Info-IBMPC@WSMR-SIMTEL20.ARMY.MIL (11/23/88)
Info-IBMPC Digest Tue, 22 Nov 88 Volume 7 : Issue 55
Today's Editor:
Gregory Hicks - Chinhae Korea <COMFLEACT@Taegu-EMH1.Army.Mil>
Today's Topics:
Daisy Wheel Printers
Disk Change Line Status failure cause of problem?
Using 1 Meg query (2 msgs)
Multiple Delivery of the Info-IBMPC Digest (2 msgs)
DOS 3.0/3.1/3.2 Differences
Directory Caching Problems
Today's Queries:
Downloadable Fonts for Hercules Card
EMS Ramdisk
TSRs and DESQview
One Diskette Copying
New Programs Available:
XUU.c (3 msgs)
----------------------------------------------------------------------
Date: Mon, 21 Nov 88 08:32:48 EST
From: carlson@gateway.mitre.org
Subject: Daisy Wheel Printers
In Info-IBMPC #54 someone asked about daisywheel printer recommendations.
I have used a Brother HR-15 (low end of the Brother line) for about
4 years and it has worked extremely well. It probably doesn't have the
durability of a Diablo, but it emulates the Diablo accurately and only
cost about $380 in 1984. An external typewriter keyboard, tractor
feed, and sheet feeder are available (standard mechanism is friction feed).
I think the keyboard retails for about $150. Warning! - this printer is
not fast, about 13 CPS. Brother does make some faster models, up to about
40 CPS.
Some software has Brother drivers, but in most cases I have just selected
the Diablo 630 driver if it is available. All of the Diablo formatting
seems to work.
When you buy Brother single strike carbon ribbons you have to make sure
you are getting the computer ribbon, not the typewriter ribbon.
The typewriter ribbon will work, but may throw off flakes of carbon
because it wasn't designed to work at the speed of a computer printer.
I haven't had any trouble with the standard Brother plastic print wheels
and I don't know if a metal wheel is available.
Bruce Carlson
carlson@gateway.mitre.org
Disclaimer: The views above are my own and may or may not agree with the
views of my employer. I have no connection with Brother, other than as a
customer.
------------------------------
Date: Sun, 20 Nov 88 20:00:16 PST
From: Ya'akov_Miles@mtsg.ubc.ca
Subject: Disk Change Line Status failure cause of problem?
Failure of non-IBM "AT-class" machine to detect diskette change...
An "AT-class" machine should detect a floppy disk signal called
"Diskette Change", that becomes true whenever a diskette is removed
from the high-capacity drive. The IBM bios contains a special call
(INT 13h, AH=16h) to determine if someone has indeed opened the door
to the floppy disk (Condition= "Media Unestablished..") I suspect
that your BIOS either fails to correctly implement INT 13h, AH=16h,
(dl=drive number 0=A:,1=B:,etc) or that there is a design defect or
hardware malfunction in your non-IBM machine...
------------------------------
Date: Mon, 14 Nov 1988 10:53 EST
From: "Paul G. Weiss" <PGW@XX.LCS.MIT.EDU>
Subject: Using 1 Meg query (2 msgs)
In the recent addition of the Microsoft System Journal they discuss how to
access the 64K of memory just *above* 1Meg. The trick is that on 286's
and 386's it is possible to enable a 21st address line, then use FFFF:10
onward to address an additional 64K. They offer a driver that allows one
to do all these neat things, called HIMEM.SYS, free to programmers by
calling (800)426-9400 (Microsoft's Information Center).
[See following message... gph]
------------------------------
Date: Mon Nov 21 13:02:45 1988
From: Gregory Hicks - CHinhae Korea <GHICKS@WSMR-Simtel20.Army.Mil>
Subject: Using 1 meg Query (HIMEM.SYS)
I contacted MicroSoft regarding the message from Paul Weiss and
received the following information in response:
As mentioned in the latest edition of MSJ, the HIMEM.SYS driver will
allow programs to access an extra 64K of memory that is addressable in real
mode, meaning that you can execute code there. Programs that use this, for
example, are Microsoft Windows and Microsoft DOS LAN Manager. HIMEM.SYS is
included with MS-DOS Version 4.00.
One thing I'm not sure the article mentioned was that this driver also
allows a method to access extended memory in blocks rounded up to a 1KB
size (an EMB, Extended Memory Block). You can allocate, free, move, and
lock one of these blocks, making it easier to manipulate than normal ex-
tended memory (from real mode, that is).
HIMEM.SYS is the Microsoft implementation of an XMM, an eXtended Memory
Manager, based on the eXtended Memory Specification 2.0. One of the nicest
things about HIMEM.SYS is that it abstracts all the OEM dependencies
regarding extended memory availibility and A20 line twiddling.
Microsoft is distributing a printed copy of this XMS specification,
along with their HIMEM.SYS driver (with MASM source code), a C library in-
terface, and test code. It's free for the asking to anyone that calls
(800) 426-9400 and asks for the "XMS, Extended Memory Specification, and
it's supplemental diskette". (It'd most likely only be useful to software
developers.) Make sure you ask for that disk as well, since it includes
all the intersting stuff:
HIMEM.SYS Microsoft's XMM: HIMEM.SYS Version 2.04
CLIB\*.* A library for calling the XMM from C, written in MASM
OEMSRC\*.* MASM source code for HIMEM.SYS
SPEC\XMS.TXT A machine-readable version of the XMS Spec
SPEC\HIMEM.TXT User documentation for the HIMEM.SYS driver
TESTS\HITEST.* A simple XMS test program with MASM source code
TESTS\TEST.* A more advanced XMS test program with C source code
TESTS\XMSTIME.EXE An EMB timing test program
To run HIMEM.SYS you need a 286 or 386 system with *some* extended
memory.
To *build* the test programs CLIB and TESTS stuff, you need Microsoft
MASM 5.1 and C 5.1. (And you need to have HIMEM.SYS installed before you
run the XMS tests...)
The only bad part about XMS is that people would sometimes confuse it
for an EXPANDED memory manager. It's not! It's an EXTENDED memory
manager, and the first 64KB of this extended memory is even more special.
If you have any difficulties or have modifications to the sources to
support a new OEM hardware configuration, please report them to MicroSoft
over the Microsoft OnLine support system.
Please excuse any typos as I've not slept in quite some time...
Disclaimer: Everything mentioned here is MY fault, not my employer's or
MicroSoft's...
Now then: Since it is a VERY expensive long distance telephone call
for me, is there anyway for me to get this package without calling?
------------------------------
Date: Mon, 21 Nov 88 13:26:28 est
From: amlovell@phoenix.Princeton.EDU (Anthony M Lovell)
Subject: Multiple Delivery of the Info-IBMPC Digest
> Info-IBMPC Digest Sat, 19 Nov 88 Volume 7 : Issue 54
At our site, most of your digest issues get posted twice.
Is this a problem in your practices?
------------------------------
Date: Mon, 21 Nov 88 19:47:08 MST
From: Gregory Hicks <GHICKS@WSMR-SIMTEL20.ARMY.MIL>
Subject: Multiple Delivery of the Info-IBMPC Digest
Well... Posting a message twice is not the usual practice.
Issues 52 and 54 were different. The mailserver they're routed to the in-
ternet through rejected this particular issue the first time around. I
broke it into two parts and re-mailed without thinking that all of the ad-
dresses did not have to be in the re-mailing.
Regret inconvience.
Gregory Hicks
PS - If you get two copies of each issue, please send me the headers.
That's the text of the message from the very first line down to the
line that says "Info-IBMPC Digest..." I'll try and figure out how
come you're getting two copies. Some people on the BITNET side of the
net are, in fact, entered in the list twice (the BITNET list that
is...) I have been unable to change the bitnet list for the past
month however.
PPS- Before you send me the headers, send me a note and tell me they're
coming. I may be able to fix it with a quick edit to the list...
Oh yes! I need a copy of both sets of headers...
(Anthony, if you'd send me the headers, I'll take it for action)
------------------------------
Date: Mon, 21 Nov 88 17:26 PST
From: "Jim Goes (goes@oregon.uoregon.edu)" <GOES@oregon.uoregon.edu>
Subject: DOS 3.0/3.1/3.2 Differences
In Info-PC Digest #54, Joe Boyd wrote...
>Numerous "getting smarter everyday" USERS are beating my staff up
>frequently to explain the EXACT differences in the most recent versions of
>MS-DOS. The history/facts of 1.x => 2.x => 3.0/1/2 are generally well
>known, but the EXACT significant differences of the above versions are
>not. Any help appreciated....
I reformatted my 40-meg hard drive in order to install DOS 3.3, only to
find the upgrade cost me over a meg in disk space! The culprit?? 4 k-byte
file clusters. MS-DOS 3.2 used 2k clusters, and I had a lot of small files
(batch files) on my disk. So the upgrade raised the minimum disk usage to
4k for all these files, and the cumulative effect was amazing. THAT'S a
"salient" difference. If anyone knows of a patch to force DOS 3.3 to use
2k clusters, I'd love to hear about it. I went back to DOS 3.2 rather than
lose the disk space.
Jim Goes | InterNet GOES@OREGON.UOREGON.EDU
Graduate School of Management | UUCP:
University of Oregon | MaBell: (503) 686-3309
------------------------------
Date: Sun, 20 Nov 88 15:08:40 EST
From: Jeff_MacKie-Mason@um.cc.umich.edu
Subject: Directory Caching Problems
First, if you have a buffers=nn statement in your config.sys, then DOS *is*
caching your disk reads. This is not a new feature.
I had the same problem with a 3.5" floppy drive that you describe. The
drive is supposed to signal to DOS that the disk has been changed when you
push the eject button. On my drive (a Toshiba, I think?), it turns out
that there is a jumper on top of the drive. The drive came with an in-
struction sheet that said on some AT-type machines it might be necessary to
move the jumper from from pins 1&2 over to pins 2&3 (or some such). Once I
did that, DOS was correctly informed that I had changed disks and I didn't
need to use the CTRL-C alert.
ciao, jm2
Jeff MacKie-Mason
Dept. of Economics
University of Michigan, Ann Arbor
Jeff_MacKie-Mason@um.cc.umich.edu
------------------------------
Date: Mon, 21 Nov 88 20:12:24 -0800
From: rtobin@ICS.UCI.EDU
Subject: Downloadable Fonts for Hercules Card
From the Sharp APL archive documentation <msdos.apl>:
> Finally, if you have a Hercules display, you can use Paul
> Jackson's 'pjackson.fnt' font definition, loaded with whatever font
> loader came with your Hercules or Hercules compatible. Paul Jackson's
> font is his own property and copyright; everything else in this package
> is copyrighted by I.P. Sharp Associates.
My hercules display card (a cloney-clone MG-150) has no such utility to
download fonts into my card. Does such a program exist? Does the simtel20
archives hold it? I am excited about this APL interpreter, but the special
APL characters do not display correctly. I have inquired to Sharp As-
sociates and to date have not gotten a response. Any info will be sum-
marized for the board. Thanks in advance.
Roy Tobin
------------------------------
Date: Mon, 21 Nov 88 10:48:29 CST
From: "Henry Mitchel" <AGRIPM%UMCVMB.BITNET@CUNYVM.CUNY.EDU>
Subject: EMS Ramdisk
Any suggestions or pointers to code that implements a ramdisk using
LIM/EEMS 3.2 memory? Public, shareware or commercial all OK. I just
haven't seen any advertisements for something like this.
Thanks in advance.
Henry Mitchel.
------------------------------
Date: Sun, 20 Nov 1988 22:24 EST
From: SGP@XX.LCS.MIT.EDU
Subject: TSRs and DESQview
Reading the TS R note via with desqview makes me pose the following
query:
Using Blaise Turbo Tools and Asynch or Turbo Power tools is there a way
to place a ISR TSR program (like BACKCOM) into extended or expanded
memory??
SG Pauker
SGP.XX.LCS.MIT.EDU
------------------------------
Date: Sun, 20 Nov 88 21:38 EST
From: <ROSSB%IUBACS.BITNET@CUNYVM.CUNY.EDU>
Subject: One Diskette Copying
I've a friend with a Toshiba 1000 who's going crazy having to constantly
switch diskettes into the one diskette drive in order to copy files. Does
anyone out there know of any public domain or commercial program that can
read larger chunks of files into RAM and then copy them? I'll relay any
responses and they'll be appreciated.
Thanks, Beverly Ross (rossb@iubacs via BITNET)
Indiana Univ. Bloomington
------------------------------
Date: Mon, 21 Nov 88 10:25:25 EST
From: Ken Van Camp <kvancamp@ARDEC.ARPA>
Subject: XUU.c
>Good point. The program below will solve the problem of the uppercase
>filenames for those on Unix. It automatically renames them. This program
>is great for fixing filenames after you have done a "mget" with FTP to get
>multiple files from SIMTEL20. The syntax is: xxu * to do all files in the
>current directory or xxu filename for only one file. Other combinations
>of wildcards are also accepted.
>
>[Although we don't normally publish programs in the Digest, this is too
>good to pass up... gph]
Well, I don't like to clog up the digests either, but the solution you pub-
lished is not too creative. The Unix utility 'dd' makes it easy to convert
cases of file names, and the shell script below even automates it nicely to
convert all files in the current directory. This isn't MS/DOS stuff, but
heh, you started it, not me! (Note: Most Unix systems have a 'dd', al-
though the path may be different than that shown below, so edit it for your
system.)
#!/bin/sh
#
# Rename all upper case-named files in a directory to the same name in
# lower case
for oldname in `ls [A-Z]*`
do
newname="`echo ${oldname} | /usr/usg/dd conv=lcase 2> /dev/null`"
if [ -r ${newname} ]
then
echo "\n!! ${newname} already exists!"
echo ">> overwrite it [y|n]-> \c"
read ans
case ${ans} in
n*|N*) echo " WARNING: ${oldname} was not renamed!" ;;
* ) echo "...${oldname} written over ${newname}"
mv ${oldname} ${newname} ;;
esac
else
mv ${oldname} ${newname}
fi
done
------------------------------
Date: Mon, 21 Nov 1988 08:48 MST
From: Keith Petersen <W8SDZ@WSMR-SIMTEL20.ARMY.MIL>
Subject: XXU.c
Ken, will your DD script fix illegal filenames, such as those with imbedded
control codes or reserved characters? XXU will. Does your script fix
names like PD1:<MSDOS.FILEDOCS>SIMIBM.ARC.14? XXU will, renaming it to
simibm.arc. Will it handle a name like PS:[NETINFO]INTEREST-GROUPS.TXT? I
think you get the point. XXU does *more* than just upper case to lower
case conversion.
Enclosed below is an improved version which deals even better with illegal
filenames.
--Keith
[The IMPROVED XXU.C has been filed in the archives. gph]
------------------------------
Date: Mon, 21 Nov 88 08:51:31 MST
From: Gregory Hicks <GHICKS@WSMR-SIMTEL20.ARMY.MIL>
Subject: XXU.C
OK! Enough! I've started a 'programmers' DUEL!
Suffice it to say, we now have three versions of the same program, each
doing the same thing, ONLY each one is BETTER than the one before!
GREAT!! Keep up the good work.
Gregory Hicks
------------------------------
************************
End of Info-IBMPC Digest
-------