INFO-IBMPC@USC-ISIB (04/02/83)
From: Dick Gillmann <INFO-IBMPC@USC-ISIB>
Info-IBMPC Digest Friday, 1 April 1983 Volume 2 : Issue 20
Today's Topics:
DOS 2.00 (2 msgs)
Experiences with a New XT
1-2-3
9600 Baud Query (3 msgs)
----------------------------------------------------------------------
Date: 30 March 1983 0905-est
From: Bruce J Nemnich <Nemnich @ MIT-MULTICS>
Subject: DOS 2.00
To: info-ibmpc @ USC-ISIB
Someone requested a summary of the differences between DOS 1.1 and
2.00. The following is paraphrased from Appendix A of the DOS manual
("DOS Version 2.00 Enhancements"):
The characters <, >, |, and \ now have special meanings and can no
longer be used in filenames.
You can create a configuration file of special commands that DOS will
read each time it starts up. The commands allow you to specify the
number of disk buffers, names of device drivers, and additional
information concerning DOS operation.
Support for one or more fixed disk devices. The disks can be divided
into partitions, each usable by a different operation system.
Utilities are included to perform disk initialization, backup, and
restore functions.
You may specify more (or less) disk buffers than the default 2.
Tree-structured directories (hierarchical storage) for both fixed
disks and floppies. Path names are specified a la Un*x using a
backslash to separate directories, i.e., "\dir1\dir2\filename".
Disk volume labels (up to 11 characters) may be given at format time.
the label will be shown in various commands including DIR.
Diskettes are formatted with 9 sectors/track, increasing their
capacity. DOS 2.00 can still read and write 8-sector diskettes,
though.
Extended DOS screen and keyboard control is provided through escape
sequences.
Redirection of standard input & output allows input from the keyboard
and output to the screen to be redirected. It is specified a la Un*x,
e.g., to put DIR output in file DIRLIST, do "DIR >DIRLIST".
Piping of standard input & output allows the standard output of one
program to be used as the standard input to another. The commands are
*not* run as separate processes; a temporary file is created to serve
as the "pipe."
The following commands are new in 2.00:
ASSIGN - re-assigns drive letters to route requests to another drive.
BACKUP - dumps one or more files from fixed disk to floppy.
BREAK - can enable ^C polling for all DOS functions (incl disk IO).
CLS - clears screen.
CTTY - allows a different primary console to be defined (e.g., remote).
ECHO, IF, FOR, SHIFT, GOTO - cmds to extend flexibility of batch.
FDISK - initializes a fixed disk.
GRAPHICS - allows Shift-PrtSc to print image of graphics screen.
MKDIR, RMDIR, CHDIR - make, remove, and change directories.
PATH - specifies a search order for .COM and .BAT files.
PRINT - prints file(s) on printer in background.
PROMPT - redefines system prompt.
RECOVER - recovers files which may have a bad sector as best it can.
RESTORE - restores files dumped with BACKUP.
SET - sets keywords & parameters accessible by commands & applications.
TREE - displays directory structure of disk.
VER - displays version number of DOS.
VERIFY - causes DOS to verify everything it writes to disk.
VOL - displays volume label of disk.
Other existing commands are enhanced in various ways. I know this
information is terse, but I am tired of typing; if anyone needs more
specific information, just ask.
--bjn
------------------------------
Date: 30 March 1983 1537-est
From: Bruce J Nemnich <Nemnich @ MIT-MULTICS>
Subject: Changing the Shell in DOS 2.00
To: info-ibmpc @ USC-ISIB
Cc: hplabs!hao!csu-cs!casterli @ UCB-VAX
Date: 27 Mar 83 20:58:30-PST (Sun)
To: Info-IBMPC at Usc-Isib
From: hplabs!hao!csu-cs!casterli at Ucb-Vax
...One feature I hope is soon supported is the ability to
change the shell at boot time.
It is supported. Include a line in your CONFIG.SYS file like
SHELL=path
where "path" is the command interpreter you want to use instead of
COMMAND.COM. I haven't tried it yet, but that's what the
documentation says (on page 9-10 of the DOS 2.00 manual).
--bjn
------------------------------
Date: 30 March 1983 1006-est
From: Bruce J Nemnich <Nemnich @ MIT-MULTICS>
Subject: Experiences with a New XT
To: info-ibmpc @ USC-ISIB
Well, I picked up my XT yesterday. I took it to my office, turned it
on, and nothing was displayed on the screen. It turned out they (IBM)
had a dip switch set incorrectly; when I corrected it, it worked fine.
I picked up 128K worth of 4164 200ns chips for $103 (that's $5.75 each)
at a place here in Cambridge and installed them on the motherboard
with no problems, giving me 256K in all. I think IBM wanted $320 to
do the same thing.
The expansion boards for the XT are *not* compatible with the PC.
Actually, based on what IBM told me, the PC boards will not fit the
XT, but the XT boards will fit the PC. Nothing like downward
compatibility. The difference is only in the size of the boards.
------------------------------
Date: 31 Mar 1983 1555-PST
Subject: 1-2-3
From: Billy <BRACKENRIDGE@USC-ISIB>
To: info-ibmPC
I have been playing with the 1-2-3 program, and I am very impressed. I
am not an accountant or manager type and have no real use for spread
sheet or data base programs, however this program is so much fun to
play with I might consider a career change.
As a programmer I have a great deal of respect for the thought that
went into the production of this piece of software. This is the first
code I really feel is worth the $500 they are asking. At last we have
an existence proof of that elusive entity known as the consistent user
interface. The program has it's own exec and all commands can be
entered by a single character or by pointing to a menu choice. At any
point extensive help is available, and the program comes with a
programmed tutorial diskette. I think a mouse will fit in this scheme
without much surgery.
1-2-3 seems to be an amalgamation of Visicalc, Dbase, Scribe, and
Emacs. The spreadsheet stuff is pretty standard. These programs have
long since eclipsed the original Apple Visicalc. 1-2-3 differs from
the standard spread sheets in that you can manipulate data base
relations as easily as you can express mathematical relationships in a
spreadsheet.
The report generation features look impressive. The program comes with
a pretty extensive looking font library, and this combined with
graphics features can probably make some pretty impressive looking
displays on the prism color printer.
I found their policy on protection interesting. They realize the era
of hard disks is here so the program is not copy protected. Instead
they rely on several pages of outrageous legal threats to keep their
customers in line.
The statement that 1-2-3 is not copy protected is a bit of a lie. They
formatted one of the tracks so that the sectors are not written in the
order of the rotation of the diskette. DOS utilities will copy any
file so written but will rearrange the order of sectors for a more
optimal access. Anyone who has run Diskcopy more than a few times can
hear something amiss when copying 1-2-3. When you copy 1-2-3 to your
hard disk or electronic disk you are still supposed to keep a copy of
the original disk on drive A. From time to time 1-2-3 will peek at
this track on drive A and time the access. Should a read ever take
less time than it feels is appropriate 1-2-3 will return to the exec.
This plus a (weakly) encrypted serial number is their protection
scheme, and it seems a pretty good compromise between the legitimate
paranoia of the games manufacturers and the stoic resign of compiler
writers who don't protect programs at all. Of course their literature
is full of a bunch of scare talk written by their lawyers about how
you can't modify so much of one byte of code or god forbid run the
thing to a terminal connected over a computer network...
------------------------------
Date: 30 Mar 83 22:31:18-PST (Wed)
To: Info-IBMPC at Usc-Isib
From: pur-ee!uiucdcs!donchin at Ucb-Vax
Subject: 9600 Baud Query
We tried running XTALK at 9600, 4800 and 2400 baud and the only speed
at which the display kept up was 2400. The fault is not at the Harris
side as it supports many terminals at 9600.
The symptoms at 9600 are missing characters, unfilled lines etc.
------------------------------
Date: 30 Mar 83 22:56:37-PST (Wed)
To: Info-IBMPC at Usc-Isib
From: pur-ee!uiucdcs!uicsl!preece at Ucb-Vax
Subject: Re: 9600 Baud query
The GRAPHICS display does have a speed problem (but not due to
character generation). The stock text display routines (in the BIOS)
wait for the vertical blanking time before displaying a character.
This drastically limits the display speed, but avoids display tearing
and snow. Some terminal programs avoid the BIOS routines and write
characters directly into the graphic display's memory. This is much
faster, but there is some ugly visual static while characters are
being written to the screen.
My understanding was that the monochrome display avoided this problem
(maybe only if you go around the BIOS routines).
scott e preece
uiuc - coordinated science lab
------------------------------
Date: 1 Apr 1983 2308-PST
Subject: Re: 9600 Baud Query
From: Dick Gillmann <GILLMANN@USC-ISIB>
To: pur-ee!uiucdcs!donchin@UCB-VAX
cc: Info-IBMPC
I think your problem is with the XTALK software. I've spent a lot of
time sweating over this in writing terminal emulator software. The
BIOS video routines for the graphics display can keep up at 4800 baud
and can almost keep up at 9600 baud. The slow part is not displaying
characters, but rather the scrolling, which takes longer on the
graphics display. You shouldn't be losing characters in any case, as
the software should buffer and use XON/XOFF if necessary. It sounds
to me like the characters are being lost in the RS232 routines. The
comm. BIOS interrupt is written very badly and won't go much faster
than 1200 baud. To go faster, the terminal emulator must handle the
interrupts itself.
PC/Intercomm (Mark of the Unicorn) can even handle 19.2K baud in
bursts, although it uses "snowy mode" (direct writes to the regen
buffers) which I find annoying. My terminal emulator VDTE (Inner Loop
Software) waits for the retrace and will keep up at 9600 baud, albeit
with the help of a little buffering and XON/XOFF.
/Dick
------------------------------
End of Info-IBMPC Digest
******************************
-------