NetNews Usenet Archive 1992 #31/NN_1992_31.iso/spool/comp/sys/ibm/pc/digest/184

Internet Message Format | 1992-12-21 | 34.7 KB

Path: sparky!uunet!zaphod.mps.ohio-state.edu!cis.ohio-state.edu!ucbvax!WSMR-SIMTEL20.ARMY.MIL!Info-IBMPC
From: [e-mail] ("Info-IBMPC Digest")
Newsgroups: comp.sys.ibm.pc.digest
Subject: Info-IBMPC Digest V92 #194
Message-ID: <[e-mail]>
Date: 21 Dec 92 10:55:41 GMT
Sender: [e-mail]
Reply-To: [e-mail]
Organization: The Internet
Lines: 858
Approved: [e-mail]
X-Unparsable-Date: Sat, 19 Dec 92 11:28:01 GMT+1

Info-IBMPC Digest           Sat, 19 Dec 92       Volume 92 : Issue 194

Today's Editor:
         Gregory Hicks - Rota Spain <[e-mail]>

Today's Topics:
                     A Win 3.1 beta success story.
                            Email to Prodigy
                    EMM386 (DOS 5) Exception codes?
                            EMM386 confusion
                            Flow Diagram sw
    HELP: my operating system is possessed!!! (NO KIDDING!) (2 msgs)
                    Help Needed w/OS/2 Service Pack
          Is Shadow RAM a motherboard or a bios feature? Help
         More Than You Wanted to Know About Distinctive Ringing
               MSDOS Descriptor Table Limitation Problem
                MSDOS virtual-file-device device drivers
              Optimal Settings for Communications (2 msgs)
       Specific DOS (3.31), CD ROM, and PKUNZIP => Crash City (!)
  The SP, Diamond SpeedStar (ET4000), and Higher Resolutions (2 msgs)
                    Watch dog timer support on ISA?

Send Replies or notes for publication to: <[e-mail]>

Send requests of an administrative nature (addition to, deletion from
the distribution list, et al) to: <[e-mail]>

Addition and Deletion requests for UK readers should be sent to:
<[e-mail]>

Archives of past issues of the Info-IBMPC Digest are available by FTP
ONLY from WSMR-SIMTEL20.ARMY.MIL in directory PD2:<ARCHIVES.IBMPC>.

----------------------------------------------------------------------

Date: 2 Dec 92 02:34:14 GMT
From: Rahul Verma <[e-mail]>
Subject: A Win 3.1 beta success story.

Hi! I bring good tidings to many of you. I have finally succeeded (it
took many reformats) to get the GA + SP + Win 3.1 + 800x600x256 +
seamless in one ball of wax. My graphics card is a Diamond Speedstar on
an 8mb 486/33. The good news is that I got word for windoze, pagemaker,
tetris for windoze, clock and Excel all running seamlessly with only a
4mb swap file!!!! It was fast. I don't know how but it was fast. The
steps I followed were as below:

1. Install The service pack. Reboot

2. Immediately after this, install the win 3.1 beta as on the beta. let
it switch you back to standard vga if it wants.

3. before rebooting, re install the service pack again.  and install
the new hires drivers again.

4. reboot.

5. edit your path in the autoexec.bat file.

6. edit your system.ini file to reflect the changes that the service
pack made in the win 3.0 files. there are 2 changes, display.drv and
svgasomething.drv and those lines should be copied from the new 3.0
.ini file. Please note that before hand win 3.0 has to be installed for
this to work.

    After this, you should have a fully working 3.1 system with
everything works. I stole most of this stuff from Alok Dhir. Thanks
Alok.

Good luck for the rest of you. MOST IMPORTANT, DON'T FORGET TO BACKUP
THE DRIVE!!!

Rahul Verma            |    This message made of 100% recycled electrons.
[e-mail]          |

------------------------------

Date: 29 Nov 92 16:49:55 GMT
From: Steve Steinberg <[e-mail]>
Subject: Email to Prodigy

[e-mail] (David Baird) writes:

> [e-mail] (Louis Schmittroth) writes:

>> I want to find out if Prodigy offers service to Alberta, and if so
>>is there a email address of an administrative unit of Prodigy where I
>>can find out what the rates are?

> The last I knew the people who ran Prodigy were not hooked into any
>other network, nor were they going to be in the future. It seems that
>either one was on Prodigy to communicate with a Prodigy user, or one
>did not communicate with the Prodigy user using Prodigy's service. Or
>to put it another way, there are no gateways between Prodigy and
>Compu$erve or the InterNet.

> Give Prodigy a call and complain about this situation. It is one of
> two primary reasons why I never signed on to Prodigy.

Prodigy _is_ planning an Internet gateway that should be available by
the end of the year.  Evidently they will demand $20/mo. for the
priviledge of e-mail. I have no idea of what their Internet address is
now. Did you try [e-mail]? Maybe info or admin or
postmaster??

=== Steve Steinberg   ==  [e-mail]  == {cmcl2,apple}!panix!ss ===

------------------------------

Date: 3 Dec 92 17:49:29 GMT
From: Tapio Sand <[e-mail]>
Subject: EMM386 (DOS 5) Exception codes?

[e-mail] (Sheldon E. Smith) writes:

>Are the Exception codes DOCUMENTED anywhere?  I've checked a few book
>stores, but none of the computer books have *anything* about EMM386
>codes.

I picked the following from a discussion last summer:

The following is a list of Intel 80386 protected mode exception errors
and their names:

Code    Meaning
----    -------
0       Divide error
1       Debugger interrupt
2       Nonmaskable interrupt
3       Breakpoint
4       Overflow interrupt
5       Array boundary violation
6       Invalid opcode
7       Coprocessor not available
8       Double fault
9       Coprocessor segment overrun
10      Invalid task state segment
11      Segment not present
12      Stack exception
13      General protection violation
14      Page fault
16      Coprocessor error

All of this was prefaced by the comment "This error should not occur on
systems that are 100 percent IBM compatible.  Because of the hardware
intensive nature of the EMM386 device driver, strict IBM hardware
compatability is required."

[e-mail]
TEX disconnects, dial again for another session.
D;+vC)?iH%"
NO.CARRIER

------------------------------

Date: 4 Dec 92 17:51:11 GMT
From: Richard Glew <[e-mail]>
Subject: EMM386 confusion

[e-mail] (S D Law) writes:

>I am having problems understanding exactly how emm386 works when used
>to hiload dos drivers.  I have a machine with a Western Digital 8003E
>ethernet card that runs pc-nfs.  The I/O base address is set to 280 and
>the Shared memory address to D000 (Hex).  which is between 640-1mb.  If
>I use mft from qemm to check out what memory is used in this upper
>memory erea I find that D000 is not clashing with anything else.  When
>I startup EMM386 from config.sys which comes with dos5 (I include the
>noems option) I find that the network on boot up hangs intermitently.
>This I assume is due to the D000 address clashing with something else.
>I have been told to exclude this address using the x option but have
                                          Correct   ^^^^^^^^

>also been told not to bother as I have included the noems option.  The
    Incorrect   ^^^^^^^^^^^^^

NOEMS means that emm386.exe does not set the 64k EMS page frame
(default is d000) which then allows EMS emulation with a portion of
your system's extended memory.  It has nothing to do with network card
page frame locations.

>hanging of the machine happens even if I am not hiloading anything,
>just starting up emm386.  Can anyone tell me what its doing and if I am
>right in assuming that the shared memory address is the culprit (another
>machine has the 3com 3c503 card running in it and it works with emm386
>because, I assume, it doesn't ask for a shared memory address thus
>doesn't use one).

>If I could get it going can anyone tell me if its possible to merge some

Assuming:
    1. you are using VGA
    2. you have a typical AT clone
    3. your WD ethernet card requires a 16k window

then set your 16k window at either c800-cc00 or ec00-efff

The c800 window sits in the area just above VGA, and the ec00 window
sits right up against the F segment.  Either way will leave you with a
single contiguous area for loading drivers high.

device=c:\dos\himem.sys
device=c:\dos\emm386.exe x=c800-cc00 i=e000-efff noems
dos=high,umb

|<---- VGA ---->|<- WD ->|<--------- Open --------->|<--- ROM --->|
A000            C800     Cc00                       F000

OR

device=c:\dos\himem.sys
device=c:\dos\emm386.exe i=e000-ebff x=ec00-efff noems
dos=high,umb

|<---- VGA ---->|<------- Open --------->|<- WD ->|<--- ROM --->|
A000            C800                     EC00     F000

* Don't forget to rework your network card setup, and to use
appropriate settings for your WD packet driver.

>of the memory blocks as there are not enough contiguous blocks to
>hiload the 70K pcnfs.sys driver.

------------------------------

Date: Tue, 08 Dec 92 10:00:40 PST
From: John Wical <[e-mail]>
Subject: Flow Diagram sw

I'm looking for DOS software that will diagram data flows.  Does anyone
know where I can find this?

John Wical
Loma Linda University Medical Center

------------------------------

Date: 3 Dec 92 02:37:46 GMT
From: Dave Oliphant <[e-mail]>
Subject: HELP: my operating system is possessed!

Last night I ftp'd a file from ftp-os2.nmsu.edu called "pspm.zip" or
something like that.  It purported to be a program to display the
different processes going on under OS/2, dos, and windoze.  The one
program called pspm2.exe seemed to work fine, but the one called
pspm.exe has taken over my machine.  I rebooted.  It came up as before,
still in control of my system.  I tried to selective reinstall from
scratch (I did not reformat the hard drive) but the offending process
was still there.  Since I only just installed OS/2 last Saturday for the
first time (actually, I have gone through the whole, complete process of
installation including reformat of hard drive three times now) I don't
really have a lot on the hard drive I need, except about 5MB worth of
stuff I downloaded yesterday and today, and which I don't want to throw
away if I don't have to.

Is there any way, short of reformating my hard drive, that I can regain
control of my system?

This is very upsetting, to say the least!

------------------------------

Date: 3 Dec 92 16:25:24 GMT
From: Bennie Harris <[e-mail]>
Subject: Re: HELP: my operating system is possessed!

[e-mail] (Dave Oliphant) writes describing problems with a
newly acquired program and OS/2.  (see previous message.)

Try rebooting from your installation disk (or boot disk if you have
one) and add the line SET RESTARTOBJECTS=STARTUPFOLDERSONLY to the end
of your config.sys. Reboot .

The application should not restart after you reboot.

I suggest getting a copy of the latest OS/2 FAQ from ftp-os2.nmsu.edu
in the /pub/os2/all/info/faq directory. The latest Rev is 2.0h. You may
also be able to find it in the /pub/uploads directory. Question (28)
explains the  SET RESTARTOBJECTS=STARTUPFOLDERSONLY line.

-- Bennie Harris

------------------------------

Date: 3 Dec 92 10:44:30 GMT
From: Lee Clarke <[e-mail]>
Subject: HELP: my operating system is possessed!

Call hobbes and get procs.exe and kill.exe. Boot from installation disk
1, then two, then escape when it tells you that you may. Run procs to
get the name of the process, the kill it. OS2 _does_ have a terminal
program. Look in the productivity folder.

------------------------------

Date: 2 Dec 92 03:02:57 GMT
From: Jason Ari Goldstein <[e-mail]>
Subject: Help Needed w/ Service Pack!!!

Hey now,

I am having a problem installing the service pack.  Everything is going
fine until it gets to disk 12 and then I get the following error:

    Error while completing user exit
        A:\FIX\OS2\PRODUCT.EXE

Does anyone know how to fix this.  It won't go past this point I just
get the red fatal error screen and all I can do is check the log, get
help, and reboot.  I really need to use my machine, this is my last
week of class and I have a lot of stuff to do.

Please send a copy of all responses so that I can take care of this.  I
am not sure how much of calling in at 1200 baud I can take so hurry!

Thanks in advance.

BTW, I have a 386DX/387 Clone with 8M RAM and 160M ESDI Drive
(Micropolis).  I am also installing off 3 1/2' disks and I have tried a
different disk 12 already.

Later...
Jason A. Goldstein

I'm a peripheral visionary...
I can see the future but only off to the sides.

Over, Finished, Gone, Done, Out.  (Finally)

------------------------------

Date: 30 Nov 92 17:00:27 GMT
From: Pradeep K Tapadiya <[e-mail]>
Subject: Is Shadow RAM a motherboard or a bios feature? Help

Greetings,

I have a 386sx with AMI bios on it. QEMM's analysis indicates that it
could get only 144K out of 384K available Shadow RAM. The 240K which is
lost is almost equal to the sum of video RAM, video ROM, disk ROM, and
BIOS ROM. I am desperately trying to find a way to get back the lost
240K of memory.

The BIOS setup lets me enable or disable shadow RAM/ROM.  No matter
what option I use, I still loose 240K of shadow RAM.  It seems to me:

1. The way (my?) motherboard is designed, there is always 384K of
Shadow RAM (provided I have atleast 1M of memory).

2. If there is any bios RAM/ROM detected, the corresponding memory RAM
region is automatically set aside for shadowing.

3. If we enable shadowing via BIOS, this region is used.  If we disable
shadowing via BIOS, this region is unaccessible and is wasted.

   The bottom line is, the more board RAM/ROM I have, the more memory
RAM is wasted if I disable shadowing.

Is this logic true? Or am I missing something?

Another question: QEMM manual says that it recognizes Shadow RAM if the
chips are manufactured by so and so companies. To my understanding,
memory chips are passive elements (besides address decoding part). What
exactly is the role of a memory chip in shadowing? Do I need to replace
my SIMMs to remove shadowing completely?

Thank you for your help.

Pradeep
[e-mail]

------------------------------

Date: 1 Dec 92 00:12:15 GMT
From: "John R. Levine" <[e-mail]>
Subject: More Than You Wanted to Know About Distinctive Ringing

[This note is from the TELCOM Digest, a moderated group on USENET.
Because of the numerous users that have more than one piece of
equipment on a single line, and given the difficulties figuring out
WHICH piece should answer the ring, I thought is was appropriate to
include in the Info-IBMPC Digest.  gph]

Lots of responses this time, as well as finding some old messages I'd
lost.  There are a few ??? still to be filled in.

 -- What is Distinctive Ringing?

It is a telephone service that assigns several different phone numbers
to the same line.  Its true name is "Multiple Directory Numbers Per
Line With Distinctive Ringing".  Each number rings in a different
cadence (single, double, and triple ring) so you can tell which number
the caller called.  The maximum number of numbers per line varies by
telco, but can be as high as four.  You can think of it as a party line
where all the parties are you.

The price is usually quite low: around here it's $3/month for the
second number and $2/month for the third.  You generally have the
option of listing the extra numbers or not.  (Note to Massachusetts
customers: NET orignally said I had to pay extra to have them unlisted,
but the DPU persuaded them that they were mistaken.)

 -- What is it good for?

All sorts of things.  One use is for people with home businesses, so
you can answer one ring "Hello" and the other "Thank you for calling
Bagel-tronics."  Another is to share a single line among several
devices such as faxes and modems.  For this use, you'll want a ring
leader, below.

It is also useful as a way to defeat Caller ID.  The C-ID number sent
on outgoing calls is always the first number so you could assign the
first number to your modem, or let an answering machine pick up calls
to the first number and tell your friends to call the second number.

Some allege that it's useful to distinguish between calls to parents
and calls to teenage children, but given the way teenagers use the
phone (call every possible number where a friend might be and talk for
hours) it's no substitute for a second line.

 -- Can I use it as a fax switch?

That's what I use it for.  I find it works better than the usual fax
switches.  It doesn't depend on answering the phone and listening for
fax tones which not all faxes generate, so it never guesses wrong.
Besides, you get a separate fax number which looks much more official.

 -- What happens if I also have call waiting?

When a call comes in, the beep is in the same pattern as the ring, so
you know which number it is.  At least, that's what's supposed to
happen.  Apparently, they sometimes forget to set up the beeps right.

 -- How do I order it?

You call up the business office, of course.  For some reason, each
telco gives it a different name.  The ones I know are:

NYNEX        Ring Mate
Bell Atlantic    Identa Ring
Southern Bell    Ring Master
Ameritech    ???
SW Bell        Personalized Ring
US West        Custom Ringing
Pac Tel        not available, see below
GTE        Smart Ring

Technically, distinctive ringing can be installed on 1A and newer AT&T
exchanges and other modern units.  A software upgrade is required (it's
almost but not quite the same as a party line) so there are many areas
in which it's still not available even though the exchange would seem
to support it.

Note for Pac Bell customers: Pac Tel has a service which they call
"distinctive ringing" which is completely different.  Their service is
a home centrex which rings differently for inside and outside calls.
If you try to order distinctive ringing, be sure you know what you're
getting.

 -- Can I automatically connect to different devices for different
rings?

Yes.  That's what a "ring leader" does.  It is a box that has a modular
cord that plugs into the phone line and several modular jacks into
which the devices plug.  When the phone rings, it listens to the first
ring cycle and then connects to one of the devices depending on which
ring pattern it was.  The ring leader itself never answers the phone --
it just connects to a phone, fax, modem, etc., which answers the call
normally.

For outgoing calls, ring leaders act as exclusion units and only let
one of the devices connect to the line at a time.  Excluded devices
hear either a busy signal or a silence.

Many vendors sell ring leaders.  Here's a summary of the ones I know
about, along with the names of the people providing reports.

I have an Autoline Plus fro ITS in Endicott NY. [phone].  It
connects up to three devices.  Mine has been entirely reliable.  I paid
about $80 but the price is more like $120 now.  (John Levine
<[e-mail]>)

Misco now has a unit that will do just that.  It is called the Ring
Decipher (misco part #fl-3622).  The only other identification in the
ad is "ASAP RD-4000".  I'd like to know who makes it, and alternate
sources.  Supposedly, it will provide a standard ring signal on any one
of the four outputs determined by the incoming ring pattern.  Price
$99.  (Bill Petrisko <[e-mail]>)

My first Lynx Automation box didn't work but their beta replacement
worked.  Four pattern model.  I bought another one that my brother had
problems with but I haven't hooked it up here yet so I dunno.  I notice
that Lechmere [local discount department store] is selling the two line
Lynx at a nonoutragous price ($69 or $79 which is about what Lynx
charges direct.  (Bob Frankston <[e-mail]>)

I use a ASAP RD4000 Ring Decipher made by Command Communications Inc of
Aurora CO.  It cost $100 at a trade show.  It decodes four different
patterns.  When one of the devices has the line, the other devices get
a busy signal if they try to access the box.  It has worked flawlessly
for the last nine months.

I had a bad experience with a box called RingMaster made by Lucas
Technologies of Beacon NY.  It failed repeatedly and the maker wouldn't
do anything about the problems until I cornered them at a trade show
and made a large noise.  They bought the box back for $100.  I had only
paid $80!  That was the only good experience with Lucas' RingMaster.
(John Adams <[e-mail]>)

The call route box (avail from Home Automation Lab at 1-800-HOMELAB) or
the RD1000 from mailorder (this is the one I have) doesn't even pass
ring voltage to the connected lines until after the end of the first
ring and it has determined which number was called; therefore, you
never even hear the phone ring unless they are calling the authorized
number.  (Carl Neihart <[e-mail]>)

Black Box's September 1992 catalog lists a product called DRD-4 that
automatically routes distinctive ring services from one incoming phone
line to up to four devices. $ 139. Phone [phone].  (Marc Kozam
<mlksoft![e-mail]>)

 -- Can I build my own ring leader?

Probably.  Several people expressed interest in coming up with a
design, but nobody's told me about one yet.

Thanks to:

[e-mail] (William Petrisko)
[e-mail] (Paul Barnett)
[e-mail] (Lars Poulsen)
[e-mail] (Rich Greenberg)
[e-mail] (William Petrisko)
[e-mail]
[e-mail] (John Adams)
[e-mail] (Terry Kennedy)
[e-mail] (Ted M A Timar)
[e-mail] (Carl Neihart)
[e-mail] (Michael Masterson)
"Wm. Bryant Faust, IV" <[e-mail]>
[e-mail] (Carl Neihart)
mlksoft![e-mail]
Steve Forrette <[e-mail]>

[TELCOM Moderator's Note: All those names can be confusing. Ameritech
(Illinois Bell, at least) says 'distinctive ringing' is when your CO
has been advised of up to ten telephone numbers which, when they call
you are to be given the red carpet; ie, they, and they alone are to
cause your phone to ring with a special cadence to let you know (for
example) the boss is calling, or your parents, etc.  On the other hand,
'Multi-line' is the service discussed in John's article where more than
one number is assigned to a single line with different ringing cadences
as appropriate.  'Starline' is IBT's 'home centrex' service which
provides a different ring for calls from within and without your
premises.  I use my 'Multi-line' service as a way for my 800 numbers to
ring in. The two short rings tell me it is my nickle paying for the
call.  :)    PAT]

------------------------------

Date: 3 Dec 92 14:56:42 GMT
From: "David L. April x5649" <[e-mail]>
Subject: MSDOS Descriptor Table Limitation Problem

I'm a fairly new MSDOS & MS Windows user and I'm not use to being
limited in certain types of memory usage.  My problem is the following:
I'm creating a linked list based on the size of information coming into
my program.  My function gets called each time a new node is added,
which in this case could result in a large linked list structure
(possibly thousands of nodes).  After many GlobalAlloc's and
GlobalLock's I eventually run out of space in the descriptor table.

I've though about allocating a large block of memory and extracting
portions of this memory to be used in each new entry in the linked
list, until such time I need more memory...  but I'm not really sure
how to handle this in the windows development environment.  It has been
recommended that I don't do selector arithmetic...

Any suggestions as well as some sample code would be greatly
appreciated.

Note:  Due to certain reason out of my control, I am running this code
in a large memory model.

David April
[e-mail]

------------------------------

Date: 2 Dec 92 08:31:00 GMT
From: Greg Huey - Aymer deGregory <[e-mail]>
Subject: MSDOS virtual-file-device device drivers

  I'm trying to write a device driver for a virtual block device, but
one that operates at a high level. I want MSDOS to handle the memory
and process management, etc, but use my code to access files, including
file access thats done 'internally'. I emphasize my virtual device is a
virtual file system -- MSDOS has no clue as to its organization below
the 'file & directory' level.  This is to say, MSDOS should never try
to read a specific sector...it should instead send a request to my
device driver to read some portion of a file.

  One might say 'just intercept the handle functions sent by the user
via interrupt x21' -- but remember, this system must work if the user
does an exec (intr x21, function x4B), for example. With the x4B
function, MSDOS handles the opening, reading, and closing of the file
itself, internally.  ie: I have no guarentee, and Im pretty sure it
doesn't, MSDOS would use intr x21, x3D to open the file it was exec-ing,
x3F to read it, and then x3E to close. It is correct that MSDOS isn't
re-entrant...in the sense that intr x21 x6C wont use intr x21 x3C,
right?

 I've looked over the device-driver chapter in MSDOS Advanced
Programming by Michael Young. The device driver in there is far too
low-level.  Is there any way I can make MSDOS use a device driver of
mine, but on a file/directory level?

Any helpfull info/suggestions/insight via email will be welcome (as
long as you dont tell me its impossible :)

Greg
[e-mail]

------------------------------

Date: 2 Dec 92 21:15:39 GMT
From: Binod Taterway <[e-mail]>
Subject: Optimal Settings for Communications

I  am  using  Kermit 3.11  (DOS  version)  on  IBM/ValuePoint  running
OS/2~2.0. I did all the  settings that I needed  for the  COM port.  I
lose a lot of characters at 19200, but so much when the speed is 9600.
However, when I  push to DOS,  I cannot seem to  recover. I run kermit
from OS/2 full  screen window,  which loads DOS automatically. I  have
not  tried  running from DOS full  screen window. Should  that make  a
difference. I think not because OS/2 automatically overlays DOS before
invoking Kermit. What do you think? Comments?  --

- Binod Taterway
  Sr. User Consultant (LUCC)
  E-mail: [e-mail]

------------------------------

Date: 4 Dec 92 23:30:09 GMT
From: Ken Bass <[e-mail]>
Subject: Optimal Settings for Communications

[e-mail] (Markolf Gudjons) writes:
>  [e-mail] writes:
>
>>I am trying to use a couple of different DOS-based communication
>>programs.  Both worked fine at 2400 bps but both lose characters at
>>9600.  What settings should I use to ensure maximum throughput?  For
>>example, what DOS settings should I use, what MODE settings and what
>>communication program settings?  Thanks.

>It's not a good idea to use Dos based comm programs to start with. Try
>and look into OS/2 apps.

>If you absolutely *want* to use them, try to locate an archive called
>nice10.* or something similar. It conatains a utility, nice10.exe,
>which can be used to increase the priority of any process started by
>it. What you want to do is something like "nice10 /c <program>" This
>will run your app. at TIMECRITICAL_PRIORITY. This will give you
>excellent throughpu tbut will also screw your system to the point where
>there isn't a lot left of OS/2s multitasking capabilities.

This is not true. I use Telix under DOS at 14.4kbps and have no
problems at all. Granted, I'm not running Windows apps in the
background, usually just reading mail or examining ZIP files Im
downloading. However, there are no good Shareware OS/2 terminal
programs worth considering that I have seen.

   ---Ken

--
Ken Bass ([e-mail])              |   Telecommunications
George Mason University                       |   Techniques Corp,
Student, Department of Electrical Engineering |   Software Engineer

------------------------------

Date: 2 Dec 92 22:52:58 GMT
From: "Timothy F. Sipples" <[e-mail]>
Subject: Specific DOS (3.31), CD ROM, and PKUNZIP => Crash City (!)

 [e-mail] writes:
>I doubt there is anything anyone can do to help me with this, but I
>thought I would pass it on, as it is kind of interesting.
>I am running MSCDEX in a specific DOS window, to access my (Sony) CD ROM
>player.  If I use PKUNZIP to unzip a file from the Simtel CD ROM to the
>hard disk, the machine crashes, with the screenful of register dumps and
>the note that the system has been halted and that I should contact my
>service representative.

>That is, I do something like:
>    PKUNZIP X:MSDOS\FILUTL\NCDC150.ZIP
>and it gets about half way through the unzipping, then, kaboom!  This is
>absolutely repeatable (100%).

>Both of the following are workarounds:
>    1) Copy the file to the hard disk before unzipping.
>    2) Use INFO-ZIP (compiled for DOS)

>Note that I cannot test this directly in a regular DOS window, since the
>CD ROM is only accessible from the specific DOS window.
>I assume this has something to do with PKUNZIP's using 386 instructions
>in a way that interferes with OS/2.  Any ideas?

>(Machine is a generic 386/33, 8 megs of ram, 200 meg IDE drive)

This problem, I believe, was a known bug and was, if memory serves,
fixed with the Service Pak (XR06055).

Timothy F. Sipples      | Read the OS/2 FAQ List 2.0h, available from
[e-mail] | 128.123.35.151, anonymous ftp, in /pub/os2/all/info
Dept. of Econ., Univ.   | /faq, or from [e-mail] (send "HELP")
of Chicago, 60637       | [Read the List, THEN post to ONE OS/2 newsgroup.]

------------------------------

Date: 3 Dec 92 16:06:16 GMT
From: "larry.a.shurr" <[e-mail]>
Subject: The SP, Diamond SpeedStar (ET4000), and Higher Resolutions

 [e-mail] (Fritsch_Wolfgang) writes:
}[e-mail] (Olaf Scherdin) writes:
}>On <17 Nov 09:15> Fritsch_Wolfgang (2:242/6.1) wrote:

[Discussion and proposals for addressing can't sync SVGA with ET4000
problem].

}Yes, I believe I faithfully tried all the stuff proposed in this
}group, certainly the stuff that is proposed in the c:\readme. It does
}not work in any of the 4 identical installations around me. People
}stare in amazement on the stable screen of *my* installation which has
}the ATI Ultra. It seems that about everyone else is happy now, judging
}from the standstill of traffic

I wouldn't say "happy."  I conjecture exhaustion and lack of new ideas.
I think that this applies both to the can't sync SVGA resolutions
problem and to the sluggish performance problems with ET4000.  I can
sync, but my windowed VDM & VIO performance stinks.  I've tried
everything suggested with limited results.  It's possible to get VDM
sessions to scroll some- what faster, but it's still pretty poor.

A lot seems to depend on whose hardware you're using.  This applies to
both sync and performance problems.  I'm using a Diamond SpeedStar and
have no sync problem, but poor performance.  At least one person using
a noname ET4000 card reports no sync problems and good performance on
his home system, but poor performance with a Diamond SpeedStar at work.

All of my video problems except for performance went away when I got a
newer version of VMODE.COM for setting up SpeedStar.  Others have
obtained the newest VMODE or equivalent and have failed to get better
results.

At least I can use SVGA resolutions and the PM performance is better
than before the SP, but I'll seriously look at XGA-2 for ISA when it
becomes available unless someone can improve the situation...
"Diamond?" "Binar?"  Hello?  Anybody out there?

Larry

}on the point of ET4000 cards. If nobody comes up with a new idea, we
}probably give up and fix our hopes to the 2.00.01 version.

Larry A. Shurr ([e-mail] or att!cbnmva!las) speaking only for myself.
EOR (end-of-ramble)

------------------------------

Date: 2 Dec 92 07:48:28 GMT
From: "Jack S. Tan" <[e-mail]>
Subject: The SP, Diamond SpeedStar (ET4000), and Higher Resolutions

[e-mail] (Fritsch_Wolfgang) writes:

>Yes, I believe I faithfully tried all the stuff proposed in this
>group, certainly the stuff that is proposed in the c:\readme. It does
>not work in any of the 4 identical installations around me. People
>stare in amazement on the stable screen of *my* installation which has
>the ATI Ultra. It seems that about everyone else is happy now, judging
>from the standstill of traffic on the point of ET4000 cards. If nobody
>comes up with a new idea, we probably give up and fix our hopes to the
>2.00.01 version.

This is what worked for me and, hopefully, it will help you:

 1.  Set the ET4000 card using VMODE under native DOS first, making
sure that the video is correct (check with VDIAG).  I have a Gateway
2000 CrystalScan 1024NI, but the monitor setting for such didn't
produce high-resolution graphics correctly.  I had to create a custom
monitor setting to get the video modes in sync.

 2.  Re-run the VMODE under a full-screen DOS session, and confirm that
the settings are the same and everything works fine.  For some reason,
trying to set the card under a DOS session alone would not set the card
properly, but first setting it under native DOS resolved the problems.

 3.  Selectively install the Tseng drivers, checking the SVGA box in
the install options (even if it is already marked).  When OS/2 asks for
disk 6 or 7, it means the OS/2 GA diskettes.  It is looking for high-
resolution fonts.  After running through most all the GA disks, then
you will pick the resolution and use the CSD display diskettes.  OS/2
will run SVGA automatically.

 4.  After installing the video, do not reboot!  Open a full-screen DOS
session, and run the appropriate VMODE statements to set the monitor
(e.g., VMODE MONITOR, VMODE VESA).  (For good measure, you may wish to
delete the \OS2\SVGADATA.PMI file.)  Now, run SVGA ON to save the
current video settings.

 5.  Exit, shutdown, and reboot.  The WPS *should* come up in the
desired resolution.

I hope this works with your system.  Good luck!

Jack Tan                                     Sattinger's Law:
[e-mail]                       It works better if you plug it in.

------------------------------

Date: 2 Dec 92 06:12:55 GMT
From: Joseph Chiu <[e-mail]>
Subject: Watch dog timer support on ISA?

I understand that Watchdog timers exist on MCA and (some?) EISA systems.

What I would like to know is if one could somehow enable the watchdog
functionality on an ISA system.

If it is just a matter of an NMI being generated, I can wire up a push
button that goes to the NMI line...  (OR, if it needs a little more, I
have tubes full of NAND gates that I could wire up into something
useful... *)

I would like to wire up watchdog-like functionality so that when I run
my "CLI / forever: JMP forever"* program, I can continue to work
without a hard shut down.

(And heck, maybe there's marketing potential for an ISA watchdog
retrofit...*)

Thanks all.

(*  LOTS of smileys where appropriate.)
--
Joseph Chiu, Dept. of Computer Science, Caltech.    [e-mail]
1-57 California Institute of Technology, Pasadena, CA 91126. [phone]
*    Now running OS/2, Windows, DOS, and UNIX (okay, well, maybe not...)   *

------------------------------

End of Info-IBMPC Digest V92 #194
*********************************
-------