Info-IBMPC@C.ISI.EDU (Info-IBMPC Digest) (06/25/87)
Info-IBMPC Digest Wednesday, 24 June 1987 Volume 6 : Issue 48
This Week's Editor: Billy Brackenridge
Today's Topics:
Review of PC MOS from Compuserve
Dos Books, Unix Shells
turbo c "rumors"
Mail order upgrades
IBMCACHE.SYS
OS/2 and 80386
TURBO'S TEXTMODE
CHEETAH puts 386 into 286 Socket
SCCS on DOS
Nasty DOS 3.2 REPLACE bug
More about TEXT BASE and NOTA BENE
System Application Architecture
Clarification of my FORTRAN question
Fortran (Several Msgs)
Today's Queries:
Bilog
X Window System
Appointment Calendars
Kermit on the Z-248
Tutorial Writing Software
More RAM for Jr.
Terrain Modeling and Statistical Package
Hercules Mono card and 80286
Wordstar 3.3 DELete/BackSpace patch?
80286 boards in AT&T 6300
INTEL-310 ABR
Turbo C Floating Point Patch
Problem formatting 1.44MB (2.0MB) diskette on PS/2 Mod 50
Any DBMSs that do Variable Length Records
FORTRAN
ECHO in MS-DOS 3.2 (Yes, Again!)
Multi-user Databases
INFO-IBMPC BBS Phone Numbers: (213)827-2635 (213)827-2515
----------------------------------------------------------------------
Date: Fri, 19-Jun-87 23:07:42 edt
From: David Farber <farber@UDEL.EDU>
Subject: Review of PC MOS from Compuserve
Notes on trying out PC-MOS/386 Release 1.01
Steve McMenamin [76530,231]
6/18/87
PC-MOS/386 ("MOS") styles itself as a multi-tasking operating system, designed
expressly for the 80386, that is upward-compatible from DOS. MOS is a product
of The Software Link, Inc. It is available in versions for one, five, and 25
users. I recently checked out the single-user version ($195) on a Compaq Desk
Pro 386 with 4MB RAM. This is a very informal report on my impressions of MOS.
I obtained MOS because I've been searching for a way to make as much RAM as
possible available to a large DBMS application I'm building. The main
constraint is that the DBMS does not exploit expanded memory, so all its RAM
must be contiguous, DOS-visible memory. The best I've done thus far without
resorting to hardware add-ons like Max-It is a 560K Carousel partition under
DOS 3.1 (544K under 3.2). In my evaluation of MOS I was specifically looking
for a way to increase this usable amount, while still retaining the multiple-
patition convenience of Carousel. I mention all this by way of letting you
know that I was _not_ especially interested in MOS's multi-tasking abilities,
or at least not as they relate to concurrent execution. Thus, I made no
special effort to test those abilities.
MOS claims to run DOS programs without modification, and with a few exceptions
I found this to be true. You can install MOS right on top of your DOS hard
disk, simply by using the .MSYS command to initialize the boot record and
then copying in the MOS files. MOS comes up with the next reboot. The file
formats seem interchangeable, and many familiar DOS sights are present:
config.sys, autoexec.bat, command.com, and others. MOS commands are _very_
similar in both form and function to DOS commands, though some are slightly
enhanced (for example, there is a built-in command line editor). If you
know DOS, you basically know MOS.
When you boot it up, MOS starts a single task, known as "partition 0", in
the first 640K of RAM. In addition to the usual inhabitants, this base memory
is home to the System Memory Pool (SMP), out of which certain memory is
dynamically allocated for use by device drivers and others. In this version,
it seemed that an SMP of 60K was adequate for up to three partitions.
One of the tech. support reps I spoke with told me that subsequent releases
would require much less SMP memory per task. The size of the SMP
requirement is important because it limits the usable memory in the base
partition. With an SMP of 60K I was able to use about 500K of the base
partition (these figures are approximate).
When you boot up, partition 0 is the accessed (and only) task. You
create additional partitions (i.e. add concurrent tasks) using the
.ADDTASK command. You specify the amount of memory you wish to make
available to the task, along with other optional information, such as
the task's ID number, the startup batch file, and so forth. Each task
is either "accessed" or in the background, and only one task can be
accessed at a time. The accessed task owns the entire display, so
there is no way to keep windows for multiple tasks on-screen at once
a la DesqView. You access a task by typing ALT-<keypad N>, where N is
the task's ID number (e.g. ALT-keypad 1 accesses task 1. If you find
that the ALT- sequence is incompatible with one of your applications,
you can temporarily disable it by typing ALT-999, which is the toggle
sequence for reenabling task switching as well.
Earlier I stated that overhead reduces the usable size of the base
partition to right around 500K. I was therefore glad to discover that
additional DOS-compatible partitions can be allocated as large as
630K, of which all but about 7K (for a copy of COMMAND.COM) is usable
by the application. The few tests I did showed the concurrent
processing to be quite smooth, with very little degradation
noticeable when I had three tasks going.
Here's a summary of what worked and what failed: Most of my everyday
tools worked just fine, including PCTools, Norton, SPF/PC, ZIM (a
DBMS), Mapmem, Turbokey, and Memwalk. Several other packages worked
to a limited extent. ProComm seemed to continue to execute disk I/O
after displaying the comm. screen on startup; then it issued a
strange error message, but proceeded to function properly. I ran
ProComm in multi-tasking mode, which performs screen I/O via the
BIOS, but one of the MOS reps said that this was not necessary, as
MOS would intercept and remap even direct video I/O from an active
background task. I haven't tried this out, so I cannot verify this
claim. ZAPCIS would carry on a CIS session successfully, but would
die -- with MOS complaining about an illegal IRQ (interrupt vector)
-- and hang the machine shortly after the session ended, but before
issuing the "chars per second" stats message. FilePath _seemed_ to
get along with MOS splendidly, but it turned out that it was only
really functional in partition 0. FilePath didn't seem to have any
effect whatever in the upper partitions. The only outright fatality
was Brief (version 1.whatever), which incurred a "General Software
Failure" immediately upon invocation. One of the tech reps said he
thought Brief did some "pseudo-multitasking" of its own, so he was
not surprised.
MOS comes with more or less the same additional facilities as does DOS:
an editor (switchable between line and visual modes), DEBUG, print spooler,
RAM disk manager, etc. There are also seemingly extensive facilities for
managing a multi-user system, including provisions for security and
network emulation. I was unable to evaluate these features as I was using
the single-user version. The documentation, on the whole, is well-written.
The vendor's technical support left quite a bit to be desired. The reps
themselves were quite helpful. The difficult part is getting _to_ them.
The Software Link demands that you have both your serial number and your
_invoice_ number on hand, evidently to prove that you purchased MOS from
them directly. Once they are satisfied, you are issued a client number,
which entitles you to a grand total of 30 minutes of telephone assistance.
If you should use your entire allotment, I was told, you would have to
"find a dealer or distributor" who could help. Maybe I've been spoiled by
the high level of customer support provided by Zanthe Information (supplier
of ZIM), but I found the whole MOS tech. support experience infuriating.
You'd think a company trying to secure a niche during the lull before OS/2
would make it a little easier for potential users to become addicted to
their product. I guess not.
Overall, I liked MOS but I can't really use it -- yet. Being able to make
over 600K of contiguous RAM available to an unmodified, non-EMS DOS
application surely does appeal to me. The smooth multi-tasking and clean,
Carousel-like partition switching also impressed me. (Of course, I'd just
finished a _very_ frustrating two weeks with DV 2.0, so perhaps I'm now
more easily pleased.) The only reason I don't switch at least one machine
over to MOS is its failure to support FilePath (or the equivalent) in the
large upper partitions. Once that happens, I'll take another look at MOS.
The next release of MOS is allegedly due in "30 to 60 days". It will
automatically be shipped free of charge to registered owners of Release 1.01.
That's a thumbnail sketch of my first encounter with MOS. If you have any
questions about it please do not hesitate to get in touch with me directly.
Steve McMenamin
------------------------------
Date: Sat, 20 Jun 87 09:57:55 EDT
From: rochester!srs!dan@seismo.CSS.GOV (Dan Kegel)
Subject: Dos Books, Unix Shells
In replay to David (telesoft!dar), who asked a lot of questions about MS-DOS:
About reference books: two good ones are the Microsoft MS-DOS Programmer's
Reference Manual (from Microsoft itself; try calling them), and a more
common book called Advanced MS-DOS. If you are serious, you will also want
the IBM PC Technical Reference Manual (from IBM; call 1-800-IBM-PCTB)
or the IBM PS/2 BIOS Interface manual ($75, same number).
About serial I/O: the default system is synchronous and useless.
You need to install a device driver that does a better job.
About Unix-like shells: you're probably better off just getting a set
of Unix-like programs (ls, rm, touch...) and using the COMMAND.COM shell
that comes with DOS. DOS can only pass 127 bytes of commandline to
exec'd programs, so shells aren't really able to do much wild card expansion.
Good luck!
- Dan Kegel (...{rutgers,seismo}!rochester!srs!dan)
Thou art gone, alas, info-ibmpc
Like a flower, thy spirit did depart
Thou art gone, alas, like the many
that have bloomed in the summer of my heart.
------------------------------
Date: Sat 20 Jun 87 13:40:07-PDT
From: Brad A. Silverberg <SILVERBERG@CSL.SRI.COM>
Subject: turbo c "rumors"
Before all you amateur detectives get too involved, be
aware that Borland has publicly acknowledged that it purchased
Wizard Systems and that portions of the Wizard C technology
have been incorporated into Turbo C.
Wizard C is *not* Turbo C. Many, many changes were made,
and the compiler was speeded up by a factor of 2.5 to 3 times,
and produces as good or better code.
Just in case you were wondering how well Turbo C is doing:
in just over a month (since May 14) -- we've shipped about
100,000 copies. I don't know what MSC's installed base,
but I'd bet (my stock option against Bill Gates') ours
is already larger!
Brad Silverberg
BORLAND INTERNATIONAL
------------------------------
Date: Sat, 20 Jun 87 16:43 EDT
From: <DSPIRO%BRANDEIS.BITNET@wiscvm.wisc.edu>
Subject: Mail order upgrades
Micky@ohio-state asked about mail order hard cards and 286 boards for
upgrading a pc. I've had good experiences with CompuAdd in Austin,
Texas (800 number is available from information). They are cheap,
and when on of the boards I bought from them burned out, they sent a
replacement right away.
The question might not be which cards are best, but rather why
someone would spend the money for the upgrade when a second 286
machine is only a bit more expensive. With prices dropping, the
marginal cost of a clone 286 based pc is just slightly greater than
starting from scratch and buying a new one.
And, then you'll have the old clunker to give to the kids!
(I don't have any financial stake in CompuAdd -- but with the way these mail
order firms are raking it in, I wish I did...) David Spiro, Brandeis U.
------------------------------
Date: 22 Jun 87 09:28 EST
From: 26-324%HARVBUS1.BITNET@wiscvm.wisc.edu
Subject: IBMCACHE.SYS
If it is not too late, please do not distribute my note about
IBMCACHE.SYS. I have been informed by IBM that there may be licensing
violations incurred if IBMCACHE is used with a system other than the one
it was originally delivered with. It still works, but I doubt you want
to publicize illegal use of IBM's or anybody's, software.
Rich Stillman
Bitnet: 26-324 at HARVBUS1
Arpa/Edunet: 26-324%harvbus1.bitnet
[In nearly 5 years of Info-IBMPC we have never advocated stealing
anybody's software. It was unclear to me whether IBMCACHE.SYS was
part of DOS 3.3 or OS/2 which isn't out yet. If it is part of DOS 3.3
I assume you can run it on an AT or PS/2 as long as you buy a copy
for each machine. If IBM considers IBMCACHE.SYS as bundled with the
hardware, they ought to know better. Courts ruled against those sorts
of tactics nearly twenty years ago.
Had this message arrived before the last digest I would have edited
it to reflect our stand against pirating software but published the
information anyway. -wab]
------------------------------
Date: Mon, 22 Jun 87 10:49:24 PDT
From: Jim Anderson <bilbo.jta@CS.UCLA.EDU>
Subject: OS/2 and 80386
You might be able to get away with using the 80386's 32 bit registers
under OS/2 if only one thread (system wide) is using 32 bit instructions.
The problem is, that OS/2 does not save the extra registers when it
switches tasks. On the other hand, it doesn't use them either.
------------------------------
Date: Mon, 22 Jun 87 00:05 EST
From: Joshua D. Males <josh%ILJCT.BITNET@wiscvm.wisc.edu>
Subject: TURBO'S TEXTMODE
Hi Folks,
I think I found a bug in Turbo Pascal V3.00. The procedure TEXTMODE, when
sent without any parameters, is supposed to return you to the most recently
used text mode. It works fine on a color monitor but really raises hell with a
monochrome. It puts the mono monitor into mono text mode but the CLRSCR
command makes minced-meat out of your window, your text color, and your
background color. I spent two days trying to figure out why my program worked
fine on color monitors, but kept on clearing the WHOLE screen and messin' up
my colors (you know what I mean: inverse, bold, normal) whenever I did a
clrscr on a small window.
Witness the following program:
(***************************************************************************)
Program Bug;
var ch: char;
procedure p1;
begin
textmode;
window(1,6,80,12);
textcolor(black);
textbackground(lightgray);
clrscr;
gotoxy(1,1);
write('Hello World!');
end;
procedure p2;
begin
textmode(7); (* I only use 7 for a mono, otherwise c80 for color *)
window(1,6,80,12);
textcolor(black);
textbackground(lightgray);
clrscr;
gotoxy(1,1);
write('Hello World!');
end;
begin (* main *)
p1;
read(kbd,ch);
p2;
end.
(***************************************************************************)
P1 clears the screen in regular colors and writes 'Hello World!' in bold
in the top left corner of the screen.
P2, however, uses a parameter of 7 for TEXTMODE, and it works. It draws an
inverse window and writes 'Hello World!' in the top left corner of the defined
window.
Has anybody out there seen this problem?
Yours Truly,
Josh Males
Jerusalem College of Technology
21 Rechov HaVaad HaLeumi
Givat Mordechai, Jerusalem
91160 ISRAEL
JOSH@ILJCT (bitnet)
------------------------------
From: yale!hsi!tankus@seismo.CSS.GOV (Ed Tankus)
Subject: CHEETAH puts 386 into 286 Socket
Organization: Health Systems Intl., New Haven, CT
Steve,
Regarding the motherboard, you might want to try Cheetah Int'l at
1-800-CHEETAH. They make the CHEETAH translator board that plugs into the
286 socket. They are supposed to release a motherboard in September and a
complete machine before the end of the year.
As far as replacing the machine, you have lots of options! Their are 386
machines available from Acer (formerly Multitech), Compaq, PC LTD, Wyse,
and many mail order houses. The Compaq and Wyse will run SCO or Microport.
I don't know about the others.
You might want to get on the 386 mailing list that was posted here.
Cheers!
-- Ed.
Net : {noao!ihnp4!yale!}!hsi!tankus
Snail: Health Systems Int'l, 100 Broadway, New Haven, CT 06511
Bell : (203) 562-2101
------------------------------
Date: Tue, 23 Jun 87 16:46:47 pdt
From: scubed!telesoft!dar@seismo.CSS.GOV (David Reisner @favorite)
Subject: SCCS on DOS
Organization: TeleSoft, San Diego, CA
SCCS was obsoleted by RCS (Revision Control System) some years ago.
Polytron ((800)547-4000) makes PVCS, which they claim is an RCS look and
act alike. There is a less expensive version missing a few features
(locks, user id on checkin, etc.) called VCS. I expect to use VCS within
the next couple of months (along with the MKS Toolkit - KORN shell and
"Unix" utilities).
-David
sdcsvax!telesoft!dar
------------------------------
Date: Wed, 24 Jun 87 10:34:28 EDT
From: jpp@ORNL-MSR.ARPA (J. L. Patton)
Subject: Nasty DOS 3.2 REPLACE bug
A problem exists using the REPLACE command under IBM's version 3.2x. When
adding files from the source to the destination (i.e. REPLACE d:\*.* d: /a),
REPLACE fails to inform the user he/she is out of disk space when that
condition exists. Instead it terminates normally and writes 0 length files
for those that would not fit! IBM was contacted and has verified this
condition. They stated that this problem has been corrected under version
3.30 and since version 3.2x would be withdrawn from the market on 7/10/87
that a patch would probably not be available. A Product Application Support
Request (PASR) has been initiated through the local IBM office to try and
coax a patch from Boca. Stay tuned!
------------------------------
Date: Wed, 24 Jun 1987 17:55 IST
From: {Amnon Aharoni Tel Aviv University} <S99%TAUNIVM.BITNET@wiscvm.wisc.edu>
Subject: More about TEXT BASE and NOTA BENE
In order to check the efficiency of Nota Bene's Text Base we have
carried out an experiment in which we have textbased Info-IBMPC
Digest from Vol. 5 Issue 70 up to Vol. 6 Issue 47, total of 84 files
sizing about 2.8 MB.
For saving as much drudgery and time as possible we have chosen to index
all words in text minus trivial and irrelevant words, for which Nota Bene
allows creating a special file of Omit List. (To get a reliable list we
textbased a preliminary sample and got some 12000 words.)
Each note in Info-IBMPC Digest was defined as an "entry" (search unit)
in the Text Base, totaling 2000 entries all in all.
There was no need to insert ANYTHING into the files themselves in order
for Nota Bene to be able to distinguish between the entries. We simply
defined the entry border with the word "Date: ", having observed that each
note has the word "Date: " in its header as an ever permanent item.
Here are the results of the actual text basing on an IBM PC/AT with original
hard disk of 20MB and 640 KB of RAM:
1. Indexing time:
3-5 minutes per Info-IBMPC issue, with an average of 7.5KB
per minute.
2. Searching time:
6-7 SECONDS per short query (3-4 keywords with any Boolean
operators and any number of occurrences, i.e., the number of
entries found as a result of a search.
3. Retrieval time:
Instantaneous, either entry by entry, or transferring to disk file.
Amnon Aharoni
Director of Information Systems
Tel Aviv University (S99 at TAUNIVM.Bitnet)
------------------------------
Date: Wed, 24 Jun 87 15:15:52 EDT
From: Ben Yalow <YBMCU%CUNYVM.BITNET@wiscvm.wisc.edu>
Subject: System Application Architecture
For those interested in IBM's Systems Application Architecture (the new
system concept linking OS/2 and the System/370 and S/3x systems), IBM
has come out with a new manual describing it (in rather general terms.)
The new manual is called "Systems Application Architecture : An Overview",
and is publication GC26-4341-0. It is available from your IBM marketing
rep, if you have one.
Ben Yalow
BITNET: YBMCU@CUNYVM
ARPA: YBMCU%CUNYVM.BITNET@WISCVM.WISC.EDU
------------------------------
Date: Mon, 22 Jun 1987 08:47:24 EST
From: Jim Ennis <JIM%UCF1VM.BITNET@wiscvm.wisc.edu>
Subject: Clarification of my FORTRAN question
Hello net,
Sorry to leave out some information from my previous posting. We have
the latest version of Microsoft FORTRAN for the IBM PC family. We are
wondering if people can recommend any more powerful packages over this
product. Also, have there been any compiler summaries in magazines lately
for FORTRAN that you can recommend.
Jim Ennis
University of Central Florida
------------------------------
Date: 22 Jun 87 12:12:02 PDT (Monday)
Subject: Fortran
From: Ready.PARC@Xerox.COM
1) Look in the July issue of PC TECH JOURNAL. There is a review
of all the available fortran compilers.
2) I have used the IBM Professional (R/M) Fortran Vers. 1.0 and
despite some niceties (error messages, extensions, etc.) the bugs
have killed me (seems large programs run into IF statement
execution problems that are predictable).
3) I have used the new version of the Microsoft fortran (Ver.
4.0, "Optimizing Compiler") and have been very pleased so far.
The system call routines and the MAKE facility makes life much
easier. However, it seems that the library installation
procedure does not work properly if the system has TSR programs
or device drivers loaded. I had to start with a CLEAN system.
Also, I have found that MAKE facility will also screw up at times
due to system resident drivers and programs. Also, there was a
bug reported to Info-IBMPC concerning nested IF statements (I
have not verified this).
Usual disclaimers.
Long live Info-IBMPC (I hope)
Steve Ready, PARC Xerox
------------------------------
Date: Mon, 22 Jun 87 21:04:33 EDT
From: Kurt Fickie (IBD) <fickie@BRL.MIL>
Subject: FORTRAN
There is a rather complete comparison of available Fortrans in
the "PC Tech Journal" June '87 . It compares 14 different
products including the big three: Lahey, Microsoft, and
Ryan-McFarland (RM).
We have some experience with Lahey and RM. Both are very good
and professional products. I liked RM, but we quickly found code
which would compile on mainframes like Cray and CDC, but bombed
or did not execute properly with RM (version 1.x). I must
confess that some of the code was very ugly with complicated
COMMON blocks and poor programming practices. I tried its
debugger and in some cases the code would starting executing
properly (much to my dismay) and other times the code was so big
with the debug option that it would no longer fit within 640K.
In desperation, we tried Lahey. All our problems went away. So
long as we followed ANSI 77 strictly, the compiler seems to
produce proper code. It compiles much faster (in one example,
RM took an hour and Lahey took 35 minutes). Furthermore, the
debugger has some nice features. The compiler produces debug
symbol tables automatically which is separate from the OBJ
files. Therefore, you can use the debug as an afterthought and
the EXE file produced on the link/edit step does not change size.
Of course, the debugger runs slower since it is reading the
tables off disk, but this never bothered us.
Lahey has a nice way of incorporating fixes. You can download
small "FIX" files from their BB and the compiler looks for this
file when it starts up. What little support we have needed has
been excellent. Our experience has been one day turnaround!
I understand the 2.0 version of RM is much improved. I hope
to try it out on code which failed previously. We did notice that
when it worked RM ran codes about 10-15% faster. The choice
might center on whether you use your PC to develop code or do you
use it for production. Also, if you do a lot of porting of
existing code, I would give Lahey the nod; if you are a good
programmer and write your own, then RM is probably the best
choice.
Kurt
------------------------------
Date: Mon, 22 Jun 87 10:17 EST
From: "GLENN EVERHART, 609 486 6328" <EVERHART%ARISIA%rca.com@RELAY.CS.NET>
Subject: Microsoft Fortran 4.0
I got the update to MS Fortran 4.0 (from 3.3) and happily made the
few edits needed to convert AnalytiCalc from 3.3 to 4.0, expecting
some speedup from the new optimizer. AnalytiCalc is a rather hefty
spreadsheet system I did which runs on PC, XT, AT, VAX, and PDP11
(and soon on Amiga). I noticed the assembly language pieces now need
to save two more registers and restore them on exit. Anyway, when I
was all done, I did timing comparisons on the old and new versions.
To my surprise, the V3.3 version was both smaller and faster than the
newer one in actual memory used - around a 10 to 15% effect at a
minimum, computing a moderate sized spreadsheet. I am not sure why
the 4.0 compiler gives larger and slower code than 3.3; I suspect it
may be the added subroutine calling overhead. Result: I have more or
less trashed the 4.0 compiler and kept 3.3.
Anyone using Microsoft (which I did to allow my code to run
on machines with and without 8087 (80x87) ) should be VERY cautious
about this upgrade and try it out before committing to it if
possible. You could gain a significant amount by keeping the older
version, or you might in SOME cases realize speedups for real
programs with the new one. As for me, I'm VERY disappointed with
V4.0.
Glenn Everhart
------------------------------
Date: Tue, 23 Jun 87 9:52:46 EDT
From: Kenneth Van Camp -FSAC- <kvancamp@ARDEC.ARPA>
Subject: Fortran Compiler Recommendations
Best recent article by far on Fortran compilers was in the June 87 issue of
PC Tech Journal, in which they reviewed 7 PC Fortran compilers in depth.
There are some excellent (and meaningful) benchmarks, and a good discussion
to boot. I highly recommend you get a copy; it may be still available from
your local bookstore. (For me, the biggest shocker of all was that there
were only two compilers -- Lahey and Microsoft -- that could handle the
2000-line benchmark the reviewer tried to compile! That should narrow down
the choices for many people.)
Also, since this group has had more discussion recently on
8088->80286 speedup boards, I might mention that the same magazine issue
also has a good article reviewing eleven of these boards.
--Ken Van Camp <kvancamp@ARDEC.ARPA>
------------------------------
Date: Fri, 19 Jun 87 13:54 N
From: <B127KLUI%HTIKHT5.BITNET@wiscvm.wisc.edu>
Subject: Bilog
We need some information on a program called BILOG. It is a program
to analyze l aten trait models. We haven't got the faintest idea if
it's a pc or mainframe package. We send a letter to the supposed
publisher, however the letter was retu rned as undeliverable. We know
the publisher is: International Educational Services, P.O. Box a3650,
Chicago, IL 60637. This address seems wrong. Anyone who can give us
some info on Bilog, please do.
kind regards,
Ruud Kluivers
Social Science Dept. Kath. Univ. Brabant
P.O. Box 90153
5000 LE Tilburg
The Netherlands
B127@htikub5.bitnet
------------------------------
Date: Sunday, 21 June 1987 14:32:40 EDT
From: Arthur.Butler@henry.ece.cmu.edu
Subject: X Window System
In the current issue of DECPRO EXTRA (6/87) in an article about the
X Window System, there is a mention of X running under MS-DOS on the
PC AT. Does anyone have any experience with this or any information
on it ? Thanks a bunch!
Arthur
(ajb@henry.ece.cmu.edu)
------------------------------
Date: Sun, 21 Jun 87 23:13:36 EDT
From: Jim des Rivieres <jeem%ai.toronto.edu@RELAY.CS.NET>
Subject: Appointment Calendars
I'm looking to purchase an appointment calendar for a person with a very
hectic schedule. I've had a look at Borland's Sidekick; it's fine as far
as it goes (not very far), only I'm looking for something with much more
functionality. E.g., moving appointments (not just insert and delete),
periodic meetings (e.g., first Tuesday of every month), alarm clock (e.g.,
beep me 5 minutes before my next appointment).
Thanks in advance for your suggestions.
---Jim des Rivieres (jeem@ai.toronto.edu)
Department of Computer Science, University of Toronto, Toronto,
Canada, M5S 1A4
------------------------------
Date: 19 Jun 87 17:19:00 EST
From: "Mark Layton" <mark@ntsc-74.arpa>
Subject: Kermit on the Z-248
After spending considerable time searching cu20b.columbia.edu 's
Kermit library, I've decided that there is no version that can use
the COM3 port on the Z-248 for communications. If anyone can point me
in the right direction, I'd be grateful.
Thanks,
Mark Layton
Naval Training Systems Center
Code 741
Orlando, FL 32813
ARPA: mark@ntsc-74.arpa
[As Kermit sources are available you could patch the sources or at least
look at the sources to find where to make the appropriate patches. -wab]
------------------------------
Date: Mon, 22 Jun 1987 17:40 CDT
From: DANIEL%TECMTYVM.BITNET@wiscvm.wisc.edu
Subject: Tutorial Writing Software
I'm looking for a Software Package running in a PC Which be able
to build Tutorials like the tutorial " Learning Show Partner " that
comes with the Show Partner in the MS-Mouse Package.
What I need is software which makes the user take actions where
the program coordinate all those actions.
My goal is to build a Bitnet Tutorial for the new Latin American
users, instead of teaching courses about it, so I need more than a
Presentation software that replaces only a slide projector.
Regards,
Daniel Trujillo
Monterey Institute of Technology
Mexico.
[Issue #33 of the digest mentioned ScriptWriter which was written at USC.
For more information contact: horowitz@usc-cseb.usc.edu 213-743-0940.
I am glad to welcome Mexico to the list of countries participating
in Info-IBMPC -wab]
------------------------------
Date: Mon, 22 Jun 87 19:58:25 EST
From: John <JOHN%NCSUVM.BITNET@wiscvm.wisc.edu>
Subject: More RAM for Jr.
Hello,
Does anyone out there know of a way to increase the amount of
memory in a PCjr to more than 128K. Thanks for any comments.
John
------------------------------
Date: Mon, 22 Jun 87 16:54:45 EDT
From: Bharat.Dave@CAD.CS.CMU.EDU
Subject: Terrain Modeling and Statistical Package
I would like to hear about any public-domain or (barring that) relatively
inexpensive software for the following:
1. Terrain modeling (as in land surveying)
2. Statistical packages
I am requesting this info. on behalf of School Architecture and Planning,
Ahmedabad (INDIA). They have a few PC-compatibles running under DOS. They
also have a mini that operates under CP/M. Any pointers would be greatly
appreciated. Thanks.
- Bharat Dave dave@cad.cs.cmu.edu (ARPA)
[Statistical packages have been frequently discussed in previous digests.
Land surveying is a new topic. -wab]
------------------------------
Subject: Hercules Mono card and 80286
From: oxy!bagpiper@csvax.caltech.edu (Michael Paul Hunter)
Date: 22 Jun 87 19:52:39 PST
I just got through writing some routines to use on a PC
with a Hercules Monochrome graphics adapter. I used the SetMd
and friends that were in the back of the Hercules manual. Lo
and behold, when I try to run the same code on an AT-Clone
(with the same Hercules card), it dies. Locks up good and
tight...need a good cold boot and all....any ideas folks...I
can get some code samples if anyone is interested. Could it
be those memory addressing modes on the 80286? (I've never
programmed it...only the 8088) Any help would be greatly
appreciated.
Michael Hunter UUCP : ...{seismo, rutgers, ames}!cit-vax!oxy!bagpiper
Box 241 ARPA : oxy!bagpiper@csvax.caltech.edu
Occidental College BITNET: oxy!bagpiper@hamlet.bitnet
Los Angeles, CA 90041 CSNET : oxy!bagpiper%csvax.caltech.edu@relay.cs.net
------------------------------
Date: Mon, 22 Jun 87 20:18:40 PDT
From: larry@Jpl-VLSI.ARPA
Subject: Wordstar 3.3 DELete/BackSpace patch?
Can someone tell me how to patch Wordstar 3.3 so that DELete does a
delete-right and BackSpace does delete-left? Thanks in advance.
Larry @ jpl-vlsi.arpa
------------------------------
Date: Tue, 23 Jun 87 13:22 N
From: <KELDERS%HGRRUG5.BITNET@wiscvm.wisc.edu>
Subject: 80286 boards in AT&T 6300
To: info-ibmpc@c.isi.edu
X-Original-To: info-ibmpc@c.isi.edu, KELDERS
Hello, I have a question about "upgrading" an AT&T 6300 (Olivetti M24)
to a 80286 or even a 80386 system. Has any one had experience with 80286
boards or the like in those machines?
Thanks in advance,
Johan Kelders
Computing Center University of Groningen, Netherlands
KELDERS@HGRRUG5.BITNET
------------------------------
Date: 23 Jun 87 07:59:39 EDT
From: dicke @ belvoir-mail1.arpa
Subject: INTEL-310 ABR
Has anyone out there been able to get their Intel-310 to autobaud detect?
Our 310's are the only systems on our network that don't have this cap-
ability, and my Intel systems administrator says ABR isn't available. Is
this really true?
------------------------------
Date: Tue, 23 Jun 87 09:40 CDT
From: "Think First, Program Later" <LANTZ%eg.ti.com@RELAY.CS.NET>
Subject: Turbo C Floating Point Patch
According to the latest PC week a patch to TURBO C which fixes the
floating point problem has been posted on CompuServe. Could someone
please post the fix to INFO-IBMPC for those of us who cannot get to
CompuServe? Thanks in advance.
BERNIE LANTZ
Texas Instruments- Spring Creek
------------------------------
Date: Tue, 23 Jun 87 18:54:01 EDT
From: G B Reilly <reilly@facman.wharton.upenn.edu>
Subject: Problem formatting 1.44MB (2.0MB) diskette on PS/2 Mod 50
Scenario: I format a 1.44 MB diskette. I try to format another.
Problem: On second diskette format get the message that either
track 0 is unreadable or media is bad.
Fix: Reboot
The problem happens every time. The fix works each time. Has anybody
else seen this?
------------------------------
Date: 24 Jun 1987 10:47-EDT
Subject: Any DBMSs that do Variable Length Records
From: MHARRIS@G.BBN.COM
I thought I was going to use Informix, but the database I'm building (on
an AT clone under DOS 3.2) consists of text fragments (5 to 10000 bytes);
variable length records seem necessary, and Informix doesn't do that. Is
there a DBMS (or some other utility) that solves the problem in a reasonable
way? I don't know much about PC DBMS software, so please assume ignorance.
Thanks.
-- Michael Harris (617) 497-3794 MHarris @ G.BBN.COM
------------------------------
Subject: FORTRAN
Date: Wed, 24 Jun 87 14:06:23 EDT
From: tgw@mitre-bedford.ARPA
Jim,
I've been using the new MS 4.0 Fortran Compiler for about a month now
and it is a vast improvement over all the older versions. Best of all
the documentation is very good. I have not used the codeview debugger
but it also looks good. Unfortunately I have not looked at Lahey
RM. I am happy with MS and plan to continue using it.
Tim Wade
tgw@mitre-bedford.arpa
------------------------------
Date: Wed, 24 Jun 87 15:12:52 CDT
From: moore@ncsc.ARPA (Moore)
Subject: ECHO in MS-DOS 3.2 (Yes, Again!)
I really need an answer to this:
Can someone tell me the areas to patch to turn ECHO OFF in batch files running
MS-DOS 3.2 ala Zenith? It's definitely *NOT* generic MS-DOS.
Thanks,
moore@NCSC.arpa
------------------------------
Date: Wed, 24 Jun 87 18:50 EDT
From: Hess@MIT-Multics.ARPA
Subject: Multi-user Databases
Has anybody used an existing database that runs on both Mac and IBM
hardware, with one machine acting as the server, and multiple users on a
mixed network doing queries?
If not, how about an opinion or twenty about a choice between the
all-IBM multi-user choices: Paradox and dBase III?
Please reply to me directly; if there is a good summary I'll post it.
Brian
------------------------------
End of Info-IBMPC Digest
************************
-------