Info-IBMPC@WSMR-SIMTEL20.ARMY.MIL ("Info-IBMPC Digest") (05/04/91)
Info-IBMPC Digest Sat, 4 May 91 Volume 91 : Issue 112
Today's Editor:
Gregory Hicks - Rota Spain <GHICKS@WSMR-Simtel20.Army.Mil>
Today's Topics:
Amiga vs. IBM
Problems with PC/XT Hard Disk (2 msgs)
Drive A problem (2 msgs)
Drive A problem; COM1 & 2 problem
Turbo Pascal 6.0 bug? (V91 #95)
interleave adjustment (V91 #95) (3 msgs)
VGA 256 colors in Turbo Pascal.
Today's Queries:
cyrillic font for WINDOWS 3.0
Help with Tallgrass TG-5025 tape unit
IDE, SCSI and ESDI disk interfaces?
Clarkson Drivers for Token Ring Nets Available?
Lunar (Chinese) Calendar
New Uploads:
ghostscript submittal
SIMIFL25 - Make SIMIBM.IDX readable but still searchable
Send Replies or notes for publication to:
<INFO-IBMPC@WSMR-SIMTEL20.ARMY.MIL>
Send requests of an administrative nature (addition to, deletion from
the distribution list, et al) to:
<INFO-IBMPC-REQUEST@WSMR-SIMTEL20.ARMY.MIL>
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: Sun, 28 Apr 91 19:43:24 EST
From: Greg Ferguson <ACDGREG@vm.uoguelph.ca>
Subject: Amiga vs. IBM
IBM architecture is archaic. I rather like IBM's but they will not
perform to the standard an Amiga will. I would instead buy an IBM
emulator for your Amiga if you need to run IBM packages. By the way,
if you are running Lotus files, VIP professional will read and write
1-2-3 files. In terms of DOS compatability, the ATOnce emulator is
great. It emulates a 286 with Hercules monochrome or CGA AND you can
run any Amiga program at the same time!! AND you can run multiple
emulators simultaneously...
Plus, you can buy the AdIDE chip for the Amiga to allow you to use a
standard IBM external hard drive for $112. The Amiga uses SCSI drives,
so they are more expensive. An Amiga also has LOTS of 24-bit colour
cards, some for $250...they are brilliant.
I hope this helps.
(By the way, I am an Amiga owner who programs IBMs for a living..)
[End of *flames* please... gph]
------------------------------
Date: Mon, 29 Apr 91 11:43:44 EDT
From: Turgut Kalfaoglu <TURGUT%FRORS12.BITNET@uga.cc.uga.edu>
Subject: Problems with PC/XT Hard Disk
I set my father's PC up a while ago with a brand new ST-238 hard disk.
However, about 1 month into use, this drive started doing something
crazy -- turn the machine on, the drive doesn't start turning|
I took the drive apart (numerous times), and if I just give a small
push to the flywheel of the drive, vrrrrnn, it takes off, and works
fine. But then you turn it off, and SOMETIMES, it won't start again
with the same problem. I then take the drive apart again, (just the
circuit board underneath), spin it by hand, and it goes.
Once I tried holding the flywheel when I turned on the machine, to see
how much resistance the motor puts. I was surprised to see that the
motor 'insisted' for one second (with me holding the flywheel), then
gave up on it (I was able to release the flywheel, and it didn't start
spinning). Then I gave it the whir, and it started fine|
The problem is not the power supply, as I tried with two different
machines, same results.
Regards, -turgut
at TURGUT@FRORS12 until May 15th, back at TURGUT@TREARN afterwards
------------------------------
Date: Mon, 29 Apr 91 13:36:00 +0200
From: S89406316%HSEPM1.HSE.NL@uga.cc.uga.edu
Subject: Problems with PC/XT Hard Disk
>Turn on the machine, the drive doesn't start turning!
According to an article I read a while ago in an issue of PC magazine
(don't remember which one, of course) this happens a lot with
particularly XT drives (if my memory serves me right).
In fact I experienced this problem exactly once, when my hard disk had
given up on me and I dismounted and examined it. (I remember cursing it
because I thought that first the controller and then the motor had gone
south, but eventually it turned out that neither of them did).
Try to turn the disk (or the computer) over 90 degrees. #include
<cliche>: Hope this helps.
Jac Goudsmit
Student, Information Technology
Hogeschool Eindhoven, The Netherlands
S89406316@HSEPM1.HSE.NL
------------------------------
Date: Mon, 29 Apr 91 13:19:03 +0200
From: S89406316%HSEPM1.HSE.NL@uga.cc.uga.edu
Subject: Drive A problem
>1. When I change the floppy in the 1.2-meg Drive A, or even
>leave the drive door open, the computer thinks the previous
>floppy is still there. If I do a DIR A:, then change the floppy
>or leave the door open, and do DIR A: again, I get the same
>directory listing.
Sounds to me like your Disk Change line (a hardware provision to make
DOS detect a disk change) is defective.
Some (most? all?) of the drives that are mounted in AT-style computers
support this Disk Change Line, which is actually a switch or a
photodiode that detects if a disk is in the drive. Drives that have
one, sometimes start spinning for a while immediately when you insert a
diskette.
If the LED that lites the photodiode has gone to the dogs, or when the
switch dies or in some other cases (short circuit or circuit break -
not very likely), the PC keeps thinking that a floppy is in the drive
when it actually isn't. when you type DIR, DOS thinks the old diskette
is still in and doesn't bother to read the new disk's directory and FAT
data. This is a dangerous situation, because DOS may write the
directory and FAT of the old disk to the new one, whenever you perform
a write operation (even a REName).
Then again, it could also be a software problem. If for one reason or
another (no particular reason comes to mind right now) the computer
started thinking that there is a drive change line when actually there
is none, the same problem arises. Did you change your CONFIG.SYS by any
chance, just before the problem started? e.g. removed "useless" device
drivers? Check you backup, then (let someone) check the hardware.
I admit, I am more into XT's, but maybe this helps. As usual, I am
typing all this without docs at hand. Must be in one of my PC mags
(>100) somewhere...
Apologies in advance if the information above is incorrect. If it is,
then someone correct me, please.
Jac Goudsmit
Student, Information Technology
Hogeschool Eindhoven, The Netherlands
S89406316@HSEPM1.HSE.NL
------------------------------
Date: Mon, 29 Apr 91 11:02:44 IST
From: Lew Golan <LEW%TAUNIVM.BITNET@uga.cc.uga.edu>
Subject: Drive A problem; COM1 & 2 problem
Two problems which may or may not be related:
1. When I change the floppy in the 1.2-meg Drive A, or even leave the
drive door open, the computer thinks the previous floppy is still
there. If I do a DIR A:, then change the floppy or leave the door open,
and do DIR A: again, I get the same directory listing.
This happens even if I boot without the AUTOEXEC.BAT and CONFIG.SYS
files.
The problem just came up a couple of days ago when I was trying to
install a program from several floppies. When the setup program
prompted me for the second disk and I inserted it, the program rejected
it as being the wrong disk.
I used PC-Tools to compare COMMAND.COM, MSDOS.SYS and IO.SYS against my
backups; they are identical. And when I booted off the backup floppy,
the problem remained.
I tried booting with a CONFIG.SYS that had only one line: BUFFERS=1 (to
get around the default 15). If the disk had several files in its
directory, the problem disappeared. But then I did a DIR of a blank,
formatted floppy, got the response on the monitor (including volume
label), opened the door, repeated the command, and got the same
response.
I do not have this problem with the 3.5" Drive B.
I have a Zeos 386/33 with an IDE hard drive. Is it possible that the
floppy drive controller is at fault?
2. I run a laser printer off COM1 and an external USR modem off COM2.
After using the modem (with Kermit or Qmodem), if I try to print from
ANY program (in either Postscript or HP emulation mode), the I get
blank pages or garbage from the printer. Somehow, COM1 seems to be
messing up COM2. The only way to get the printer to work properly is to
reboot the computer after using a communications program.
I have one I/O card with two serial ports. But even when I replaced it
with two separate serial cards, the problem persisted. Booting without
the AUTOEXEC.BAT and CONFIG.SYS files did not help.
Lew Golan, Tel Aviv University <lew@vm.tau.ac.il>, <lew@taunivm.bitnet>
------------------------------
Date: Mon, 29 Apr 91 10:06:04 MZT
From: bax@tudgw2.tudelft.nl (Jan Adriaan Bax)
Subject: Turbo Pascal 6.0 bug? (V91 #95)
Markus Boie suspects he found a bug in Turbo Pascal 6.0, having to do
with type casting.
>Date: 10 Apr 91 19:19 +0200
>From: "Markus F. Boie" <boie%dkrz-hamburg.dbp.de@RELAY.CS.NET>
>[...] but most of the time I tried to hunt down something of which in
>the beginning I thought was my fault. Now I am not that sure if it
>isn't a bug. Please have a look at the following code:
>FUNCTION TimeDiff(Time1, Time2: DateTime): LongInt;
>{ DateTime is a pre-defined record with variables of type WORD }
>VAR t1, t2: LongInt;
> h, m : LongInt;
>BEGIN
>{ The following gives rather strange (wrong) results when for example
>Hour*3600 or exceeds the boundaries of the WORD type. If I was a
^^^^^^^^^^
>compiler I would look at the left side of the becomes symbol
^^^^^^^^
I completely agree with you, but Turbo Pascal doesn't! The type casting
of the right hand side to the type of the lhs is done AFTER the
computation of it.
> to determine the possible value range and then work with variables
>of THAT type. Not with what happens to be the type of the first
>variable after the becomes symbol! }
>WITH Time2 DO
> t2:= Hour*3600 + Min*60 + Sec;
Because all of your variables and constants in the rhs of the
assignment are of type word, the expression will be computed in short
(2 byte) integers, including overflows etc.
>{ This works great, but with some extra operations that seem completely
> unnecessary to me. }
>WITH Time1 DO BEGIN
> h:= Hour;
> m:= Min;
In the above statements Hour and Min are type casted to longints.
> t1:= h*3600 + m*60 + Sec;
>END;
>TimeDiff:= t2 - t1;
>END; { TimeDiff }
In this code part the variables h and m are of type longint, so the
complete expression is computed in longint, yielding the correct
result. You may avoid the "unnecessary" operations with the following:
t1 := longint(Hour)*3600 + longint(Min)*60 + Sec;
Strictly spoken, only the first typecast is needed, because TP uses
longint computations as soon as one longint occurs in the expression.
>Do I miss something? Are there any options to set to make TP6 work like
>expected? Imagine how many errors like these are never been found!
Arjen Bax.
------------------------------
Date: 28 Apr 91 22:04:19 GMT
From: jaapv@accucx.cc.ruu.nl (Jaap Verhage)
Subject: interleave adjustment (V91 #95)
In digest V91 #95, Adriano writes:
>Is there anyone who has or knows of a reliable, compatible, safe
>utility to perform the optimization of hard disks' interleave factor? I
>have tried the Norton's one and another called IAU. Both seem to have
>compatibility problems on my machine, which is an Amstrad PC1640, 100%
>compatible apart from this. My hard-disk is really an add-on Olivetti
>"hard-card", 20M, don't know what kind of controller. Its interleave
>was reported to be 5 or 6. Another question to anyone who bother to
>reply: is it safe to change the interleave to 1? Thank you for replies
>(if any),
Many utilities will do what you want, like IAU that you mentioned, as
well as HDTEST, which I got from Simtel 1 or 2 years ago. However,
they rely on MFM-encoded hard disks. I tried them on a hard card, maybe
like yours (Miniscribe was this one's manufacturer), and it turned out
to be RLL-encoded. My hard card had a low level formatting program in
it's controller ROM, reachable via DOS' DEBUG:
after DEBUG's "-" prompt, you type: "g=c800:5", if "C800" is the
controller's address. It could also be "CA00". The ROM program then
asks you if you want to format disk 0 or disk 1 (0 if it's the only
one), and at which interleave you want this done. After this finishes,
you can key in any bad sectors/tracks from the list that should have
come with the disk. After finishing, you are back in DEBUG which you
leave with "q".
Then comes DOS' FDISKing and FORMATting, and finally restoring your
data. If your disk doesn't carry a defective track list, you can use
something like PC Tools' surface analysis utility, part of PC Compress,
to check the disk's surface. Let it run overnight to make a thorough
test.
I've tried various interleaves and got the disk to be three times
faster than it was in data transfer rate as measured with Coretest,
also available from Simtel. Strangely (to me), the best performing
interleave was 4:1, not 1:1. The disk sits in a 8 MHz Olivetti M28 (AT
compatible), so I thought the processor wouldn't have trouble keeping
up with the data stream. Maybe it's the disk controller? No idea.
Anyway, for my disk, the only way to change the interleave turned out
to be with it's low level format program. You may try the same; let me
hear if it works.
Regards, Jaap.
Jaap Verhage,
Academic Computer Centre,
State University at Utrecht, Holland.
jaapv@accucx.cc.ruu.nl
------------------------------
Date: 29 Apr 91 10:11:00 EDT
From: "SI02::KELLYL" <kellyl%si02.decnet@v3.hanscom.af.mil>
Subject: Interleave Adjustment
Adriano Santoni asked about a program to reset the interleave of hard
disks. The best I have found is Sprinrite II. Also performs other
useful functions like a non-destructive low level format. Don't have
the book with me at the moment so you would want to check on
compatibility.
In regards to your question about is it safe to change the interleave
to 1:1 - Its safe but may not be desirable. The interleave is a
balancing act between the system and the disk that depends on how
quickly data can be moved. Just setting the interleave to 1:1 does not
mean improved performance. In fact, I have several systems where
loosening the interleave from 3:1 to 4:1 has IMPROVED throughput.
Sprinrite also tests each interleave setting to determine the best on.
- Larry Kelly
------------------------------
Date: Mon, 29 Apr 91 00:11 EDT
From: Danny Taub <DATAUB@vaxsar.vassar.edu>
Subject: Interleave Changing
In Info IBM-PC Vol 91 #95, Adriano Snntoni
<SANTONI@IMIMNVX.IRFMN.MNEGRI.IT> writes:
>Is there anyone who has or knows of a reliable, compatible, safe
>utility to perform the optimization of hard disks' interleave factor? I
>have tried the Norton's one and another called IAU. Both seem to have
>compatibility problems on my machine, which is an Amstrad PC1640, 100%
>compatible apart from this. My hadd-disk is really an add-on Olivetti
>"hard-card", 20M, don't know what kind of controller. Its interleave
>was reported to be 5 or 6. Another question to anyone who bother to
>reply: is it safe to change the interleave to 1? Thank you for replies
>(if any),
First, let me say that in my experience, Spinrite by Steve Gibson is
one of the best Low Level HD Formatting utilities around. (I have no
affiliation with them). Second, in my experience Hard Cards are not
supposed to be formatted by a low level formatter. They come
preformatted and are usually not supposed to be touched. Finally,
setting the interleave to 1 is only good if all your hardware supports
it! If the PC is slow, it may send the HD spinning until it can
"catch" the info it needs. Setting the interleave to 1 will not make
for a faster system. If you decide to get Gibson's Spinrite, the
program gives an excellent explanation of interleave.
Danny Taub
Dataub@Vassar.Edu
------------------------------
Date: Sun, 28 Apr 91 20:06:32 PDT
From: 6500boo%ucsbuxa@hub.ucsb.edu (William Bushing)
Subject: VGA 256 colors in Turbo Pascal.
Luis,
There is a 256-color BGI driver for Turbo Pascal that worked just fine
when I tried it a few months ago. It did 640x480, 800x600 and 1024x768
on my Diamond Speedstar (ET-4000 based). Try getting it from either
Simtel20 or WUarchive.
Bill
William W. (Boo) Bushing
6500boo@ucsbuxa.bitnet
6500boo@ucsbuxa.ucsb.edu
Marine Biotechnology Lab
Univ. of Calif. Santa Barbara
------------------------------
Date: Mon, 29 Apr 91 09:30 +0200
From: <STEFFENS%DMSWWU5P.BITNET@CUNYVM.CUNY.EDU>
Subject: cyrillic font for WINDOWS3.0
Dear netlanders,
does anybody know from where - PD, SW or COMMERCIAL - I can retrieve
cyrillic fonts for the use with WINDOWS 3.0 applications? The best
would be if you can point to a source of plain windows fonts, i.e.,
fonts in the usual pointsizes (8,10,12,14,18...), working without the
need of the ATM or the Bitstream Face-Lift. But if you know of cyrillic
fonts working under ATM this would also help me.
Thanks a lot in advance
K.Steffens STEFFENS@DMSWWU5P.BITNET
STEFFENS@CERNVAX
STEFFENS@VXWA80.CERN.CH
------------------------------
Date: 28 Apr 91 22:15:51 EDT
From: KLBY@VAX5.CIT.CORNELL.EDU
Subject: Help with Tallgrass TG-5025 tape unit.
I have a Tallgrass Technologies TG-5025 unit. This is a nice sized box
with a 24 megabyte ST-412 interface hard disk drive in it and a 60
megabyte tape drive with an unknown interface. The tallgrass unit has
a board in it that has cables going off to both the hard disk and the
tape drive. There is a male db25 on the back that I am assuming used
to go to some sort of interface that lived inside the computer. If you
know anything about this box please drop me a note. I have some
questions.
1) Is the signal going into the board inside the tallgrass box SCSI?
2) What interface does the tape drive use? (ie scsi, qic-02 etc..)
--Matthew
------------------------------
Date: Mon, 29 Apr 91 10:50 GMT
From: Mark Walters <MSERVMRW@vax.oxford.ac.uk>
Subject: IDE, SCSI and ESDI disk interfaces?
Dear All,
I'm looking at the specs. for possible Novell server and wondered if
somebody could explain to me the differences (pros/cons) of the
different hard disk interfaces; IDE, SCSI and ESDI.
The sort of size that I'm thinking of is 300+ Mb. Would one of the
above type be more sutable than the others?
Thanks in advance for any help.
Mark Walters
MSERVMRW@UK.AC.OXFORD.VAX
------------------------------
Date: Mon, 29 Apr 91 08:25:21 EDT
From: jfjr@mbunix.mitre.org (Freedman)
Subject: Clarkson Drivers for Token Ring Nets Available?
I am aware of the Clarkson drivers for various ethernet cards on a
PC. Is there anything equivalent for a token ring network? I would
appreciate any information. Thank you.
Jerry Freedman,Jr
------------------------------
Date: Sun, 28 Apr 91 22:56:17 EDT
From: jyoo@elbereth.rutgers.edu (J-S Yoo)
Subject: Lunar (Chinese) Calendar
What I want to find out is a lunar-to-solar date conversion
program. That is, when the lunar date is entered, the output will be
the matching solar date. Thanks.
------------------------------
Date: Sun, 28 Apr 91 17:51:44 -0400
From: nate@NEUTRON.LCS.MIT.EDU (Nate Liskov)
Subject: ghostscript submittal
I have updated pd1:<msdos.postscript>GS211.ZIP on SIMTEL20 with:
pd1:<msdos.postscript>
GS211A.ZIP Ghostscript v2.1.1 exec. for Herc, EGA, VGA
This contains the required support file FONTMAP that was missing from
the original posting.
Best regards... Nate Liskov
nate@neutron.LCS.MIT.EDU
------------------------------
Date: 28 Apr 91 17:47 +0200
From: "Markus F. Boie" <boie%dkrz-hamburg.dbp.de@RELAY.CS.NET>
Subject: SIMIFL25 - Make SIMIBM.IDX readable but still searchable
I have uploaded to SIMTEL20:
pd1:<msdos.filedocs>
SIMIFL25.ZIP Make SIMIBM.IDX readable but still searchable
Markus
boie@dkrz-hamburg.dbp.de
------------------------------
End of Info-IBMPC Digest V91 #112
*********************************
-------