Info-IBMPC@USC-ISIB.ARPA (Info-IBMPC Digest) (01/20/86)
Info-IBMPC Digest Sunday, 19 January 1986 Volume 5 : Issue 6
This Week's Editor: Billy Brackenridge
Today's Topics:
New Year's Resolutions
Patching Madness!
Prolog-V
Modula-2
Free Concurrent Euclid PC Compiler on Floppies
FORMAT from Micro Computer Concepts
PC to Burroughs Mainframe Communication
Turbo EDITASM
Turbo Editasm, Epsilon and DISNDATA
Epsilon time_and_day function
DeSmet C and Make
NEC V20 on Zenith 158 at 8Mhz problems
AT BIOS IRQ2 bug
IBM-EGA
David Nilsen's Networking query
Rename Library Routine in Microsoft C 3.0 (3 Msgs)
Detecting the Presence of ANSI.SYS
Aviation Software?
Tax Software
8 bits with MEX and MEX-PC
PC-MORE Version 1.3 Now Available
Query: Victor SpeedPac 286
Request Information on Epson HS-80 Ink Jet Printer
Hard Disk Backup
Net address of Satellite Software International?
PC, MIDI, MPU-401, MPS, Sequential Multi-Trak
Zenith 158 and EMS boards
lint from Gimpel
----------------------------------------------------------------------
Date: 19 Jan 1986 13:34:50 PST
Subject: New Year's Resolutions
From: Billy <BRACKENRIDGE@USC-ISIB.ARPA>
To: info-ibmpc@USC-ISIB.ARPA
The editors of INFO-IBMPC met last week. Several changes have taken
place.
First we unanimously adopted a resolution to cut off our feed of
messages from usenet. If you want net.micro.pc messages forwarded to
you ask your local unix guru. This information is easily available
everywhere and we are just repeating a service by carrying these
messages.
The old digests are now on file by month rather than quarter.
Jaunary.83 would contain messages from January of 1983. This makes
archive files easier to FTP and allows us to use Tops-20 utilities as
these files are all now within a Tops-20 address space. Here is a
list of the files and their corresponding issue numbers:
DIGESTS-86 (latest month)
August.82 V1 #1 #10
September.82 V1 #11 #16
October.82 V1 #17 #23
November.82 V1 #24 #30
December.82 V1 #31 #34
February.83 V2 #6 #11
March.83 V2 #12 #19
April.83 V2 #20 #26
May.83 V2 #27 #33
June.83 V2 #34 #39
July.83 V2 #40 #46
August.83 V2 #47 #60
September.83 V2 #61 #70
October.83 V2 #71 #80
November.83 V2 #81 #93
December.83 V2 #94 #104
January.84 V3 #1 #14
February.84 V3 #15 #24
March.84 V3 #25 #35
April.84 V3 #36 #48
May.84 V3 #49 #60
June.84 V3 #61 #69
July.84 V3 #70 #79
August.84 V3 #80 #89
September.84 V3 #90 #99
October.84 V3 #100 #107
November.84 V3 #108 #112
Dexember.84 V3 #113 #119
January.85 V4 #1 #9
February.85 V4 #10 #18
March.85 V4 #19 #40
April.85 V4 #41 #52
May.85 V4 #53 #65
June.85 V4 #66 #76
July.85 V4 #77 #87
August.85 V4 #88 #99
September.85 V4 #100 #111
October.85 V4 #112 #127
November.85 V4 #128 #134
December.85 V4 #135 #143
Progress on making a new index moves slowly. We have had many offers
but not from anyone who already has complete archives at their local
site. I have to extract the headers before someone else can do the
final sort and edit. Breaking the archives into smaller files is a
start.
Dick Gillmann is rejoining us as an occasional editor.
Those of you who have made program submissions recently may wonder
where they are. They are all in Koji's mailbox. We are looking to get
Koji some help in this area.
We are unresolved as to what to do about sending back digests and
library files to unconnected networks. We never have responded to
requests from usenet hosts unless they were closely connected to the
arpanet, and we are getting increasingly fed up with sending things
via mail to csnet and bitnet. The only reason these nets don't have file
access to our archives is because the political administrators of
these networks are unwilling to implement gateways that can handle
file transfer. I fail to see why we should have to send files as mail
because the users of these connected networks are unwilling or unable
to put pressure on the administrators of their nets to allow them
file access to ARPA hosts. In the mean time some requests will get
answered and others won't depending on the mood of who is editing at
the moment. This sounds harsh but only reflects the reality of the
last year or so as requests have grown beyond our capability to handle
them.
If someone had a mail based server that could automaticly respond to
these requests, this could solve the problem for the short term.
It doesn't look like we can find a replacement source of official IBM
announcements or the price list of IBM products. The last was very
handy, and was used here and other sites regularly as purchasing
departments always liked having the official IBM part numbers even if
they bought PC equipment elsewhere. I'd be particularly interested to
see the technical specs of the Risc/AT system IBM is going to
announce this week, but since we have been cut off from IBM it looks
like we will have to read the newspapers the same as everyone else.
Thanks again to all those who took the time to answer queries and submit
programs to the library. Particular thanks go to Jim Celoni our guest
editor on Christmas day and Joel Goldberger for the continued support
of the ISI computing resources.
------------------------------
Date: Thursday, 16 January 1986 21:24:05 EST
From: Joe.Newcomer@a.sei.cmu.edu
Subject: Patching Madness!
No, not patching code, but patching RS232 connections. Well, it
turns out that I have the expected two serial I/O ports on my PC; a
Mouse Systems serial mouse, a terminal, a 2400 baud modem, a 1200
baud modem, a connection to another local machine, a connection to
another site in the house where I can take a terminal, a serial
plotter, etc.
Now the fun part. Sometimes I want to run my terminal to the modem; other
times I want the PC (for Kermit purposes). The plotter hooks to the serial
port sometimes. The portable computer uses the serial port for FTPoid
transfers to the PC, and sometimes I use the portable from the porch in the
summer as a terminal. Sometimes I have to use the 1200baud modem, and most
times I can use the 2400baud modem. ARGGHHHH!!!!!!!!......
\ ....
After I went slightly bonkers attempting to hook the plotter up, I threw up
my hands in despair, and started looking at RS-232 switches. Outrageous!
hundreds of dollars, and they are only n-to-1 connectors for various tiny
values of n (say, 5 at most). I want a crosspoint switch...
Well, it turns out that I got what I need, so I pass the info on to others
in the PC digest. I went out and bought a Tascam PB-32H patch bay at
my local keyboard&synthesizer shop. Discounted, it cost $108. I added
another $35 of Tascam patch cords. I went to Radio Shack and bought
about 10 24' "speaker extension" cables, RCA-to-RCA.
This patch bay has 16 pairs of connectors. The front panel has 1/4" phone
plug sockets, the rear has RCA-type audio sockets. I then made up connectors
of the following flavor:
RS-232 DB25 RCA connectors
<----- 2 -------------------------------------------- ---->
<----- 3 -----+ +----------------------------------- ---
| |
+--(----------------------------------- ---->
<----- 7 --------+----------------------------------- ---
Cutting a 24' cable in the middle gave me two 12' connectors at a lower cost
than buying two 12' RCA-to-bare-wire cables. The sex of the DB25 connector
(or for the /AT, DB9 connector) was chosen as appropriate. Each pair was
marked carefully by using electrician's adhesive numbers. Each pair was
given a number (e.g., 23) and each wire was marked as pin 2 or pin 3
(based on the DB25 pinout) so each end was labelled, for example, 23-2
and 23-3. Thus, I now converted all my RS232 connections to appear
at the patch panel, which is placed conveniently bolted below the shelf in
front of me.
Now, the really great feature of the PB-32H patch panel comes into play. The
jacks are "normalled", so that the top jack of the pair is normally
connected to the bottom jack of the pair if no cable is inserted in each.
So instead of each pair being "thing out" and "thing in", the upper jack is
"thing out" and the bottom jack is "thing it normally wants to connect to".
So I have, for example, "COM1 Out" directly above "Modem in" and "Modem out"
directly above "COM1 In". "Mouse out" above "COM2 In". etc. So in normal
operation I don't have any cables plugged in at all. But now I want to
connect the plotter up. So I plug a cable into "Plotter out" and "COM1 In"
and "Plotter In" and "COM1 Out". Works great. And I still have 7
positions left for expansion...
Obviously, this works only if you are interested only in data in/out, but
if you have to deal with RTS/CTS type stuff, the generalization is obvious
but you have to be more careful in patching (don't patch one device's data
and another device's control lines, for example, into a port and expect
the handshaking to be meaningful).
Shop around for the patch panel; it lists at much higher than I paid for it, as
do the cables, but music shops frequently discount such items. I got something
like 40% off. (If you don't know where to start, try synthesizer shops. Look
for dealers who handle Korg, Yamaha, Casio professional keyboards, guitars,
mixing boards, pro and semipro audio equipment, that sort of thing).
While I realize this is not IBM-PC specific, it was just too good of an idea
not to share more widely.
------------------------------
From: Peter Boeijink <boeijink@ark.uucp>
Subject: Prolog-V
Date: 16 Jan 86 01:58:54 GMT
In the BYTE magazine of december Chalcedony Software offers a Prolog
interpretor. They have two versions: PROLOG-V for $69.95 and PROLOG-V-Plus
for $99.95. Both are running on MsDos and PcDos systems according to
the ad. I own a Sanyo MBC-555 with MsDos 2.11. Can a Prolog-V user on
the net tell me:
- whether the packet is running on the MBC-555;
- what his/her experience is with the packet.
mbc555(msdos).
prologV(msdos).
quality(minimal_as_prologC).
price(reasonable).
user_friendly(not_as_prologC).
run :- prologV(TYPE), mbc555(TYPE).
experience :- quality(minimal_as_prologC), user_friendly(not_as_prologC),
price(reasonable).
buy :- run, experience.
?- buy.
Thanks in advance, Peter Boeijink.
------------------------------
From: Peter Boeijink <boeijink@ark.uucp>
Subject: Modula-2
Date: 16 Jan 86 02:00:20 GMT
In the BYTE magazine of december Interface Technologies offers a Modula-2
compiler kit (the M2SDS packet). In exchange for an original Turbo Pascal
disk they sell it for $50.88. It should run under MsDos according to
the ad. I own a Sanyo MBC-555 with MsDos 2.11. Can a Modula-2 user on
the net tell me:
- whether the packet is running on the MBC-555;
- what his/her experience is with the packet.
Thanks in advance, Peter Boeijink.
------------------------------
From: CSRI Distribution Manager <distrib%Toronto@csnet-relay>
Subject: Free Concurrent Euclid PC Compiler on Floppies
Date: 14 Jan 86 15:32:15 GMT
The compiler group at the Computer Systems Research Institute,
University of Toronto has produced a version of the Concurrent
Euclid (ConEuc) programming language that runs on 8086 MS-DOS systems.
(Our previous versions run on Vaxes, PDP-11s, 370s, SUNs, etc.)
ConEuc is being used at the University of Toronto in many high-level
undergraduate and postgraduate courses.
Concurrent Euclid is a systems programming language developed from
the Euclid programming language. It combines systems programming
features with concurrency, monitors, a highly structured syntax and
fully defined semantics. This allows the development of software that
is verifiable and efficient.
ConEuc has been used to write: compilers (for example, itself, Turing),
operating systems (TUNIS, a UNIX compatible operating system),
compiler writing tools (Syntax/Semantics Language), networking systems
(HUBNET), etc. There are 100s of universities and commercial
enterprises that already use ConEuc on machines such as the VAX (VMS
and UNIX), IBM 3033, 68000s and PDP-11s.
We have now developed an MS-DOS version that we've used successfully
on IBM-PC's, Eagles, TI-PC's, etc. We are interested in
distributing it to people who are willing to evaluate a new
and potentially useful tool in exchange for a copy of it. The version
we will be distributing in this beta-test will require an 8086-family
PC (yes 8088's, 80188's, etc. are OK) with at least 256K RAM and a hard
disk. We will be shipping some documentation (not everything everyone
will want I'm sure, but enough) and 3 floppies that come with an
INSTALL.COM that will place all the necessary files on your hard disk
in directory \CONEUC.
Our budget will allow us to produce 100 copies of the compiler
distribution to be given away without cost to serious software
people who are willing to send us written comments about the quality of
the distribution and the compiler. We are supplying a graphics
interface for those with the colour graphics options. The distribution
contains a complete compiler, paragrapher, and full-screen editor.
System and graphics interfaces are also supplied.
You should also have a copy of the book "Concurrent Euclid, UNIX,
and TUNIS" by R.C. Holt (Addison-Wesley).
This is only a preliminary version; all rights remain with the
University of Toronto. You may not distribute or sell the compiler
without direct written consent from the University of Toronto.
--
CSRI Distribution Manager {decvax,ihnp4,allegra}!utcsri!distrib
University of Toronto distrib@Toronto.CSNET
Room SF2002 distrib%Toronto@csnet-relay.ARPA
10 King's College Road
Toronto, Canada M5S 1A4 (416) 978-6985
------------------------------
Date: 16 Jan 86 14:48:09 +0100
From: XBR1D90F%DDATHD21.BITNET@WISCVM.WISC.EDU (D90F@BR1.THDNET)
Subject: FORMAT from Micro Computer Concepts
I heard rumors about a FORMAT utility by Micro Computer Concepts (Mercer
Islands, MA), that is supposed to increase harddisk-capacity by up to
50 percent.
Does anybody have any experiences with it or closer information about it?
Volker Weber, Darmstadt, W-Germany
Please reply to XBR1D90F%DDATHD21.BITNET@WISCVM.WISC.EDU
------------------------------
From: hqb <hqb@gatech.csnet>
Subject: PC to Burroughs Mainframe Communication
Date: 16 Jan 86 20:14:43 GMT
I seem to recall there are at least two companies that specialize in this -
one is called MasterLink (or something like that), I can't seem to recall
the name of the other one. Last I heard, one package cost ~$495, and the
other was about $95. It pays to shop around. If you can get hold of
Datapro publications, they're both listed in the volume that discusses
micro to mainframe communication.
Both the above-mentioned packages emulate a somewhat extended version
of the Burroughs poll-select terminal - you can have several sessions
going at once, and switch between them at will, or allocate windows to
each.
Hope this helps.
--
Henry Bibb
School of Information & Computer Science, Georgia Tech, Atlanta GA 30332
CSNet: hqb @ GATech ARPA: hqb.GATech @ CSNet-Relay
uucp: ...!{akgua,allegra,rlgvax,sb1,unmvax,ulysses,ut-sally}!gatech!hqb
------------------------------
From: CJS@psuvm.bitnet
Subject: Turbo EDITASM
Date: 15 Jan 86 14:52:07 GMT
Here's a partial answer from my own query: I found a review of this
assembler and a couple others in December 1985 PC Tech. Journal. They
weren't entirely happy with it, but I think I'll buy it anyway.
I'd appreciate any first-hand reports.
Christopher J. Sacksteder, The Pennsylvania State University,
Computation Center, 122 Computer Building, University Park, PA 16802
Bitnet: CJS at PSUVM
UUCP: ...!allegra!psuvax1!cjs@psuvm.BITNET
------------------------------
Date: Fri 17 Jan 86 22:28:19-EST
From: Nathaniel Polish <POLISH@CS.COLUMBIA.EDU>
Subject: Turbo Editasm, Epsilon and DISNDATA
Turbo Editasm is an interesting product. I do a fair amount of
assembler work -- mostly drivers for MS-Pascal -- and I hate the long
cycle using the Microsoft MASM and LINK programs. Turbo Editasm
(TASM) is, however, not the answer. The lengthiest part of the
development cycle is the link not the assembly, especially with
version 4.0 of MASM. TASM does not speed linking at all.
TASM has some very serious flaws. The editor is very limited. While
configurable, the editor limits you to one keystroke for each command
so you can not make it look like a lot of other editors. Also, TASM
has the only editor that I have seen in years which uses the BIOS for
screen updates. It is, therefore, very slow. In addition, the cursor
flickers randomly due to the way TASM uses the clock interrupts to
display the time. The other major flaw is that the assembler itself
is somewhat broken. I could not declare "esc" and several other
labels for reasons not explained in the documentation. Also, I gave
up when TASM accepted "out 0350h,al" This is, of course, illegal as
any port number over FF must be specified through the DX register :
"mov dx,0350h ... out dx,al" The assembler did not flag an error and
generated code for "out 00,al"
The interface to the system is of course a straight clone of the
Borland Turbo Pascal interface. This is fine if you write standalone
code; I, however, need to link so the interface is useless. I find
that I much nicer way to work is to use Epsilon and assemble using
MASM 4.0 in Epsilon's process facility. It is easy enough to write
something in EEL to jump you into the code on error messages.
This brings me to a major gripe about Epsilon. The smart people who
wrote Epsilon managed to detect control up and down arrows - the BIOS
does not. They clearly had to go to bare metal to read the keys for
this. Great idea. They, however, do not clean up after themselves so
anything else reading the keyboard is clobbered. I use Sidekick and
Lightning but I must use it outside Epsilon. Also, if you run a
process and the program hangs (during development this is common) you
can not exit the editor. Further, I can not control-alt-del since
Epsilon traps this sequence and informs me that I should exit the
process! If I hit control-alt-del I don't want some smart-ass program
protecting me. I think that they are just showing off that they
control your keyboard better than you do. Any fixes out there?
[Todd Doucet (TD9H@CMU-CC-TE) of Lugaru has asked that I point out
that version 3.0x of Epsilon is compatible with the Borland resident
routines. I have argued the same point as regards Ctl Alt Del, however,
since I discovered "super ^C" ^U^C I haven't felt the need for a real
Ctl Alt Del. Todd would respond in person but CMU won't let mail out
from that account. -wab]
Also, someone asked about the DISNDATA disassembler. Yes, it does
seem to trace program flow and it considers all unreferenced code as
data. This is great for everything except jump tables and interupt
handlers which are never directly referenced. The real reason to get
DISNDATA is that it generates MASMable code. I have never seen a
disassembler which does this. I disassembled three assembler programs
of mine. One did not re-assemble the first time due to a missing
"word ptr" but then assembled nicely. The others re-assembled fine.
One program ran, one ran and did not work and the third crashed the
machine. In all, not too bad. It is the best disassembler that I
have used and I recommend it.
Nat
------------------------------
Date: 17 Jan 1986 11:34:19 PST
Subject: Epsilon time_and_day function
From: Craig Milo Rogers <ROGERS@USC-ISIB.ARPA>
To: Knight@SRI-NIC.ARPA
The Epsilon time_and_day() routine takes a pointer to the
time_info structure as an argument, not the time_info structure
itself. In your example try "time_and_day(&date_time)".
Craig Milo Rogers
------------------------------
From: Mike Schwartz <mykes@3comvax.uucp>
Subject: DeSmet C and Make
Date: 16 Jan 86 19:04:51 GMT
I bought a copy of DeSmet 'C' for the IBM PC for $109. I highly recommend
the compiler, because it is very fast, doesn't complain too much, generates
good code, and comes with some usable source code. Also, I got some extra
disks from them, including source to a MAKE program. At work, I use Lattice
'C', but given a choice, I would be using DeSmet in a jiffy. I do use the
MAKE program all the time, and it works real well. It is not a full Unix
make, but it is very simple to use, and extremely effective. It might be
worth getting DeSmet 'C' just for the MAKE program, becuase you end up with
no only MAKE, but a great 'C' compiler too. DeSmet 'C' is advertised in
DDJ by a company called CWARE. Just a quickie review.
mike schwartz @ 3Com Corp.
------------------------------
Date: Fri, 17 Jan 86 11:20:00 PST
From: Doug Lind <entropy!lind@uw-beaver.arpa>
Subject: NEC V20 on Zenith 158 at 8Mhz problems
I tried to install the 8 Mhz version of the NEC V20 chip on my Zenith
158. Works fine at 4.77 Mhz, but goes down in flames at 8Mhz. The NEC rep
said the Zenith may use different timings than the IBM. Has anyone out there
successfully installed this chip at 8 Mhz ?
------------------------------
Date: 17 Jan 86 14:40:41 PST (Friday)
From: Colvin.PA@xerox.com
Subject: AT BIOS IRQ2 bug
To: Joe.Newcomer@a.sei.cmu.edu
I would like to make a correction about the IRQ2 line on the AT
expansion bus being connected to IRQ8, it is actually connected to IRQ9
(in fact it has been renamed IRQ9 on the bus). It still gets redirected
to IRQ2 in software.
Craig Colvin
(Colvin.pa@Xerox.COM)
------------------------------
From: johnl@ima.uucp
Subject: IBM-EGA
Date: 16 Jan 86 17:00:00 GMT
There are two good sources of information on the IBM EGA card.
1 - The "Options and Adapters Technical Reference" manual. This manual
is very expensive, $125 US, but contains hardware descriptions of all of
the PC peripherals that IBM sells. The copy you buy will probably not
have the section on the EGA, but there is a card to send in and they will
send you updates that include it. The section on the EGA looks like it
was typeset and proofread by somebody who didn't speak any English, but it
will do. It includes a listing of the EGA's BIOS.
2 - The IBM Personal Computer Seminar Proceedings, volume 2, number 11.
This is a series published for PC software developers and their ilk. You
can write to IBM PC Seminar Proceedings, 4629, PO Box 1328, Boca Raton FL
33432 and they'll send it, or you can try calling +1 305 998 2000 and be
put on hold for a long time.
John Levine, ima!johnl
[A similar response came From: Jerry Lotto <lotto@harvard.EDU> -wab]
------------------------------
Date: Sat, 18 Jan 86 01:01 EST
From: Miyata@MIT-MULTICS.ARPA
Subject: David Nilsen's Networking query
One of my concerns is accessing the various services and resources
that may be distributed within a 'network', all of which probably is
not co-located at a single 'host'. If my PC was directly connected
(via RS232, for example) to my 'preferred' host at work, my PC could
essentially act as a virtual terminal when it needed to access
non-PC-resident services; unfortunately, this host does not handle a
large number of asynchronous lines so not everyone in the office can
participate in this interconnect strategy. If there are several
hosts that serve as repositories for information I access regularly
or to which I archive information, I prefer transferring files to and
from my PC, rather than a more indirect mechanism. A small company
that relies on PC's for information processing, may have a several
databases distributed amongst several PCs. A more sophisticated
database or financial application may use distribution principles to
coordinate the queries and updates in an interactive and real-time
mode, while still maintaining consistency and accuracy; I'm not sure
there is an easy solution for this type of interworking with the
RS232 point-to-point strategy.
I don't think that handling high data rates, or 'easiest/best way to
go', and perhaps even expense is at issue; although this may be true
in the 'low-end' market space (home?). Particularly when businesses
and users are probably thinking and using -286 based machines and if
we take into account futures, -386 based machines (ie, what we once
thought of as the low-end is pushing on the mini/midi-range) . I
think what is at issue is the degree of interworking and
communications, resource sharing, and distribution of services that
the market seems to be demanding.
With respect to networking PC's, there are several forces at play -
IBM and the defacto standards it leaves in it's wake (NETBIOS, IBM PC
Network, Token Ring), currently existing environments (both large -
eg, Internet, Decnet, SNA, etc. and small(er) - 3Com, Ungermann Bass,
Novell), acceptance and solidification of international standards
(ISO), more specialized environments (GM/MAPs), etc. I think in the
end and along the way, we will want them to coexist and interwork.
------------------------------
From: Frank Whaley <micropro!sagan!frank@lll-crg.ARPA>
Subject: Rename Library Routine in Microsoft C 3.0
Date: 17 Jan 86 19:08:11 GMT
One of my favorite pasttimes is arguing that Microsoft is wrong.
I was going to talk about all of the other libraries I have which use the
(BSD?) "rename(old, new)" form, but instead I open the April 30 X3J11 draft
to section D.9.4.2 and find:
int rename(const char *old, const char *new);
I would guess that Microsoft felt that this function falls into the same
category as other "replacement" functions (strcpy(), memcpy()) where an
algebraic notation (a = b) as opposed to algorithmic notation (b => a)
is used. Some of us think in a noun/verb fashion ("a is replaced by b")
while others prefer verb/noun ("move b into a"). Oops, I'm arguing...
But I like standards, and code to fit them.
...Frank Whaley, MicroPro Product Development
UUCP: {decvax!decwrl | ucbvax}!dual!
{hplabs | glacier}!well!
seismo!lll-crg!
ihnp4!ptsfa!micropro!sagan!frank
ARPA: micropro!sagan!frank@lll-crg.ARPA
"I'm told there are better programs [than WordStar],
but I'm also told there are better alphabets."
--William F. Buckley Jr.
------------------------------
From: Guido van Rossum <guido@boring.uucp>
Subject: Rename Library Routine in Microsoft C 3.0
Date: 17 Jan 86 23:51:55 GMT
In a previous article with the above subject I noted a strange error in
Microsoft C's library. I received several letters, none very informative,
most to the tone of "Microsoft blew it again".
That's not fair! While they may have slipped here, my general feeling
towards their 3.0 C compiler and library is extermely positive.
We received it just a week ago, and I have already decided to junk our
(2.15) Lattice compiler completely. I have used it to compile some
20,000 lines of C, and with one exception it compiled flawlessly.
The exception was a compiler assertion error, apparently caused by a
too complicated expression (it was really HUGE). I can live with that.
About two thirds of this code had earlier been used with Lattice, and MSC
proved to be *extremely* compatible. They support almost all MS-DOS-
specific functions and #include files in the Lattice library.
The only thing that bit me (besides the rename bug) was the fact that
Microsoft has signed characters -- my fault for not writing portable code.
The other third had hardly ever left UNIX, and was full of things like exec,
chdir, stat, chmod and signal. It was very easy to port, because Microsoft
implemented almost all well-known UNIX system calls that can possibly
be emulated on MS-DOS to some extent. My main problems with porting it
were changing assumptions about file names (I was parsing paths and
only took '/' as a delimiter) and finding a replacement for fork/exec/wait
sequences -- not too difficult, since Microsoft provides functions spawn*
that combine these three calls.
But the best surprise was the execution speed of my program (even though
I directed all optimizations to saving space): sometimes twice as fast!
I could go on, but these are my actual experiences until now. The manuals
are also of very good quality, and I am looking forward to using the overlay
feature of the new linker. Next, I might try porting Gosmacs!
P.S.: please don't send any more mail concerning rename.
Guido van Rossum, CWI, Amsterdam (guido@mcvax.UUCP)
[As one with lots of Pascal code I am particularly delighted with MSC and
its support for a Pascal interface. This was a great deal of work without
much "splash". Microsoft promised continued support for Pascal and have
kept their word. As a seldom C programmer I am particularly delighted
with real error messages. This is the first C compiler I have seen
that gives you a clue as to what your syntax error might be. -wab]
------------------------------
Date: 17 Jan 1986 12:31:55 PST
Subject: Rename Library Routine in Microsoft C 3.0
From: Craig Milo Rogers <ROGERS@USC-ISIB.ARPA>
My copy of the Unix System V Interface Guide, Spring 1985
Issue 1, does not include a rename() call. Since I have a fair
amount of code which expects the BSD calling order, here's a nice
kludge to get around the problem:
1) Extract the rename module from the C library (or libraries,
if you use more than one memory model). This is something
on the order of: LIB LLIBC *RENAME
2) Use DEBUG to change the "rename()" routine to a "renamx()"
routine. This edit is valid for my copy of the large
model library; expect different offsets for the other
models and/or other releases.
debug rename.obj
-e 109
xxxx:0109 65.78
-e 187
xxxx:0187 65.78
-w
3) Rename "rename.obj" to "renamx.obj".
4) Create a new rename() which takes its arguments in the
"right" (read "Berkeley") order:
int rename(old, new)
char *old, *new;
{ return renamx(new, old); }
5) Compile the new rename() routine.
6) Update the library, something like: LIB LLIBC -+RENAME+RENAMX
On second thought, though, this isn't too hot. There may be
code in the Microsoft library which expects rename() to take its
arguments in Microsoft order rather than Berkeley Unix order. So,
how about making a header file containing:
#define rename(old,new) (renamy(old,new))
Create a "renamy()" function and add it to the library:
int renamy(old, new)
char *old, *new;
{ return rename(new, old); }
Then, #include the new header file in anything you write that uses
Berkeley-style rename(). The main disadvantages are: 1) locating and
editing the appropriate files, and 2) the resultant source-code
incompatabilities.
Let's hope that Microsoft (or AT&T) clears up this misunderstanding
soon.
Craig Milo Rogers
------------------------------
From: Ross Greenberg <greenber@phri.uucp>
Subject: Detecting the Presence of ANSI.SYS
Date: 16 Jan 86 17:53:57 GMT
A quick hack:
Send the ANSI sequence to position your cursor to some location
{x,y}.
Send the ANSI sequence to read current cursor poition. If
they match, you're running ANSI, else you're not.
ross m. greenberg
ihnp4!allegra!phri!greenber
------------------------------
From: Lee Rosenbaum <leer@ixe5m.uucp>
Subject: Aviation Software?
Date: 17 Jan 86 15:29:38 GMT
I am interested in any information on software for
the IBM-PC that will do aviation flight plans
and/or access aviation weather.
If you know of or have any good software please
drop me a line.
Lee Rosenbaum
ixe5m!leer
IX 1M-313
(312) 979-0694, 8-367-0694
------------------------------
Date: Sat, 18 Jan 86 12:40:58 EST
From: Jacob Moskowitz <JMSK@MC.LCS.MIT.EDU>
Subject: Tax Software
Any reccomendations out there for good software for Lotus-123 or IBM-PC
DOS that handles federal income tax ?
Thanks.
------------------------------
Date: Sat 18 Jan 86 01:37:00-MST
From: Ron Fowler <RFOWLER@SIMTEL20.ARPA>
Subject: 8 bits with MEX and MEX-PC
In response to the recent inquiries regarding 8th-bit display using MEX,
here is a patch and some supplementary information:
For CP/M MEX version 1.14, do "POKE $4822 $0 $0" ... this will delete
the code that zeroes the 8th bit.
For the commercial versions (both CP/M and MSDOS), a patch is impractical,
since there have been several major releases of the program, and an equal
number of minor releases. As an alternative, we're adding a fixed-position
patch point in all versions, effective with release 1.61, to be released
18 January (we can't unconditionally pass the 8th bit, since many users de-
pend on the parity bit being removed for use with many timesharing
services). Now that won't automatically fix all those copies out in the
field, so to make it easier to get your copy replaced, we'll waive
the customary $10 update charge, providing:
1) You return your original distribution disk in a
re-useable container
2) You include return postage.
3) Mark the package "
3) Mark the package "Eighth-bit update"
This will also update your software to the recently released version 1.6
(which is a pretty good deal, since 1.6 includes the Kermit and YMODEM
protocols, as well as a bunch of other new features).
One final consideration: someone mentioned that MEX filters some con-
trol characters even when filter is off. Actually, this is only partially
true: the characters in question are SO and SI (0EH and 0FH) ... these are
the Compuserve protocol start and end-protocol signal characters. If you
disable Compuserve protocol ("STAT CIS OFF"), MEX will no longer "eat"
these characters. ---Ron Fowler, NightOwl Software, Inc.
------------------------------
Date: Sat, 18 Jan 1986 15:54 MST
From: Keith Petersen <W8SDZ@SIMTEL20.ARPA>
Subject: PC-MORE Version 1.3 Now Available
Just obtained from net.sources and now available via FTP from SIMTEL20:
Filename Type Bytes CRC
Directory PD:<CPM.PCDOS>
PCMORE13.LBR.1 BINARY 47872 6D98H
The LBR contains PCMORE13.C, PCMORE13.DOC and PCMORE13.EXE.
PC-More is a pager program similar to the UNIX* more program. It is
designed to work on the IBM Personal Computer, PC XT, and PC AT, as
well as compatibles. The purpose of PC-More is to display a file one
screenful at a time for easy viewing. The PC-DOS "type" command can
do this, but the text scrolls too fast for most people to read.
Beginning with version 2.0, PC-DOS supplied its own version of a more
filter, but it is extremely limited in its usefulness, since it can
only display one screenful at a time. People who are used to the UNIX
version of more expect to be able to move through the file in any
increment they choose, not just a screenful at a time. PC-More
corrects this problem.
History:
Ver 1.3 New clear screen function installed, written by Harry McGavran.
All references to crt_cls() and crt_mode() eliminated. This new
clear screen function works on either a monochrome or color
monitor.
Nextfile() rewritten to understand that "-" on the command line
represents stdin. The " case 'f': " part of wait() was modified
in the same way.
New functions cpi() and opi() added to allow stdin to be
reopened to the console when it is redirected to a file or a
pipe. This allows the user to execute a "!" or a "e" when a
pipe is present. Of course, you still can't edit the pipe.
The knowledge to implement cpi() and opi() was provided by
Brandon S. Allbery.
Nextfile() modified to find the file's size and put it in the
external variable x_fs. Wait() also modified so that when it
constructs the prompt, it uses x_fs to compute what percent of
the file has been read. When the prompt is printed, this
percentage is included as part of the prompt.
Wait() modified to clear the screen when a file is rewound.
Ver 1.21 This version fixed a bug which prevented PC-More from working on
a color monitor. Under previous versions, only machines
equipped with a monochrome monitor would run PC-More properly.
The problem was a hardware-dependendent function call to
crt_mode(). The problem was fixed by determining the type of
monitor in use and using the correct function to clear the
screen.
Ver 1.2 This was the initial release to USENET. Other earlier versions
were too buggy for a general release.
------------------------------
Date: Sat, 18 Jan 86 14:46:08 PST
From: Rich Patterson <bilbo.rp@LOCUS.UCLA.EDU>
Subject: Query: Victor SpeedPac 286
Hi,
I was wondering if anyone out there has used one of these little
boards yet ?
Rich Patterson
Logical: lcc.rp@LOCUS.UCLA.EDU
lcc.rp@UCLA-CS
{ucivax,trwrb}!lcc!rp
{ihnp4,randvax,sdcrdcf,ucbvax,trwspp}!ucla-cs!lcc!rp
ICBM: 34 00'57''N 118 27'01''W
------------------------------
Subject: Request Information on Epson HS-80 Ink Jet Printer
Date: Sat, 18 Jan 86 10:11:14 -0500
From: Dick Edmiston <edmiston@CSNET-SH.ARPA>
Has anyone looked at the new low cost Epson ink jet printer,
the HS-80, enough to give any opinions about it? I am interested
in a quieter version of the FX-85.
Dick
------------------------------
Date: Sat, 18 Jan 86 17:14:42 EST
From: Dan_Bower%RPI-MTS.Mailnet@MIT-MULTICS.ARPA
Subject: Hard Disk Backup
A group I work with in Washington has the following protocol for backing up
their XT's hard disks. First, they are running an implementation of
IBM PC Network, with one AT (64 mb hard disk) and 4 XT's. The AT has a
Tallgrass streaming cart tape (TG 4060). Each XT has a 'workspace' (a shared
subdirectory) on the AT's hard disk. To do a backup, the XT owner uses
BACKUP over the network as follows.
C>D:
D>MD BACKUP (D: is typically the local drive designation for the AT's
shared space.)
D>CD BACKUP
D>E:BACKUP C:\*.* D: /S (E: is a shared, read only space on the AT for
programs. After issuing this command and responding
to its prompts, the user waits for up to 10 minutes
for BACKUP to do its thing.)
After this, the user sends a message to the poor woman using the AT to mount
a tape in the Tallgrass machine, perform the backup to tape, and remove the
backup files from the AT.
The process is reversed (after the swearing stops and the dust settles :-)
whenever a restore is needed.
Note, if they were using a Bernoulli box, they could share its drive(s) on
the net and take the AT -> tape step out of the cycle. I'm not sure if
Fastback works with PC Network resident, but if it does, it would speed
up the backup process further.
------------------------------
Date: Sat, 18 Jan 86 00:56 EST
From: LBAFRIN%clemson.csnet@CSNET-RELAY.ARPA
To: info-ibmpc@usc-isib.arpa
Subject: Net address of Satellite Software International?
Does anyone know if SSI, the makers of WordPerfect, are reachable on the net?
And if so, what is their address? I've found a bug in WordPerfect 4.1, and
I want to tell them about the bug and some other suggestions I have. Yet,
I don't relish the idea of going through Snail Mail to get to them. Can
anyone help?
-- Larry Afrin
Dept. of Computer Science
Clemson University
================================
Please send replies, if any, to:
lbafrin@clemson.csnet or
lbafrin%eureka@clemson.csnet or, as a last resort,
any reasonable-looking string with
"lbafrin", "eureka", and "clemson" in it
I disclaim everything anybody ever said about anything.
------------------------------
From: Ward Christensen <ward@chinet.uucp>
Subject: PC, MIDI, MPU-401, MPS, Sequential Multi-Trak
Date: 18 Jan 86 22:26:08 GMT
I have purchased a Sequential Circuits Multi-Trak, and the appropriate
MIDI hardware/software from Roland: MPU-401, MIF-IPC PC interface card,
and MPS software.
MPS is very impressive, but doesn't seem to be a good match to my
particular synth and requirements. I bought the Multi-Trak because it
can simultaneously play 6 instruments [or timbres if you prefer]. It
sends "change program", which MPS records, and when playing back, the
appropriate program [meaning 'instrument' or 'timbre'] setting is done.
By playing with record and playback MIDI modes on the Multi-Trak,
I can cause a part to be recorded on a particular MIDI channel; and by
setting it up correctly it will play on only one voice.
I have two problems: (1) that the MPS software doesn't "show" the
"change program" MIDI function, although it DOES remember it and "play
it back". Thus its difficult to know where/when an instrument change
occurs. (2) when recording one instrument, for example, a bass, then
recording the next component, MPS plays back the bass, but the "select
a random oscillator" way the M-T works, causes it to pick the track that
has the bass, 1 in 6 times. This chops off whatever note I'm playing.
Anyone have this setup, or any comments? Thanks in advance.
------------------------------
Date: Sat, 18 Jan 86 21:52:21 est
From: BACS Access <access%indiana.csnet@CSNET-RELAY.ARPA>
To: info-ibmpc@usc-isib.ARPA
Subject: Zenith 158 and EMS boards
I would like to know if anyone has had any experience with using
any extended or expanded memory boards with the new Zenith 158 series
of computers (the 151's new version, IBM PC compatible). I have
attempted using a Jram 2 board in a fully populated 158 with 640k
using appropriate settings to expand the memory. Also tried a starting
address of 256k after removing appropriate memory from the mother board.
The board and software works fine in an IBM PC and Leading Edge Model M.
Has anyone found a 2meg memory board that will work in the Zeniths?
Please either send mail directly or post to the net. Thanks...
Robert Cole 812-335-0910
ihnp4!iuvax!access
access@iubacs Bitnet
------------------------------
Date: Sun, 19 Jan 86 12:46 EST
From: Yedidyah Langsam <YDL%BKLYN.BITNET@WISCVM.WISC.EDU>
Subject: lint from Gimpel
I have had very good experiences with PC-LINT from
Gimpel Software
3207 Hogarth Lane
Collegeville. Pennsylvania 19426
215-584-4261
------------------------------
End of Info-IBMPC Digest
************************
-------