kermit.columbia.edu/kermit.columbia.edu.tar/kermit.columbia.edu/e/v12.2

Internet Message Format | 1990-08-06 | 21KB

From cmg  Fri Aug  3 14:50:25 1990
Return-Path: <cmg>
Received: by watsun.cc.columbia.edu (5.59/FCB)
    id AA29240; Fri, 3 Aug 90 14:50:25 EDT
Date: Fri, 3 Aug 90 14:50:24 EDT
From: Christine M Gianone <[e-mail]>
To: Info-Kermit
Subject: Info-Kermit Digest V12 #2
Reply-To: [e-mail]
Queries-To: [e-mail]
Message-Id: <[e-mail]>

Info-Kermit Digest         Fri, 3 Aug 1990        Volume 12 : Number 2

Departments:

  ANNOUNCEMENTS -
        Another Update of the MS-DOS Kermit 3.02 Test Release
        Version 8.12 of PRIME Kermit Now Available
        New Documentation Available for VAX/VMS Kermit-32 3.3.126
        New Test Release of DEC Rainbow Kermit

  MS-DOS KERMIT -
        MS-DOS Kermit 3.0x and Microsoft Windows 3.0
        MS-DOS Kermit 3.0 Hints
        MS-Kermit Server and Hayes 2400bps Modems?

  KERMIT-370 -
       IBM Mainframe Kermit and Trailing Blanks
       IBM Kermit 4.2 and tn3270

Digest submissions may be sent to [e-mail],
requests for addition to or deletion from the Info-Kermit subscriber list to
[e-mail] or to [e-mail].

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: Wed Aug  1 19:56:09 1990
>From: Christine M. Gianone <[e-mail]>
Subject: Another Update of the MS-DOS Kermit 3.02 Test Release
Keywords: MS-DOS Kermit 3.02

>From Joe Doupnik.  Changes since the version announced in Info-Kermit V12 #1
are listed below.  Items 38 and 39 will be of interest to Honeywell users for
supporting connections to the Honeywell DPS-6 (see article by Frank Dreano in
Kermit News #4) -- they remove the need for a custom version of MS-DOS Kermit
to communicate with the the Honeywell systems.  Please report any problems to
Joe at [e-mail] or [e-mail].

30. Add correction to ensure NUL and DEL are passed through when printing
    transparently. msxibm

31. Revise REPLAY filespec a little. Previously at the end of the file any
    key would exit Connect mode. Now all keys remain active so that screen
    dumping and screen rollback function; ALT-X and Control-C (emergency
    bailout) exit Connect mode. msxibm

32. Internal revisions to details of handling 75/1200 baud split speeds.
    Previously, filling the receiver buffer would cause an XOFF to be sent
    at 1200 baud rather than at the correct 75 baud. Now the XOFF character
    is sent at 75. A pseudo-queue is present to prevent conflicts of using
    the timer chip for beep or regular 75 baud output when the receiver
    interrupt occurs (and resulted in the XOFF being sent at interrupt level
    rather than at task level). The queue makes the XOFF be sent at task level
    and at the proper 75 baud rate. Dan Norstedt had the idea, this queueing
    stuff is my solution. 75 baud output is done as a software UART. msxibm

33. Details here and there to save code segment space, from Dan Norstedt.

34. Add terminal type of VT100. This performs identically to a VT102 except
    the response to the type inquiry from a host is ESC [ ? 1 c (a VT100).

35. Found and fixed a protocol problem present in MSK 3.0 to date. If an I
    packet is sent to a server Kermit which does not understand I packets
    the resulting Error packet from the server is to be taken as a forgivable
    error, rather than otherwise being a fatal one. The version 3.0 and 3.01
    code forgot to forgive. I packets are sent by GET and REMOTE commands.
    Fixed on 19 July. mssser.

36. Remove VTxxx response to CSI [ .. R. That should not have been present.
    Thanks to Ted Medin for discovering its presence.

37. Add 80/132 column support for Paradise Autoswitch EGA Mono board, thanks
    to David Zielke of Duke Univ. msyibm

38. Add two new commands

    SET SEND DOUBLE-CHAR <number between 0 and 255>
        which sends this character twice in a row in outgoing packets.
        The doubling occurs as packet bytes are being sent to the
        communications channel and hence does not occur during packet
        construction. This accomodates communications devices which
        need doubling to let through a single copy. The doubling
        will not appear in the packet log. Padding characters are
        not affected by this command since padding is to be discarded
        at the other end anyway.

    SET RECEIVE IGNORE-CHAR <number between 0 and 255>
        which ignores this character in received packets. This
        accomodates communications devices which may insert a
        character into a stream, such as a Line Feed inserted before a
        Carriage Return. The ignored character will not appear in the
        packet log. Note that NUL (0) and DEL (127) are normally
        ignored even without this command.

39. Add terminal type of HONEYWELL to the emulator. This is a VT100 with
    two special responses. A received ENQ character, decimal 5, causes the
    emulator to send the 14 characters

        '7813  P GC  A' 03h

    and receipt of  ESC y  causes the emulator to send the 27 characters

       ESC '[8p  OT' 03h ESC '[y7813  P GC  A ' 03h

    where the quote marks (') above are not sent but are present so we can
    see the spaces which are sent; other spaces are for reading convenience.
    03h is Control-C, a binary value 3. The terminal ident response is VT100.
    These Honeywell VIP7809 emulation features are from Frank Dreano,
    [e-mail].

40. Modify the internals for SET KEY LK250 to result in checking for an
    active external LK250 driver and if found activate the LK250 usage in
    Kermit. If the driver is not found the message "?LK250 keyboard external
    driver is not active." is displayed. Note that to obtain the default
    LK250 key mappings the user must say SET KEY CLEAR after the above
    command, and then add any local definitions. Tnx to Terry Kennedy. msuibm

41. In command WAIT <timeout> \CD \CTS \DSR   let the backslash character
    be optional, to merge syntax with C Kermit 5A. mssscp

42. Add further Honeywell terminal type modifications from Frank Dreano:
    - change invocation ESC y to be ESC [ y     (see item 39 above)
      This will cause loops if the host echoes the response.
    - add  ESC `  (accent grave) clears screen and homes the cursor.

43. Revise slightly the query for the active Code Page, and the table of
    character sets for use by SET FILE CHARACTER-SET, to be system dependent
    so that MS-DOS in Japan can be used more effectively. That MS-DOS lacks
    the concept of Code Pages. Thanks to Hirofumi Fujii for the suggestions.

44. Add another host invocation control sequence to the text terminal emulator.
    Reception of  CSI Pn;...Pn ~  invokes a macro named PRODUCT, if it
    exits, and sets variables \%1, \%2, and so forth up to \%9 to be the ascii
    version of the numeric parameters Pn above. Omitted parameters are taken
    to be zeros. Only as many variables are set as numeric parameters, and
    script item ARGC is one larger than this quantity. If macro PRODUCT does
    not exist only the \%x items are changed. Connect mode is not reentered
    automatically if the macro exists. This is very similar to TERMINALR and
    TERMINALS, except for the variable setting part.
    The invocation sequence may change so don't write it into applications
    until Kermit is released again. It is not invocable from the keyboard,
    because of the needed parameters. mssset.asm, msyibm.asm, mszibm.asm.

45. Correct a keyboard translation problem. Control codes could be translated
    to printable characters for a few National Replacment Character sets,
    because that was where the special graphical symbols are located on the
    IBM PC display adapter. This has been corrected to never translate control
    codes.

46. Slightly revise script INPUT and REINPUT command sensing of timeouts to
    do the time reading after each serial port character reading operation.
    This might cure complaints of REINPUT 0 reporting premature FAIL status.

47. If disk reading fails while sending a file tell the receiver to dispose
    of the file with a Delete status, rather than the previous status of
    just end-of-file. There is no other simple way of informing the receiver
    of such a local problem. Other files in a group will be processed normally.

  More changes are expected.

------------------------------

Date: Fri Aug  3 10:46:17 1990
>From: Christine M. Gianone <[e-mail]>
Subject: Version 8.12 of PRIME Kermit Now Available
Keywords: PRIME Kermit

>From Matthew Sutter of Lincoln National Corporation in Fort Wayne, Indiana,
USA, comes version 8.12 of PRIME Kermit, an upgrade of version 8.00 that was
prepared by John Horne of Polytechnic South West in Plymouth, England.

The major feature of version 8.12 is support for local-mode (dialout)
operation using the AMLC lines.  CONNECT, GET, FINISH, and BYE commands have
been added, as well as the script commands INPUT, OUTPUT, CLEAR, and PAUSE.

This version also contains fixes to repeat character processing and sliding
windows, but the sliding window feature still needs more work in order to
interoperate with MS-DOS Kermit 3.0 or C-Kermit 5A.  This work is underway,
watch Info-Kermit for further announcements.

The new files are in kermit/d/prime8.* on watsun, and PRIME8.* on CUVMA.
There is, as yet, no new documentation.  Thanks to Matt for his work and for
contributing it to the Kermit collection!

------------------------------

Date: Wed Aug 1 18:29:49 1990-EDT
>From: Christine M. Gianone <[e-mail]>
Subject: New Documentation Available for VAX/VMS Kermit-32 3.3.126
Keywords: VAX/VMS Kermit

The VAX/VMS Kermit-32 chapter of the Kermit User Guide has been updated to
include all the information about the new release, and the help text has also
been updated.  Thanks to Jonathan Welch of the University of Massachusetts for
his help in clarifying the actions and interaction of the SET FILE BLOCKSIZE
and SET FILE TYPE commands, and for proofreading the manual.

The files for the new manual are VMSHDR.MSS and VMSMIT.MSS (Scribe source),
VMSMIT.DOC (plain ASCII text) and VMSMIT.PS (Postscript).  The help text is in
VMSMIT.RNH (Runoff source).  Tape B (kermit/b, etc etc).

------------------------------

Date: Thu Aug 2 13:01:34 1990-EDT
>From: Christine M. Gianone <[e-mail]>
Subject: New Test Release of DEC Rainbow Kermit
Keywords: MS-DOS Kermit, Rainbow Kermit, DEC Rainbow

>From Robert Weiner, a new version of Kermit for the DEC Rainbow, based on
the system-independent parts of the July 22nd MS-DOS Kermit 3.02 test release.
The Rainbow version lacks many of the IBM-specific enhancements that appeared
in MS-DOS Kermit 3.0, like international character sets, instantaneous screen
rollback, etc, but these MAY be added before the final release of this
program.  The system-independent features -- script programming enhancements,
sliding windows, long packets, VT200/300 emulation (it's still at the VT102
level), etc.

Thanks to Rob for keeping the Rainbow up to date!  The files are in
kermit/a/mstrb1.boo ("BOO" file to be un-boo'd into an .EXE file using any of
the MSBPCT files) on the Internet or MSTRB1 BOO on BITNET KERMSRV at CUVMA.
For internet binary-mode FTP access only, also kermit/bin/mstrb1.exe.

------------------------------

Date: Thu, 2 Aug 90 11:18:30 PDT
>From: [e-mail] (William Hall)
Subject: MS-DOS Kermit 3.0x and Microsoft Windows 3.0

Just tried MS-DOS Kermit 3.0 in enhanced mode Windows.  I began by running a
DOS shell then running Kermit.  Did not bother with a PIF file.  I changed the
settings to run from exclusive to background.  You can get to these setting
with ALT-SPACE, then moved it to a Window.  For fun, I ran vttest on UNIX and
checked the character set, especially the one for DEC special graphics.  They
looked as expected.

Then I tried a long file transfer.  I made the window iconic, and the file
transfer ran in the background to completion with no errors.  At the end, the
bell beeped to let me know it was done.

So it works and quite well to the extent that I tested it.

--Bill

------------------------------

Date: 1-AUG-1990 20:33:04.67
>From: Joe Doupnik <[e-mail]>
Subject: MS-DOS Kermit 3.0 Hints
Keywords: MS-DOS Kermit 3.0 Troubleshooting

[Ed. - This is extracted from a reply by Joe to a user who reported a
nonfunctioning Kermit on a Compaq computer.]

One common item is an internal modem which fails to follow the normal rules of
how to behave like a regular serial port.  Another is two devices are having a
small battle for an IRQ line or a memory (port) address.

MS-DOS Kermit 3.02 (not 3.2) is strictly development software, as indicated in
the note file MSR302.UPD and on the startup banner.  The REMOTE command
problems you report fall into the category of fixed, broken, and then fixed
again (sigh).  In all of the broken cases Kermit would retry many times and
that takes up to a minute or so to finish.  While that is happening one side
can send an XOFF to the other and forget to send a matching XON later; Kermit
waits 10 seconds to figure out this situation when it needs to send and breaks
through regardless.  So, this "flow control deadlock" can add to the waiting
interval, but it's better than being told to never speak ever again.

Another situation, which happened to me a couple of years ago, was the poor
hard disk controller had a conflict with another board and every once in a
while the machine would hang waiting on the hard disk for 10 minutes or more.
Usually I had to push the big Red switch. When Kermit has to go to the hard
disk then these kinds of things can happen.

Generally, I try to sort out the problem by trying Connect mode between
machines right away. If that works then I know the communications line part is
basically ok.  If packets fail then I say SET DEBUG ON and SET DISPLAY SERIAL
to watch what's happening during a file transfer or REMOTE command (REMOTE WHO
is my favorite "protocol" quicky, no disk access).

Let's suppose the Compaq is in sad shape.  My suggestion is to backup the hard
disk contents and ensure they are correct (say to DOS VERIFY ON to help).
Then do a low level format of the drive (yow!) and a DOS FORMAT with the
correct version of DOS.  Put things back on the hard disk and keep your
fingers crossed.  Double check the settings of the boards too.  Unload the
likely mountain of favorite TSR's and see if things work.  If they do then
start adding them back one at a time and retesting.  Make sure that CONFIG.SYS
has BUFFERS=20 or so and FILES=20 or so.  Watch out for Desqview because it
has troubles remembering that a communications program needs to service
interrupts from the serial port.  Don't mix DOS files from different
releases/manufacturers.

Ensure that parity is the same on both machines, NONE being better than any
other flavor (Kermit does it's own checking in a much more secure way).  Don't
use MODE COM1:....,P with that ",P" (for Printer waiting) on the same port as
the communications channel because DOS will grab the port and won't release
it.  Watch for mice drivers nibbling on the same serial port.  In short, take
the machine down to bare metal and rebuild it slowly from a consistent set of
original DOS floppies (put write-protect tabs on them).

Did I forget to mention virii?

------------------------------

Date: Tue, 02 May 89 16:54:34 CDT
>From: Barry Pederson  <[e-mail]>
Subject: MS-Kermit Server and Hayes 2400bps Modems?
Keywords: MS-DOS Kermit Server Mode

Hi, we've been using MS-Kermit for a while now, and it's been incredibly
useful, but we're having a slight problem.  We have some IBM model 30-286
machines connected to Hayes 2400bps external modems.  They work fine except
when we try to have the modems answer the phone while Kermit is in server
mode.

We have a script set up to set the speed to 2400, parity to none, then we
output ATS0=1\13 to the modem, and go into server mode.  Sometimes it works,
but a lot of the time, when the phone rings the modem will drop down to
1200bps.

In a previous Kermit Digest, I thought I saw something about C-Kermit having
problems with Hayes modems, does MS-Kermit have a known incompatibility
between it's server mode and a Hayes modem?

I suspect that Kermit is working fine, but the Hayes is doing something screwy
(we've tried MS-Kermit 2.31 and 2.32/a).  I'd really appreciate any
suggestions on how to keep the Hayes from slowing down.

Thanks.

Barry Pederson    University of North Dakota School of Medicine
     [e-mail]     or    [e-mail]

[Ed. - When MS-Kermit is in server mode, it times out every 30 seconds or so
and transmits a negative acknowledgement packet (the familiar "^A# N3").  It
is likely that the NAK packet is confusing the modem.  The workaround is to
give MS-DOS Kermit the SET TIMER OFF command before putting it in server mode,
which is safe as long as the other Kermit will do timeouts (some, like IBM
mainframe CMS Kermit, cannot).  In MS-DOS Kermit 2.32/A or later, you can
disable server command wait timeouts without turning the timer off, by using
the command SET SERVER TIMEOUT 0.]

------------------------------

Date: 6-Jul-90 22:03:17-GMT
>From: Randy [e-mail]
Subject: IBM Mainframe Kermit and Trailing Blanks
Keywords: IBM 370 Kermit, VM/CMS Kermit, Blanks

We're running Kermit CMS and MS Kermit to transfer files with a record length
of 1500 bytes. What's happening is that the records have trailing blanks.
Kermit strips these blanks off. We need them to remain intact for input into a
commercial package we are running under DOS. Is there any way to keep the
blanks from being stripped? I know that you can specify LRECL and RECFM at the
CMS level but that is for files being received in CMS.  We need the same
capability for MS-DOS Kermit.

Randy Schirmer
West Chester University, West Chester, PA
215-436-1094, RSCHIRME@WCU

[From John Chandler - When the file type is TEXT and the recfm is F,
Kermit-370 strips off trailing blanks from each record before sending a file.
In order to keep trailing blanks in a downloaded file, you can convert it to
RECFM V in CMS before sending it, e.g., by using COPYFILE with the RECFM V
option.  This is true of versions from 4.1 onward.]

------------------------------

Date: Wed, 20 Jun 90 08:33:25 -0500
>From: C Lance Moxley <[e-mail]>
Subject: IBM Kermit 4.2 and tn3270
Keywords: IBM 370 Kermit, tn3270

I just assembled Kermit/370 on an IBM 3081 running VM/SP 5.  The program
assembled just fine.  But, I have no idea how to set things to work through
tn3270. I've used MS-DOS Kermit for quite a while now and also am very
comfortable with C-Kermit, but this seems to be alot different. I would
appreciate any help you can give on this.  I also have the ability to use 3708
protocol converters, but I don't think there is any way I can put them in
line-by-line mode. I would really prefer to use tn3270. If you are the wrong
person to ask this to, please direct me to the right person. Also if there are
any examples anywhere, let me know. I couldn't find anything on watsun.

Thanks for a very fine product. The other versions that I use are the best of
their kind. Especially the new C version.

                                Thanks,
                                Lance

[From John Chandler - Actually, no, I haven't used tn3270 myself, but Bob
Babcock has -- and from a Unix system only.  It should have been essentially
the same as your attempt, but it worked.  The special considerations are only
two: you must SET CONTROLLER SERIES1 by hand, and you cannot (apparently)
upload using long packets.  The system in question was a CONVEX, as opposed to
a SUN, but I think the tn3270 would have been the same.]

------------------------------

End of Info-Kermit Digest
*************************