Info-IBMPC@B.ISI.EDU (Info-IBMPC Digest) (08/03/86)
Info-IBMPC Digest Saturday, 2 August 1986 Volume 5 : Issue 74
This Week's Editor: Billy Brackenridge
Today's Topics:
Rootpath.c
LEX Source
ATPERF & ATDISK
Testing File Existence in Turbo Pascal
DeSmet C 2.41 no malloc Bug
Number Nine Board
PD Assembler
Orchid's 286e Accelerator Board
Micro Stock Software Summary
Stock Program for Micros
Volume Labels (2 Msgs.)
dCLOCK Clock (2 Msgs.)
Security (3 Msgs.)
Numerical Recipes: The Art of Scientific Computing
Large Priam Disks for AT
ProDesign II
Today's Queries:
VDISK & PC3278/9 -- INT 15H Conflict
Multi-User AT
Epson JX Driver for MS Windows Wanted
Tek 4105 or DEC VT240 Emulators for PC
EGA Printscreen
dBase III Bug
Crossword puzzle maker wanted
Expansion Box: XT or AT Bus
Serial I/O expansion
Accessing COM1 from Lattice C 3.00
PC-DOS V3.2 -- What are the problems???
Bootable Disks
IBM PC Professional Graphics Display and Controller
Micro Emacs 3.7
Epsilon EEL Language Question
CMI Drives
How to Drive Multiple Graphics Screens from One AT?
PC-like Keyboard for AT Needed
IBM Logo
Assembly Language Routines package
Large Screen Video Display
"direct.h" Needed for grep.c
----------------------------------------------------------------------
Date: Thu 17 Jul 86 16:09:24-EDT
From: Robert Lenoil <LENOIL%DEEP-THOUGHT@EDDIE.MIT.EDU>
Subject: Rootpath.c
To: lotto%harvard@EDDIE.MIT.EDU
ROOTPATH.C Converts a path name argument to root based
canonical form rootpath determines the current
directory, appends the path argument (which may
affect which disk the current directory is
relative to), and qualifies "." and ".."
references. The result is a complete, simple,
path name with drive specifier.
-Robert Lenoil
[ROOTPATH.C has been added to the library. -wab]
------------------------------
Date: 24 Jul 1986 19:08:54 PDT
Subject: LEX Source
From: James H. Coombs <JAZBO%BROWNVM.BITNET@WISCVM.WISC.EDU>
LEX.C A Lexical Analyzer Generator. From UNIX converted
to DeSmet C.
James H. Coombs <JAZBO%BROWNVM.BITNET@WISCVM.ARPA>
7/22/86
------------------------------
Date: 1 Aug 1986 08:53:04 PDT
From: Anonymous@B.ISI.EDU
Subject: ATPERF & ATDISK
ATPERF.C and ATDISK.C are two of the August 1986 PC Tech Journal
PC-AT test programs. The combination ATPERF.C and TIME.ASM
check AT performance, while ATDISK.C and PDISKIO.ASM check disk
performance. I didn't type the 80287 or EMM code in the former
as I have neither on my machine I didn't type the disk write
code in the latter as I didn't want to accidentally clobber the
hard disk.
[ATPERF.C Has been added to the INFO-IBMPC library. This file contains
ATDISK.C and the assembly language routines as well. -wab]
------------------------------
Date: Sat, 2 Aug 86 11:00:14 PDT
From: Mike_Gertsman%UBC.MAILNET@MIT-MULTICS.ARPA
Subject: Testing File Existence in Turbo Pascal
There is a minor bug in the Exists function that has been given
in the Turbo Pascal reference manual since the product was first
released. The problem is that the function creates a file
handle when it performs a reset. If the file exists, and is not
explicitly closed after the call, repeated calls to the function
will eventually produce I/O error F3 (too many open files).
The following is a modified version of the function which avoids
this error:
function Exists(Filename : Filename_Type):boolean;
var
Fil : file;
begin
Assign(Fil, Filename);
{$I-}
Reset(Fil);
{I+}
Exists := (IOresult = 0);
Close(Fil); { statement added to Borland version }
end;
------------------------------
Date: Mon, 28 Jul 86 17:32:54 edt
From: SCHEUTZOW%gitpyr%GATECH.CSNET@CSNET-RELAY.ARPA,
MICHAEL J <gt6294b%gitpyr%GATECH.CSNET@CSNET-RELAY.ARPA>
Subject: DeSmet C 2.41 no malloc Bug
I disagree with the recent Bug Report that said that DeSmet C will
overwrite the static variables if malloc() is used. I'm also using
version 2.41. I ran the program:
static char buf[1024];
main()
{
char *a;
printf("Buf loc %d\n", buf); /* bad programming! */
while(a = malloc(5 * 1024 * sizeof(char)))
printf("New buf %d\n", a);
}
It looped 12 times when run (reflecting the 64K data limit), and
none of the buffers overlapped. I think Jim should check for a
programming error, i.e corrupted pointer or maybe a segment wrap-
around (doing any math on pointers?).
I hope you find the problem.
Mike Scheutzow "There is nothing wrong with the 8086 that
Georgia Tech treating it like a PDP-11 wouldn't cure."
gt6294b@gitpyr -RBJ-
------------------------------
Date: Sun, 27 Jul 86 23:26:18 pdt
From: Richard Jennings <nike!lll-crg!jennings@ucbvax.Berkeley.EDU>
To: nike!B.ISI.EDU!Info-IBMPC
Subject: Number Nine Board
At a recent Intel seminar they had a Number Nine Board demo with
their graphics chip running in a COMPAQ.
I am holding off on an EGA for the same reason.
Rich.
arpa -> jennings@lll-icdc.ARPA
uucp -> seismo!lll-crg!jennings@icdc
------------------------------
Date: Mon, 28 Jul 86 07:12:40 cdt
From: mlw@ncsc.ARPA (Williams)
Subject: PD Assembler
John Duimovich asked about a Public Domain assembler...suitable for
teaching. CHASM (a CHeap ASseMbler) is available in the public domain
and includes tutorial materials. It's on most BBS systems for download.
I think we may have a copy here that can be ftp-d and I'm sure it
exists somewhere in the simtel20 archives. Don't know for sure if
it's hard to get through a BITNET connection...
Mark L. Williams
mlw@ncsc.arpa
------------------------------
Date: Mon, 28 Jul 86 14:37:34 PDT
From: rick@sdcsvax.ucsd.edu (Rick Randall)
Subject: Orchid's 286e Accelerator Board
Just my 2 cents for the accelerator board audience...
Got Orchid's 286e (1M on board, standard -- no wait states), and I after some
trial and error finding out which software caused problems, I like it!
I won't bother advertising the Norton SI rating, because although it is very
high, I don't really believe in it as a good measure. But it is *fast* for
the practical applications I indulge in. I guess it is "better than an AT"
for many applications, as claimed, but I have little experience on an AT, so
I can't judge. It sure is one heck of a lot faster than my native PC-1,
however. It's lovely to see the screen update almost instantaneously.
Has Ramdisk, spooler, and cache software, all of which appear to work well.
I had problems with the printer using my older AST spooler, but they cleared
up with Orchid's. I also had mysterious problems with the keyboard until I
identified the culprit as one of my ram-resident programs (NU-EPSON), which
doesn't seem compatible. Also, I was using the 1DIR (DOS shell) program that
came delivered with my QUBIE hard-disk, but found that it would not run in
the 286e mode -- apparently it accessed the controller card directly in some
way not supported, and decided I didn't really have a QUBIE. Solved this by
upgrading to the "generic" 1DIR for $25. [I wouldn't be without 1DIR, by the
way!]
There is very little else in the way of problems to report...I am getting an
occasional "missed keyboard interrupt" problem, which for various reasons I
am convinced is not due to software; I have not yet gotten through to Orchid
for assistance. It is more of a nuisance than anything else -- it pops you
back into the host (8088) mode with a message left on your screen, and you
have to re-boot the 286. Hope this is something like a jumper switch that
could be in a better position... Other than this, I have been *very* pleased
with my new machine.
I notice that the board comes with a fairly large number of "socketed" chips;
these, I presume, are either ROMs or proprietary chips and are capable of
being easily upgraded. I like this.
In order to switch between the -286 and the 8088 modes you have to run a
DOS-level program. In other words, there is no "hot-key" or hardware
switch to switch modes.
One of the advantages of going with an accelerator board is that you
can still run programs that must run on the 8088 -- i.e. that will
not run on an AT. This was the main reason I decided not to upgrade
to the AT or an AT clone. Another reason was that I couldn't see
sinking any more $$ into a 286 configuration when the -386 machines
will be coming out soon!
I have found getting through to Orchid somewhat frustrating. They
have an interesting "phone-menu" system: when you dial on a
touch-tone phone, there is a message that prompts you to dial one of
several 2-digit numbers accord- ing to the type of assistance
required. Dialing the number truncates the message and gets you to a
second layer of messages appropriate to your prob- lem, etc.
Finally, you have the precise area you want to talk to, you dial
their number...and it rings 6 or 7 times...precipitating a final
message indicating that you can dial another number to leave a
message. When this number also rings and ring, you can wait for an
operator -- if he/she answers! Modern technology to the rescue -- at
least it's more fun than getting a busy signal (although more
expensive, at long-distance rates). Last week, all the technical
personnel were at a picnic; this week, they haven't yet returned my
call. Are they busy or something?
The above diversion is not meant to discourage buyers; the product seems
very sound. Dissenters are welcome to chime in.
Rick Randall EECS Department
University of California, San Diego
decvax\ rick@sdcsvax.ucsd.edu
ihnp4 >---> sdcsvax ---> rick
ucbvax/ rick@sdcsvax.uucp
------------------------------
Subject: Micro Stock Software Summary
Date: Fri, 01 Aug 86 16:11:48 -0800
From: Scott Menter <escott@BONNIE.UCI.EDU>
Thanks much to everybody who replied to my request for microcomputer
software for stock market analysis. I'm excerpting here the replies;
they were very helpful, and I'm very grateful. (Sorry for the length,
but I didn't want to have to make any judgments about what parts of
these letters would interest everyone, and what parts not).
+------------------------------------------------------------------------+
E. Scott Menter Internet: escott@ics.uci.edu
UCI Systems Support Group UUCP: ...!ucbvax!ucivax!escott
"...Say, Foz, about my vocals..."
+------------------------------------------------------------------------+
------- Forwarded Messages
Subject: Re: Stock Program for Micros
From: SAC.HQSAC-ACMI@e.isi.edu
In 'Better Investing', a publication of the National Association of
Investors Corporation (NAIC), there is an ad for a program called
EvalForm 3.0. The Investment Club I belong to uses version 2.0
with a CP/M based machine. Although the program is designed to
support NAIC methods to evaluate a stock it is an excellent tool
for the investor. Data is gathered from either Standards and Poor's,
Value Line or both and entered into the program. The printout provides
buy/sell/hold price ranges, best price to buy, upside/downside, and
several other valuable stats. The report is about two pages long with
very little wasted space.
For more info write to :
INVESTOR'S SOFTWARE
BOX N
BRADENTON BEACH, FL 33510
(704) 743-2109
(813) 778-5515
and identify your computer.
For info on the NAIC or how to start your own investment club write to:
NAIC
1515 E. ELEVEN MILE ROAD
ROYAL OAK, MICHIGAN 48067
Marc Frederick
HQ SAC/ACCA, OFFUTT AFB NE 68113
================================================================
From: jhs@mitre-bedford.arpa
Organization: The MITRE Corp., Bedford, MA
Without a doubt, the micro with the most stock market forecasting
software is the IBM PC/XT/AT and clones. I don't have specific
recommendations on forecasting packages, but there is so much of it
that finding it shouldn't be a problem.
The Atari ST will, it is rumored, be able to run most IBM software
with an adapter to be announced soon (or is it already announced?).
However, your friend should probably just go buy an AT clone such as
the Zenith Data Systems Z248, or the IBM PC/AT itself, or comparable
units from AT&T, NEC, ITT, WANG, Leading Edge and many others. The
only reasons for using the Atari ST for this purpose would be to save
a lot of money and to have a 68000 processor to play with. From the
way you posed the question, neither goal seems especially applicable.
Buying a straight AT clone would undoubtedly minimize the unforeseen
complications in using standard software.
For a color monitor, I would suggest the NEC JC-1401P3A "Multisync"
monitor, which has very high resolution and is one of the few that
you can use for word processing without feeling bleary-eyed. (The
monochrome text display is the sharpest I have seen on a color
monitor.) Princeton Graphics is also said to have a very good
monitor. The Atari monitor might also be a good choice at a lot
lower price. Any of these would justify one of the hi-res graphics
cards that are available for the AT clones.
- -John Sangster
jhs@mitre-bedford.arpa
------- Message 3
From: chabris_c%h-sc4@harvard.harvard.edu
I can't discuss it since I haven't actually used it, but I hear good
things about the ISGUR Portfolio System from Batteries Included, for
the Atari ST and MS-DOS. It uses the GEM interface and includes an
excellent communications program, IS-Talk. Hope this helps.
Christopher F. Chabris
15 Sterling Road
Armonk, NY 10504
CompuServe: 73277,305
UUCP: ...!harvard!h-sc4!chabris
ARPA: chabris%h-sc4@harvard.ARPA
BITNET: chabris@HARVUNXU.BITNET
AT&T: (914) 273-8828
------------------------------
Date: Mon, 28 Jul 86 07:08:53 cdt
From: mlw@ncsc.ARPA (Williams)
Subject: Volume Labels
Bill Crocca asked (a while ago) about changing disk(ette) volume names
via Turbo Pascal...I can't address the Turbo Pascal part of the question,
but I dug up some info on the volume label itself which may help or may
inspire somebody else on the net to provide a definitive response...
The volume label is a special directory entry. The 11 characters that
ordinarily represent the filename.ext are used as one field for the volume
name. Using DEBUG to look at my hard drive showed that the volume name
wasn't in what I took to be any particular place and I guess that's right --
the file attribute field (the 12th byte of the directory entry) had a value
of hex 28. That maps out to the attributes "archive-system-volume label",
so I guess the entry could be anywhere in the directory.
I seriously doubt that normal DOS services will let you toy with the
volume name once it's established. I tried a plain "rename" command
against the whole volume name and the volume name with the last 3 char-
acters as an "extension." No go. It would probably work if the file
attributes were changed to the appearance of a standard entry, then
the renaming were done, and the attributes changed back. I have no
idea what facilities Turbo Pascal provides for any of this.
I hope that either gives you some help, ideas, or inspires other res-
ponses. Perhaps Bill can tell us how he ultimately meets this challenge.
Mark L. Williams
mlw@ncsc.arpa
------------------------------
Date: 1986 Jul 28 22:28 EST
From: Bob Babcock <PEPMNT%HARVARDA.BITNET@WISCVM.ARPA>
To: <crocca.wbst@xerox.COM>, <Info-IBMPC@B.ISI.EDU>
Subject: Volume Labels
In reply to a question from Crocca.wbst@Xerox.COM regarding the method
for writing disk volume labels from Turbo:
If you know how to execute DOS calls from Turbo, you're almost there.
It is not necessary do anything dangerous like writing directly to the
directory. I have successfully labeled disks using a C language
program as the interface with DOS. The same technique works for
floppies, hard drives and RAM disks. The only thing you can't do this
way is write a volume label with lower case or other illegal
characters.
First point: there are bugs which make MS-DOS functions 4EH and 4FH
unsuitable for reading volume labels. You can't find the label at all
if it's the first entry in the directory, and labels with embedded
blanks are not returned correctly. You need to set the disk transfer
address with function 1A, then use function 11H with DS:DX pointing to
an extended FCB with the attribute byte set to 8.
If the disk does not have a label, you can create one with function
16H. If the disk is labeled, you can change the label with 17H, the
rename function. After either of these, I think you also need to
close the file with 10H. (My code does it after a rename; I don't
remember if this is really necessary.)
Do not, under any circumstances, attempt to delete a volume label with
function 13H. You may trigger a bug which exists in some, if not all
versions of MS-DOS 2.11: deleting a volume label on a disk which has a
hidden file (such as the BIOS) as the first directory entry will
damage the FAT entry for that file. This will probably not be
immediately evident until you write additional files on the disk, at
which point you will find that some of them are cross-linked. I first
ran across this bug in a disk labeling program supplied by Sanyo. I
disassembled the program to see how it worked and built the faulty
algorithm into a disk indexing program I was writing at the time.
Before I realized what was happening I had zapped 14 disks. The bug
was reproducible on a Compaq, so it isn't just a Sanyo DOS problem.
Bob Babcock
pepap@cfa1.BITNET
pepmnt@harvarda.BITNET
------------------------------
Date: 1986 Jul 28 22:34 EST
From: Bob Babcock <PEPMNT%HARVARDA.BITNET@WISCVM.ARPA>
To: <mlw@ncsc.ARPA> , <hess@mit-multics.ARPA> ,
Subject: dCLOCK Clock
Responding to an inquiry from mlw@ncsc.ARPA regarding the dCLOCK
clock/calendar board:
I put a dCLOCK clock/calendar board in my slightly IBM compatible
Sanyo 555 a few months ago and it worked fine. The driving software
came in the form of lists of numbers and a BASIC program to convert
them into two .COM files, a technique sometimes used in PC Magazine.
One program is put into your AUTOEXEC.BAT file to set the DOS date and
time from the dCLOCK, the other transfers the DOS date and time into
the dCLOCK when you need to set it. I didn't actually have to type
these things in, I got them off a bulletin board run by the company
which sold them to our users group, along with a device driver which
forces all time checks to the dCLOCK. This isn't necessary on a
machine which keeps reasonable time, but it's nice on the Sanyo which
runs its software clock fast to compensate for the fact that it stops
during disk accesses. Unfortunately, the device driver doesn't allow
setting the dCLOCK. Meanwhile, I recently added a memory board which
has a clock on board, and interferes physically with the dCLOCK, so
I'm no longer using it. I never had any problems with the dCLOCK
while I was using it (a claim I wish I could make for my current
clock).
If you are buying more than one of these boards, you can probably
negotiate a better price. My Sanyo users group got them for around
$45 each.
The board is made by Microsync. Usual disclaimers apply.
Bob Babcock
pepap@cfa1.BITNET
pepmnt@harvarda.BITNET
------------------------------
Date: Tue, 29 Jul 86 15:49:21 pdt
From: decwrl!amdcad!cae780!ubvax!lance@ucbvax.Berkeley.EDU (Lance Keigwin)
Subject: dClock Clock
A couple of people asked about experiences with dCLOCK, a
real-time clock-calendar that fits directly onto the system's
motherboard in the 8088 socket. It does not occupy an
expansion slot, nor does it interfere with other cards.
We bought a few of these for some XT's that didn't need all
the features provided in today's multi-function cards. The
price was very attractive ($60 + $2 shipping), and the idea
deserved investigation. We were pleasantly surprised!
The dCLOCK in effect, fits *under* the 8088. Installation was
quite simple. The package included instructions for keying in
two BASIC programs of about 50 lines each which produced 2 .COM
files for reading and setting the clock (that was annoying, but
it helps keep costs down and makes shipping this tiny clock
easy -- it comes in a little tube).
As far as functionality is concerned, it works just like any
other battery-operated real-time clock. You set the time once,
and read the clock when you reboot. We have had no problems
with it.
Oh, you can get one of these little marvels from:
Microsync
Computer Products Dept.
P.O.Box 116302
Carrollton, TX 75011
(214)492-5265
------------------------------
Date: Mon, 28 Jul 86 11:09:19 cdt
From: mlw@ncsc.ARPA (Williams)
Subject: Security
Michael Domino asks about preventing unauthorized access to the hard
drive...
There are a few different ways to handle this, so here are some ideas.
Some are good, some not so hot.
1. Buy a key-lock for the system.
2. Use some other device (furniture, etc.) to provide physical
security.
3. Encrypt your hard drive information. There are a few software
packages that will do this.
4. Use a program like ALTER that lets you hide subdirectories and files.
Users who don't know what to look for will have trouble finding
anything.
5. Buy a security card for the PC (my favorite). A few vendors (I think
MISCO was one) sell a shamefully big card with lots of empty space
that forces a logon process whenever the machine is turned on -- no
matter which drive the system will be loaded from. Usernames &
passwords can be changed & several can be resident at once...
If one thought far enough ahead, I suppose one could get a Bernoulli Box
or equivalent and put the cartridges in a safe when they weren't in use.
You might not even need a hard drive!
Mark L. Williams
(mlw@ncsc.arpa)
------------------------------
Date: Tue, 29 Jul 1986 09:35 EDT
From: willie sanchez <PCZBH%CUNYVM.BITNET@WISCVM.ARPA>
Subject: Security
In response to the query by ugebrow%buffalo.csnet@CSNET-RELAY.ARPA
on protection of hard disk drives or complete system protection for
pc's, there are at least two such systems in the public domain/shareware
world.
For protection of just your hard drives while still allowing boots
and thus access to the pc's computing power, I use PC-LOCK by Johnson
Computing in Newport News, Va. It works through the config.sys file
so it is not easy to circumvent. You are prompted for a password at
boot time and if not correct will not allow u access to hard disk
drives. If u boot from a: or b: drive and attempt to access any hard
disk will give u an "invalid device" message, thus allowing use of
the pc's computing power while protecting hard drives. An authorized
user can hit the left hand shift key three times successively and
lock up the machine until their return at which point entering of the
initial password will unlock the machine (processing, if any will
continue while machine is locked!). Johnson Computing's address is
20 Dinwiddie Place, Newport News, Virginia 23602. It is available
through various bbs's. If you have trouble obtaining it please
advise, since i have a c copy of it. It is distributed as shareware
and a per pc charge of $17.50 is asse ssed for commercial use.
For locking up a pc, hard drive and computing resources, SYSLOCK can
be used. It also works through config.sys but can be circumvented by
booting with floppy. Software comes with documentation on disabling
the a: drive for booting and example batch programs to redirect
instructions to a: drive once you have booted. It is also available
on bbs's as public domain software. It is coms from J.C. Kilday
Associates, Peaks Island, ME 04108.
My personal opinion is that PC-LOCK is preferable. Please contact me
if u are unable to locate a bbs with this software.
------------------------------
Date: Tue, 29 Jul 86 15:05:40 cdt
From: moore@ncsc.ARPA (Moore)
Subject: Security
I just downloaded PC-LOK11.ARC from Trenton State BBS (609-771-2829), which
seems to do what you're looking for: if you boot from the hard drive you're
forced to supply a password; if you boot from a floppy you get 'invalid drive
specification' if you try to access C: w/o first supplying a password.
Jim Moore
NCSC
Panama City FL
------------------------------
Date: Mon, 28 Jul 86 14:52:53 edt
From: "Craig C. Douglas" <ccd%DUKE.CSNET@CSNET-RELAY.ARPA>
Subject: Numerical Recipes: The Art of Scientific Computing
To: marvel-h%osu-20@CSNET-RELAY.ARPA
This regards Howard Marvel's query about FORTRAN packages for computing:
... I particularly need good
matrix inversion (MicroWay offers an assembler program for this -- any
good?), random number generators, and distribution functions (normal,
central F, non central F, chi-square, etc.). These may be in assembler
or Fortran.
There is a new book from Cambridge University Press called Numerical Recipes:
The Art of Scientific Computing by William H. Press, Brian P. Flannery, Saul
A. Teukolsky, and William T. Vetterling which contains routines to do the
following (according to the book flap in front of me):
- solution of linear equations, matrix inversion, calculation of
determinants
- singular value decomposition
- interpolate and extrapolate polynomials and rational functions
- construct and evaluate splines
- integrate functions on open and closed intervals, improper
integrals, adaptive Romberg integration, Gaussian quadrature
- fit and evaluate Chebyshev polynomials
- accelerate the convergence of series
- compute special functions, including complete and incomplete gamma
and beta functions, Bessel functions, and elliptic integrals
- generate uniform, normal, Poisson, and binomial random deviates
- sort, rank, and index arrays
- find roots of generic functions and polynomials, and solve nonlinear
sets of equations
- find minima and maxima using Brent's method, variable metric methods,
linear programming, and simulated annealing method
- calculate eigenvectors and eigenvalues
- do fast Fourier transforms, sine transforms, cosine transforms,
power spectrum estimation, and convolution and deconvolution
- maximum entropy estimation and linear prediction
- calculate multidimensional Fourier transforms
- do statistical tests, including t-test, F-test, chi-square test,
- Kolmogorov-Smirnov test, contigency table analysis, and
nonparametric statistics
- fit by linear and nonlinear least squares and with robust methods
- Runge-Kutta and Bulirsch-Stoer integration for ODE's
- do two-point boundary problems by shooting and relaxation methods
- solve PDE's by simultaneous over-relaxation and ADI methods
- ... and much more.
All of the routines in the book are available on floppy in FORTRAN or Pascal
from the publisher. There are also example books and routines available
separately. The relevant ISBN numbers are
ISBN 0-521-30811-9 book $ ?
ISBN 0-521-31330-9 Numerical Recipes Example Book (FORTRAN) $18.95
ISBN 0-521-30956-5 Numerical Recipes Example Book (Pascal) $18.95
ISBN 0-521-30958-1 Numerical Recipes FORTRAN diskette, V1.0 $19.95
ISBN 0-521-30955-7 Numerical Recipes Pascal Diskette, V1.0 $19.95
ISBN 0-521-30957-3 Numerical Recipes Example Diskette (FORTRAN) $19.95
ISBN 0-521-30954-9 Numerical Recipes Example Diskette (Pascal) $19.95
[I assume these are IBM-PC format. Microsoft compatible?? -wab]
I paid $34.50 for the book a week ago (I do not know the list price). All of
these items have publication dates of 1986.
New York and California residents have to add local sales taxes. New York and
Canadian residents can order by calling 914/235-0300 and the rest of us by
calling 800/431-1580 (VISA/MC orders). Checks and VISA/MC are accepted by
mail to Cambridge University Press, Order Department, 510 North Avenue,
New Rochelle, NY 10801.
------------------------------
Date: Wed, 30 Jul 86 00:22:08 EDT
From: johnl%ima.UUCP@CCA.CCA.COM
Subject: Large Priam Disks for AT
We have a 60MB Priam "Inner Space" disk drive. It came with a device
driver that let us use the space not visible to the IBM BIOS. The
BIOS thinks that it's a type 12 drive, and can see a 32MB disk
partition. Priam's driver lets you divide up the drive any way you
want, in our case turning the rest of the disk into a 26MB partition.
I'd expect the same scheme to apply with the 71MB drive.
Priam's software is quite nice, and makes it easy to format and partition
the disk as desired.
John Levine, Levine@YALE.EDU or ima!johnl
------------------------------
Date: Fri 1 Aug 86 00:59:14-EDT
From: Joseph M. Newcomer <Joe.Newcomer@C.CS.CMU.EDU>
Subject: ProDesign II
I needed a good graphics system, and after having suffered through
PC Paintbrush and other really bad systems, I decided to break down
and get ProDesign II ($299, American Small Business Computers, 118 South
Mill Street, Pryor, OK 74631).
Well, there's good news and bad news. The good news is pretty good and
the bad news isn't so bad, so I rate this product rather highly.
It is not too hard to learn, although I still must keep the manual
nearby even after a couple dozen reasonably complex drawings. Like
most PC products, it reflects a horrid fascination with function keys
and the cursor pad, which makes it very hard to use (the quality of
life would be greatly improved if function keys and cursor pads
disappeared. Give me EMACS-like key binding ability!) But all in
all it is not as bad as many such programs. The one major defect it
has is that you configure it with a horribly elaborate program that
requires you reconfigure EVERYTHING even if it is the same. Thus
there are times when I'd like to do the drawing on my Hercules card
(because of the quality of the resolution) and then display it on my
CGA to check out the colors (even though that loses because there
aren't enough colors...) but this requires a full configuration. Or
set it up to print sometimes on my Canon color ink-jet printer, then
on my FX-80+, etc. A royal pain. I'd like to have startup-loadable
drivers instead of these silly configuration programs (I'd be happy
to exit and restart to change devices. PC Paintbrush is the biggest
lose ever, requiring a system reboot to reconfigure the device! Fine
for PCjrs but not the way to fly on an /AT).
The good news is that it really does some quality drawings. It has
ways of setting real-world units and putting down unit-calibrated
lines between two points, which they call "auto-dimensioning". I sat
here with a ruler and a map and "digitized" a map to our house, and
it went quite smoothly. I've done a number of drawings of things
like panel layouts, rack panels, etc. and they work real well. It
drives my 6-color Radio Shack plotter and FX-80+. My Canon color
ink-jet printer doesn't appear in the list, but may be compatible
with some other one there. Another piece of good news is the folks
there are quite responsive to suggestions, requests for help, etc.
and really try to help.
It has a nice multilayer capability, with up to 20 layers per
drawing. Great for providing different views of an object. I've
used this feature quite a bit.
It comes with a nice "quick reference" sheet that I set up in front
of me and has the same information in online help.
Minor gripes include the fact that it doesn't seem possible to change
the line width or type (solid or dotted) once the line has been
placed. The rotation of objects requires bizarre placement of points
(one of those 3, 4, 5 or 6-point affine transforms, I think. I've
found it easier to erase and redraw objects). The status line
doesn't display all the interesting status, but does display status
many other programs don't including relative and absolute coordinates
in terms of the user's calibration. Erasing a single line segment is
impossible; it always erases everything connected to a point, so you
can't select and erase a single line, only things connected to a
point.
It suffers badly from not having the EMACS philosophy about repeats
or arguments. If I want to go 8 steps over I have to hit the key 8
times; instead of typing ^U8^F I have to hit the cursor right key 8
times. No keyboard macros either. If I want to activate or
deactivate a layer of the drawing or change the current layer I'm
working on I have to go into a "layer menu", use the cursor keys to
get to a layer and select it on or off. I'd also like to be able to
package several layers up with a particular name and select by name
(one view is layers 2,7,9 and 13; another is 1,3,4,5 and 12, and
finally one is 6, 10 and 11). Needs an exchange-point-and-mark
option to get me back to where I started the line. A mouse is nearly
useless if you care about exact dimensions and grid alignment. If I
could just get key motions bound to the keyboard instead of those
silly cursor keys and get other functions without using silly
function keys, I would be able to use it significantly faster than I
currently do.
It has no "area fill" techniques like the Macintosh, and cannot
represent colors by textures on a monochrome display. Area fill is
to a solid color only and is a pain to use. If it weren't for the
auto-dimension feature and color I'd use a Mac instead, but the
ability to work to color compensates for a lot.
I talked with them and passed some of these ideas on and they said
"just wait until you see the LATEST version...out in a couple
months". I think this is one of the nicer CAD packages around.
I've been very happy with what I've produced using it, and can say
that unlike a lot of packages I've bought this is one I don't regret
buying; even with the problems it is still a pleasant package to use,
and worth the price (in fact it may be underpriced but I'm not going to
tell them...)
As usual, I don't have any affiliation with these people except as a
pleased-if-not-absolutely-totally-satisfied-in-every-tiny-detail
customer. But then I'm real picky about graphics software. This
ranks very high. If I could just get the selection and grouping
capabilities of McDraw as well it would be super.
------------------------------
Date: july 30 1986
From: <$29%DHDURZ2.BITNET@WISCVM.ARPA>
Subject: VDISK & PC3278/9 -- INT 15H Conflict
We have problems in operating the VDISK-driver (PC-DOS 3.0, 3.1)
and the PC3278/79 terminal-emulation-adapter-card simultaneously.
A probable cause may be that both, the VDISK-Driver and the 3278/79-
emulation-program, use INT 15H.
Does anyone have a suggestion for the circumvention of this
problem ???
Peter Stede, $29@DHDURZ2.BITNET
Heidelberg
West-Germany
------------------------------
Date: Wed, 30 Jul 86 10:00:34 ADT
From: wdw%ACADIA.BITNET@WISCVM.ARPA (Bill Wilder)
Subject: Multi-User AT
A lot of faculty wordprocessing at our university is done on a multi-user
micro (CP/M-like OS and WordStar). We would like to move up in the world and
provide WordPerfect access to faculty without the expense of individual
micros for each faculty member.
We would very much like to hear from anyone with experience using a
multi-user AT running DOS (not XENIX). What type of PC-like terminals do you
use? How is performance? Any problems running software (WordPerfect in
particular)?
Thanks for any leads you can give us.
------------------------------
Date: Wed, 30 Jul 86 13:10:01 GMT
From: Keith Dale <kdale@bbncc-eur.ARPA>
Subject: Epson JX Driver for MS Windows Wanted
A friend of mine has an IBM PC/AT with an EGA monitor and an Epson
JX-80. He would like to be able to print color pictures created with
Draw under Windows. When he's running Symphony under Windows there's
no problem, as Symphony has it's own driver; however, his Windows
distribution disks included no JX drivers. Does anyone know where he
can obtain one?
I saw that in Vol.5,#37 of Info-IBMPC Gabe Newell from Microsoft
stated that device drivers were posted on the Windows RoundTable on
GEnie ("the new General Electric Information Services network").
Perhaps someone who has access to that BBS (I don't) could
download/upload to an FTP-able site?? I have only Arpanet access.
Is this a reasonable request? Legal? Any reply would be welcome!!
Keith M. Dale
BBN Communications Corp.
------------------------------
Date: Wed 30 Jul 86 10:35:05-CDT
From: Bob Paver <CP.PAVER@MCC.COM>
Subject: Tek 4105 or DEC VT240 Emulators for PC
I'm looking for pointers to products that emulate either a Tektronix 4105
or a DEC VT240. We're using Leading Edge and Sperry PC's with a mixture
of Hercules, Tecmar, and IBM-clone graphics boards.
If you're aware of any, please let me know. Even better would be anyone's
personal experience with one. Do they work well? Good color? Resolution?
and so on.
Thanks,
Bob Paver (512) 834-3316
Microelectronics and Computer Technology Corp. (MCC)
9430 Research Blvd, Echelon Building #1
Austin, TX 78759-6509
ARPA: paver@mcc.com
UUCP: ut-sally!im4u!milano!paver
------------------------------
Date: Wed, 30 Jul 86 21:28:23 PDT
From: crash!pnet01!gof@nosc.ARPA (Jerry Fountain)
To: crash!noscvax!info-ibmpc@usc-isib
Subject: EGA Printscreen
We just purchased a Video-7 Vega (EGA compatible) + NEC Multisync for
our Engineering system at work as an upgrade of the normal CGA. The
results are great BUT we can no longer do screendumps (in EGA mode).
Graphics.com is only for the CGA modes. Does anybody have a PD EGA
dump program or know where I can get one?
Jerry Fountain
ARPA: crash!pnet01!gof@nosc
------------------------------
Date: Thu 31 Jul 86 13:00:23-PDT
From: Steve Dennett <DENNETT@SRI-NIC.ARPA>
Subject: dBase III Bug
A few months back I thought I saw some articles in various pc
magazines and online about a serious bug that turned up in the new
dBase III. However, on going back, I can't find any of them. Has
anyone heard about this bug, specifically, what it does? Since we
installed dBase III on one of our hard disks, it has died twice.
Thanks.
Steve Dennett
dennett@sri-nic.arpa
------------------------------
Date: Thu, 31 Jul 86 16:05:09 EST
From: Jim Ennis <JIM%UCF1VM.BITNET@WISCVM.ARPA>
Subject: Crossword puzzle maker wanted
Hello net,
I am looking for a crossword puzzle generator for the IBM PC family of
computers. This is meant to be used to generate puzzles for high school
language students in Spanish and Latin derivatives. I would appreciate
any pointers to a free (hopefully) or shareware or commercial product.
Jim Ennis
University of Central Florida
Computer Services
PO BOX 25000
Orlando, FL
32816
BITNET: JIM@UCF1VM.BITNET
ARPA: JIM%UCF1VM.BITNET@WISCVM.ARPA
CSNET: ENNIS@UCF-CS.CSNET
------------------------------
Date: Fri 1 Aug 86 00:16:27-EDT
From: Joseph M. Newcomer <Joe.Newcomer@C.CS.CMU.EDU>
Subject: Expansion Box: XT or AT Bus
I'm in the market for expansion boxes for PC class machines. I need two
PC or PC/XT compatible boxes and one PC/AT compatible box, although I
can easily limit the /AT expansion to have only PC and PC/XT compatible
cards with no serious loss. Any recommendations for good products or
warnings about turkeys?
Configuration 1: Expand PC with serial line card, Appletalk card
Configuration 2: Expand XT with tape cartridge, second full-height hard disk.
Configuration 3: Expand AT with 1/2 height 360K floppy, full-height hard
disk, tape cartridge, multi-serial-I/O.
------------------------------
Date: Fri 1 Aug 86 00:22:55-EDT
From: Joseph M. Newcomer <Joe.Newcomer@C.CS.CMU.EDU>
Subject: Serial I/O expansion
I need to add multiline asynchronous RS-232C serial I/O capability to
an AT-class machine. Minimum of 4 lines, very desirable to have 8, on
one card. Application will require special software be developed, so
COMx compatibility is second-order effect (does not need to be same
device, or have same register interface, etc.). *MUST* come with
sample source code for initialization and use in interrupt driven and
polled modes! I've spent too many hours in the Z80 world trying to
figure out how to initialize one of those multiprotocol chips when all
I wanted was 9600 baud async 8 data, no parity, 1 stop bit...
Recommendations welcome!
------------------------------
Date: Fri, 01 Aug 86 12:22:48 +0300
From: CHANA WOLF <CHANA%HBUNOS.BITNET@WISCVM.ARPA>
Subject: Accessing COM1 from Lattice C 3.00
I am trying to write a dumb terminal emulator in Lattice C and am having
trouble accessing COM1. I don't have the Cfood Smorgasbord. I tried
setting the communications parameters with pokes to ports (hex) 3fb and
3f8,3f9 (baud rate), and then accessing the input either as stdaux, as
a file, and directly from the receiver buffer register. Nothing seems
to work. Any suggestions?
Chana Wolf
Computation Center, Hebrew University of Jerusalem
chana@hbunos.BITNET
[Why not use the numerous packages offered in the info-ibmpc free lending
library! COM_PKG2.ASM is my favorite. -wab]
------------------------------
Date: Fri 01 Aug 1986 08:42:01 EDT
From: <DJEX@LL.ARPA>
Subject: PC-DOS V3.2 -- What are the problems???
Cc: djex@ll
PC-DOS V.3.2 -- What are the real problems??? Recently we
received some new PC/XT's with new keyboards and half height
360's. We are using the Tall tree Jram3's for the memory
expansion card. Everything goes fine until we try to format a
floppy disk. It appears to go through the format ok, but hangs
up before it gives the # bytes message. Without the Tall tree
board in, the format command works correctly. Anybody have a
handle (or Patch) on this problem?
I will be happy to do a summary of PC-DOS V3.2 problems and
solutions, so please "send me your problems (and solutions)".
Dan Enxing
MIT Lincoln Laboratory (L-234)
244 Wood St.
Lexington, MA. 02173
-or-
AUTOVON 478-5980 (ask for ext. 4177)
-or-
(617) 863-5500 x4177
-or-
djex@LL.ARPA
------------------------------
Date: Fri, 1 Aug 86 09:10:28 cdt
From: moore@ncsc.ARPA (Moore)
Subject: Bootable Disks
A coworker recently came to me with AT&T MS DOS 2.11 floppy in hand
and said that diskettes created with FORMAT B:/S aren't bootable, and
it's true. I tried on her machine (PC computable Monarch) and my
XT, and the dreaded non-system disk message comes up. I ran a
COMPare against the original, bootable COMMAND.COM, IBMBIO.COM, and
IBMDOS.COM files successfully. There must be something that /S or
SYS do *besides* simply copying those files. Any clues?
Jim Moore
NCSC
(904) 235-5158 (voice *and* data)
------------------------------
Date: Fri, 1 Aug 86 12:06:45 wst
From: munnari!swanee.oz!edmund@seismo.CSS.GOV (Edmund Lai)
To: munnari!info-ibmpc-request@usc-isib.ARPA
Subject: IBM PC Professional Graphics Display and Controller
We would be most grateful if someone can give us some details
regarding the IBM Professional Graphics Display Controller for the
PC.
We bought the hardware together with the graphics development
tool kit from a local IBM dealer in Perth, Western Australia and
were told that we can address each individual pixel on its 640x480
screen for our image processing work. However, a few thousand dollars
and a few months later NO ONE from the dealer or IBM Australia could
provide us with the information on how to do it (sound familiar? |-)
Someone from IBM Perth did mention about a book published by IBM
which contains the technical details on the workings of the controller
card. But he could not remember the title or order number. Does
anybody know?
Any information will be useful. Thanks in advance.
Edmund Lai
Dept. of Electrical & Electronic Engineering,
University of Western Australia.
ARPA : edmund%swanee.oz@seismo.css.gov
UUCP : ...!{seismo,mcvax,ucb-vision,uks}!munnari!swanee.oz!edmund
------------------------------
From: george@EDN-VAX.ARPA (George Sullivan)
Date: 30 Jul 1986 1444-EDT (Wednesday)
Subject: Micro Emacs 3.7
I seem to remember an announcement indicating that version 3.7 of microEmacs
had been completed. If so where can I ftp it from? Any other news about
it would also be appreciated.
Thanks in advance.
George Sullivan
Defense Communications Engineering Center
Reston, VA.
------------------------------
Subject: Epsilon EEL Language Question
Date: Mon, 28 Jul 86 20:52:46 -0700
From: Randy <randy@uw-bluechip.arpa>
I have a question about programming in eel.
To quote the rather scanty (should I say brief?) Epsilon manual:
Quote:
All loaded variables, commands, EEL subroutines, key tables, and keyboard
macros have their definitions stored in the name table, which is internal
to Epsilon....Entries can be retrieved by name or numeric index.
Endquote. (p. 130)
Does anyone know how to reference the name table? Say I have the numeric index
(or the name) of a variable. How do I determine it's value? I want to do
something like print the value of variable given it's index:
bprintf("Variable's value is: %d\n", Name_Table[index]);
How can this be done?
Thanks,
Randy Day.
UUCP: {decvax|ihnp4}!uw-beaver!uw-june!randy
ARPA: randy@washington
CSNET: randy%washington@csnet-relay
------------------------------
Date: Fri, 1 Aug 86 23:20:34 edt
From: Nathaniel Polish <polish%lexington@columbia.edu>
Subject: CMI Drives
Does anyone have any experience with the newer CMI drives? In
particular the 30 meg model 6640. BCE is advertising them in Computer
Shopper real cheap and also they suggest that "quantity discounts
available". The claim is that the 20 megs in the old ATs were bad but
these are fine. They are hard to sell because of the bad name. Any
thoughts? Anybody want to put together a bulk order, please send mail
to me immediately.
Also, these are for ATs with 39msec av access time.
Nat Polish@cs.columbia.edu
------------------------------
Date: 27 Jul 1986 13:06-EDT
Subject: How to Drive Multiple Graphics Screens from One AT?
From: MHARRIS@A.BBN.COM
I am still trying to find a way to drive several (different) graphics screens
using a single AT. All the EGA & clone cards (e.g. Orchid) seen so far are
hard wired to the same memory address, i.e. two of these cards will conflict.
I'm talking to the higher-priced folks this week (e.g. Matrox, Number Nine),
but I need to drive four screens -- and four of those cards cost a bunch.
The right choice still feels like EGA/clone cards. Do you know a way to put
their memories at different addresses? Or are you a manufacturer who can
help?
Or does anyone know a BETTER solution?
Thanks. -- Michael Harris
MHarris @ BBNA
617-497-3794
------------------------------
Date: 28 Jul 1986 19:52:42 EDT
Subject: PC-like Keyboard for AT Needed
From: Norman Lyons <LYONSN@A.ISI.EDU>
I just purchased a PC-AT clone. It's a good machine, but I have one
problem with it. The keyboard is very sensitive, it is easy to
unintentionally generate strange strings of characters. This problem
is even more acute for me because I suffer from Lou Gherig's Disease.
My hands are too stiff to handle such a keyboard.
I would like to continue to use my PC keyboard with the AT. The plug
fits, but the units are incompatible. Does anybody know if it is
possible to make them compatible by swapping chips around? If so,
how? Are there any firms that do this? Beyond this, does anyone
know of a keyboard like the PC keyboard that is compatible with the
AT?
Thanks in advance.
Norman R. Lyons LYONSN@ISIA (Arpanet)
Naval Postgraduate School 2872p@navpgs (Bitnet)
Monterey, California
------------------------------
Date: Mon, 28 Jul 86 21:52 EDT
From: "Roger C. King" <RCKing@MIT-MULTICS.ARPA>
Subject: IBM Logo
Can someone give me a pointer on where to find Logo Computer Systems
Inc., the originator of IBM Logo?
Roger King
------------------------------
Date: Mon 28 Jul 86 17:07:46-EDT
From: EDICK@TL-20B.ARPA
Subject: Assembly Language Routines package
To: info-ibmpc@B.ISI.EDU
Does anyone know of any assembly language routines packages compatible with
MS Fortran to allow interactive programming e.g. setting cursor location,
clearing screen.
Thanks.
------------------------------
Date: Tue, 29 Jul 86 12:41 EDT
From: slade.wbst@Xerox.COM
Subject: Large Screen Video Display
I need to display the screen of a PC to a large audience in a classroom
and/or lecture hall. Normal PC monitor would likely be connected to a
MDA (monochrome) with Hercules graphics board. ( Also curious about CGA
and EGA solutions).
What alternatives are there and how readable, i.e. text, is the
technique?
Michael Slade
arpanet: slade.wbst@xerox.com
------------------------------
Date: Tue, 29 Jul 86 16:18:44 EDT
From: George Keller (IBD) <keller@BRL>
Subject: "direct.h" Needed for grep.c
In order to compile grep.c with C86 in my machine, I need "direct.h."
Where is it found? If small, could someone mail it to me at
<keller@brl>. Thanks.
George Keller
------------------------------
End of Info-IBMPC Digest
************************
-------