running HIMEM.SYS v2.06 and do not have this problem.
>If you have an older version of HIMEM, perhaps you should upgrade.

The pre-Windows3 version(s) of HIMEM.SYS will not work with windows.
Windows will complain that the driver is outdated and tells you to
install _its_ HIMEM driver.

Many other responses to my initial article were posted and/or sent
e-mail.  Several people reported never having seen the problem.  I have
seen it show up on two different machines (a 12 MHz 286 and a 20 MHz
386).  I did leave out one fact which was sort-of implied, but not
directly stated.  The problem occurs when I try to load 4DOS as the
primary command shell _and_ install SmartCache that comes with Windows
3.0.  SmartCache loads first (no matter how one arranges the sequence
of lines in the CONFIG.SYS file), then 4DOS tries to use the same
memory, and the first attempt to access the swapped memory that is
supposed to be in extended memory by 4DOS causes the system to halt.
The conflict seems to arise from the way in which Windows uses extended
memory.  As I had said earlier, the solution (for me) was to tell 4DOS
_not_ to use extended memory, either by making it swap to disk (/S:D)
or not to swap at all (/S:N).

The other problem, that SETUP doesn't run properly, is not a major
problem since, as was pointed out by someone, it isn't a frequent
operation, but when I _do_ have to run it, it's nice to know that there
is a way it can be done properly.  I couldn't even get Windows 3.0 to
install (unless I bypassed the scan-disk-for-applications portion of
the setup procedure) on my home 286 until I replaced the
'shell=4dos.com...' line with a 'shell=command.com...' line in my
CONFIG.SYS.  Others have reported similar problems to me since my
initial article.  It may be bios specific (AWARD), or it may be
something else.  Fortunately, hardware changes (like changing the
display mode from 640x480x256 to 800x60x16) _will_ run properly no
matter which command interpreter is running.

Again, I think (especially since I heard this from others) that the
problem is real, and not due to "something else," as has been
suggested, but I am willing to acknowledge that the conflict may
involve a particular hardware/ bios/device driver combination.

Comments?

-- 
 --Jeff Hartung--  	
 Disclaimer: My opinions only, etc., etc., BLAH! BLAH! BLAH!...
 InterNet - hartung@crl.ucsd.edu    or    ps299bx@igrad1.ucsd.edu
 UUCP - ucsd!crl.ucsd.edu!hartung	 BITNET - hartung@ucsd