Info-IBMPC@USC-ISIB.ARPA (Info-IBMPC Digest) (02/22/86)
Info-IBMPC Digest Friday, 22 February 1986 Volume 5 : Issue 26
This Week's Editor: Eliot Moore <Elmo@USC-ISIB>
Today's Topics:
Genius VHR (2 Messages)
ARC5xx
Removing Non-DOS Partition on AT
LINK /E vs EXEPACK Update
Interrupt Handlers in Turbo (V5 #20)
Floppy Access Problem on 10 MHz AT
Diablo <=> Lotus on Wang PC
MS-DOS Find First File/Find Next File functions
Max Files Open
Multitasking a la TopView Query
Color & Mono Cards in AT
Public Domain Backup Program Wanted
----------------------------------------------------------------------
Date: Thu 20 Feb 86 23:54:46-PST
From: Richard Pattis <PATTIS@WASHINGTON.ARPA>
Subject: Genius VHR
To: info-ibmpc@USC-ISIB.ARPA
I am seriously considering buying a Genius VHR for my AT system. This monitor,
discussed in PC Mag's "Best of 85" article (page 112, Vol 5 No 1) supposedly
crisply displays 66 lines x 80 columns in portrait mode. It has 730 pixels
across by 1004 pixels down. Unfortunately, I can't seem to find one installed
locally, such that I could actually see it and try it out.
I am interested in hearing of anyone owning such a monitor, or anyone owning or
knowing of a similar product. Mostly, I want to use the bigger screen for
editing (with Epsilon) and previewing TeX output. I'll appreciate any pointers.
Richard Pattis
------------------------------
Date: 21 Feb 1986 12:52:25 PST
Subject: Re: Genius VHR
From: Craig Milo Rogers <ROGERS@USC-ISIB.ARPA>
To: Richard Pattis <PATTIS@WASHINGTON.ARPA>, info-ibmpc@USC-ISIB.ARPA,
mockAPETRIS@USC-ISIB.ARPA
In-Reply-To: <12185063836.21.PATTIS@WASHINGTON.ARPA>
I have one of the Genius VHRs. I have mixed opinions about it.
The large screen is good for editing, terminal emulation, etc. I haven't
written or acquired any software to use the graphics plane yet. The
BIOS-replacement supplied with the monitor was very buggy, and I've never
received a fixed version or the sources so I could fix it myself, even
though I've asked them and been promised results three times.
The worst problem that I've encountered is heat. The controller
board and monitor use an awful lot of power; my office is noticeably
warmer than it was before I installed the Genius in my AT. Furtheremore,
the board is heat sensitive. When we've had complete air conditioning
failures the whole screen will go crazy. When we have partial failures
perhaps one or two characters, in certain columns, will have some
flickering bits.
Another problem I've noticed is that occasionally when editing
with Epsilon the whole screen will go blank. Everything will reappear
when I hit the next keystroke. I don't know whether Epsilon or the
Genius is at fault here.
I've written my own programs to initialize the Genius (set its
mode bits, clear the screen, etc) because the ones supplied with the
monitor/controller didn't do it the way I wanted.
In white-on-black mode the image is pretty crisp and flicker-free.
There's a problem with the brightness and contrast controls... it's
hard to get the intensity level up high enough. Furthermore, I think
there was an intensity problem with software that uses inverse video
(or something like that), but I haven't tried that in a while.
Based on my experience I would not recomment this display. I
would consider changing my mind if the manufacturer provided better
customer support.
Craig Milo Rogers
------------------------------
Date: Fri, 21 Feb 86 18:34:55 -0100
From: hans@oslo-vax (Hans A. ]lien)
To: info-ibmpc@usc-isib.ARPA
Subject: ARC5xx
For the last few days, ARC on SIMTEL20 has been
PD:<MSDOS.ARC-LBR>ARC51.COM.1
This is a special "programme" containing both ARC.EXE and ARC.DOC (of v.5.10).
About 90K altogether, packed into a file of 58,880 bytes.
Make sure to unpack by running ARC51.COM in your current directory.
The accompanying .DOC file states the availability of the source code,
but nothing is said on distribution policy of sources AND/OR .ARC file
format descriptions. I am interested in both, and maybe other
programmes (like DEARC.PAS) being able to unpack .ARC5 archives.
Please send your info to INFO-IBMPC, or maybe better, directly to me,
and I will summarize for the Digest. Tnx!
Hans
P.s.: Maybe ARC52.COM has become available by the time this reaches you,
but that's ok, isn't it?
------------------------------
Date: Fri, 21 Feb 86 11:13:39 PST
From: Nicholas B. Mason <mason%cod@nosc.ARPA>
To: info-ibmpc@usc-isib
Subject: Removing Non-DOS Partition on AT
In V5 #21 of the digest someone wrote in that they were trying to
remove a non-dos partition from their PC-AT harddisk (a PC-IX
partition I think). I had a similar problem. To remove the non-dos
partition they need to use the "Advanced Diagonstics for the IBM
Personal Computer AT" diagonstic disk version 2.01 or later. On this
disk there is a routine that will reformat the hard disk. After this
is done you have to use "fdisk" and the dos "format". The disk can
now be used for dos. Note that the diagonstics disk that comes with
the AT does not do the above. You must have the advanced diagonstics.
Nick Mason
NOSC San Diego , CA
ihnp4 \
akgua \
decvax ------------------!sdcsvax!noscvax!mason
dcdwest /
ucbvax /
------------------------------
Date: Thu, 20 Feb 86 14:43:18 PST
From: Ya'akov_Miles%UBC.MAILNET@MIT-MULTICS.ARPA
To: info-ibmpc@USC-ISIB.ARPA
Subject: LINK /E vs EXEPACK update...
Just took delivery of my Microsoft MASM 4.00 update today. It appears
that the new EXEPACK.EXE ver 4.00 supplied with this MASM is more
efficient than the EXEPACK.EXE ver ?.?? supplied with my Microsoft
"C" version 3.00, and yes, the (NEW) utility EXEPACK.EXE ver 4.00
produces smaller files than the LINK option /E does. It appears that
the LINK option /E packs files the same way as the old EXEPACK.EXE
which was distributed with Microsoft "C" version 3.00
As an aside, the new EXEPACK.EXE ver 4.00 supplied with the MASM 4.00
will successfully pack many .EXE files that the old EXEPACK.EXE failed
to pack correctly...
My apologies for the previous message...
------------------------------
From: mcvax!ukc!jmh@seismo.CSS.GOV
Date: Tue, 18 Feb 86 13:47:21 GMT
To: info-ibmpc@usc-isib.arpa
Subject: Interrupt Handlers in Turbo (V5 #20)
A few comments on Alastair Milne's interesting article on Turbo interrupts
from V5 #20. I don't claim to be a Turbo/8086 guru, so please be kind if
this is all a load of old cobblers. All comments apply to Turbo 3.01A.
1. Turbo does indeed place a prelude in front of the first user instruction
in each procedure. This procedure is concerned soley with saving
the old stack frame pointer and setting up a new stack frame, and for
a procedure with no parameters and no local variables goes something
like
push bp ; Save old frame pointer
mov bp,sp ; Mark start of new one
push bp ; save new frame pointer
jmp L1 ; jump around space for local
; typed constants
L1: <start of your code>
Should there be any local variables, SP is decremented to make
way for them at L1 before your code is started.
So you can (though see below) install interrupt handlers directly
as hinted on p214 of the Turbo 3.0 manual. Additionally, if you
load DS from a typed constant you can get at global variables.
2. It is only necessary to inform the 8259 of interrupts you wish
to receive *which are generated by external hardware*. To handle
interrupts generated by e.g. INT instructions the 8259 does not
need to be involved. Oh, and it isn't necessary to re-enable
interrupts before leaving an interrupt - the IRET will do that
for you, as it pops the flags register to its previous state.
A ghastly problem.
The following applies to interrupt routines handling interrupts
that occur from within MSDOS. I am chiefly thinking of external
interrupts (e.g. those from the serial port - time-consuming
sections of MSDOS enable interrupts for at least some of their
duration), but I believe there may be INT instructions from within
MSDOS that e.g. call screen print routines (INT 5 for IBM PCs).
The problem is that when MSDOS is doing its stuff it
switches to an internal stack which is VERY short on space. In
fact it is so short on space that it is unwise to rely on it having
room for anything more than the flags register and return address
saved when the interrupt occurs. This, unfortunately, means that
what you need to do is to switch to your own stack as the first
thing your interrupt procedure does, and then save all the registers
etc. Restore all registers on the way out and then switch
back to the previous stack before doing the IRET. Nasty, innit ?
Finally, note that procedures called from within interrupt procedures
with the stack in an unexpected place must have their calls compiled
with the K- switch to prevent Turbo assuming the stack has run into
the heap and squealing accordingly.
Persons interested in picking their way through unpleasant code are invited
to mail me for a copy of a memory-resident INT ?? handler which uses
interrupts from the serial chip to provide full-speed serial facilities.
Jim
..!mcvax!ukc!jmh.UUCP
------------------------------
Date: Thu, 20 Feb 86 23:45:45 EST
From: "Paul L. Kelley" <PLK@MC.LCS.MIT.EDU>
Subject: Floppy Access Problem on 10 MHz AT
To: INFO-IBMPC@USC-ISIB.ARPA
A few days ago I reported that the "Change Diskette Parameter
Table's motor startup delay" code sent to INFO-IBMPC by Glenn Connery
(connery%bnrmtv.UUCP at BRL.ARPA) did not fix an access failure
problem for the 1.2 Mbyte floppy on my 10 MHz AT.
Several people suggested that I use FIXDSK.ASM from the
INFO-IBMPC directory at USC-ISIB. I am happy to say that this has
solved the problem.
Paul Kelley (PLK@MIT-MC)
------------------------------
Date: 21 Feb 1986 19:51:38 EST
From: DASG@USC-ISID.ARPA
Subject: Diablo <=> Lotus on Wang PC
To: info-ibmpc@USC-ISIB.ARPA, abn.20e-27@USC-ISID.ARPA
cc: MEDCOM18@KOREA-EMH.ARPA
Two reasons may be causing your hangup. With the Wang PC, Lotus uses
its own PRINTER DRIVER and you have to insure that you have your
printer configured properly in Lotus. You do this by going to the
configuration menu in lotus thusly /WGDPI. For serial you select port
2 for lotus which throws you into serial mode and asks you for the
baud rate--probably 1200. The other problem you may have is the
parity setting. Lotus actually requires that your serial printer be
set to a specific parity. Dont remember what that is right now but it
is in the Lotus manual. You say the PM012 is a Diablo. Are you sure?
We were were told the PM014 55cps daisy was also a Diablo but later
came to find that it is is really a Data Products Printer which does
not allow for parity changes.There is no way to compenstate with Lotus
other than to save your file to a print file and print it from DOS or
the Wang menu. By the way, the Local Wang reps here were astounded
when they found that their "Diablo" PM014 was really a data
products--a very heavy but otherwise good printer--albeit at least
$1000 too expensive. Curious part of this is that Data Products
apparently decided to at least use the Diablo printing mechanism since
we have no problem with generic diablo ribbons. Hope this helps. If
you need more info call me autovon 22-51633.
--gary swallow
DASG-AMZ, Pentagon
------------------------------
Date: Fri 21 Feb 86 00:06:52-EST
From: Bard Bloom <BARD@XX.LCS.MIT.EDU>
Subject: MS-DOS Find First File/Find Next File functions
To: info-ibmpc@USC-ISIB.ARPA
I'm writing a directory editor (which is to SWEEP as a good screen
editor is to EDLIN). I'm using DOS function calls 4E and 4F to read
the directory (from Turbo Pascal). The program can read directories
on Rainbows, but apparantly can't on IBM PC/AT's. Any ideas why not?
Please reply to BARD @ XX.LSC.MIT.EDU if possible.
Vasty thanks,
Bard Bloom
------------------------------
Date: Fri, 21 Feb 86 13:39:38 mst
From: cib%f@LANL.ARPA (C.I. Browne)
To: info-ibmpc@usc-isib.ARPA
Subject: Max Files Open
I have just come across one of the DOS quirks which limits any process
to opening 20 files. Not concurrently, but ever. To add insult to this
limitation, DOS takes at least three off the top for itself, so that
a program can only open (and close) 17 files. This fillip is rather a
severe limitation on some text retrieval programs.
NOTE: FILES, in the config.sys file, is not relevant to this situation.
Can any dos wizard suggest a graceful way around this limitation?
Thank you.
cib@lanl.arpa
cib@lanl.uucp
------------------------------
Date: Friday, 21 February 1986 15:50:57 EST
From: Bob.Sproull@vlsi.cs.cmu.edu
To: info-ibmpc@usc-isib.arpa
Subject: Multitasking a la TopView Query
Does anyone know of a multitasking executive for the PC that is like
or (preferrably) emulates the multitasking features of TopView but
is MUCH smaller. Multiplexing the display is NOT necessary. Thanks.
------------------------------
Date: 20 Feb 86 14:01 PST
From: Ghenis.pasa@Xerox.COM
Subject: Color & Mono Cards in AT
To: info-ibmpc@USC-ISIB.ARPA
Is there supposed to be a problem in intalling both a monochrome and a
color graphics adapter in a PC/AT? I know some software like Lotus 1-2-3
and DoubleDOS allow you to run with both types of displays and flip
between the two or show different screens on each. However, I can't get
it to work right. Is having both cards plugged in something you can only
do on an XT?
Any info will be appreciated. Thanks in advance.
------------------------------
Date: Thu, 20 Feb 86 10:18:09 EST
From: John Shaver STEEP-TMAC 879-7602 <jshaver@apg-3>
Subject: Public Domain Backup Program Wanted
To: info-ibmpc@usc-isid
Is there a Hard disk baskup program in the public domain which is easily
reached by FTP?
Thanks
John
------------------------------
End of Info-IBMPC Digest
************************
-------