SY.FDC@CU20B.COLUMBIA.EDU (Frank da Cruz) (12/29/87)
Info-Kermit Digest Mon, 28 Dec 1987 Volume 6 : Number 28
Today's Topics:
Announcing IBM Mainframe VM/CMS Kermit Version 4.0
MacKermit Status?
Version 0.9(36)b4 of Macintosh Kermit Available for Testing
Kermit News
Request for Kermit Information
VAX/VMS Cluster and Kermit Problem
Non-Bug Report, MS-Kermit, CP/M Kermit (Kaypro)
Kermit 3.79 on Apple 2c
IBM 370 Mainframe UTS24-Kermit?
Kermit & Everex Modems?
----------------------------------------------------------------------
Date: Mon, 1987 Dec 21 12:08 EST
From: (John F. Chandler) PEPMNT@CFAAMP.BITNET
Subject: Announcing IBM Mainframe VM/CMS Kermit Version 4.0
Organization: Harvard/Smithsonian Center for Astrophysics
Keywords: IBM 370 Kermit, VM/CMS Kermit, CMS Kermit
Xref: IBM Mainframe, Also see IBM 370
Xref: VM/CMS Kermit, Also see VM/CMS Kermit, IBM 370
This is to announce CMS Kermit Release 4.0 for IBM 370 series mainframes
running the VM/CMS operating system. The program is now a member of the
generic family Kermit-370 and appears in the Kermit distribution under a new
prefix: all CMS-specific files begin with IKC, while generic Kermit-370 files
begin with IK0 (I K Zero). Kermit-CMS no longer consists of a single source
file. Instead, the source is split into sub-files, some generic and some
CMS-specific. The separate pieces are to be recombined into a single
composite source (or made into a macro library) for installation. See the
file IKCKER.INS for instructions. Generally, the files formerly known as
CMSKERM.* have been replaced by new IKCKER.* files. The companion TSO Kermit
Release 4.0 is still in the testing and debugging stage, but should be
available soon. Anyone interested in helping to get Kermit-TSO ready for
release should contact John Chandler <PEPMNT@CFAAMP.BITNET>.
Below is a list of the more important additions in Version 4.0:
--- generic features ---
1. Code reorganization into generic 370 and system-specific sections.
2. Optional separate translation tables for counteracting the system
conversion of terminal I/O.
3. New GIVE command for saving a modified translation table.
4. A new, RAW debug mode for recording the packet traffic as actually
sent and received on "GRAPHICS" and "SERIES1" devices.
5. Preservation of the case of commands as typed, with uppercase
conversion of only those words that must be uppercase.
6. New SET MARGIN command for limiting the width of a file to be sent.
7. Settable tab stops for Kermit's conversion of tabs to spaces
(alternative to the default 1, 9, 17, etc.).
8. Replace SET SERIES1 subcommand with new SET CONTROLLER. Support for
multiple terminal controller types.
9. New DIRECTORY and HOST subcommands following Kermit standard.
10. Combination of file-attribute SET subcommands (FILE-TYPE, LRECL, and
RECFM) into a new group SET FILE.
11. Separate retry limits for initial and subsequent packet exchanges.
12. Pad binary records on disk with nulls, rather than blanks.
13. Automatically tune packet length when sending long packets according
to heuristic optimum based on sparse Poisson statistics, provided
that transmission errors do occur.
14. Expand STATUS report to include the number of files in the last
transfer, throughput statistics, heuristic optimum packet length
(when long packets are enabled), and the reason for any file
rejection based on A-packets.
15. New command TDUMP NAMES to display the list of files sent in the
last transfer.
16. Add file creation date to A-packet repertoire.
17. REMOTE COPY and REMOTE RENAME commands to a server at the other end.
18. Allow long packets through a 7171 with VTAM.
19. New type D-BINARY for binary files with undelimited variable-length
records.
20. SET 8-BIT-QUOTE. Allow 8-bit data where possible via SET PARITY.
21. SET SYSCMD, so that Kermit can be told to try "illegal" subcommands
as host system commands instead of just rejecting them.
22. SET PROMPT subcommand.
23. Do not forget parameters specified by the other Kermit in I-packets.
24. Keep track of truncated records during a RECEIVE operation and
report the count in STATUS; also call truncation an error after
everything is received.
25. SET HANDSHAKE subcommand to alter or suppress handshake character
Kermit-370 sends out after each packet.
--- CMS-only features ---
26. System commands issued through Kermit via the subcommands CMS or
HOST are automatically passed on to CP if (a) CMS rejects them and
(b) IMPCP is set ON.
27. Kermit subcommands may be executed directly from CMS EXEC's.
28. Reject files known (via A-packets) to be too big for available
storage.
29. Bypass user translation tables and set TERMINAL SCROLL CONT for
protocol mode on TTY lines.
30. KERMBOOT avoids the loading problem (VIRTUAL STORAGE CAPACITY
EXCEEDED) due to large GLOBAL TXTLIB's and preserves the untokenized
command line so that Kermit may be given mixed-case or long words as
part of the initial commands.
[Ed. - Thanks, John! And thanks to all who helped put this new program
together. This is a kind of milestone in Kermit development. It should allow
the many IBM mainframe operating systems to run a common, advanced version of
Kermit. All that's necessary is for some volunteers who are expert in
MVS/TSO, DOS/VSE, MUSIC, MTS, GUTS, and the various other 370 OS's to step
forward and fill in the system-dependent modules for their systems (as John
points out, the TSO version is nearly complete, but still needs some testing
and debugging). If you want to volunteer to help, please contact John
directly, cc to Info-Kermit@CU20B. The files are in KER:IK*.* on CU20B,
available via anonymous FTP, or on BITNET as IK* *, available from KERMSRV on
host CUVMA, and replace the files whose names started with CMS.
By the way, some of you may have seen an article in Digital News, November 30,
called "Advanced Kermit Version Available Soon for VAXs" (p.10). It reported
on part of our talk at Tokyo DECUS, in which we described John's portable IBM
mainframe Kermit. The editors at Digital News felt that putting "VAX" in the
headline would give the article more of a DEC slant, but it's obviously
misleading.]
------------------------------
Date: Sat, 28 Nov 87 13:59:06 PST
From: Dwayne Virnau... <INFO-MAC-REQUEST@SUMEX-AIM.STANFORD.EDU>
Subject: MacKermit Status?
Keywords: Macintosh Kermit
Greetings from the INFO-MAC moderator, keeper of the sacred INFO-MAC
archives. Files in the archives are available to anyone on the ARPAnet or
BITNET, and specifically to some 10,000 people who subscribe to the INFO-MAC
mailing list. The latest version of Kermit for the Macintosh I have is
0.8(34), which does not work well (at all?) on the Mac2. So, a few
questions:
What is the latest version of Kermit for the Macintosh? How can I obtain a
copy of it? I have FTP access, and would prefer to avoid tape or disk
distribution.
And on a slightly different note, what is the proper way to make this
request? Since I am INFO-MAC I field hundreds of questions from people who
expect me to know every detal of the Macintosh, I apologize if INFO-KERMIT
is not the proper address. But I am also hoping you might be able to direct
me.
Many thanks,
Dwayne Virnau...
Moderator, Info-Mac
[Ed. - A new MacKermit was sent to us by Matthias Aebi of the Eidgenossische
Technische Hochschule in Zuerich, who also wrote Kermit for the Lilith
workstation, announced in Info-Kermit V6 #25, November 6, 1987. Since then,
Paul Placeway at Ohio State has been working on it. See the next message for
news about this.]
------------------------------
Date: Mon, 21 Dec 87 16:35:01 EST
From: paul@ohio-state.arpa (Paul Placeway)
Subject: Version 0.9(36)b4 of Macintosh Kermit Available for Testing
Keywords: Macintosh Kermit
Following is a BinHex 4-ified copy of MacKermit 0.9(36)b4. I have made
several changes to Matthias' code:
The C-Kermit part of 0.9(36)b4 is based on the (absolutly vanilla) 4E(067)
code. I didn't have to edit any of the ckc* files a bit (ckmpro.c is just a
warted ckcpro.w with the standard Mac patch).
The backslash-number characters in the key macro code are now done in Octal
(just like C does). Also, one can do control characters symbolically:
\^A --> Control-A.
Macro strings are stored as Pascal strings throughout now, so that one can
make a macro that includes NUL (ASCII 0).
MacKermit now has MultiFinder support. It understands how to give away time
slices, has the SIZE (-1) resource, and will do background file transfers
(timeslicing multiple times per packet). (In the process, the file xfer
dialog became modeless).
It has the FOND resource in the source files now (so SEs should be happy).
I added the extra stuff to the emulator so that VAX TPU will get along with
it.
Clayton Elwell added the ANSI insert multiple characters command
(ESC [ n @) to the emulator.
I also added a few extra bells and whistles:
. flashing/nonflashing cursor option
. visible bell option
. extra status indicators in the file xfer dialog,
. the mouse cursor is hidden on every key typed,
. the menu command characters option default is now dependant on the type of
keyboard that the user is running (set for keyboards with a CTRL key, unset
for all others), and
. there is now a default set of key macros and modifiers (arrow keys do VT100
arrow key commands, BACKSPACE -> DEL, backquote -> ESC, command-backquote
and control-backquote -> plain backquote).
Have fun,
-- Paul
[Ed. - Thanks, Paul! (And Matthias too!) This new version is supposed to run
on any Macintosh, and to correct the various problems that many of you have
reported not only with 0.8(34), but also some of the newer prereleases. It
should run on the entire Mac family with no special fussing about fonts or
other minutia. But it probably will not fit into an original 128K Mac (does
anyone still have such a thing?) The new Mac Kermit is still not a finished
product, however. Some finishing touches are required to the key definition
feature, and in some other areas too. And the manual has yet to be written.
Comments and reviews are welcome, and hopefully we'll have a final release
(maybe this time it'll actually be called 1.0!) soon, complete with source in
MPW C. The program and documentation (such as it is) are in KER:XKM936.* on
CU20B, and KXK936 * on CUVMA.]
------------------------------
Date: Thu, 3 Dec 87 10:53:25 EST
From: Phil Ritzenthaler <ritzenth%andy.bgsu.edu@RELAY.CS.NET>
Subject: Kermit News
Keywords: Kermit News, Newsletter
Is this newsletter still being printed? I just had Vol.1 No. 1 pass across
my desk (slow, aren't they!) and I was wondering if there have been other
issues?
[Ed. - Yes, we just mailed out Volume 2 Number 1 (the second issue). Anyone
who was on the mailing list for the first issue should also receive the
second one. Plus the thousands of people who have ordered Kermit from us
since August 1986, or who have requested to be added to the mailing list.
Meanwhile, this time (unlike the first) we have an on-line version, available
for FTP'ing, etc. It's in KER:NEWSV2.N1 on CU20B, or NEWSV2 N1 on CUVMA.]
------------------------------
Date: Fri, 4 Dec 87 09:23:36 mst
From: modular!earley@arizona.edu (Joe Earley)
Subject: Request for Kermit Information
Keywords: Kermit Protocol
Would you please send me some information about Kermit. I've heard good
things about it and am hoping Kermit's capabilities exceed those of a package
we have developed in-house.
Specifically, we are interested if Kermit can do the following:
o do file transfers which leave intact most VMS file header information,
o do file transfers between Unix and VMS and retain file attributes,
o handle arbitrary packet sizes to take advantage of clean lines,
o have a 'talk' mode to do remote logins,
o do automatic dialing and login to the remote host,
o do session logging,
o encoding of nonprintable characters to get around PBX's which act
upon control characters,
o do file transfers in an unattended batch mode.
We have only been on the net for about one month. Our only access to the
outside world is through arizona. We cannot directly get to an archive site
that I know of, so we can't get any previous information put into the Kermit
news group. Thanks for any information you can give us.
Joe Earley, Modular Mining Systems, Tucson, Arizona 85714
USENET: {ihnp4,allegra,cmcl2,noao}!arizona!modular!earley
INTERNET: modular!earley@arizona.edu
[Ed. - Kermit protocol, as defined, can do everything you ask. Actual
implementations, on the other hand, vary according to which features they
include. Let's look at your list:
o Do file transfers which leave intact most VMS file header information?
Currently, no. If you want to preserve VMS file headers, you have to hexify
the VMS files first and dehexify them on the other end, using a special
utility that comes with Kermit. Then you get the entire RMS FILES-11 FAB
structure, or whatever it's called. Future versions of VMS Kermit may
transfer this information directly, using Kermit's File Attribute mechanism
(as PDP-11 Kermit currently does).
o Do file transfers between Unix and VMS and retain file attributes?
Currently, no. In the future, VMS and Unix Kermit will be built from common
C-language sources, and should be able to handle file attributes.
o Handle arbitrary packet sizes to take advantage of clean lines?
Normal Kermit packets are 96 bytes long at most. Extended-length packets (a
different format) may be up to about 9K in length. Many Kermit programs
support this option, including C-Kermit for VMS and Unix. Some
implementations (see the CMS Kermit announcement above) even vary the packet
size according to prevailing line conditions.
o Have a 'talk' mode to do remote logins?
Yes, most Kermits -- including practically all PC or microcomputer Kermits --
include terminal emulation. So do VMS and Unix Kermit.
o Do automatic dialing and login to the remote host?
These features are found in some Kermits, but not all. Unix Kermit includes
both a "dial" command (with accompanying modem control) and a script language.
MS-DOS Kermit includes a script language, etc.
o Do session logging?
Most Kermit programs that perform terminal emulation can also do session
logging.
o Encoding of nonprintable characters to get around PBX's which act
upon control characters?
This is a hallmark of the Kermit protocol. It encodes all packets as lines
of printable text.
o Do file transfers in an unattended batch mode?
Yes. This is a built-in part of the Kermit protocol.]
------------------------------
Date: Tue, 15 Dec 87 04:46:33 PST
From: fayr%armory.DEC@decwrl.dec.com (Rich Fay SPO-103/1 POLE 1-6)
Subject: VAX/VMS Cluster and Kermit Problem
Keywords: VAX/VMS Kermit
My name is Rich Fay from Digital. I work at the Springfield Plant. We are
having a problem using KERMIT with our VAX cluster as described below:
We are using VMS 4.6 and have the last 3 versions of VMS Kermit including
V3.3.111. All of them exhibit the same problem I will now describe. We are
using MSKermit 2.29 and CPM ROBIN KERMIT V 4.05, (this problem only occurs
when logging in through a modem, or a Hardline that is connected through an
interface that uses modem protocols).
Sample Session follows;
dial up and logon
$Kermit
VMS Kermit-32 version 3.3.111
Kermit-32>set file type binary
Kermit-32>server
Kermit Server running on VAX/VMS host. Please type your escape sequence to
return to your local machine. Shut down the server by typing the Kermit BYE
command on your local machine.
12:12:58.14 %KERMIT32-F-TIMEOUT, device timeout
12:12:58.75 %KERMIT32-E-RECERR, Receive error - !AS
12:12:59.14 %KERMIT32-F-TIMEOUT, device timeout
12:12:59.56 %KERMIT32-E-RECERR, Receive error - !AS
12:12:59.96 %KERMIT32-F-TIMEOUT, device timeout
12:13:00.38 %KERMIT32-E-RECERR, Receive error - !AS
It seems that this problem only exists on vax systems that belong to a
cluster. I also have an RT11 system running KERMIT version 3.53 with RT11
verison 5.4 and am using a DZV11M on an 11/23+, the DZ is connected to a hard
line to the VAX. This system exhibits the same problem as the Robin and the
Rainbow. From home, I dialed up a system that is not on the cluster and all
works very well. I then try dialing up a system on the cluster and get
immediate failure.
I have posted this problem to the Digital KERMIT notes conference and noone
seems to have a solution to our problem. We are hoping that you will be able
to shed some light on this problem. Is there some special hardware/software
setups that must be done on the cluster to make KERMIT work properly??
Thanx in advance for any help you can offer to this problem.
Regards,
Rich Fay
Return address is: FAYR@HEFTY.DEC.COM.
[Ed. - As noted previously, VMS Kermit-32 (the Stevens version, written in
Bliss) is a "stable" product (to use the corporate euphemism). This message
was circulated to various VAX/VMS cluster sites, and they reported no such
problem. Can anybody reproduce it or suggest a cure? Meanwhile, development
on the new C-language VMS Kermit continues.]
------------------------------
Date: Sun, 20 Dec 87 15:15:39 PST
From: jeh@pnet01.cts.com (Jamie Hanrahan)
Subject: Non-Bug Report, MS-Kermit, CP/M Kermit (Kaypro)
Keywords: CP/M Kermit, Kaypro Kermit, MS-DOS Kermit
I've used both CP/M Kermit and MS-Kermit between my Kaypro 8 and my brand-new
HyperTurboWhizzoSchizo-AT clone to the VAXes at work and various other large
machines, but until recently, I'd never tried transferring files between my
machines; so when it came time to move some dBASE II databases to the AT
clone, I approached the project with some trepidation.
The AT follows the RS232C standard of using a male connector for a DTE port,
but the Kaypro (also wired as a DTE) does not, so I needed a gender changer
in addition to my modem eliminator cable (which has two females). Before
attempting file transfers I decided it'd be best to just CONNECT the two
machines and see if characters typed on one would appear on the screen of
the other.
This worked the first time.
Time to try transferring a .DBF file. Put the AT in RECEIVE mode, and SEND
from the Kaypro...
This, too, worked the first time. Wait a minute, things aren't supposed to
be this easy... I'll run this .DBF through dBASE III+'s II-to-III converter
and see if it hiccups on anything.
Looks perfect. The rest of the .DBFs, index files, and program files
went just as easily.
You people do good work!!!!
(I'm sure this is old hat to you, but with all the bug reports you must get,
I figured you'd appreciate something a little different...)
Merry Christmas,
--- Jamie Hanrahan
(uucp: {cbosgd | hplabs!hp-sdd | sdcsvax | nosc}!crash!jeh)
(arpa: crash!jeh@nosc.mil)
(internet: jeh@crash.CTS.COM)
[Ed. - Thanks, Jamie! It's always nice to get reports like yours.]
------------------------------
Date: Mon, 21 Dec 87 06:45:04 PST
From: adelman@LBL.Gov (Kenneth Adelman)
Subject: Kermit 3.79 on Apple 2c
Keywords: Apple II Kermit
> I have been using Apple Kermit 3.79 on an Apple 2c and find it works great!
> Terminal emulation and transfers at 2400 baud with no missing characters,
> VT100 emulation including keypad, the works! But there's one minor
> problem...
>
> While characters received from a remote system over the modem come through
> OK at 300, 1200, and 2400 baud, those generated by the modem itself are
> often garbled.
>
> Has anyone seen this problem? Is it a Kermit or a 2c problem? At first I
> thought it was a matter of not responding to interrupts in time, but since
> it even happens at 300 baud this doesn't seem likely.
This sounds like a familiar problem. In order to save a few dollars, the Apple
//c's serial interface was clocked with a signal which was already present on
the motherboard and very close to the right frequency rather than adding a few
dollar crystal and clocking the 6551 ACIA at the right frequency. As a result,
the Apple interface runs a few percent slower than the advertised baud rate,
and won't talk to some modems. This problem is not present on any of the other
Apples, and perhaps Apple fixed it on the later //c's by adding the crystal.
I seem to recall someone saying that Apple would fix the problem if you could
convince your dealer that it exists. Apparently the serial interface talks to
Apple's modems just fine. The other solution would be to find a 6551 ACIA spec
sheet which would tell you what frequency crystal you need and what pins on
the 6551 to connect it across. Presumably you would need to cut one trace and
add the crystal.
Ken
------------------------------
Date: Wed, 16 Dec 87 15:44:47 PST
From: senderow%janus.Berkeley.EDU@berkeley.edu (Dan Senderowicz)
Subject: IBM 370 Mainframe UTS24-Kermit?
Keywords: UTS24 Kermit, IBM 370 Kermit
Does anyone have a working version of Kermit for a 370 mainframe running
UTS24? Thanks. Dan.
[Ed. - C-Kermit has code in it to support UTS24 (a kind of half-duplex Version
7 Unix for IBM mainframes from Amdahl Corp), and this code was used at several
sites in the past, but apparently it does not work at Berkeley. Can anybody
who's still running UTS24 help? This is one case where "portable 370" Kermit,
announced above, probably does not apply.]
------------------------------
Date: 8 Dec 87 07:58:24 GMT
From: amdcad!amdcad.AMD.COM!indra@ames.arpa (Indra Singhal)
Subject: Kermit & Everex Modems?
Keywords: Everex Modem, Internal Modem, MS-DOS Kermit
I just began subscribing to this group. I was told that there
had been a posting about a quirk in Everex modems that interfered
with proper Kermit operation on their modems. If any on of you
has a copy of the proceedings of the discussion, please
e-mail to me.
Thanks...
-I said so... & said it for myself. Indra K. Singhal
{ucbvax,decwrl,allegra}!amdcad!indra or amdcad!indra@decwrl.dec.com
[Ed. - MS-DOS Kermit 2.29 for the IBM PC family did indeed have trouble
dealing with Everex or Hayes half-card internal modems. Version 2.29B and
later fix the problem. Try the current 2.29C release in KER:MSTIBM.*.]
------------------------------
End of Info-Kermit Digest
*************************
-------