cmg@WATSUN.CC.COLUMBIA.EDU (Christine M Gianone) (09/12/90)
Info-Kermit Digest Tue, 11 Sep 1990 Volume 12 : Number 4
MS-DOS KERMIT:
Next Test Release of MS-DOS Kermit 3.02
Kermit on Novell Networks
Feedback on MS KERMIT 3.02 of July 20
Kermit Problems - V3.01
MS-Kermit and VAX ALLin1
Question about Kermit 3.01 and the 128-byte history buffer
Kermit & WordPerfect under PCSA
Re: MSKERMIT <-- TCP/IP --> Unix-Host
Digest submissions may be sent to Info-Kermit@WATSUN.CC.COLUMBIA.EDU,
requests for addition to or deletion from the Info-Kermit subscriber list to
Info-Kermit-Request@WATSUN.CC.COLUMBIA.EDU or to KERMIT@CUVMA.BITNET.
Kermit files may be obtained over networks and by mail order. On the
Internetwork, use FTP to log in to host WATSUN.CC.COLUMBIA.EDU, a SUN-4/280
running UNIX (SUNOS 4.1), IP host number 128.59.39.2. Login as user anonymous
(note, lower case), any password, and GET or MGET (MULTIPLE GET) the desired
files. The Kermit files are in directories kermit/a, kermit/b, kermit/c,
kermit/d, and kermit/e. Test versions are in kermit/test. Binaries are in
kermit/bin (use ftp in binary mode). You can also get Kermit files over the
BITNET/EARN network; to get started send a message with text HELP to KERMSRV,
the Kermit file server, at host CUVMA. For detailed instructions, read the
file kermit/a/aanetw.hlp (AANETW.HLP on KERMSRV). To order by mail, request a
complete list of Kermit versions and an order form from Kermit Distribution,
Columbia University Center for Computing Activities, 612 West 115th Street,
New York, NY 10025 USA.
----------------------------------------------------------------------
Date: Mon, 10 Sep 90 16:03:23 EDT
From: Christine M Gianone <cmg@watsun.cc.columbia.edu>
Subject: Next Test Release of MS-DOS Kermit 3.02
Keywords: MS-DOS Kermit 3.02
63. Correct a small problem leading to confusion in terminal emulation if an
arriving control sequence is terminated prematurely by another. msscmd.asm
64. Add command REMOTE SET FILE COLLISION UPDATE, meaning the file receiver
on the other side should process files with the same name as "updates"
(whose meaning is system dependent) to existing files. A MS-DOS Kermit
server does not recognize the command. mssser.asm
65. Another small adjustment to item 62. Move the long interval retrying of
sending characters to be at the final launching point for BIOS comms
channels. The intent is to compensate for a particular X.25 channel which
rejects new characters with a failure indication; that channel is known
to be very sluggish. This change permits character retries for Connect
mode and scripts as well as packets. msscom.asm, msxibm.asm
66. Ensure that the graphics terminal emulation writes characters as opaque,
when so selected, for black and white displays. msgibm.asm
67. Accomodate LATIN5/Cryillic character set of Code Page 866 as much as
possible. The command SET TRANSFER CHARACTER-SET has the new option
LATIN5, and if selected will force FILE CHARACTER-SET to be CP866.
Similarly, if the FILE CHARACTER-SET is selected to be the new option
CP866 then it will force TRANSFER CHARACTER-SET to be LATIN5.
Warnings are issued for both cases.
When sending a text file with TRANSFER CHARACTER-SET LATIN5 the file
attributes packet will send "I6/144" for the file character-set
identification (the "C" attribute) and no translations will be done.
When receiving such a file no translation will be done, regardless of
the currently active FILE CHARACTER-SET. Actually the file should be
rejected unless the user has stated SET UNKNOWN-CHARACTER-SET to KEEP,
but the code gets very messy if the rule is enforced at this time.
The reason is for the coupling of commands is CP866 has many Cryillic
characters in the right, high bit set, section which have no equivalents
in other Kermit character sets and regular Code Pages. Hence, when LATIN5
or CP866 is selected there can be no translation of characters between
CP866 and either the other (Western European) Code Pages or the LATIN1
character set. Thus LATIN5 or CP866 selects a Transparent file transfer
but with a LATIN5 "I6/144" character set identification.
For terminal emulation SET TERMINAL CHARACTER-SET TRANSPARENT will prevent
unwanted translations and also convert the C1 controls area into printable
characters found in CP866.
Files affected are mssfil.asm, mssset.asm, msyibm.asm.
68. Add communications enhancement to using the Int 14h serial port
interceptor TES from Interconnections Inc (distributed with Novell
NetWare for VMS). The new command
SET PORT TES <optional host name>
invokes TES and make a connection to the indicated host. This competes
with SET PORT BIOSn because they use some common operations.
<optional host name> Action
host-name Connect to that host, and disconnect from a previous
TES host if such a connection were active.
* Show all available TES hosts but do not connect.
nothing Use the current TES host name to reestablish an
existing connection. Most useful when switching
between TES and a real serial port. If no host name
exists then a list of available hosts is shown.
Keyboard verb \Knethold is used to invoke the TES command interpreter
while in Connect mode (to revise connections usually), and HANGUP breaks
an existing connection. Sending a BREAK (ALT-B or \Kbreak) performs the
same function for TES as \Knethold. Use this function when manual control
of sessions is desired, such as creating multiple sessions managed by
TES rather than by Kermit.
These new TES operations couple closely with version 2.x of TES and will
not work with earlier ones. Instead use SET PORT BIOS1 and the TES hot
key with older TES's. If the TES hot key is used then \Knethold may not
be recognized; use the hot key again to control TES. If the hot key has
not been used then TES material is shown on the Connect mode screen.
SHOW COMMUNICATIONS will show the current host name, as read back from
TES itself. The apparent serial port identifier on the Connect status
line is the letter "I".
Note that placing a connection on hold and later returning to it may
require pressing the \knethold or \kbreak keys and typing RESUME to make
the TES command interpreter return control to the data stream. It's
apparently a problem in the interpreter.
If Kermit is exited completly then the active session is terminated, but
held sessions are retained. Thus, to preserve sessions after Kermit has
exited use \knethold or \kbreak to get the TES command interpreter and
place the connection on hold manually. The session may be resumed as above.
I am testing with TES version 2.1 beta so both sides have growing pains.
Modifications are all in file msxibm.asm
69. Make an attempt to pass macro arguments \%0..\%9 through (around) a
TAKE command to keep some consistency with MS-DOS Kermit 3.00 and 3.01.
For example,
DEFINE DIAL Take Hayes.tak which has dialing commands
DIAL 555-1212 we want the number passed down to the
interior of Hayes.tak as \%1.
where file Hayes.tak (from the diskette with Chris' book) starts off as
COMMENT - Initial setup
def errstop echo Error: \%1\13,def \%1,hang,stop
set speed 2400 ; *** Change to 1200 ...
<more stuff in here>
echo \13Dialing \%1. Wait...\13\10 ; Dialing for the first time,
goto dial ; so skip the "Redialing" message
COMMENT - Dialing loop
:REDIAL
echo \13Redialing...\13\10 ; Give this message when redialing
:DIAL
output ATDT\%1\13 ; Dial the number
<more stuff follows>
Another example:
DEFINE xxx Set term VT302,take x.tak,connect
xxx A B C
where file x.tak holds
echo \%0 \%1 \%2 \%3
This displays XXX A B C and then enters Connect mode as a VT320 terminal.
70. Add a new element to command line variables, taken from C Kermit 5A.
Numerical variables ARGC, COUNT, ERRORLEVEL, and VERSION can be used as
the ascii representation of their values by the new text substitution
operator
\v(numerical variable)
Soon this will be expanded to include other "named variables" such as
TIME, DATE and so on which are also internal quantities in Kermit.
Examples:
ECHO The version number of this program is \v(version) today.
which displays
The version number of this program is 302 today.
And DEFINE See ECHO \%0 - the num var is \v(\%1) here.
See argc
which displays
SEE - the num var is 2 here.
See version
which displays
SEE - the num var is 302 here.
The expression \v(numerical variable) is replaced by it's ascii value.
This may be used in any command. The current syntax of IF = ARGC 2 ...
is retained for compatibility purposes but in new scripts it should be
replaced by IF = \v(argc) 2 etc.
If the item within the parentheses is not one of the above numerical
variables, or if a right parenthesis does not end the variable name
then the entire expression from \v( to the current point is discarded.
Examples of bad syntax:
ECHO testing \v(versions) stuff
or
ECHO testing \v(version stuff
both of which display
testing stuff
All the changes for this item are in file msscmd.asm
More changes are expected.
[Ed. - The Cyrillic and \v() items are not final. An actual translation
between CP866 (Alternative Cyrillic) and ISO 8859-5 will probably be
installed before 3.02 is released. \v(name) will probably be changed to
require the name of a builtin variable (which need not necessarily be
numeric) within the parentheses.]
------------------------------
Date: Thu, 09 Aug 90 10:22:53 EDT
From: Doug Salane <SALJJ%CUNYVM@cuvmb.cc.columbia.edu>
Subject: Kermit on Novell Networks
Keywords: MS-DOS Kermit, Novell Network
I am trying to use Kermit for station to station communication on my Novell
Network. This is an ELS Level II Network which uses Arcnet cards in a bus
topology. I would like to use Kermit to transfer files between machines
without relying on the server. Is this possible?
I have tried using the SET PORT NETBIOS and SET PORT NOVELL after loading the
Novell IPX workstation shell. I get an error message that says the Network is
not available.
I'd appreciate your help. Thanks.
[From jrd - Doug, what's needed is NetBios running on top of IPX and NET3/4.
It's in the Novell NetWare collection. After that just follow the instructions
in the Kermit documentation about SET PORT NET <name>. I use Kermit that way
here, and it is also the official AT&T method of talking to an AT&T Unix
machine from a DOS PC. SET PORT NOVELL(NASI) uses the NASI/NACS async server,
not what you want at all.
The short form of the documentation says start the Server Kermit
first (SET PORT NET, Server). No destination address is needed by the server.
Then start a client Kermit with SET PORT NET <server's NetBios name> and
issue file transfer or REMOTE commands.]
------------------------------
Date: Sat, 21 Jul 90 14:23:19 PDT
From: leland@scdt.intel.com (Leland Bruns )
Subject: Feedback on MS KERMIT 3.02 of July 20
Keywords: MS-DOS Kermit 3.02
I am forwarding some feed back on the test version of MS KERMIT version 3.02
and some comments on version 3 in general. Please forward to the appropriate
persons. Thank you.
ENVIRONMENT
First, my environment is a Wedge Technology 386-based PC-AT clone, with a
Video-7 VGA board. I am compiling using MicroSoft's MASM version 5.01. My
hosts are VAX VMS and Ultrix machines. The PC's operating system in PC-DOS
version 3.30.
[From jrd - similar to my own environment: DELL 310, Video-7 VGA, VMS host,
PC-DOS 3.30]
I have recently copied MS KERMIT version 3.02 from the /kermit/test directory
at watsun.cc.columbia.edu. My comments below are based on the versions dated
July 15 and July 20. (I see there is a new version this morning, July 21, but
have not had the chance to port or test this version.) I am comparing the
test version 3.02 to the original 3.00 distribution version.
TEKTRONIX EMULATION PROBLEMS
I see that there are significant changes to the MSGIBM.ASM module for
Tektronix emulation, and have concentrated on looking for differences there.
The one difference I would like to report is in the way GIN input apparently
works in the new version. In the original 3.00 version, when a key was
entered at the PC, the host would receive the ASCII code for the character
entered, the 4 character xy coordinate, and one final character (I can't tell
easily from the host software I have to work with what that final character
is, as the software discards it).
The new version 3.02 test version apparently does not send the GIN report to
the host until the user has entered the desired key (at which time the
cross-hair cursors disappear) AND an additional carriage return. Not too
surprising, the host receives an additional character which must be ignored to
continue properly.
I believe that the behavior of the original version 3.00 of MS KERMIT is
correct, while that the new test 3.02 version is incorrect. The user should
not have to enter both the desired key plus a carriage return for each GIN
input point. I would very much like to see the original behavior restored.
[From jrd - advice is most welcomed because my Tek docs seem to differ
about what should happen (they are not consistent either).]
FEEDBACK & ENHANCEMENT REQUEST
Since I'm taking the time to report the problem above, I'd like at the same
time to pass on some positive feedback on MS KERMIT version 3 and an
enhancement request. I very much like the work that has gone into MS KERMIT
between versions 2 and 3. KERMIT is unique in its universal availability, its
faithful terminal and graphics emulation, and public availability of source
code.
The latter feature, availability of source code, is important to me as I have
one minor hardware incompatibility for which I must modify each KERMIT release
as it becomes available. Specifically, I have dedicated hardware on all four
COM ports, and on IRQ2 thru 5 (see the comments below). In order to make use
of my 9600 Baud Telebit Trailblazer modem on COM3, IRQ5, I must create a copy
of MS KERMIT modified to use IRQ5 rather than IRQ4.
I don't think my situation is too uncommon these days, as more and more
powerful PC's become available and are equipped with more and more hardware,
requiring use of COM ports 3 and 4. I frequently see queries posted asking
how to get around the restrictions imposed by the current MS KERMIT. I think
many users would benefit the ability to specify the interrupt number and
address on a per port basis...just as the newer MS KERMIT versions now allow
specification of the COM port address. Failing that, the next best thing
would be to allow the user to define the interrupt numbers and addresses in a
straightforward manner in the MSXIBM.ASM source (vs. the convoluted encoding
required at present...see my difference file below).
[From jrd - But.... If one is not using all the COM ports simultaneously then
one can use two at once if they are on different IRQ lines. Just ask Kermit to
touch the competing ports so their interrupt driver chips are taken off the
bus (as decent software should have done anyway). I am reluctant to move far
from IRQ 3 and 4 because of the interference with other peripherals (IRQ 5 is
the hard disk controller on XTs). I can't easily change things in v3.02
anwyway because of the necessary revisions to the data structures. Btw, I too
have a Trailblazer (T2500), and I avoid most serial port conflicts via a $30
RS232 A-B-C-D box; COM1 is comms, COM2 is the mouse.]
Once again, let me say how much I appreciate the work going into MS KERMIT
version 3. Thanx in advance for your consideration of the suggestions above.
[From jrd - Thank you for the suggestions and the compliment.]
Leland M. Bruns
Intel Corporation
(408) 765-4219
leland@scdt.intel.com
------------------------------
Date: Wednesday, 25 Jul 1990 13:06:12 EDT
From: "Larry F Armbruster" <U22C5%WVNVM@cuvmb.cc.columbia.edu>
Subject: Kermit Problems - V3.01
Keywords: MS-DOS Kermit 3.01
I have experienced several problems using Kermit. The configuration
of my system is:
PS/2 Model 50Z with an AT&T Voice Data Multiplexer dialing into a
DECserver 200 V3.0 (BL33) - LAT V5.1 and accessing VM/CMS through
a DECnet/SNA Gateway into a SIM3278 session manager. Also at times
I access a VAX cluster.
I have redefined the PS/2 function keys to emulate the VT100/200/300 series
of terminals.
When I emulate a VT320 - the session messages (ie SYS AVL, etc) writes to
the top line of the screen thus wiping out the top line of the screen.
When I emulate a VT102 - the session messages are at the bottom of the screen.
This is only an annoyance and I can live with emulating the VT102.
The serious problem is the inability to use Kermit file transfers on either
CMS or VAX. My computer center is aware of the CMS problem and attribute it
to the session manager. However we cannot explain the VAX problem. The VAX
Kermit Host (or maybe Mskermit) issues a parity error. This is the case no
matter what parity, duplex, or flow control is used. Any suggestionor help
is greatly apperciated.
[From jrd - since the pathway between the PC and either IBM/CMS or the VAX
goes through two external communications boxes one supposes that one or the
other box is adding its own parity, or more likely, is unable to handle a
stream of characters as long as a packet. Otherwise MS-DOS Kermit does run
fine to VAXen through common LAT boxes. Logging packets on both ends and
viewing the log files will show which machine is experiencing the troubles.]
------------------------------
Date: Wed, 15 Aug 90 08:08 CDT
From: DEC-WOLF <CC_BRYSON%SWTEXAS@cuvmb.cc.columbia.edu>
Subject: MS-Kermit and VAX ALLin1
Keywords: MS-DOS Kermit 3.01, VAX/VMS, ALLin1
Hello all,
I was wondering if there were other Ms-Kermit (3.01) users who are using PCs to
communicate with a VAX and who use ALLin1, who are also experiencing problems
with printing over the printer port. We have determined that everything works
fine when the Vax thinks the terminal type is a VT200. If the terminal type is
NOT VT200 series than the printing is erratic or non-existent. After
researching the problem I dicovered that Allin1 is sending the following escape
sequences for the two terminal types:
VT20O Esc [ 5 i
VT100 Esc [ 5 i Esc [ 6 i
VT300
The Esc [ 5 i is the proper sequence to open the printer port for a VT100,
VT200, or VT300 terminal. After calling Digital, I was told that (Esc [ 6 i)
was the escape sequence to open the printer port for the VT125 terminal.
Evidently the sloppy programmers at DEC check for VT200 ONLY and send the VT100
and VT125 sequences for other terminals. It seems that either Kermit or the
printers are confused by the (Esc [ 6 i) sequence. I know that Dec should just
straighten their act out but I was wondering if anyone knew how to work around
this problem from the Kermit end. Could the (Esc [ 6 i) sequence be ignored or
stripped out by Kermit? We plan on just changing the terminal type to VT200 on
the Vax end but it seems a bit kluging to me (normally the Vax and terminal
communicate the terminal type between each other at login).
Bill Bryson
User Services Analyst
CC_BRYSON@SWTEXAS
P.S. Our systems personel could also patch allin1 to change the printer
scripts (i.e. not send ESC [ 6 i) but feel it is unlikely that they will (it is
considered a no-no to "mess" with Digital software).
[From jrd - MS-DOS Kermit v3.02 consumes the ESC [ 6 i and does not send it
to the printer. Yes, it's a carry over and should not be there. Here is the
entry in the file msr302.upd describing changes to version 3.02 -
15. Correct bug in transparent printing within text emulator which corrupted
back to back escape sequences sent to the printer.]
------------------------------
Date: Fri, 27 Jul 90 17:47 EDT
From: John W Manly <JWMANLY%AMHERST.BITNET@CUNYVM.CUNY.EDU>
Subject: Question about Kermit 3.01 and the 128-byte history buffer
Keywords: MS-DOS Kermit 3.01, CLEAR Command
Hi there. Another question about MS-Kermit and the use of the history buffer
in Take or Script files.
Could anyone fill me in on the relationship between the INPUT statment, the
REINPUT statement, the CLEAR command, and the 128-byte history buffer?
In particular, the manual states that REINPUT only accepts additional input
from the serial port until the history buffer is full. How does one then
clear it out? A CLEAR command presumably does so. Does an INPUT operation
clear it as well, or clear it only if it's full, or anything like that?
Any help would be greatly appreciated. Thanks.
- John W. Manly
BITNET: JWMANLY@AMHERST Systems Manager
PHONE: (413)-542-2526 Amherst College
[From jrd - The CLEAR command does indeed clear things: the history buffer
and the serial port receive circular buffer (preceeds the history buffer).
I agree that a separate script history buffer clearing command would be
beneficial. In the meanwhile, the history buffer can be made either larger
or smaller via the DOS command
SET KERMIT=INPUT-BUFFER-LENGTH value
value is number of bytes (default 128)]
------------------------------
Date: Tue, 31 Jul 90 14:44 EDT
From: Pete Kanaitis <X979PK0P@VB.CC.CMU.EDU>
Subject: Kermit & WordPerfect under PCSA
Recently, we have been looking at WordPerfect 5.0 under VMS and have tried
the MS-Kermit macros and DCL command procedure (ftped from watsun) to set
the IBM PC keyboard to do WP bindings. This seems to work fine under any
asynch connection, but does not seem to work with MS-Kermit running under
PCSA. Under PCSA, when the terminalS and terminalR macros are envoked,
the key bindings are set, the connect command gets issued, but the original
terminal session gets disconnected! (Since we use virtual terminals, we
can re-connect back to the WordPerfect session, and resume, but only after
we re-login in)
Here's an example of what is happening:
MSKERMIT.INI has the following:
set port decnet singer
set terminal wrap on
define terminalS take C:\wp30.ini, connect
set key \2344 \KTerminalS
Then from the PC:
C>kermit
IBM-PC Kermit ...
MS-Kermit>connect
Welcome to Singer...
Username: FOOBAR
Password:
$ @PUBLIC:PCWP
What type of Graphics adapter are you using (ega, cga, vga)? vga
$ WP FILE.EXT
Defining Function Keys for WordPerfect Products....
(After a few seconds Kermit reconnects..)
(Nothing happens. When I hit any key...)
Welcome to Singer...
Username: FOOBAR
Password:
You have the following disconnected process:
...
Connect to above listed process [YES]:
(Now you are back in WordPerfect)
(When you exit..)
Save Document? (Y/N) Yes
Exit WP? (Y/N) Yes
WP key definitions cleared...
(screen clears, back to normal)
Does anyone have any ideas?
Thank you...
Pete Kanaitis
Manager, Computational Resources
Allegheny-Singer Research Institute
X979PK0P@VB.CC.CMU.EDU (Internet)
X979PK0P@CMCCVB (Bitnet)
[From jrd - When WordPerfect starts up it requests a long Color Palette
report from the VT300 terminal. Kermit sends the report fine, but some
communications channels can't handle 200 some odd byte bursts. Also, the
VAX must have SET TERM/HOSTSYNC so that the VAX can XON/XOFF the terminal.
There is really nothing I can do, because on my VAX some paths just can't
cope, even when I insert massive pauses between characters. Sometimes
it's the length of the DEC Color Palette report which is the root problem.
My test copy of WordPerfect/VMS has exceeded it's 6 month lifetime, but
I recall that I could use PCSA/DECnet-DOS fine this way. The only "but" is
the VMS command SET TERM/HOSTSYNC must be stated.]
------------------------------
Date: Wed, 18 Jul 90 14:34:55 -0400
From: mregeste@amo.Wichita.NCR.COM
Subject: Re: MSKERMIT <-- TCP/IP --> Unix-Host
Keywords: MS-DOS Kermit, TCP/IP, Unix
In article <1318@kullmar.se> you write:
>In <1990Jul10.134437.18854@duc220.uni-duisburg.de> hz231gr@duc220.uni-duisburg.de (Gressel) writes:
>
>>Hello,
>
>>we tried it with an Interlan-NI5210-Card, no success!
>
>In Kermit News number 4, June 1990 Joe R. Doupnik and Christine M. Gianone
>writes:
>
>TCP/IP is a very widespread networking method linking machines locally and
>around the world. ...
I would like to add that Wollongong's Pathway Client Plus also supports
the INT 14 service. We use both Kermit and Procomm Plus Network Version
and they work great!
Mark Regester Information Systems & Services, NCR Peripheral Products Division
NCR:654-8340 <M.Regester@Wichita.NCR.COM>
(316)636-8340 <uunet!ncrlnk!ncrwic!m.regester>
FAX:636-8889
[From jrd - Yeah! A WINning way.]
------------------------------
End of Info-Kermit Digest
*************************