SY.FDC@CU20B.COLUMBIA.EDU (Frank da Cruz) (11/14/85)
Info-Kermit Digest Wed, 13 Nov 1985 Volume 3 : Number 29
Today's Topics:
Kermit the Frog
Kermit the Book
New Kermit-11 Coming
Kermit for NEC APC-3
C-Kermit 4C(057), Mackermit 0.8(33), and 2.9BSD
Okstate's Kermit Server Active Again
Binary File Transfer Between C-Kermit and VMS Kermit-32
Kermit Problems, Notes and Praises
Osborne Executive Kermit?
Kermit & MUSIC SP
----------------------------------------------------------------------
Date: Tue 12 Nov 85 15:26:44-EST
From: Frank da Cruz <SY.FDC@CU20B.COLUMBIA.EDU>
Subject: Kermit the Frog
The new (December 1985) MACWORLD Magazine features a cover story on Kermit
-- no, not Kermit the file transfer protocol, but Kermit the Frog, the
proprietor of a whole new line of Muppet-based educational software.
"Kermit" is a trade mark for some of this software, like "Kermit's
Electronic Story Maker." This is one among the many good reasons why "our"
Kermit should (and can) not become a commercial product.
------------------------------
Date: Wed 13 Nov 85 17:02:22-EST
From: Frank da Cruz <SY.FDC@CU20B.COLUMBIA.EDU>
Subject: Kermit the Book
I'm writing a Kermit book for Digital Press. I hope it will be out in
Spring or Summer 1986, and I hope the price will be relatively low (it
can't be set yet, because the manuscript isn't done, but there is general
agreement that it should not be a high-priced item). I hope the book will
be a big improvement over the Kermit User Guide and Protocol Manual; it
will contain all the information from these manuals, plus a lot more --
background in data communications, file organization, etc; more cohernet
organization, lots of illustrations, tables, code fragments, etc etc. But
it will lack the detailed descriptions of the various Kermit programs found
in the User Guide; that information will continue to be supplied along with
each program. Anyhow, the idea is for the book to "work" for everyone from
the utter novice to the Kermit programmer, pehaps in conjunction with a
program-specific handout, and to promote Kermit a little more as a serious
protocol and maybe encourage future Kermit program contributors to stick to
the "standard" command syntax a little better and provide code examples to
make it easy for them to include server mode, 8th-bit quoting, etc etc.
Preparing the manuscript, plus doing my "real job," is keeping me pretty
busy, which partly explains the lag between Info-Kermits (the rest of the
explanation is that there hasn't been a lot of activity recently anyway).
Occasionally I might bother someone for a bit of specific information to
round out some table or other. Here's one tidbit I haven't been able to
dig up anywhere -- what multiplexing technique is used by VA-3400 1200bps
modems (frequency division?) and at what baud rate does it transmit (some
1200bps modems actually transmit at 600 baud). Here's another one -- does
anyone know the etymology of "D-Connector" or "DB-xx" (as in DB-25)?
Also, if anyone wants to fill in the following questionnaire and send it
back to me, I'd be most grateful. This is just for purposes of filling in
some illustrative tables, contrasting the diverse communication and
file architectures of various machines and OS's. I've already got info
for the common ones, like DEC-20, VAX/VMS, IBM VM/CMS, MS-DOS, etc etc, but
am looking more for the "strange" ones -- HP minis, DG minis, P-E minis,
Prime computers, mainframes of Burroughs, Honeywell, Sperry, CDC, etc.
Machine, Model(s):
Operating System:
Version of OS following info applies to, if it makes a difference:
Machine Word Size:
Text Byte Size, and how many bits within byte are used for a character, and
what happens to any leftover bits:
Directory structure (flat file system, 1 level of directory, hierarchical):
Filespec format (indicate punctuation and max length of each field, e.g.
DEV:[DIR]NAME.EXT;n with DEV=6, DIR=12, NAME=8, EXT=3, n=numeric generation.
Text code: (7-bit ASCII, 8-bit extended ASCII, 8-bit EBCDIC, etc):
Normal record separation technique for text files (CRLF, LF, CR, RCW, fixed
block, etc):
EOF indication (nearest character, word, block); what happens at EOF if
EOF not recorded exactly (e.g. CP/M ^Z trick):
Asynchronous Communications (But first please indicate if the system
normally prefers some other style, like IBM-style synchronous block mode
terminals):
Duplex (full, half):
Flow Control (half duplex line turnaround handshake char, XON/XOFF, ENQ/ACK,
ETX/ACK, etc):
Required or default parity:
Special or unusual characteristics worth noting about file system,
text representation, or communications:
Thanks to anyone who responds! Also, any suggestions of a general (or
specific) nature would be most welcome, especially from people who have had
problems with some aspect of Kermit that could have been avoided by better
documentation.
- Frank
------------------------------
Date: 8 Nov 1985
From: BRIAN@UOFT02.BITNET
Subject: New Kermit-11 Coming
There currently is a version 2.38 of Kermit-11 available. The version will
only be available via Bitnet or dialup to UOFT02 until December 85 as I am
waiting for modifications from some other people regarding M+ v3 and also
PRO/TMS support. As always, the file K11CMD.MAC has the edit trail.
How to get it:
Bitnet:
from VM/CMS: CP SMSG RSCS MSG UOFT02 KERMSRV DIR
CP SMSG RSCS MSG UOFT02 KERMSRV SEND K11*.*
from VMS Jnet: $ SEN/REM UOFT02 KERMSRV SEND K11*.*
Dialup:
(419) 537-4411
Service class VX785A
User: KERMIT
Password: KERMIT
Source and hex files are in KER:, binaries are in KERBIN:
[Ed. - I imagine Brian will be sending the successor to 2.38 to Columbia
when it's ready; will announce it when it arrives.]
------------------------------
Date: Fri, 8 Nov 85 13:55:55 est
From: Phil Johnson <johnson@LL-XN.ARPA>
Subject: Kermit for NEC APC-3
Since you had the source but not the binary for NEC APC-3 Kermit,
I built the binary for you.
Phil Johnson
[Ed. - Thanks! It's in KER:MSVAP3.BOO ("boo" file) and KB:MSVAP3.EXE
(8-bit binary).]
------------------------------
From: Vic Abell <abe@purdue-asc.ARPA>
Date: 8 Nov 1985 1438-EST (Friday)
Cc: abe@purdue-asc.arpa, ach@purdue-asc.arpa, acu@purdue-asc.arpa
Subject: C-Kermit 4C(057), Mackermit 0.8(33), and 2.9BSD
C-kermit 4C(057) does not work properly with 2.9BSD, as distributed
by Berkeley. It may work with the Harvard, Seismo distribution, but
I haven't tested that.
One problem is that the Berkeley 2.9BSD distrubution does not support
the 4.2BSD timeval and timezone structures and the functions that
surround them. Thus, all of the #if tests in ckutio.c that assume
2.9 to 4.2 equivalence in that area must be undone.
A second problem is that the protocol rule for <rdata>Z in ckcpro.w
tests for a return value from the function reof() in ckcfns.c Reof()
does not return a value. The protocol rule works on 4.2BSD out of
happenstance - apparently the return value register happens to be
non-negative. Under 2.9BSD it doesn't work - apparently because
the return value register happens to be negative.
I am enclosing diffs that show the changes necessary to eliminate
these two problems. The diffs for ckcpro.w also include a fix to the
<rfile>B protocol rule that I reported earlier to this mailing group.
It prevents the EOT ack from being lost in interactive mode when the
terminal is switched from cbreak to cooked. Note that the sleep time
was raised to five seconds on the slower 11/70s.
Vic Abell, abe@asc.purdue.edu
[Ed. - Thanks! Diffs omitted, appended to KER:CKUKER.BWR, will be
included in the next release.]
------------------------------
Date: Sun, 10 Nov 85 17:42:19 CST
From: Gregg Wonderly <gregg%okstate.csnet@CSNET-RELAY.ARPA>
Subject: Okstate's Kermit Server Active Again
Due to an obscure bug in some modifications made to the KERMIT SERVER at
OKSTATE, the server has not been able to service GET requests from users.
This problem has been located, and fixed. We appologize for any headaches
caused (We had a few here in finding this one). Thanks to those who pointed
the problems out initially.
Gregg Wonderly
Department of Computing and Information Sciences
Oklahoma State University
UUCP: {cbosgd, ea, ihnp4, isucs1, mcvax, uokvax}!okstate!gregg
ARPA: gregg%okstate.csnet@csnet-relay.arpa
------------------------------
Date: Fri, 1 Nov 85 21:55:26 pst
From: daver@cit-vax.ARPA (David Robinson)
Subject: Binary File Transfer Between C-Kermit and VMS Kermit-32
I am having troubles transfering files to and from a SUN-2/170 running
Ckermit 4C(57) and a VAX/VMS3.7 running Kermit-32(66).
Transfering both ways one side or the other is mapping ALL ^J's to ^M's
which reaks havic on my binary data. I have set file type to binary on
both machines with no success. When I set file type fixed on the VAX (A
suggestion from a friend) The file makes it thru the data transfer part but
dies on the final handshaking with an error "unimplemented server command".
We are constrained to use only the VAX as a server, our connection allows
logins only on the VAX side.
Any and all help on this problem would be appreciated.
"How to get raw binary files from Ckermit to kermit-32?"
David Robinson
daver@cit-vax.arpa
[Ed. - I've heard complaints like this before, but don't have a VMS system
to track them down with, and the folks at Stevens Inst of Tech are badly
bogged down in other work for the time being. Anybody out there can help?]
------------------------------
Date: Wed 6 Nov 85 21:25:42-EST
From: Christopher Lent (LENT@CUPHOA.CCNET)
Subject: Kermit Problems, Notes and Praises
PROBLEMS:
*****
MSRVRB1.BOO - MS-DOS Rainbow-100 boot file.
The first line of KER:MSVRB1.BOO contains:
KER:MSRB10.EXE
rather than
MSVRB1.EXE
A minor oops but it could give a first time user a problem.
[Ed. - Thanks, I fixed it.]
*****
MSSDEF.H - MS-DOS KERMIT macro assembler (MASM) source header file.
Perhaps a stronger warning should be included that KER:MSSDEF.H
must be renamed to MSDEFS.H before compiling.
[Ed. - Thanks, I put warnings in appropriate places.]
*****
MSIBMP.EXE - Compiling your own:
I compiled MSIBMP.EXE on an IBM/PC-XT using V2.0 MASM with the suggested /S
option and have experienced no problems with the resultant V2.28 IBM-PC KERMIT.
*****
notes:
+++++
MSIBMP.EXE - IBM/PC KERMIT through a DECServer 100:
I have been using KERMIT V2.27 and now V2.28 through DEC's new
DECServer 100 terminal server. All I can say is WOW!! The configuration
consists of a VAX-11/780 with VMS 4.1 running KERMIT-32 V3.1.066 and IBM/PC's,
and XT's running PC-DOS 3.1 and KERMIT V2.27 and V2.28.
Kermit works fine with the initial settings on the terminal server.
The server doesn't interfere with the KERMIT protocol. KERMIT file transfers
exhibit no visible loading of either the terminal server or VMS, even at 19.2K
baud. The local CWD command has greatly aided doing incremental PC/XT backups
to corresponding VMS directories.
+++++
Note - Converting "BINARY" Text files on VAX/VMS created by KERMIT-32.
Most users doing backups of PC's to our VAX use the KERMIT-32 SET
FILE TYPE BINARY option so the .COM and .EXE files will transfer properly.
This results in a minor problem when text file transferred as "BINARY" files
are to be edited or revised on the VAX. A workaround procedure for this
follows using the age-old editor TECO. Since TECO's calling sequence has
changed in VMS 4.x two versions are given.
VMS 3.x:
$ ! Untested
$ TECO :== $SYS$SYSTEM:TECO.EXE
$TOP:
$if "''p1'" .eqs. "" then inquire p1 "Binary file to be converted to text"
$if "''p1'" .eqs. "" then goto TOP
$WRITE SYS$OUTPUT "Converting file ''p1' to text format."
$! the line after edit/teco 'p1' should contain EX<ESC><ESC>
$! where <ESC> is the escape character (decimal 27, octal 33, hex 1B)
$TECO 'p1'
EX$$
$write sys$output "File ''p1' converted to text format."
$exit
VMS 4.x
$! Works!
$TOP:
$if "''p1'" .eqs. "" then inquire p1 "Binary file to be converted to text"
$if "''p1'" .eqs. "" then goto TOP
$WRITE SYS$OUTPUT "Converting file ''p1' to text format."
$! the line after edit/teco 'p1' should contain EX<ESC><ESC>
$! where <ESC> is the escape character (decimal 27, octal 33, hex 1B)
$edit/teco 'p1'
EX$$
$write sys$output "File ''p1' converted to text format."
$exit
Usage:
If the above file is called CVT.COM in the current directory
simply type type following to convert the "BINARY" text file FILE.BIN
$ @CVT.COM FILE.BIN
and the new version of file.bin will be in VMS's default text file format.
Beware:
If the "BINARY" text file has lines which exceed 255 characters,
VMS's EDIT/EDT editor will truncate these lines during editing.
+++++
Praises:
#####
Ckermit - general praises
Nice Job!!! I was previously using the version now relegated to the
to the documentation. The new version work fabulously on our AT&T 3b5
computer. I've even managed to get a version running on a V6 UNIX system.
The V6 version was compiled using a modified V7 compiler so I doubt it would
be of any use to others. With quite a bit of emulating V7 system calls,
I produced a working version of Ckermit. Just a warning, in split I/D spaces
on the PDP-11, some V6 system calls do NOT work.
#####
Final comments:
Kermit is fabulous. I now try to get it to friends as a matter of
course. At first they're a little confused as to why they need Kermit, but
about a month after they come back and say "Wow, how can I get Kermit for
machine XYZ-123/X?" and we figure out how to load the initial copy.
I must admit I'm a convert. When I originally read about Kermit in
BYTE, I thought "Oh, yet another ASCII file transfer package". Reading on I
thought, "Hmm, sounds pretty easy to implement". Then I put these thoughts on
the back burner for a few years.
In March of 1985, I found Kermit source on a FIDO BBS system and tried
it on the PC. It worked wonderfully. They also had 'C' source for a Unix
Version of Kermit. I loaded it on our 3b5 and our V6 system and it worked with
only minor modifications. The BYTE articles made the debugging much more
simple.
Keep up the good work!
Chris Lent
lent@cuphoa.ccnet
Care of:
oc.pedhem@cu20b
------------------------------
Date: Wed, 6 Nov 85 20:38:02 pst
From: George Cross <cross%wsu.csnet@csnet-relay.arpa>
Subject: Osborne Executive Kermit?
Can someone inform me of the status of the Osborne Executive version
of Kermit? Browsing through the mail archives (83/84 Epoch) one finds
a flurry of interest when the OE was alive but it seems to have faded
out. My questions are:
1. Does the Osborne 1 version work on the Osborne Executive?
2. Does the Generic CPM version work on the Executive? If so how?
I don't receive this list regularly, so a mail reply to me would be preferred.
Thanks,
George R. Cross cross@wsu.CSNET
Computer Science Department cross%wsu@csnet-relay.ARPA
Washington State University faccross@wsuvm1.BITNET
Pullman, WA 99164-1210 Phone: 509-335-6319 or 509-335-6636
------------------------------
Date: Fri, 08 Nov 85 08:57:21 cet
From: PCSC%WUVMD.BITNET@WISCVM.ARPA
Subject: Kermit & MUSIC SP
Does anyone know if Kermit-MUSIC 1.0 is compatible with the new
release of MUSIC (MUSIC SP)? I realize that it contains its own
communications program PCWS, but sites with several operating systems
like ours would prefer to use one program (Kermit!) for all
micro to mainframe communication if possible.
Michael Palmer, Washington University Computing Facilities
BITNET address: PCSC@WUVMD
------------------------------
End of Info-Kermit Digest
*************************
-------