Info-IBMPC@B.ISI.EDU (Info-IBMPC Digest) (10/11/86)
Info-IBMPC Digest Friday, October 10, 1986 Volume 5 : Issue 93
This Week's Editor: Phyllis O'Neil
Today's Topics:
CED Command Editor
DOS Command Line Editors
Environment Space
Music Symbol Processor Query
Multi-Language Word Processing
Curriculum for Computer Trade School
NEC Equivalent to Intel 8087 Math Co-Processor
XT Compatible Motherboards
Keyboard Query
----------------------------------------------------------------------
Date: Wed, 8 Oct 1986 12:00 EST
From: Juan Noyles <V4024%TEMPLEVM.BITNET@WISCVM.WISC.EDU>
Subject: CED Command Editor
While reading the digest today, I read an article about using superkey for
the command stack. I must say that the stack was exactly what I bought
superkey for, but I happened across CED, a public domain command editor
that is so much easier to use that I strongly recommend it to anyone who
has superkey for that purpose, or is looking expressly for a command
editor. It has two stacks for DOS commands and for user programs that
use the keyboard input DOS routine (OAH), adjustable stack sizes (down
to 128 bytes), command synonyms and much more.
The trouble is, I don't know how to make it available to the network!
Here at Temple we have some PC's attached to the Network indirectly via
RS232. They function as terminals using kermit. I know of no one here
that has been able to transfer a file successfully.
If anyone can explain how this is down, I will post the binary and doc
in either archive format or straight ASCII.
[INFO-IBMPC Has a "source language only" policy. We don't post binary
files. -wab]
Juan Noyles
V4024@TEMPLEVM.BITNET
------------------------------
From: "Roger Fajman" <RAF%NIHCU.BITNET@WISCVM.WISC.EDU>
Date: Thu, 09 Oct 86 17:35:26 EDT
Subject: DOS Command Line Editors
I highly recommend the Tallscreen package. It is a resident program
(actually a device driver), that adds to DOS functions such as
. Reviewing text that has scrolled off the screen
. Reusing past commands without retyping
. Returning to past directories
. Entering multiple commands on a line
. Screen editing for DOS commands
. Screen editing of screen data
. Ability to mark a block of screen data for saving to
disk, printing, or freezing in place on the screen
. Definable keys for DOS
Tallscreen costs $49.95 from Qualitas, Inc., 8314 Thoreau Drive,
Bethesda, MD 20817, telephone 301-469-8848.
It was reviewed in PC Tech Journal several months ago.
I have no financial interest in the product.
------------------------------
Date: Thu, 9 Oct 86 10:58:54 CDT
From: zinzow@uxc.cso.uiuc.edu (Mark Zinzow)
Subject: Environment Space
With everybody writing fancy C programs to get at the environment
space it seems some simple patches have been forgotten. So, for my
two cents worth, here is my standard blurb on the subject.
Heavy use of long paths and environment variables quickly uses up the
available environment space. When DOS is booted the environment space
is not fixed until you load a memory-resident program. Unfortunately
DOS considers AUTOEXEC.BAT to be a memory-resident program so it's
already too late to increase the environment space when the time comes
when you would normally fill it!
To solve this the environment space must be enlarged first. In Dos
3.x this can be accomplished with the SHELL command in CONFIG.SYS as
follows: SHELL=C:\COMMAND.COM /P/E:nn where nn is the number of 16
byte paragraphs (nn= 10..62 or 160-992 bytes) in DOS 3.00 and 3.10 or
bytes in the range of 160 to 32768 in DOS 3.20 (Default 160, rounded
mod 16)
In DOS 2.xx it is necessary to patch the object code in COMMAND.COM
where the environment space is allocated during the boot process.
There are two good techniques for this on PC-SIG disk #307. I quote
the one from the file ENVIRO.PAT here:
The following is a patch for COMMAND.COM (IBM-PC) to allow for a
larger environment area. Currently, only 223 bytes total are
allowed for environment string storage, with all data over 223 bytes
truncated. Use the following patch to increase the size of this
data area:
1) DEBUG COMMAND.COM (load COMMAND.COM into debugger)
2) S 100 4600 BB 0A 00 B4 48 CD 21 (search for allocation call)
3) Replace the 0A with the new area size factor (in paragraphs).
The string searched for in step 2 is a call to allocate space for the
environment area:
BB 0A 00 MOV BX,0AH ; area size factor=10 decimal
B4 48 MOV AH,48 ; allocation function number
CD 21 INT 21 ; perform allocation call
The area size factor is calculated as (A+4)*16, with A being the
area factor. The default value of 10 decimal yields an environment
area of 224 bytes, increasing the value to 20 decimal (14H) will
yield an area of 384 bytes; increasing the value to 30 decimal (1EH)
will yield 544 bytes.
In DOS 2.10, the area factor byte is plugged in at location ECF.
In my Zenith MS DOS 2.11 COMMAND.COM file this byte was at location 1025.
Mark S. Zinzow ARPA: zinzow@uiucuxc.CSO.UIUC.EDU
Research Programmer BITNET: MARKZ@UIUCVMD.BITNET
Computing Services Office UUCP: ihnp4!pyrchi!uiucuxc!zinzow
University of Illinois at Urbana-Champaign
150 Digital Computer Laboratory
1304 West Springfield Avenue
Urbana, IL 61801
(217) 244-1289
------------------------------
Date: Thu, 9 Oct 86 14:12:35 -0300
From: Yaniv Hakohen <yaniv%taurus.BITNET@WISCVM.WISC.EDU>
Should we call this `yals' ?
Yet another MS"DOS" ls. First I wrote it, then they told me that someone else
had done it too. But our machine has already lunched on that source, so I
couldn't compare. The source compiles with a plain-vanilla configuration
of the Microsoft C v3.0 compiler. All Unix options are implemented, appart
from those that don't make sense (user & group id's, symbolic links etc.)
[UNIX shell archive in YALS.ARCH --pmo]
Yaniv Hakohen
E-Mail: yaniv@taurus.bitnet
MailMail: Computer Science Dept., Tel-Aviv University, Tel-Aviv, ISRAEL.
------------------------------
Date: Thu, 9 Oct 86 21:27 CDT
From: Wilkinson@HI-MULTICS.ARPA
Subject: Music Symbol Processor Query
I am enquiring for a friend as to the availablility of public domain
software for 'music' (vs word) processors. IE a program that would
manipulate musical symbols on a screen and also output to a printer
using the IBM or some other graphics set. Being totally naive in this
area, I gather that composers use pencil and paper to write their
music(?).
Richard Wilkinson {Wilkinson@HI-MULTICS.ARPA}
------------------------------
From: nike!tslu@oliven.olivetti.com (Teh-Shang Lu)
Subject: Multi-Language Word Processing
> The most popular CHINESE CHARACTER PACKAGE so far is the KCCDS
> system. You can mix English with Chinese characters, and it
> supports lots popular software packages: LOTUS 1-2-3, DBASE III,
> .....etc,
I am sorry KCCDS is a typing error, it should be KCMDS.
>> We looked at KCCDS. While a nice enough package, it has drawbacks.
>> What killed it for us was its size -- since it takes about 280K to
>> run, wasn't enough memory left over to run our application under
>> it.
I have been running KCMDS with some large programs for about
9 months, so far I don't have any problems yet. My clone PC has
512K and lots of clone dealers are selling PC with a standard 640K
nowadays. Besides, the memory size KCMDS uses CAN be specified.
>> While it does indeed allow you to run Lotus 1-2-3 (for instance),
>> don't plan on having a spreadsheet of any reasonable size.
In IBM PC, a byte is used to represent a single character,
and each character has a corresponding display attribute. I am
not sure about the display attribute in KCMDS, but each Chinese
character takes only two bytes. Thus the spreadsheet size you can
have is between 67% to 100% of the spreadsheet size you can have
in English.
>> ...... While KCMDS may be the 'most popular',
>> DALT is the most widely-used -- over 40,000 installations (in China,
>> Korea, Japan, and U.S. academic and government). It requires a
>> Hercules graphics card, and at $965 it's not cheap, but...
KCMDS has been on the market for less than a year. In Taiwan
alone, it has over 10,000 installations, or 60% of the market.
The population in China, Korea, Japan, and U.S. is around 100 times
the population in Taiwan.
The price of KCMDS is $140/$170 for PC/AT. I agree a
Mercedes 450 is better than a Datsun, with 6 times the price.
>> To begin with, it handles 11 languages (Chinese, simplified Chinese,
>> Japanese [both Kana and Katakana], Korean, Thai, Russian, Greek,
>> English, Arabic, Hebrew)
KCMDS is designed for the bilingual market, the philosophy
behind bilingual software is to provide a bridge between a non-
prevailing language with the world's prevailing language -- English.
I know people who understand 3 or 4 languages, but 11 is
probably attractive for the United Nations or some reserach institutes.
Manufacturer of KCMSD has helped develop other English/Whatever
packages, the I don't think they have intention to combine bunch of
non-prevailing languages.
>> -- and actually does translation (both
>> transliteration and grammatical processing), rather than merely 'word
>> processing'. Additionally, it disambiguates input: if you enter
>> "He looked at the girl with a telescope", it will ask (in whichever
>> language was used for the input) who has the telescope -- the girl or
>> the observer?
That is a good feature, I do not think Lotus 1-2-3 or dBase III
can do that either.
Again, KCMDS is a GENERAL PURPOSE software, it is a bridge of
Chinese to many popular softwares: Lotus, dBase, Lattice C, Turbo
Pascal, PE, Multiplan, Sidekick, ... and many others. Does DALT
do that too ?
>> Next, it's hardware-based (comes with a 1/2-card), so it's not only
>> faster but doesn't eat RAM. In fact, it'll run on a 256K machine,
>> not that you'd want to do so.
Hardware based may NOT necessarily mean it is faster. If the
fonts are stored in EPROM, its access time may be 2 times slower
than if the fonts are in RAM. KCMSD also runs on 256K machines.
Back to the software/hardware issue, I do not think Lotus or dBase
will be as popular if a "1-2-3 Hardware Card" or "dBase Hardware Card"
is required to run the 1-2-3 or dBase.
>> Rather than using stored bit-maps, characters are generated
>> algorithmically ...
KCMDS designers chose the bit-mapped technique because they
want KCMDS to be upgradable to survive the fast technology
evolution -- laser printing, desktop publishing, and more memory.
Most languages with non-alphabetical graphic type characters are
very delicated proportioned, only bit-mapped technique can survive
the laser printing (or even the Epson LQ-800) technology.
KCMDS is available from: Jumbo Software
264 North Whisman Road, Suite 16
Mt View, CA 94043
------------------------------
Date: Thu, 9 Oct 86 17:50:25 pdt
From: Chris Kagen <nike!isieng!chrisk%isieng@seismo.CSS.GOV>
Subject: Curriculum for Computer Trade School
I am posting for a friend who is designing a curriculum for a
private computer training school. The school specializes in trade
applications of computer science. She seeks an existing turnkey
educational package that includes:
Modular, self-paced lessons
Hardcopy text accompanying software for IBM PCs
Lessons for these applications:
Word processing
Bookkeeping
Accounting
Computer-aided design (CAD)
This is a pretty big order; anything even close is of interest. She
would appreciate names, addresses, or phone numbers of people who are
developing such programs.
Thank you for your time and attention.
Chris Kagen {decwrl,allegra,sun,ut-sally,hplabs,
Integrated Solutions, Inc. shasta,topaz,lll-crg}
1140 Ringwood Court . . . !pyramid!isieng!chrisk
San Jose, CA 95133
(408)943-1902
------------------------------
Date: Thu, 9 Oct 86 23:21:53 pdt
From: "Ya`akov N. Miles" <bd%dac.triumf.cdn%ubc.csnet@CSNET-RELAY.ARPA>
Subject: NEC Equivalent to Intel 8087 Math Co-Processor
Does anybody know where and for how much I can purchase an 8 mHz math
co-processor for my NEC "V20" upd-70108 processor? I would prefer
a NEC math co-processor, but there are various contradictory rumours
as to the availability of the NEC co-processor chip. Please forward
any information (distributor, price, availability) to me at
BD@TRIUMFCL - on BITNET
I particularly want the NEC math co-processor, because it contains
special code which "handshakes" with the NEC "V20" and therefore works
faster. I think that it is also CMOS, so it should not run hot...
------------------------------
Date: Fri 10 Oct 86 16:33:06-EDT
From: Thomas S. Wanuga <WANUGA@XX.LCS.MIT.EDU>
Subject: XT Compatible Motherboards
Message-ID: <12245757151.39.WANUGA@XX.LCS.MIT.EDU>
I'm looking for an XT compatible motherboard with dual speed operation
(4.77 and 8 Mhz), and 0 wait state memory. Any pointers to companies
that manufacture such a board would be greatly appreciated. I'll
summarize any responses to the digest.
Thanks,
Tom Wanuga
wanuga@xx.lcs.mit.edu
------------------------------
Date: Thu 9 Oct 86 13:57:37-CDT
From: George Reimonn - Hospital Finance <G.HF3GRXX@CHIP.UChicago>
Subject: Keyboard Query
We have a couple of IBM-XTs with the new keyboards --
seperate numeric keypad and arrow keys. One user (usually in
WORD or 123) does something that stops the arrow keys from
working. Rebooting clears the problem, but we can't reverse the
lockup from the keyboard or stop it in the first place. We also
can't figure out what is causing it.
Anyone know what might be causing the problem? Is this
a user error or a bug?
George Reimonn, U of Chicago Hospital Finance Dep
------------------------------
End of Info-IBMPC Digest
************************
-------